Methods & Products Posters

Landsat is a joint USGS and NASA space program for Earth Observation (EO), which represents the world’s longest running system of satellites for moderate-resolution. The European Space Agency (ESA) has acquired Landsat data over Europe, Northern Africa and the Middle East during the last 40 years.

A new ESA Landsat Multi-Spectral Scanner (MSS), Thematic Mapper (TM) and Enhanced Thematic Mapper Plus (ETM+) processor has been developed. This enhanced processor aligns historical Landsat products to the highest quality standards that can be achieved with the current knowledge of the instruments. The updated processor is mainly based on the USGS algorithm; however it has some different features that are detailed in this paper.

Current achievements include the processing and availability of approximately 860 000 new TM/ETM+ high-quality products between 1983 and 2011 from the Kiruna (S), Maspalomas (E) and Matera (I) archives;  Matera includes data from the Fucino (I), Neustrelitz (D), O’Higgins (Antarctica), Malindi (Kenya), Libreville (Gabon) and Bishkek (Kyrgyzstan) ground stations.

The products are freely available for immediate download to the users through a very fast and simple dissemination service (at: https://landsat-ds.eo.esa.int/app/) and through ESA’s browsing system, EOLI. The remaining MSS data, dating back more than 40 years, will gradually become available during 2016.

The ESA Landsat processor algorithm enhancement, together with the results of the ESA archive bulk-processing data regarding production, quality control and data validation are herein presented.

The Microwave Radiometer (MWR) flown on Envisat, ERS-1 and ERS-2 provides a nearly uninterrupted time series of microwave observations over a period almost 21 years between 1991 and 2012. This dataset is complementary to other microwave datasets. Despite its nadir-only coverage it provides an opportunity to independently provide constraints on total column water vapor (TCWV) and cloud liquid water path (LWP).

Here we report on our efforts towards a fully inter-calibrated and validated physical retrieval of TCWV and LWP for MWR. We will address issues related to satellite inter-calibration, homogeneity of the derived time series of brightness temperatures, observation-simulation biases, and provided first results of fully physical retrievals. An outlook on planned activities will be given and the importance of the results in light of the upcoming Sentinel-3 mission will be discussed. 

This paper presents the scientific and technical aspects of the Level 2A (atmospheric/topographic correction) for the Sentinel-2 Simplified Level 2 Product Prototype Processor (S2 SL2 PPP). Design aspects are partly fixed by the ESA as main customer. The developed chain based on ATCOR is used for the estimation of the following products: Atmosphere type, Bottom of atmosphere reflectance (including cirrus detection and correction), Aerosol optical thickness, and Water vapor.

Being a mono-temporal correction chain ATCOR requires a selection of the spectral bands for the estimation of Aerosol type, Aerosol optical thickness based on the dense dark vegetation method and Water vapor based on the atmospherically pre-corrected differential absorption method as well as an estimation of the best parameter set for these methods. The parameter set was estimated by a sensitivity analysis on a simulated top and bottom of atmosphere radiance/reflectance data based on radiative transfer simulations. The aerosol type is estimated by the comparison of the path radiances ratio to the ground truth path radiances ratio for the standard atmospheres, namely rural, urban, maritime, and desert. Aerosol optical thickness map and Water vapor map are initially estimated on the 20m pixel size data, then the maps are interpolated to the pixel size of 10m and the 10m reflectance data are estimated. The cirrus cloud map is created by the cirrus 1.38 µm band thresholding to the thin, medium, thick cirrus and cirrus clouds. Cirrus compensation is performed by correlating the cirrus band reflectance to the reflective region bands and subtraction of the cirrus contribution per band.

Validation of the chain is performed given the top of atmosphere data (as input) and bottom of atmosphere products (the reference). Estimated reflectance is assessed given the ground truth reflectance, Aerosol optical thickness is validated given the AERONET measurements, cirrus correction is validated using a pair of Landsat-8 scenes acquired for the same area with a small time difference. One scene is contaminated by cirrus cloud that has to be restored, while the other is cirrus free and used as reference. A comparison of the estimated products is also performed with an alternative atmospheric correction chain – FLAASH. A good agreement with the ground truth measurements as well as the alternative atmospheric correction chains is reached.

The GLISTEREO mission concept aims to derive submesoscale information using rapid repeat sunglint observations over fronts and eddies of the global ocean.

The objectives of the mission are:
-Derive precise surface deformation at fine resolution (order 1 km, 75 km wide swaths) with repeat sunglint observations of submesoscale surface roughness contrasts at scales from about 50 m to 200 m.
-Simultaneously derive wind speed and direction from sunglint, to identify the wind-forced component over submesoscale ocean features.

Here we discuss the mission concepts which are:
-Fine resolution, panchromatic-like cameras, inclined with different incidences, carried on 2 (or 3) closely following small satellites.
-Polarization could improve derivation of wind speed and direction from sunglint.
-Two (or three) payloads flying in close proximity along same orbit track can image same sunglint areas with different geometry angles at slightly different times (sufficient) to enable measurable offsets between identified features, and then used to derive surface current vectors and precise surface deformation, simultaneously with wind speed and direction.

On average less than half of the carbon dioxide (CO2) emitted by human activities stays in the atmosphere, with the remaining balance taken up by global oceans, terrestrial vegetation and soils. There are substantial uncertainties associated with the location, strength and durability of these natural components of the carbon cycle. The tropics and sub-tropics are regions of key importance for the global carbon cycle. These diverse and productive ecosystems are subject to rapid environmental change due to extensive deforestation and urbanisation, with consequent changes in hydrology and regional carbon balance. However, CO2 flux estimates of the tropics and southern-hemispheric sub-tropics are poorly constrained by the existing network of surface measurements. Consequently, how the tropics affect the global carbon budget remains poorly characterized compared to other ecosystems on the planet, and currently we cannot determine with confidence if tropical ecosystems are a net source or a sink of carbon.

We present the Tropical Carbon Mission (TCM) concept that will address the unmet technical challenge to provide observations of CO2 over the tropics with the precision and frequency that are required by scientists and policy makers. TCM will be launched in a low-inclination orbit of 35 degrees, which will precess within tropical latitudes to maximize the frequency and coverage cloud-free scenes over and downwind of tropical landmasses. The TCM comprises three instruments developed from current technology: (1) a short-wave IR (SWIR) spectrometer that will measure CO2, carbon monoxide (CO), methane (CH4), and oxygen (O2); (2) a co-boresighted aerosol spectrometer; and 3) a cloud imager. The maturity of the technology minimizes the overall risks and costs to the mission. Aerosols represent the largest source of systematic error in the retrieval of CO2 at SWIR wavelengths.  We adopt a multi-view approach, building on work developed by MISR, to improve the characterization of biomass burning atmospheric aerosols prevalent over tropical latitudes.

The primary science objective of TCM is to reduce the overall uncertainties in the magnitude and distribution of tropical CO2 fluxes, and to improve our understanding of the tropical carbon cycle. By better estimating tropical fluxes, we also improve the efficacy of existing surface measurement networks to help estimate extra-tropical CO2 fluxes. The secondary science objectives of TCM are to: 1) reduce the uncertainties in the magnitude and distribution of CO and CH4; and 2) improve source attribution of observed variations in CO2 by using concurrent measurements of CO and CH4. The tertiary science objective of TCM is to complement global survey CO2 measurements from low-Earth orbiting instruments such as GOSAT and NASA-2.

One of the main objectives of the Sentinel-3 mission is to measure sea- and land-surface temperature with high-end accuracy and reliability in support of environmental and climate monitoring in an operational context. Calibration and validation are thus key criteria for operationalization within the framework of the Sentinel-3 Mission Performance Centre (S3MPC).

Land surface temperature (LST) has a long heritage of satellite observations which have facilitated our understanding of land surface and climate change processes, such as desertification, urbanization, deforestation and land/atmosphere coupling. These observations have been acquired from a variety of satellite instruments on platforms in both low-earth orbit and in geostationary orbit. Retrieval accuracy can be a challenge though; surface emissivities can be highly variable owing to the heterogeneity of the land, and atmospheric effects caused by the presence of aerosols and by water vapour absorption can give a bias to the underlying LST. As such, a rigorous validation is critical in order to assess the quality of the data and the associated uncertainties.

The Sentinel-3 Cal-Val Plan for evaluating the level-2 SL_2_LST product builds on an established validation protocol for satellite-based LST. This set of guidelines provides a standardized framework for structuring LST validation activities, and is rapidly gaining international recognition. The protocol introduces a four-pronged approach which can be summarised thus: i) in situ validation where ground-based observations are available; ii) radiance-based validation over sites that are homogeneous in emissivity; iii) intercomparison with retrievals from other satellite sensors; iv) time-series analysis to identify artefacts on an interannual time-scale. This multi-dimensional approach is a necessary requirement for assessing the performance of the LST algorithm for the Sea and Land Surface Temperature Radiometer (SLSTR) which is designed around biome-based coefficients, thus emphasizing the importance of non-traditional forms of validation such as radiance-based techniques. Here we aim to present application of the protocol to LST data produced during the commissioning of Sentinel-3A.

We present the first results from the GLORIE campaign (GLObal navigation satellite system Reflectometry Instrument Experiment) that took place in June and July 2015 over the southwest of France.

The goal of this measurement campaign was to assess the capabilities of airborne dual-polarization GNSS reflectometry for the estimation of land parameters such as soil moisture,vegetation biomass and water reservoir height.

Six flights took place in a 3-week span, with a total of 15 hours of raw data recorded thanks to the  L-band dual polarization GLORI GNSS reflectometer aboard the SAFIRE ATR-42 aircraft.

During these flights several flyovers of lakes (Etang de Biscarosse, Lac de Cazaux) were performed to assess the altimetry performance of the instrument. Reflected signal code delay and phase unwrapping methods were used, together with precise positioning techniques in order to estimate lake height. The processing of the data lead to centimeter level accuracy over calm waters.

The TerraSAR-X mission has been contributing to Copernicus Space Component Data Access (CSCDA) Data Warehouse (DWH) project since 2010. For this purpose, the mission has established required interfaces to the ESA Coordinated Data Access System (CDS) and the Copernicus Service Projects (CSPs). The mission subsystems involved in CSCDA interfaces are TerraSAR-X Commercial Service Segment (TSXX) and TerraSAR-X Ground Segment (TSX G/S). TSXX is operated by Airbus DS Geo GmbH and TSX G/S by DLR e.V.

In Phase 1 of the project, catalogue browsing/ordering and future tasking/ordering for TerraSAR-X data relied on the human-to-human interface, where ESA contacted TerraSAR-X Customer Service via phone or email. For Phase 1, TerraSAR-X Copernicus Contributing Mission (TSX-CCM) also implemented necessary adjustments and add-on workflows in TSXX in order to be compatible with CSCDA specific scenarios, ordering, product delivery, and reporting, which can be considered as the initial steps for machine-to-machine interface between TSXX and CDS.

For Phase 2, TSX-CCM upgraded the interface baseline and developed a machine-to-machine interface based on HMA (Heterogenous Mission Accessibility) standard between TSX CCM and ESA CDS. HMA Catalogue Browsing Service is deployed in TSX G/S, and HMA Catalog Ordering in TSXX. These services, operational since the end of 2011, have enabled ESA to browse the TerraSAR-X catalogue and automatically order TerraSAR-X archive products by using CDS user interfaces, i.e. without need for any human interaction at TSX CCM side.

In Phase 3, TSX-CCM will introduce the HMA Sensor Planning Service (SPS) and HMA Future Ordering Service for future acquisition requests in addition to the existing HMA Catalogue Browsing and HMA Catalogue Ordering Services implemented in Phase 2. These will enable ESA to task and order TerraSAR-X future acquisitions via the machine-to-machine interface. In this way, ESA can automatically get feasibility analysis about the possible acquisitions for a desired Area of Interest (AOI) and time period, submit, and monitor future orders 24hours/7days via CDS user interfaces. Since no human interaction is needed, these services are expected to be especially useful in emergency cases, where there is a need of short service response time and immediate satellite tasking capability.

This paper gives a background about the past activities performed by TSX-CCM, provides a detailed overview of ongoing activities, and discusses lessons learnt as well as future opportunities in terms of HMA interfaces for the TerraSAR-X Mission.

Conventional and along-track interferometric (ATI) Synthetic Aperture Radar (SAR) sense the motion of the ocean surface by measuring the Doppler shift of reflected signals. Together with the water displacement associated with ocean currents, the SAR measurements are also affected by a Wind-wave induced Artefact Surface Velocity (WASV) caused by the velocity of Bragg scatterers and the orbital velocity of ocean surface gravity waves. The WASV has been modelled theoretically in past studies but has been estimated empirically only once using Envisat ASAR.

Here we propose, firstly, to evaluate this WASV from airborne ATI SAR data and secondly, to examine the best inversion strategy for a dual-beam concept to retrieve accurately both the ocean surface current vector (OSCV) and the wind vector in the frame of an OSC satellite mission.

The airborne ATI SAR data were acquired in the tidally dominated Irish Sea using a Wavemill-type dual-beam SAR interferometer. A comprehensive collection of airborne Wavemill data acquired in a star pattern over a well-instrumented site made it possible to estimate the magnitude and dependence on azimuth and incidence angle of the WASV. The airborne results compare favourably with those reported for Envisat ASAR. In light winds (5.5 m/s), the wind-wave induced contribution to the measured ocean surface motion reaches up to 1.6 m/s upwind, with a well-defined 2^nd order harmonic dependence on the direction with respect to the wind.

The inversion strategy points to the need for accurate measurement of both the backscatter amplitude and the Doppler information (either as a Doppler centroid frequency anomaly for SAR DCA, or as an interferometric phase for ATI) as well as the need for dual polarization capability (VV+HH). Preliminary inversion results show that the retrieval accuracy for OSC velocity better than 10 cm/s can be achieved but that the OSC accuracy is strongly sensitive to the wind direction relative to the antennas orientation.

Sea Surface Temperature Products from Sentinel-3 Sea and Land Surface Temperature Radiometer are planned from the Marine Centre, part of the Sentinel-3 Payload Data Ground Segment, located at the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT). The Marine Centre together with the existing EUMETSAT facilities provide a centralised operational service for operational oceanography as participation in the European Commission’s Copernicus programme.

Activities towards operational oceanography over the next twenty years at EUMETSAT also include the continuation of the Mandatory Programmes (MSG, EPS) and future (MTG, EPS-SG). All include ocean observations of Sea Surface Temperature.

An overview of the oceanography activities at EUMETSAT relating to Sea Surface Temperature will be described. This includes information on activities at EUMETSAT central facilities; the EUMETSAT Ocean and Sea-Ice Satellite Application Facility (OSI-SAF); and those in preparation for the Sea and Land Surface Temperature radiometer (SLSTR) on the Copernicus Sentinel-3 satellite.

The presentation will include further information on the SLSTR SST product, which has been developed together with ESA and industry. Since the Sentinel-3 launch is currently scheduled for late 2015, the presentation will give a preliminary overview of any available SST results from Calibration and Validation activities performed at EUMETSAT. These activities contribute to the Sentinel-3 Mission Performance Framework, in collaboration with the European Space Agency’s Mission Performance Centre and the joint ESA/EUMETSAT SLSTR Quality Working Group.

In addition the presentation will describe further work in support of the continuation and evolution of sea surface temperature products from SLSTR. This includes collaboration of near real time in situ datasets for SST validation, and activities towards better drifting buoys for satellite validation.

There are currently no Ice Surface Temperature products planned in the initial core operational baseline for Sentinel-3 SLSTR, although requirements for IST have been gathered by the European Space Agency. The presentation will present work and studies towards sea-ice surface temperature and marginal ice zone temperatures from SLSTR.

Details on how to find further information will be presented, and opportunities described on how to participate in the Sentinel-3 Validation team for Temperature.

Faraday rotation is the change in polarization of an electromagnetic signal as it propagates through the ionosphere.  At L-band (1.4 GHz) this change can be significant and it can be an important consideration in the retrieval of soil moisture and ocean salinity.  Faraday rotation can be quite variable and the error in values derived from models can be large, especially over the ocean where in situ data is limited.  These are among the reasons that all of the recent L-band radiometers launched in space (SMOS, Aquarius and SMAP) included radiometric channels to estimate the rotation angle in-situ.  This can be done using the ratio of the third Stokes parameter to the second Stokes parameter as proposed by Yueh [Yueh, 2000] and demonstrated with measurements from Aquarius [Le Vine et al, 2013].  

The retrieval of the Faraday rotation angle from SMOS polarimetric channels has been difficult to verify because of the relatively noisy measurements themselves.  As a result alternative approaches have been reported [Corbella et al, 2015; Vergely et al, 2015].  In this manuscript the noise in the image “snap shot” is reduced by first calculating the TEC (total electron content) associated with each pixel in a given SMOS “snap shot” and averaging the value of TEC.  The TEC is relatively slowly varying in contrast to the Faraday rotation which is strongly dependent on the orientation of the propagation path from the satellite to pixel on the surface.  Noise is further reduced by including several overlapping snap shots in the average.  The average value of TEC (i.e. a single value for each snap shot) is then used to re-compute the Faraday rotation for each pixel in the snap shot.

To verify the results, the Faraday rotation computed in this manner has been compared with theory and with the results obtained by Aquarius during close overpasses of the two sensors.  Theory consists of using the TEC provided by the International GNSS Service [http://www.igs.org] and the local Earth magnetic field.  Results have been obtained for 7overpasses using the latest version (V620) of the SMOS data.  Good agreement is obtained in the southern hemisphere (a case of summer in the south).  In this case, the values obtained from Aquarius and from SMOS and derived from theory, are in good agreement.  The results for the northern hemisphere are not as good.  In the northern hemisphere, Aquarius and theory are again in good agreement; however, the values of TEC obtained from SMOS are consistently low and lead to a low value of Faraday rotation angle.  At the time of preparation of this abstract the reason for the poor agreement in the northern hemisphere is unresolved.

S.H. Yueh, “Estimates of Faraday rotation with passive microwave polarimetry for microwave remote sensing of earth surfaces,” IEEE Trans. Geosci. Remote Sens., Vol 38 (#5), pp. 2434-2438, September, 2000.

D.M. Le Vine, S. Abraham, C. Utku and E.P. Dinnat, “Aquarius Third Stokes Parameter Measurements: Initial Results”, IEEE Geoscience and Remote Sensing Letters, Vol.10, no. 3, pp 520-524, May 2013.

I. Corbella, L. Wu, F. Torres, N. Duffo and M. Martin-Neira, “Faraday Rotation Retrieval Using SMOS Radiometric Data”, IEEE Trans. Geosci. Remote Sens., Vol.12, no. 3, pp 458-461, March, 2015.

J-L Vergely, P. Waldteufel, J. Boutin, X. Yin, P. Spurgeon and S. Delwart, “New Total Electron Content Retrieval Improves SMOS Sea Surface Salinity”, Journal Geophysical Research: Oceans”, Vol 119, pp 7295–7307, October, 2014, doi:10.1002/2014JC010150.

Resolving 3-dimensional measurements of surface displacement from interferometric SAR (InSAR) requires the fusion of multiple data sources. The present study aims to combine differential InSAR (DInSAR) and multiple aperture interferometry (MAI) to produce a time series analysis documenting the surface dynamics of the Blåmannsisen Glacier, Norway, in an attempt to address the lack of InSAR derived 3-dimensional time series analyses in the wider body of contemporary research, which have previously been limited to 1-dimensional analyses.

Measurements of surface displacement along the satellites line of sight (LOS) can be resolved from the interferometric phase between image pairs. Deriving measurements along the satellites azimuthal track using MAI, requires filtering the constituent images of an interferometric pair using azimuthal common band filtering, producing forward and backward looking subaperture images. The subaperture images are then interfered, and through through a complex conjugation of the forward and backward looking interferograms, the azimuthal displacement can be retrieved. The full 3-dimensional measurement of surface displacement is then produced from a weighted least squares adjustment of LOS and azimuthal displacement measurements retrieved from both ascending and descending orbits, the differing look geometries from each orbit path allowing for the vertical component of displacement to be resolved.

The accuracy of DInSAR and by extension MAI, is limited by the coherence between image pairs, reduced revisit times decreases the risk of temporal decorrelation. The short revisit period of the Sentinel-1 platform, and the coming launch of Sentinel-1B resulting in increased coverage, offers an excellent opportunity to develop and apply these interferometeric techniques. Results produced from Sentinel-1 data will be compared against Tandem-X data, which is known to have a high coherence over the study area¹, to explore the effectiveness and accuracy of a fully interferometric approach to resolving 3-dimensional measurements of surface displacement.

Martone, M., Bräutigam, B., Rizzoli, P., Gonzalez, C., Bachmann, M., & Krieger, G. (2012). Coherence evaluation of TanDEM-X interferometric data. ISPRS Journal of Photogrammetry and Remote Sensing, 73, 21-29.

The aim of the paper is to provide an overview of the PRISMA (PRecursore IperSpettrale della Missione Applicativa) mission and the related scientific foreseen applications.

The PRISMA program is aimed to the development, qualification and launch of a space mission based on an advanced hyperspectral sensor and his pre-operations via the companion mission control and data processing ground segment. The project started in 2007 and it is now close to the completion of the design C-phase  with a launch foreseen for 2018.

PRISMA is an earth observation system with innovative electro-optical instrumentation which combines an hyperspectral sensor with a panchromatic, medium-resolution camera, capitalizing ASI investments in the field of "small missions" (e.g. AGILE), hyperspectral payloads (e.g. Hypseo, Joint Hyperspectral Mission/JHM), satellite platforms (MITA/PRIMA) and centers for tracking and remote sensed data processing centers (COSMO-SkyMed and CNM – Multi-mission National Centre).

The user segment offers a full range of mission products comprising the following level 0, 1 and 2 products, for both hyperspectral and panchromatic data.

Within this framework, five scientific studies have been conducted to undertake research on some specific hyperspectral applicative themes and procedures for the processing of hyperspectral data.

Sentinel-3 and Jason-3 are scheduled to launch in December 2015. Significant wave height measurements from the Synthetic Aperture Radar Altimeter SRAL on board Sentinel-3, and from the Poseidon-3B altimeter on board Jason-3 will be validated as soon as the data are available. Preliminary assessment results will be given, based on comparisons with buoy data and with other in-flight altimeters, presumably Jason-2, Cryosat-2 and Saral.

Abstract submission subject to data availability.

In the frame of the Copernicus program of the European Commission, Sentinel-2 offers multispectral high-spatial-resolution optical images over global terrestrial surfaces. In cooperation with European Space Agency (ESA), the Centre National d’Etudes Spatiales (CNES) is in charge of the Image Quality (IQ) of the project, and so ensures the CAL/VAL commissioning phase during the months following the launch.

Sentinel-2 is a constellation of 2 satellites on a polar sun-synchronous orbit with a revisit time of 5 days (with both satellites), a high field of view - 290km, 13 spectral bands in visible and shortwave infrared, and high spatial resolution - 10m, 20m and 60m. The Sentinel-2 mission offers a global coverage over terrestrial surfaces. The satellites acquire systematically terrestrial surfaces under the same viewing conditions in order to have temporal images stacks. The first satellite has been launched in June 2015. Following the launch, the CAL/VAL commissioning phase is then lasting during 6 months for geometrical calibration.

This paper will first point on the resolution of the roll oscillation issue observed on first Sentinel-2 images, and which was having great impact on geometric performances.

Then it will provide explanations about Sentinel-2 products delivered with geometric corrections. This paper will detail calibration sites, and the methods used for geometrical parameters calibration, and will present linked results. The following topics will be presented: viewing frames orientation assessment, focal plane mapping for all spectral bands, results on geolocation assessment, and multispectral registration.

At last, as multi-temporal series of images taken under the same viewing conditions are available, there is a systematic images recalibration over a same reference which is a set of S2 images, which geometry has been corrected by bundle block adjustment, and produced during the 6 months of CAL/VAL. This reference is called the Global Reference Image (GRI). Points will be taken between this reference image and the Sentinel-2 acquisitions, the geometric model of the image will be corrected, in order to ensure the good multi-temporal registration. This paper details the production of the reference over Europe during the CAL/VAL period, and then details the qualification and geolocation performance assessment of this GRI. It presents its use in the Level-1 processing chain and gives the assessment of the multi-temporal registration.

The SAR and especially the polarimetric SAR radar (PolSAR) images are affected by a noise called speckle which deteriorates image quality. Its nature is multiplicative and corrupts both amplitude and phase data, which complicates image interpretation, degrades segmentation performance and reduces the detectability of targets. This is why polarimetric filtering is a necessary treatment prior to analysis that allows to reduce the undesirable effect of speckle, therefore, obtain an improved image quality and extract information in order to properly interpret the results.

Many polarimetric filtering algorithms have been developed to reduce speckle and improve image quality. Different wavelet transform methods have shown good results in many applications to solve a variety of image processing problems as compression and filtering. In the field of PolSAR filtering, speckle reduction by stationary wavelet transform SWT has shown its effectiveness in providing a good compromise between smoothing homogenous areas and edge preservation in heterogeneous areas.

In this paper, we present a polarimetric speckle filtering method based on wavelet thresholding by applying the stationary wavelet transform SWT. We propose to enhance the wavelet thresholding, hard and soft threshoding, by computing the new proposed directional coefficients for wavelet coefficients improvement to reduce speckle without destroying the polarimetric information. The main goal of this algorithm is to detect edge regions and no-edge regions and to classify significant coefficients. Then apply a suitable image thresholding and modify the wavelet coefficients to obtain a better quality filtered image which satisfies the requested criteria. In this study, the Span, the polarimetric covariance matrix elements and the complex channels are taken into account.

The methods are applied to different polarisation data, the three polarimetric E-SAR images (HH, HV and VV) acquired on Oberpfaffenhofen area located in Munich, Germany, in P-band and the fully polarimetric RADARSAT-2 images (HH, HV, VH and VV)  acquired on Algiers, Algeria, in C-band. These methods are compared with results obtained by the refined Lee filter.

To evaluate the performance of each filter, we based it on the following criteria: smoothing homogeneous areas, preserving structural characteristics of objects in the scene and maintaining the polarimetric information. Visual evaluation is included and statistical parameters are calculated by using  the standard parameters for this type of treatment, such as mean, standard deviation and coefficient of variation, as well as other parameters necessary for the evaluation of PolSAR images to verify the requested criteria and an image classification is used to show the reduced speckle effect and also a comparative study is performed to validate the studied methods and determine the advantages and disadvantages of each filter.

Data quality is the aptitude of a product to answer user needs. User needs are not homogeneous and are dependent of each application field (e.g. cartography, forest monitoring, photo-interpretation). Therefore, as an example, the ability to detect and quantify changes in the Earth’s environment depends on sensors that can provide calibrated (known accuracy and precision) and consistent measurements of the Earth’s surface features through time.

Calibration and Validation correspond to the process of updating and validating on-board and on-ground configuration parameters and algorithms to ensure that the product data quality requirements are met.

Sentinel-2 has been launched on 23^rd of june 2015 and the “Mission Performance Center” (MPC) is responsible for performing the calibration (CAL) of the multi-spectral MSI sensor and validate (VAL) the quality of Sentinel-2 products. These operations are achieved through products processing and analyses at different levels. The CAL / VAL operations performed at level-1 are called L1CAL or L1VAL operations. They are dedicated to the Expert Support Laboratories (ESL) and they can be related to either geometry or radiometry.

The radiometric calibration activities allow determination of the parameters of the radiometric calibration model, which aims to convert the electrical signal measured by the instrument, transformed in digital counts, into the physical incoming radiance lighting the sensor. The nominal calibrations are based on the exploitation of the on-board sun diffuser images (relative gains calibration, absolute radiometric calibration) or images acquired over ocean at night (for dark signal calibration). Some other calibrations are foreseen only in case of contingency (crosstalk, refocusing). At the moment, the radiometric calibration is very close to what was expected before launch and the discrepancies observed are weak and under control. In particular, SWIR focal plane is subject to predictable electronic cross-talk which is corrected by a ground postprocessing but also to Random Telegraph Signal (RTS) affecting some pixels in band 12 that must be monitored.

Geometric calibration activities allow the determination of all Ground Image Processing Parameters (GIPP) of the geometric calibration model which aims to ensure better geometry is maintained for Sentinel-2 images (orientation of the viewing frames, lines of sight of the detectors of the different focal planes). These parameters have been estimated before launch and the purpose of the geometric calibration activities is to take into account any update of these parameter values that may occur. Moreover, to meet the multi-temporal registration and absolute geo-location requirements, a worldwide reference image called GRI is generated and used for automatic extraction of GCP for systematic refinement of the geometric model. The GRI generation process will probably take 2 years from satellite launch, but Sentinel-2 already offers an excellent geolocation performance. Obviously, in case of contingency, any geometric calibration parameter is likely to be assessed by MPC L1CAL.

This poster describes on one hand the tools and methods that are used to calibrate the instrument, and on the other hand the radiometric and geometric performances of the MSI sensor. The discrepancies with respect to ground predictions are also presented.

An experiment was carried out in Tuscany (Italy) for investigating the potential of GNSS signal in estimating forest biomass. The importance of forest monitoring is universally recognized due to the role played by forests and other semi-natural ecosystems in regulating the carbon cycle as alternatively sources and sinks of CO₂. Spatial explicit Information at different scales on forest status is therefore tremendously requested, since several applications can benefit from this information: from bioenergy production and sustainable forest management to the detection of land-use change and assessment of carbon stocks.

Forests may be characterized in terms of biophysical parameters (Leaf Area Index - LAI, Above Ground Biomass - AGB, Net PrimaryProductivity - NPP), which provide direct information on the productivity, structure and amount of forest resources. Usual forest biomass measurements are based on local surveys by selecting homogeneous small sample plots inside alarger area and by identifying areas representative of the entire forest. The possibility of measuring these parameters by using a receiver placed inside the forest plots is extremely appealing and therefore the use of a GNSS receiver was investigated in the framework of an ESA Project (GNSSBio AO/2-1610/14/NL/CVG-EGEP). The impact of vegetation on GNSS received signal is characterized in terms of signal attenuation and de-polarization.This concept was validated in previous experimental campaigns where the receiver was looking down from a ground based or an airborne platform (LEiMON and GRASS experiments, respectively). LEiMON focused on the analysis of the scattering characteristics of GNSS signals reflected off land surfaces, whereas GRASS determined the sensitivity of GNSS–R signals to soil bio-geophysical parameters. During both projects, the reflectivity coefficients at both polarizations were found to be sensitive to soil moisture and surface roughness changes. In addition, monotonic reduction in the reflectivity was observed with AGB, showing limited saturation for high biomass, differently to what happens with conventional L-bands radars.

The Oceanpal instrument provided by Starlab allows full-polarimetric GNSS-R data acquisition and processing. This new instrument is composed of three subsystems: the Oceanpal Antenna Rig (OAR), the Oceanpal Radio-Frequency Unit (ORFU), and the Oceanpal Data Management Unit (ODMU). The new OAR features two up-looking antennas for the reception of the direct signal, that are a GPS L1 with RHC and LHC polarizations, which carried out observations of two direct received GNSS signal, one in clear sky view and one below the canopy. Both receiver antennas are identical, presenting the same orientation, and performing simultaneous power measurements. By means of this approach, the attenuation and depolarization due to the vegetation were measured. The attenuation introduced by the vegetation layer, (or one-way loss factor), can be initially defined as the ratio between the power received,i.e transmitted, below the vegetation canopy and the free-space signal received underidentical conditions but without an intervening canopy between the GNSS transmitter and the receiver. Thus, the attenuation can be measured by means of the signal received at RHCP below the vegetation and the direct signal received at RHCP  in free space. On its turn, depolarization can be measured by means of the LHCP signal received below the vegetation and the clear-sky signal at RHCP.

The GNSSBio experiment was carried out in April, June and September 2015, in some poplar plots close to Florence, in Italy, characterized by different biomass values, ranging from <100m³/ha up to >600m³/ha. According to the acquisition plan, GNSS acquisitions were carried out for 3 hours continuously in different points of the canopy. In each poplar plot, 3 GNSS measurements were carried out in order to check the spatial variability of the canopy. The three measurements were carried out in the same time interval in order to receive the same satellites. Ground-truth measurements of soil moisture and vegetation parameters (tree height, diameter, LAI) were carried out during the GNSS data acquisitions. Measurements of LAI were carried out by using fish-eye photos collected in the same points of the antenna measurements, in order to have a clear picture of the tree structure.

The preliminary analysis of the datasets allowed confirming a marked attenuation of the signal collected under the canopy with respect to the one collected in open air. The direct overlapping of the received signal over the fish-eye pictures pointed out the different effect of the branch dimensions and positions.

The final goal of any Earth Observation product is to provide a variable of interest with the highest accuracy as possible. Validation analyses provide the means to inform about product accuracies by using independent reference source of data, which is assumed to be the ground truth. The level of maturity of an Earth Observation product is given, among other aspects, by the validation analyses is subjugated to. The Land Product Validation Subgroup of the Committee on Earth Observation Satellites (CEOS) defined four levels of validation stages according mainly on the amount and type of sampling involved. The current paper presents a probability sampling design to cover the Globe through several years with pairs of Landsat images, therefore hopefully achieving CEOS Validation Stage 3 for Burned Area products. The sampling design addresses methodological challenges to make sure it can be properly implemented in the Fire Disturbance (fire_cci) project Phase II of the European Space Agency’s Climate Change Initiative and in eventual joint inter-agency efforts. The sampling is implemented at global scale from 2003 to 2014. The sampling unit is delimited spatially by the Thiessen scene areas (TSAs) which are specifically constructed for use with Landsat WRS-II frames. The TSAs provide a partitioning of the globe for selecting a sample of non-overlapping areas that allows for convenient computing of unbiased estimators. The sampling unit is delimited temporally by the dates of two Landsat images acquired at the same TSA in two consecutive revisit times separated by 16 days or less. Reference data will be generated in a sample of units following the standard protocol defined by the CEOS Cal-Val which is based on the spectral change of two consecutive Landsat images. The stratification of the sampling is designed to make sure it can be adapted to additional years and that there is sufficient sampling in each of the major Olson biomes with special focus on regions with high BA, the category of interest for BA products. Therefore, three levels of stratification are designed, yearly, biome and burned area (BA). Each sampling unit is assigned to a year according to the first day of the unit’s temporal period and to a biome according the one most present in terms of spatial extent. The third level of stratification is implemented within each year-biome base on the BA extent provided by the Global Fire Emissions Database (GFED) version 4. The 80^th percentile of BA extent on each year-biome is used to assign the sampling units to the low and high BA strata. Similarly as in fire_cci project Phase I, reference data is planned to be processed for 100 sampling units for each year. The difference is that while the sample at global scale of Phase I was limited to one year, the Phase II aims to cover at least twelve years. The preliminary sample consists in 1200 sampling units, with a minimum of five units in each stratum and a sampling intensity proportional to the BA extent.

Fast and accurate cloud screening complemented with associated uncertainties is of great interest for a wide range of satellite data applications. Majority of existing algorithms feature binary logic and provide a cloudy/clear state of a pixel. Some of them additionally introduce categorized information about the classification by reporting a confidence flag. The information about the confidence level of provided physical quantities is required to construct an error budget of higher level products and to correctly interpret final results of a particular analysis. Regarding the generation of products based on satellite data the common input consist of a cloud mask which allows discrimination between surface and cloud signals. Further the surface information is divided between snow and snow-free components. At any step of this discrimination process a misclassification in cloud/snow mask propagates to higher level products and may alter its usability. Within this scope the Probabilistic Cloud Mask (PCM) algorithm suited for the Advanced Very High Resolution Radiometer (AVHRR) was developed (Musial et al., 2014). It generates classification probabilities for three categories: cloudy, clear, and snow. As opposed to majority of available techniques which are usually based on decision-tree approach in the PCM algorithm all spectral, angular and ancillary information is used in a single step to retrieve probability estimates from a pre-computed Look Up Tables (LUTs). Moreover, the issue of derivation of a single threshold value for a spectral test was overcome by the concept of multidimensional information space which is divided into small bins by an extensive set of thresholds. Generic formulation of the probability and simplicity of the PCM algorithm combined with bitwise vector processing allowed to successfully apply this methodology to low stratiform cloud detection (Musial et al., 2014). Recently, the PCM algorithm was extended (xPCM) to generate other cloud products such as: cloud type, cloud physical properties, cloud phase, cloud top temperature, pressure and height. It was proven to be at least 4 times faster than analogous PPSv2014 package developed within the scope of Nowcasting Satellite Application Facility (EUMETSAT NWC SAF) without the loss of accuracy.

During the SPOT-5 Take5 experiment an extensive imagery was collected over 5 sites in Poland that allow to train xPCM algorithm against improved cloud flag embedded in the level 2 product. Some of the scenes will be further used for validation purposes. As a result quality indicators will be derived to assess accuracy of the xPCM algorithm and to perform intercomparison with a standard SPOT cloud flag. Finally the benefits of xPCM application to Sentinel-2 data will be discussed with example featuring MultiSpectral Instrument (MSI) imagery (if available).

Musial, J. P., Hüsler, F., Sütterlin, M., & Neuhaus, C. (2014). Probabilistic approach to cloud and snow detection on Advanced Very High Resolution Radiometer (AVHRR) imagery. Atmospheric Measurement Techniques, 7(3), 799

Musial, J. P., Hüsler, F., Sütterlin, M., Neuhaus, C., & Wunderle, S. (2014). Daytime low stratiform cloud detection on AVHRR imagery. Remote Sensing, 6(6), 5124-5150

Sentinel-2 is a polar orbiting satellite constellation of two (initially A and B units) carrying each one an optical imaging sensor called MSI (Multi-Spectral Instrument).
Sentinel-2A was successfully launched on 23 June 2015. Commissioning operations have proceeded until 15 October 2015. Sentinel-2B is scheduled for end of 2016.

The Sentinel-2 Ground-Segment (GS) is composed of the Flight Operation Segment (FOS), responsible for all flight operations of both Sentinel-2 spacecraft units, and the Payload Data Ground Segment (PDGS), responsible for payload and downlink planning, data acquisition and processing of the Sentinel-2 satellite data.

The Mission Performance Centre (MPC) is the element within the PDGS in charge of covering the following functionalities:
- Calibration (CAL) – to update on-board and on-ground configuration data in order to meet product quality requirements.
- Validation (VAL) – to assess the quality of the generated data products.
- Quality Control (QC) – to routinely monitor the status of the sensor and to check if the derived products meet the quality requirements along mission lifetime.
- Data processors and tools corrective and perfective maintenance (PTM) – to manage the updates of the processors, quality control tools, calibration tools and all auxiliary files.
- End-to-end system performance monitoring (E2ESPM) – to monitor the performance of the Sentinel-2 system operations (including the instrument) and assessing the system.

All these activities have been validated during commissioning operations and are now ready for the routine phase. Cross-comparison with results of the commissioning team ensures the continuity of the mission performance.

In particular, the calibration procedure has been consolidated, and MPC calibration results have been compared to those of the commissioning team. The MPC validation team has verified the very good radiometric performance of the MSI. Two decontamination operations have been performed so far, and led to a recovery of the SWIR band signal to its nominal value. During the first year in orbit, systematic monitoring and decontamination operations are foreseen in order to cope with the effects of post-launch outgassing.

The instrument noise is closely monitored with several methods. The first in-flight assessment is compatible with the pre-flight estimations. A relatively important electronic cross-talk has been shown to affect the water vapour band B10, whose signal is relatively low compared to other SWIR bands. This electric effect is however easy to correct and the correction has been implemented in the processing chain. Another important point for the SWIR channel is the monitoring of Random Telegraph Signal (RTS) noise. A procedure has been established to keep track of affected pixels during the mission lifetime.

The geometric performance of the instrument has been assessed by the validation team and the first results are in line with those of the commissioning team. After fine-tuning of the attitude determination algorithm, the final geolocation performance is expected to be well below one pixel. In the meantime, a Global Reference Image (GRI) has been elaborated. The European GRI is in progress, and the world GRI will be started in the near future. The GRI will be used to improve the geolocation of L1C products.

Validation and calibration activities of L2A products have also started. Adjustment of the cirrus correction and classification algorithms is in progress. Results from airborne and field measurement campaigns have been gathered and will be analysed to validate BOA reflectances.

Overall, the Sentinel-2 mission is proving very successful in terms of product quality, but also in terms of timeliness and availability, thereby fulfilling the promises of the Copernicus program.

The consortium in charge of the Sentinel-2 MPC activities is composed of:
- CSSI, Toulouse, France: Prime Contractor, ESL Coordinator, ESL Level-1 Products Radiometry, Processors Maintenance Manager
- Argans, Plymouth, UK: Technical Manager, Operation Manager, ESL Level-1 Products Radiometry Validation
- ACRI ST, Sophia Antipolis, France: IT Manager, Operator
- Elecnor Deimos, Madrid, Spain: Tools Maintenance Manager, OGCD/OBCD Manager, Operator
- GMV, Madrid, Spain: Web Site Manager, ESL Level-1 Products Geometry Prototyping
- Airbus Defence & Space (Astrium), Toulouse, France: ESL Level-1 Products Leader
- IGN, Toulouse, France: ESL Level-1 Products Geometry Calibration (GRI)
- Thales Alenia Space, Cannes, France: ESL Level-1 Products Validation Leader
- ONERA, Toulouse, France: ESL Level-1 Products Radiometry Validation
- Telespazio, Toulouse, France: ESL Level-2A Products Leader
- DLR, Berlin, Germany: ESL Level-2A Products Validation

In the framework of its Earth Observation Programmes the European Space Agency (ESA) carries out ground based and airborne campaigns to support geophysical algorithm development, calibration/validation, simulation of future spaceborne earth observation missions, and applications development related to land, oceans and atmosphere.

ESA has been conducting airborne and ground measurements campaigns since 1981 by deploying a broad range of active and passive instrumentation in both the optical and microwave regions of the electromagnetic spectrum such as lidars, limb/nadir sounding interferometers/spectrometers, high-resolution spectral imagers, advanced synthetic aperture radars, altimeters and radiometers. These campaigns take place inside and outside Europe in collaboration with national research organisations in the ESA member states as well as with international organisations harmonising European campaign activities. ESA campaigns address all phases of a spaceborne missions, from the very beginning of the design phase during which exploratory or proof-of-concept campaigns are carried out to the post-launch exploitation phase for calibration and validation.

We present examples of several campaigns in the Arctic, tropics and the Antarctic, from sea-ice, land ice to forest biomass, soil moisture and gravity measurements. Campaigns such as CryoVEx2014, AfriSAR2015/16, PolarGap2015/16 and S-1 Supersite in support of CryoSat, Biomass, GOCE and Sentinel-1 respectively will be described, with particular emphasis on their measurement methods and delivered products. All the campaigns' final reports and datasets are archived in the ESA campaigns archive [https://earth.esa.int/web/guest/campaigns] at the end of each campaign where they can be freely accessed. The ESA campaigns archive is continuously populated with new campaign datasets and presently 66 datasets are available on the archive.

The ESA Sentinel-3 satellite, within the Copernicus programme, will be the second satellite to operate a SAR mode altimeter. The Sentinel 3 Synthetic Aperture Radar Altimeter (SRAL) is based on the heritage from Cryosat-2, but will be complemented by a Microwave Radiometer to provide a wet troposphere correction, and will operate at Ku and C-Band to provide an accurate along track ionospheric correction. Together this instrument package will allow accurate measurements of sea surface height over the ocean, as well as measurements of significant wave height and surface wind speed. 

SCOOP (SAR Altimetry Coastal & Open Ocean Performance) is a project funded under the ESA SEOM (Scientific Exploitation of Operational Missions) Programme to characterise the expected performance of Sentinel-3 SRAL SAR mode altimeter products, in the coastal zone and open-ocean, and then to develop and evaluate enhancements to the baseline processing scheme in terms of improvements to ocean measurements. There is also a work package to develop and evaluate an improved Wet Troposphere correction for application to Sentinel-3 SRAL data, and provide recommendations for use.

At the end of the project recommendations for further developments and implementations will be provided through a scientific roadmap.

In this presentation we provide an overview of the SCOOP project, highlight the key deliverables and discuss the potential impact of the results in terms of the application of delay Doppler (SAR) altimeter measurements over the open ocean and coastal zone.  We also present the initial results from the first phase of the project, which involves a review of the current “state of the art” for SAR altimetry, establishes the “reference” delay Doppler processing and echo modelling /retracking and agrees the specifications for the Test Data Sets to be evaluated in the first part of the project. 

In partnership with the European Commission and in the frame of the Copernicus program, the European Space Agency (ESA) is developing the Sentinel-2 optical imaging mission devoted to the operational monitoring of land and coastal areas. The Sentinel-2 mission is based on a satellites constellation deployed in polar sun-synchronous orbit. Sentinel-2 offers a unique combination of global coverage with a wide field of view (290km), a high revisit (5 days with two satellites), a high spatial resolution (10m, 20m and 60m) and multi-spectral imagery (13 spectral bands in visible and shortwave infra-red domains). The first satellite, Sentinel-2A, has been launched in June 2015.

The Sentinel-2A Image quality Commissioning Phase has been delegated to CNES by ESA. It has started immediately after the Launch and Early Orbit Phase and has continued six months after the launch. This paper provides first an overview of the Sentinel-2 system and a description of the products delivered by the ground segment associated to the main radiometric specifications to achieve. Then the paper focuses on the radiometric results obtained during the in-flight commissioning phase. The radiometric methods and calibration sites used in the CNES image quality center to reach the specifications of the sensors are described. A status of the Sentinel-2A radiometric performances at the end of the first six months after the launch is presented. We will particularly address in this paper the results in term of absolute calibration, pixel to pixel relative sensitivity, MTF and SNR estimation.

Geoscience Australia has permanently deployed 40 trihedral corner reflectors in Queensland, Australia, covering an area of approximately 20,000 km². The array of corner reflectors was constructed as part of the AuScope Australian Geophysical Observing System (AGOS) initiative to monitor crustal deformation using Interferometric Synthetic Aperture Radar (InSAR) techniques. The array includes 34 corner reflectors of 1.5m, 3 reflectors of 2.0m and 3 reflectors of 2.5m inner leg dimensions. Through the design process and the precision manufacturing techniques employed, the corner reflectors are also highly suitable for calibration and validation of satellite-borne Synthetic Aperture Radar (SAR) instruments.

Nine of the 1.5m corner reflectors in the AGOS array had their Radar Cross Section (RCS) individually characterised at the Defence Science and Technology Organisation’s outdoor ground reflection range, prior to permanent deployment in Queensland.  The RCS measurements for the corner reflectors were carried out at X and C-band frequencies for both horizontal and vertical transmit-receive polarisations, and at a range of elevation and azimuth alignments.

The field performance of the AGOS corner reflectors has been studied using SAR data from a range of satellites including Sentinel-1A. This study focuses on the calibration of the Sentinel-1A satellite presenting results from exercises undertaken both at Geoscience Australia and the European Space Agency’s Mission Performance Centre as part of the satellite commissioning and routine phases. Radiometric calibration results in conjunction with geometric calibration and validation results for Sentinel-1A products from Stripmap and Terrain Observation with Progressive Scans (TOPS) modes are presented in this paper.

The current configuration for most corner reflectors in the AGOS array is set to serve calibration requirements for a broad range of SAR missions on ascending orbital passes. However, the design allows for mission-specific corner reflector alignment if needed, as in the case of the 2.5m and 2.0m reflectors which have specifically been aligned to support calibration of the L-band SAR instrument on ALOS-2.  Results reported in this paper could inform the need for re-configuring one or more corner reflectors in the array to specifically support ongoing calibration of the Sentinel-1A and B satellites.

The permanently deployed AGOS corner reflector infrastructure presents an opportunity for independent calibration and comparison of SAR instruments on current and future satellite missions, and is considered an important Australian contribution to the global satellite calibration and validation effort. 

Multisensor image fusion is the process of combining relevant information from two or more satellite images into a single image. In remote sensing applications, the increasing availability of space borne sensors gives a motivation for different image fusion algorithms. Several situations in image processing require high spatial and high spectral resolution in a single image. Most of the available equipment is not capable of providing such data convincingly. Image fusion techniques allow the integration of different information sources. The fused image can have complementary spatial and spectral resolution characteristics. However, the standard image fusion techniques can distort the spectral information of the multispectral data while merging. Many methods exist to perform image fusion. The very basic one is the high pass filtering technique. Later techniques are based on Discrete Wavelet Transform, uniform rational filter bank, and Laplacian pyramid and neural networks.

We suggest a method for fusion of data from MSI/Sentinel2 and OLCI/Sentinel3. In the visible range MSI measures radiance with 10 m resolution at 490, 560 and 665 nm; with 20 m resolution at 705 nm; and with 60 m resolution at 443 nm. In the visible range OLCI measures with 300 m spatial resolution at 400, 412, 443, 490, 510, 560, 620, 665, 673, 681, 708 nm. The data from the visible from both sensors is fused to get products with values of remote sensing reflectance wavelengths of OLCI and with spatial resolution of 20 m using the Neural Network.

The results of the fusion are utilized in the multi-spectral MCC and feature tracking algorithms to identify weak ocean currents as well as for retrieval of water quality parameters including chlorophyll-a, suspended mineral matter and dissolved organic carbon in coastal Case-II waters.

The term 'retracking' refers to fitting an expected waveform shape to the actual radar altimeter observations, with the parameters describing the shape, position and scaling of the model allowing the retrieval of useful physical information. Conventionally, this is performed on a single waveform at a time, and the shape parameters from multiple along-track estimates then consolidated into lower frequency (typically 1 Hz) observations. From physical considerations one would expect the true shape parameters to be slowly-varying within the 1-second interval, but this is not taken into consideration. In this present work, sponsored by ESA's Sea Level CCI programme, we evaluate an approach that analyses 21 or more waveforms at once, modelling the gradual changes in the shape parameters. Such an approach would have been impractical a decade ago, because of the greater number of unknowns to be determined, but modern processors with advanced techniques for locating a minimum in multi-dimensional space make this approach feasible.

Having shown that the technique works well for the open ocean, where (in the absence of rain cells or surface slicks) ocean waveforms change gradually, we are know testing it in the coastal zone, where sheltering by headlands and shoaling bathymetry may make both wind and wave conditions vary appreciably. We specifically explore ERS and Envisat passes that run close to well-maintained tide gauges. The ones selected cover a range of conditions from the relatively enclosed Irish Sea to La Coruna and Brest which are exposed to North Atlantic swell.

The Copernicus programme for long-term monitoring of the Earth's land, ocean and cryosphere is underway with already two ESA-built Sentinel satellites in orbit, and has two Sentinel-3 missions due for launch in the next 2 years. The Sentinel-3 Mission Performance Centre (S3MPC) has already been set up to develop the processing chain for all the instruments and to assess the instruments' health and data quality, as well as maintain calibrations over months to decades. The S3MPC's effort on the Surface Topography Mission (STM) brings together contributions from many Expert Support Laboratories (ESLs).

The initial focus will be on the integration of low level sensor data and housekeeping telemetry from the altimeter (SRAL) and the microwave radiometer (MWR). Subsequent over-ocean analysis will compare the data from both these instruments with similar ones on other contemporaneous missions, and will fine tune (as needed) the parameters for the SAMOSA retracking, and the algorithms for wind speed, water vapour and sea state bias. Transponder measurements on Crete will determine the absolute bias of the range and provide monitoring of the stability of the backscatter strength. New techniques are being developed to extend out from the coast the reference datum information that is available from levelled tide gauges. Metocean information (wave height, wind speed and water vapour) will be assessed by comparison with operational ECMWF model prediction, and via comparison to global buoy data. The performance of the various different waveform retrackers will be evaluated over both land-ice and sea-ice to understand any differences with respectto Cryosat-2, thus enabling the datasets eventually to be merged for climate-scale studies. Lakes, such as Issyk-Kul in Kyrgyzstan will provide further opportunities to constrain any errors in range bias. Once sufficient data have been acquired, the performance of the Sentinel-3A and Sentinel-3B altimeters over rivers and lakes will also be assessed. The overarching aim of all the validation work by the ESLs is to give users confidence to use the S-3 Surface Topography Mission data over all terrains and whether looking at short-time scales or as part of long-term climate variations.

Telespazio VEGA and Pixalytics have undertaken a project focused on understanding a future market for NovaSAR-S with a particular focus on flood mapping; funded through a call for pathfinder projects from UK’s Centre for Earth Observation Instrumentation and Space Technology (CEOI-ST). The principle objectives were to develop a simple simulator that could produce S-band Synthetic Aperture Radar (SAR) datasets from a terrain model, supported by data mining of Envisat S-band Radar Altimeter (RA) data alongside land surface characteristics.

The simulator has initially been developed using the approach adopted in the Next ESA SAR Toolbox / Sentinel-1 Toolbox (S1TBX) to encompass terrain variations (Lui et al. 2004, PERS). As an outline, Envisat DORIS data is used to determine the satellite position with ASAR metadata providing the SAR geometry, and then the return signal is determined from a combination of a terrain or elevation model, RA data, Landsat and Corinne land cover information. Together these inputs allow for a simulation of a SAR image that is influenced by both the geometry and surface type. The test sites correspond to S- and X-band data collected during airborne overflights during the 2014 AirSAR campaign (a UK partnership between Airbus DS, NERC and the Satellite Applications Catapult) so that simplistic simulator can be intercompared with other simulations e.g., results from using the S1TBX and AirSAR data transformed into a satellite SAR simulation using the ENVI SARscape toolbox.

It’s envisaged that the resulting simulated data and the simulator will not only aid early understanding of NovaSAR-S, but will also aid the development of flood mapping applications.

The modelling of soil, hydrologically and ecologically relevant parameters of the land surface is only viable, when the elevation data that is used for 3D-modelling characterizes the non-vegetated bare ground rather than the canopy surface or other objects on the ground. Most of the available digital elevation model (DEMs) products that are generated from remote sensing acquisitions (e.g. SRTM, ASTER GDEM, TanDEM-X DEM) represent the top of canopy surface and thus are not adequate for such purposes. Apart from urban areas, this problem is mainly associated to forest areas.  Here, the correction of the DEM to a digital terrain model (DTM) needs information about the canopy height which might be taken from ground observations or LIDAR campaigns. However, both data sources are not available for large areas. There are a number of studies that model canopy parameters from SAR or InSAR data but there is a lack of knowledge in the derivation of (InSAR) height from satellite SAR data in forest areas.

The presentation introduces a new concept for the derivation of a high resolution DTM from DEMs with a focus on the forthcoming DEM data from the TanDEM-X mission of the German Aerospace Center (DLR). The approach exclusively makes use of operationally available SAR satellite data in order to provide a tool that is not restricted by cloud cover which is especially prevalent in the forests of the mid and northern latitudes and the tropics. It combines data from the present national (DLR: TerraSAR-X, TanDEM-X; CSA: RADARSAT-2) and European missions (ESA: Sentinel-1) of the space agencies to build a workflow for the adaptation of the surface height to ground height without prior knowledge of the absolute object height.  The height modelling builds on a categorization of forest stands and quantification of canopy parameters at the stand level from an extensive set of intensity and coherence measures of the X- and C-band SAR satellite data.

The talk will provide insight into the concept of SAR-based DEM correction and the study design. It will deliver first exemplary results for a test site in central Germany.

In the framework of the European Union Copernicus programme, the European Space Agency (ESA) has launched the Sentinel-2 (S2) Earth Observation (EO) mission which provides optical high spatial resolution imagery. A new tool named, S2-RUT, (Sentinel-2 Radiometric Uncertainty Tool) has been integrated as part of the Sentinel Toolbox. It allows the estimation of the radiometric uncertainties associated to each pixel using as input theLevel 1C (L1C) top-of-atmosphere (TOA) reflectance images provided by ESA. That is, for each pixel in the TOA reflectance factor image, a radiometric uncertainty will provide an indication of a certain probability that the reflectance factor value is within a certain range. The probability range provided in this tool is ~68.27% (i.e. k=1). In [1], it was discussed the radiometric analysis, an initial validation of the uncertainty combination and the preliminary software design of this tool.  Here we focus on the latter since a first implementation of thetool is readily available to the users as part of the Sentinel Toolbox.The S2-RUT code integration in the Sentinel Toolbox has been implemented through the SeNtinel Application Platfom (SNAP).  For the tool uncertainty calculation, it is necessary to extract auxiliary information from the metadata and quality masks integrated in the L1C product as well as the codification of the uncertainty image and the metadata generation. Several constraints and compromises were introduced in order to make this integration feasible. For example, we are currently investigating how the S2-RUT could modify the L1C products so that the results can be directly embedded as part of them. This first implementation looks at how to overcome these type of challenges as well as how to minimise the memory and processing requirements. Next iterations of the tool will be looking a refinement of the tool based on the community feedback and the gradual integration of more complex contributions and features. Some of these additional features are required for the contextualisation of the radiometric uncertainty and its propagation to higher-level products. These features were initially described in[1] and separated as warning flags and uncertainty type classification. The codification of a second byte of the uncertainty image was proposed for its implementation. Here we will present the current status of the research and implementation of such additional features.

[1] Javier Gorroño, Ferran Gascon, and Nigel P. Fox, “Radiometric uncertainty per pixel for the Sentinel-2 L1C products,”SPIE Europe Remote Sensing, International Society for Optics and Photonics, 2015.

The Microwave Radiometer (MWR) flown on-board Envisat, ERS-1, and ERS-2 has provided a time series of global microwave observations over a period of nearly 21 years between 1991 and 2012. A successor instrument with very similar characteristics will be carried on-board the Sentinel-3 suite of satellites which should extend the availability of the MWR observations from late 2015 well into the mid-2020s and possibly beyond.

In order to improve the quality and usefulness of MWR observations, the European Space Agency (ESA) has initiated the project “ERS/Envisat MWR Recalibration and Water Vapour Thematic Data Record Generation” (EMiR) as part of their Long-Term Data Preservation (LTDP) activities. In this context, fully re- and inter-calibrated data records of MWR brightness temperature and derived total column water vapour (TCWV) have been generated (see ELP 2016 companion contribution by Bennartz et al.).

Here, we provide a basic evaluation of the newly generated fundamental (FDR) and thematic data records (TDR) derived from MWR observations, primarily for climate studies. First, we perform a number of plausibility checks to identify potential quality issues. We then compare MWR-derived TCWV with corresponding products of both space based and non-space based origin compiled and pre-processed in the context of the GEWEX Water Vapour Assessment (G-VAP). We also analyse to what extent the improved information on TCWV from MWR benefits the correction of ocean altimetry observations. Finally, we examine how far the MWR-derived TCWV thematic data record meets the needs of the Global Climate Observing System (GCOS) and how it can potentially contribute to the Global Space-based Inter-calibration System (GSICS).

At the time of writing, the Sentinel-3 launch is planned for the end of December 2015. Therefore by the time of the Living Planet Symposium about 4 months of measurements will have been collected and processed by the Sentinel-3 ground segment.

EUMETSAT, as part of the Sentinel-3 Mission Performance Framework, is involved with the calibration and validation of the Sentinel-3 sensors, OLCI, SLSTR, SRAL, and MWR. This presentation will focus on EUMETSAT's contribution to the validation of the Level 2 altimetry products and the calibration of the range, significant wave height and wind speed measurements made by the SRAL altimeter.

The Sentinel-3's SRAL altimeter is very similar to the SIRAL altimeter of CryoSat-2, except of the capability of the later to provide SARin measurements. New is that SRAL functions 100% of the time in SAR mode, which is vastly dissimilar to the conventional low-resolution mode (LRM) altimetry. Therefore, the validation of the products and the calibration of the measurements deserves a large amount of scrutiny, eventually demonstrating the unique and game-changing capabilities of SAR.

This presentation gives an analysis of the Level 2 altimetry products generated by the Sentinel-3 ground segment. Some of the focal points are:

Conformity of the products to the expected format specification.

Analysis of the geophysical corrections and orbits provided on the products against those determined off-line by RADS.

Comparison of the quality of near real time (NRT), short time critical (STC), and non time critical (NTC) products.

Cross-calibration of the measurements of sea level, wind speed, and wave height against other altimeters currently in operation: Jason-2, SARAL/AltiKa, CryoSat-2, and possibly Jason-3.

Comparison of the performances in LRM and SAR mode; or this purpose, EUMETSAT has proposed a special LRM/SAR mode operation on ascending/descending tracks during the phase E2 ramp-up, which will likely fall after the Living Planet Symposium.

All these analyses are supported by the Radar Altimeter Database System (RADS), a global altimeter database and analysis tools jointly developed and supported by EUMETSAT, NOAA, and TU Delft.

The main objective of this study was to assess the informational potential of the red edge reflectance spectral region (680 – 780 nm) for improving quantitative estimations of vegetation biophysical parameters, namely leaf chlorophyll content (Cab) and leaf area index (LAI).  The study was carried out in two parts: i) retrievals from airborne hyperspectral data including the red edge inflection point position (REP) and ii) retrievals from Sentinel 2 MSI simulated data together with REP information. Potentials of red edge region were assessed using real remotely sensed image data (acquired, for instance, during the SEN3EXP/ESA campaign with a CASI instrument) and data simulated for four vegetation types (citrus orchard, maize field, spruce and beech forest stands) by the Discrete Anisotropic Radiative Transfer (DART).

At first, an exhaustive bibliographic review focusing on the relationship between the red edge reflectance and the vegetation biophysical parameters of interest (LAI and Cab) was performed. Also methods for the REP reconstruction and estimation from hyper-/multispectral data, including the presence of a double inflection point, were reviewed and implemented. Once selected, the appropriate REP estimation algorithm was applied to the full database of synthetic DART simulated and real airborne hyperspectral data, and Sentinel 2 spectral data. Within the latest task, aiming at simulation of Sentinel 2 MSI data from the airborne hyperspectral data, an innovative method of continuous reflectance spectra reconstruction has been implemented. This method was tested on the DART spectral database of Sentinel 2 reflectance bands as well as on Sentinel 2 images generated from real airborne data. The red edge and REP reconstruction methods and the related results will be presented.

One of the objectives of the CEOS/IVOS working group is to support the establishment of the RadCalNet network of test sites for radiometric calibration of earth observation sensors. Three sites located in France (LaCrau), USA (Railroad Valley playa), and Baotou (China) are already equipped, managed individually by Space agencies (CNES, France, NASA/USA, Academy of Opto Electronics, China).

This project is the ESA contribution to the addition of a fourth site in the network. It covers the identification of the site and its characterization, its equipment with an automatic photometer able to measure sky radiance, sun irradiance and surface radiances and its operation, and the data release on the RadCalNet portal (https://web.magellium.fr/action).

The first task was to identify the site according to a long list of criteria allowing the characterization of the surface and atmosphere components.

Spatial homogeneity, cloud coverage, temporal variability, and flatness of the sites are the criteria that have been studied to allow the selection of potential sites.

For this study, 6000 Landsat 8 tiles, MODIS aerosol products, MODIS albedo products, ECMWF cloud coverage, and SRTM tiles have been used to perform the analysis at a global scale. Results have been also examined according to the site accessibility to regularly perform maintenance operations. They have been classified according to notations provided by each criterion to provide a list of potential sites.

Among the list, Gobabeb site located in Namibia has been selected to be the fourth RadCalNet site.

There is a feature that is evident in all the images obtained by the probes sent to explore our planetary system: all bodies having a solid surface, without distinction, show, with greater or lesser evidence, the marks left by the geological processes they undergone during their evolution. the most common of these signs are impact craters.

Craters formed by the impact of small cosmic bodies (asteroids, comets, meteoroids) have dimensions ranging from some meters to hundreds of kilometers. The kinetic energy of the impactor, which velocity is in general of the order of tens km/s, is released in fractions of a second, generally in a explosive way generating complex phenomena that transform not only the morphology of the surface involved in the impact, but also the mineralogy and crystallography of the impacted material. Even our planet is not immune to these impacts. At present, more than 180 geological structures recognized of impact origin are known on Earth.

In this article we aim to show how these impact structures on Earth's surface are seen from space. To do this we use the images obtained by the COSMO-SkyMed satellite constellation.

COSMO-SkyMed is the first dual use (civilian and military) Earth Observation system  conceived and designed in order to create a service for the provision of data, products and services that meet international standards, with a wide range of applications, such as risk management,  scientific, commercial and military applications. This system, funded and managed by ASI (Italian Space Agency) and MoD (Ministry of Italian Defense), consists of a constellation of four satellites in Low Earth Orbit. Each satellite is equipped with a Synthetic Aperture Radar (SAR) working in X band (fra parentesi ci metterei anche lunghezza d’onda e/o frequenza) , therefore able “to see” through the clouds and in absence of sunlight. A ground infrastructure completely dedicated to the management of the constellation ensures the data collection, archiving and distribution.

COSMO-SkyMed, which is able to capture nominally up to 1800 images a day, allows the global coverage of the Earth at any operating lighting and weather conditions (day/night) and provides high spatial resolution images with fast response times (36 hours in the current nominal operational mode).

Starting from 2010, ASI proposed, in collaboration with the  Astrophysical  Observatory of Turin, the realization of an  Encyclopedic Atlas of Terrestrial Impact Craters using COSMO-SkyMed data. This is the first atlas of terrestrial impact craters in X band. To observe these impact craters all sensor modes were used, according to the size of the analyzed crater.

The project includes research of any new features that could be classified as craters and, for the sites whereby  it is considered necessary, the implementation of a geological survey on site to validate the observations.

In this paper the status and an overview of the Encyclopedic Atlas of Terrestrial Impact Craters using COSMO-SkyMed data, currently under publication, is provided.

The Radiometric Calibration Network (RadCalNet) portal (hosted at https://web.magellium.fr/action) is the infrastructure created by the RadCalNet Working Group under ESA contract to distribute the data acquired by a network of instruments for the absolute radiometric calibration, intercalibration, and validation of Earth-observing sensors.

At this time, there are four calibration test sites equipped located in Gobabeb (Namibia), LaCrau (France), Railroad Valley Playa (USA), and Baotou (China) which collect BOA reflectances and atmospheric parameters. Therefore, a processing of the data is performed by NASA to estimate TOA reflectances in 30min intervals for a nadir view between 9:00 and 15:00 local time. The data could be downloaded directly from the portal for RadCalNet WG.

The portal provides also the documentation relative to the sites, the processing, data format, and data quality control, and calibration inter-comparison reports.

Although the principle of satellite radar altimetry is fairly simple, properly accounting for all the geophysical corrections across separate missions makes it complex to arrive at a homogenous dataset. To become a fundamental Climate Data Record (CDR) for sea level, mm-level signals need to be extracted from measurements taken from an altitude of 700 to 1400 km. 

Data from ten altimeter missions are presently available in the Radar Altimeter Database System (RADS), forming the basis for a prototype Level 2 sea level CDR. The 24 years of "reference missions" (TOPEX/Poseidon, Jason-1, and Jason-2) are complemented by "mesoscale missions" (Geosat, GFO, ERS-1, ERS-2, Envisat, and SARAL) and the "polar mission" CryoSat-2. The latter two groups increase the spatial coverage of sea level change and start yielding stability comparable to the reference missions through some recent developments in corrections, like orbits and ionospheric corrections. By the time of the Symposium, the early mission data of Sentinel-3 and Jason-3 are likely already added to the RADS database of analysis, calibration, and validation.

RADS was first developed at the Delft University of Technology in the late 1990s in order to better compare and combine altimeter data from ERS-1, ERS-2 and TOPEX/Poseidon into a single database. At that time, limited resources required condensing the data to the most essential information such as sea level, wave height, wind speed and time and location of the measurement. 

Since 2001 development and maintenance has continued at NOAA's Laboratory for Satellite Altimetry, while TU Delft has provided a web interface, an rsync server, and a mailing list server. And since 2013, EUMETSAT has taken over most of the development while NOAA is updating the database with the latest altimeter data, in near-realtime, around the clock.  About 62 institutes in 23 countries are mirroring part or all of the data set, while many others access the web site to make data requests on demand. Currently, RADS provides a multitude of additional variables needed to convert the original satellite range measurement into a climate-quality sea level record. RADS includes alternative variables to allow the user to assess the possible influence of model errors on sea level, and to correlate its variations with those in wind speed, wave height, and sea surface temperature. 

After beta testing by some of our expert users, the developers are proud to release to the general public RADS version 4, which includes the following developments:

Fortran 95 code which allows easier expansion to new uses and parallelisation.

Compatibility with the RADS version 3 data base and user code.

Documented software library for users to build on, in order to use the data set directly for their own applications, and share their code with other users. 

A number of convenient tools to analyse the Level 2 data set.

Full documentation of all data fields contained in the data record.

Improved attribute fields in the netCDF data files for CF-1.6 compliance and documentation. 

Web-based netCDF data selection.

Source code freely available on GitHub, as well as bug/issue tracker (https://github.com/remkos/rads).

User forum through e-mail list.

The attached figure shows the global sea level time series as derived from the individual altimeter satellites from 1991 until present, with an average rate of 3.1 mm/year.

The scope of this presentation is to feature the G-POD SARvatore service to users for the exploitation of the CryoSat-2 data, which was designed and developed by the Altimetry Team at ESA-ESRIN EOP-SER (Earth Observation – Exploitation, Research and Development). The G-POD service coined SARvatore (SAR Versatile Altimetric Toolkit for Ocean Research & Exploitation) is a web platform that allows any scientist to process on-line, on-demand and with user-selectable configuration CryoSat-2 SAR/SARIN data, from L1a (FBR) data products up to SAR/SARin Level-2 geophysical data products. The Processor takes advantage of the G-POD (Grid Processing On Demand) distributed computing platform to timely deliver output data products and to interface with ESA-ESRIN FBR data archive (155’000 SAR passes and 41’000 SARin passes). The output data products are generated in standard NetCDF format (using CF Convention), therefore being compatible with the Multi-Mission Radar Altimetry Toolbox and other NetCDF tools.

By using the G-POD graphical interface, it is straightforward to select a geographical area of interest  within the time-frame related to the Cryosat-2 SAR/SARin FBR data products availability in the service catalogue. The processor prototype is versatile allowing users to customize and to adapt the processing, according to their specific requirements by setting a list of configurable options. After the task submission, users can follow, in real time, the status of the processing. From the web interface, users can choose to generate experimental SAR data products as stack data and RIP (Range Integrated Power) waveforms.

The processing service, initially developed  to support the development contracts awarded by confronting the deliverables to ESA’s, is now made available to the worldwide SAR Altimetry Community for research & development experiments, for on-site demonstrations/training in training courses and workshops, for cross-comparison to third party products (e.g. CLS/CNES CPP or ESA SAR COP data products), and for the preparation of the Sentinel-3 Surface Topography Mission, by producing data and graphics for publications, etc. Initially, the processing was designed and uniquely optimized for open ocean studies. It was based on the SAMOSA model developed for Sentinel-3 Ground Segment using CryoSat data. However, since June 2015, a new retracker (SAMOSA+) is offered within the service as a dedicated retracker for coastal zone, inland water and sea-ice/ice-sheet. In view of the Sentinel-3 launch, a new flavor of the service will be initiated, exclusively dedicated to the processing of Sentinel-3 mission data products. The scope of this new service will be to maximize the exploitation of the upcoming Sentinel-3 Surface Topography Mission’s data over all surfaces. The service is open, free of charge for worldwide scientific applications and available at https://gpod.eo.esa.int/services/CRYOSAT_SAR/.

More info can be read at:

http://wiki.services.eoportal.org/tiki-index.php?page=GPOD+CryoSat-2+SARvatore+Software+Prototype+User+Manual

Sentinel-2 is a multispectral, high-resolution, optical imaging mission, developed by the European Space Agency (ESA) in the frame of the Copernicus program of the European Commission. It is based on a constellation of 2 satellites: Sentinel-2A launched on 23 June 2015 and entering now in its routine phase after 9 months of commissioning; and Sentinel-2B scheduled for end of 2016.

As part of the Payload Data Ground Segment (PDGS), the Mission Performance Centre (MPC) is the element within the PDGS in charge of covering the following functionalities:

-          Calibration (CAL) – to update on-board and on-ground configuration data in order to meet product quality requirements.

-          Validation (VAL) – to assess the quality of the generated data products.

-          Quality Control (QC) – to routinely monitor the status of the sensor and to check if the derived products meet the quality requirements along mission lifetime.

-          Data processors and tools corrective and perfective maintenance (PTM) – to manage the updates of the processors, quality control tools, calibration tools and all auxiliary files.

-          End-to-end system performance monitoring (E2ESPM) – to monitor the performance of the Sentinel-2 system operations (including the instrument) and assessing the system.

This poster concentrates on the S2A image quality Validation activities performed within the MPC since S2A launch, making an up-to-date status of L1 product quality and providing details on the methods applied.

All these activities have been validated during commissioning operations and are now ready for the routine phase. Cross-comparison with results of the commissioning team ensures the continuity of the mission performance.

The L1 performance validations performed cover:

 Level 1 Radiometric Validation:

-          Equalization Validation

-          Absolute Radiometry Vicarious Validation

-          Absolute Radiometry Cross-Mission Validation

-          Multi-temporal Relative Radiometry Vicarious Validation

-          Inter-band Relative Radiometric Uncertainty Validation

-          SNR Validation

-          Pixel Response Validation

-          MTF Validation

-          RTS monitoring

-          GPP/PDGS products comparison

Level 1 Geometric Validation:

-          Geolocation Uncertainty Validation

-          Multi-spectral Registration

-          Uncertainty Validation

-          Multi-temporal Registration Uncertainty Validation

-          Global Reference Images Validation

-          GPP/PDGS products comparison

The validation activities are performed within the MPC frame, combining automatic or semi-automatic analysis performed by tools and operators at MPC Coordinating Centre (MPC/CC) and deeper analyses performed by the MPC Expert Support Laboratories (MPC/ESL), entities providing scientific, expertise and engineering support to the mission.

During the commissioning phase, nominal calibrations have been performed and main anomalies investigated and corrected  (pointing oscillations due to solar array rotation, attitude restitution), leading to a high geometric and radiometric quality of the product, closely monitored by the MPC validation team.

Overall, the Sentinel-2 mission is proving very successful in terms of product quality thereby fulfilling the promises of the Copernicus program.

The ESLs in charge of the Level 1 validation activities are:

-          Thales Alenia Space, Cannes, France: ESL Level-1 Products Validation Leader

-          Argans, Plymouth, UK: ESL Level-1 Products Radiometry Validation

-          ONERA, Toulouse, France: ESL Level-1 Products Radiometry Validation

-          CSSI, Toulouse, France: ESL Level-1 Products Radiometry & Geometry Validation (also Prime Contractor, ESL Coordinator)

The other entities involved in MPC/CC activities are:

-          ACRI ST, Sophia Antipolis, France: Technical Manager

-          Argans, Plymouth, UK: Operation Manager, Operator

-          Elecnor Deimos, Madrid, Spain: Operator

The European Space Agency’s  (ESA’s) Climate Change Initiative (CCI)  Programme, producing harmonised datasets from long term Earth Observation satellite data records for a number of Essential Climate Variables (ECV’s), provides a solid basis for climate science and modelling, for specialist application development and ultimately for European and global policy making.

The first phase of the CCI set up thirteen different ECV projects and addressed the development and validation of algorithms to meet GCOS ECV requirements, and produced and validated time series of multi-sensor global satellite products for climate research and modelling. During Phase 2 (2014-2017) the aim is to extend initial ECV products to full mission time series, improve their quality to meet climate user needs, and maximise scientific impact.

The resulting Climate Data Records (CDRs) represent a major investment of science, funding and personal effort, therefore access to those products by a broad user base is a key element of programme success.

The Phase 1 datasets, and early Phase 2 datasets, are currently being distributed through the individual ECV teams, and access procedures and interfaces vary.  To complement and unify the work of the individual teams and to maximise the visibility and uptake of ECV data in the climate data user community within and beyond the CCI,  a new ESA CCI project has started,  to create a central open data portal and metadata catalogue for the ESA CCI project.

The CCI Open Data Portal project will provide a single point of harmonised access to disseminate mature and validated ECV products to a widened user base and engage with the user community, with key requirements of user-friendliness, reliability and long-term availability. The Portal will facilitate free, open and easy access to ECV data products generated through the CCI, with its improved programme web site, development and operation of a central data archive, and access to data through multiple interfaces and protocols. Support to ECV teams for maintaining their individual web sites and submitting their data to the Obs4MIPs (Observations for Model Intercomparisons) activity will also be provided via the Portal service.  In addition, an analysis of requirements for a future ESA  exploitation platform for climate, beyond the immediate CCI Open Data Portal, will be undertaken.

The Portal is envisaged as a living, dynamic environment benefiting both external users and CCI partners and is being developed in parallel with the CCI Toolbox, offering potential opportunities for links and synergies between the two.

This paper describes progress and plans for the development and operation of the CCI Open Data Portal, and outlines plans for its evolution in response to the evolving needs of the climate community. The project is being led by Telespazio VEGA UK Ltd, with partners from STFC, CGI, Reading University and Brockmann Consult.  

The long time-series of observations made by the Along Track Scanning Radiometer (ATSR) missions represents a valuable resource for a wide range of research and EO applications.

The ATSRs are multi-channel imaging radiometers with the principal objective of providing data concerning global Sea Surface Temperature (SST) to the high levels of accuracy and stability required for monitoring and carrying out research into the behaviour of the Earth's climate.

The first ATSR instrument, ATSR-1, was launched on board ESA’s European Remote Sensing (ERS) satellite, ERS-1 in July 1991. ATSR-1 was followed by ATSR-2, launched on ERS-2 in April 1995 and the Advanced Along Track Scanning Radiometer (AATSR) on ENVISAT in February 2002. Whilst the AATSR mission has now come to an end with the loss of ENVISAT in April 2012, the data from these three missions provides a significant 20+ year data set of global observations in the Short Wave Infra-Red (SWIR) and Thermal Infra Red (TIR) regions. In time, this data set will also be augmented by observations from the fourth ATSR-type instrument, the Sea and Land Surface Temperature Radiometer (SLSTR), to fly on ESA's Sentinel-3 satellite.

With the advent of ESA’s Long-Term Data Preservation (LTDP) programme, thought has turned to the preservation and improved understanding of such long time-series, to support their continued exploitation in both existing and new areas of research, bringing the possibility of improving the existing data set and to inform and contribute towards future missions.

For this reason, the ’Long Term Stability of the ATSR Instrument Series: SWIR Calibration, Cloud Masking and SAA’ project, commonly known as the ATSR Long Term Stability (or ALTS) project, is designed to explore the key characteristics of the data set and new and innovative ways of enhancing and exploiting it.

Main areas of investigation in Phase 1 (July 2013 to June 2015) have been:

Exploring a new approach to the assessment of Short Wave Infra-Red (SWIR) channel calibration.

The development, implementation validation of a new method for Total Column Water Vapour (TCWV) retrieval.

A study of the South Atlantic Anomaly (SAA) using ATSR data.

Radiative Transfer (RT) modelling in support of TCWV algorithm development.

Prototyping of a tool to provide AATSR observations with their location in the original instrument grid.

Topics for Phase 2 (Sept 2015 to Sept 2016) include:

The development of a strategy for the retrieval and archiving of historical ATSR mission documentation.

An extension of the ATSR calibration analysis work.

An extension of the work on TCWV to include retrievals over land

The development of potential new methods for cloud masking

Continuation of the development on the AATSR GBT-UBT-Tool and an extension to include the analysis of issues associated with the Sea and Land Surface Temperature Radiometer (SLSTR) on the Sentinel-3 platform.

This paper provides an overview of these activities and illustrates the importance of and issues associated with preserving and understanding ‘old’ data for continued use in the future.

The Radar Altimeter Database System (RADS) provides access to harmonized, validated, and cross-calibrated sea level data acquired by all satellite altimeters. The data are provided at a sampling frequency of 1 Hz, which is obtained by applying a so-called boxcar filter to the original 10 Hz, 20 Hz or 40 Hz data. It is, however, well known that the boxcar filter is far from being an optimal choice; in particular, artifacts may be introduced if additional filtering is applied in later data processing steps. In our presentation, we will evaluate the choice of this filter design. In doing so, we will first show how the choice of the boxcar filter affects the 1 Hz CryoSat data stored in RADS by comparing these data to the 1 Hz data obtained by applying more suitable low-pass filters, such as a Butterworth filter. Second, we will evaluate the consequences of the current filter design when successive filtering is applied. Finally, we will assess the impact a more suitable filter design will have on regional gravity field modeling using radar altimeter data.

This communication deals with the SHAPE study that was kicked off on 14 September 2015. SHAPE stands for Sentinel-3 Hydrologic Altimetry Processor prototypE. The team, the objectives, the work breakdown structure, the methodology, the technical approaches, the first results as well as the status and the upcoming milestones of the project will be presented.

This study is part of SEOM, Scientific Exploitation of Operational Missions, an ESA programme element which aims at expanding the international research community, strengthening the leadership of the European EO research community and addressing new scientific researches.

This Research and Development study not only intends to make the best use of all recent improvements in altimetry but also clearly pushes for major breakthroughs that should boost the scientific use of altimetry data (and especially “SAR” mode data) in hydrology. The stakes are high in the context of climate change, as scientists need to improve their analyses of water stocks and exchanges over wide geographical regions. The study focuses on three main variables of interest in hydrology: river stage, river discharge and lake level, which are part of the Terrestrial Essential Climate Variables (TECV) defined by GCOS. It also is the scientific step towards a future Inland Water dedicated processor on the Sentinel-3 ground segment.

The main characteristics of the project will be summarized. Cooperation with the scientific community will be encouraged. Project documents available at the website (ATBD for example) will go through a critical review outside the project team so as to collect feedback. Valuable feedback will be taken into account so as to provide a new processing chain that should be capable of providing water heights of unprecedented quality, making it possible to couple it with the hydrological dynamic and semi-distributed model HYPE (Hydrological Predictions for the Environment). This model has been developed by SMHI and will be used to assimilate study's new "Alti-Hydro" Products to assess the added value of space altimetry in the forecast and hind-cast of river discharge, river water level, and lake water level for a varied range of study areas.

Time series of satellite observations at high spatial (10-30 m pixel size) and temporal (7-15 days) resolutions are needed in research, administration and business for mapping and analysing various phenomenon on the land surface such as land use and land cover (LULC), LULC change, crop breeds/species, crop yields, long term trends, phenology and many more. Thanks to the Landsat program, which opened up its >30 years archive in 2008, and to the recently launched Sentinel-2 satellite, the remote sensing community has now a wealth of information at its disposal to work with time series. However, the main factors limiting the use of these data are:

Missing data due to prolonged cloudy periods;

Data affected by clouds and cloud-shadows;

 Cloud-free observation not equality spaced in time;

Sensor saturation due to snow coverage;

Atmospheric effects adding path radiance and reducing the transmissivity of the atmosphere;

Sensor failure such as the Landsat-7 scanning line problem (SLC-off data) and shut-down of Landsat-5 in 2012;

Heterogeneity of information and surface conditions.

The above-mentioned problems are further aggravated by the fact that most effects depend on the area (e.g. mountainous and coastal areas with little usable data). Mosaicking and temporal compositing is often proposed as a solution to overcome these issues. The idea is to combine several images in order to fill in the missing data. This concept has its limitations, as two individual points in time can undergo a significant change. This phenomenon is well documented in remote sensing of vegetation, where the crop growth and phenology can undergo a rapid transformation driven by climatic events. Furthermore, a variation in atmospheric conditions can impede usability of the data over short spatial and temporal scales. Suggested solutions such as histogram matching are often not sufficient and can result in sharp borders between two images.

The scope of this paper is to introduce a novel algorithm to create spatially and temporally consistent and continuous time series of cloud-free, bottom-of-the-atmosphere (BOA) multi-spectral images. The algorithm generates smoothed and gap-filled spectral data, based on inputs from the Landsat Surface Reflectance Climate Data Record (Landsat CDR). The provided quality assessment flags are taken into account to identify high quality observations (pixels not contaminated by clouds and cloud-shadows or missing data). To prepare for the soon available Sentinel-2 data, the algorithm was implemented, tested and validated with Landsat-5, -7 and -8 satellite data. Sentinel-2 data will be integrated as they become available.

The novelty of the approach stems from its use of templates (i.e. information derived from the neighbourhood of the pixel to be filtered that can be used to complement high quality observations) and the use of a state-of-the art Whittaker smoother (largely exploited for the filtering and gap-filling of MODIS NDVI data).

The algorithm can produce any desired temporal resolution (weekly to monthly). Contrary to traditional smoothing algorithms, which only work reasonably well for vegetation indices (Xiao et al. 2015), the developed procedure is applied for each spectral channel individually. Examples of raw and filtered time series are provided in Figure 1 for the near-infrared spectral band.

Once the cleaned and gap-filled multi-spectral data are obtained, it is relatively straight forward to derive vegetation indices and to apply other spectral transforms. For instance, from the filtered reflectance in the six reflective spectral channels of Landsat, the NDVI (Normalized Difference Vegetation Index), NDWI (Normalized Difference Water Index), fAPAR (fraction of absorbed photosynthetically active radiation) and Tasselled Cup Transformation (TCB) were calculated according to the equations described in Rouse et al. (1974) (NDVI), Gobron and Taberner (2008) (fAPAR), Gao (1996), Jackson et al. (2004) (NDWI), Shi and Ding (2011) (TCB). A pixel-based example is reported in Figure 2. We also investigated the use of physically-based radiative transfer models (RTM) for retrieval of LAI and other bio-physical vegetation variables.

The method was tested for a study area encompassing Lower Austria and it was validated by comparing our smoothed and gap-filled data to high quality datasets, which we excluded from our filtering process. An example of RGB (Near-infrared, Red, Green) image composite is shown in Figure 3 for the test site for the year 2009 (temporal resolution of gap-filled data is 15-days). Overall, our results are very satisfying and match the validation dataset. Only occasionally the time series reveal artifacts mainly originating from the cloud mask, which does not flag all atmospheric effects or shadows. Also during periods of prolonged cloud cover, not enough high quality input data might be available to derive high quality outputs. We expect to resolve most issues, when Sentinel-2 data becomes available and will be integrated into the tool.

The  availability of   harmonized Sentinel 2 - MSI  / Landsat 8 - OLI time series, at 10 m spatial resolution,  fit for the purposes of  many downstream services  involved in land applications is becoming more and more urgent.

In the context of pixel based analysis, the harmonization topic address finally two fundamental stages. The first one is the consolidation of Sentinel 2 data / Landsat 8 data in order to associate QA information to time series data quality and also to recover missing measurements for certain dates because of occultation. The second ones, is the temporal densification of Sentinel 2 data with Landsat 8 data.

The project has in first focus on the development and testing of methods for signal reconstruction. Instead of restoring a one date measurement by using the closest past measurements and the closest future measurement (interpolation), a new approach has been proposed. This approach takes benefits of the temporal structure of a time series by measuring distance between input time series and reference ones in order to find the most appropriate one date measurement to be recovered (distance based method).

The different methods have been benchmarked based on a Sentinel 2, Landsat 8 and RapidEye data. From data validation point of view, accuracy and uncertainty metrics have been systematically computed when consolidating time series with the different methods and when one date measurements are occulted.

By comparing methods, the accuracy remains the same, it is nonetheless important to note that the uncertainty is significantly reduced with distance based method. Beside, sensitivity to several parameters when consolidating with time series is analyzed.

The second stage of the project is focus on the merging of Sentinel 2 and Landsat 8 time series including specific work on radiometric calibration alignment and also unmixing due to spatial resolution difference.

The ESA Landsat MSS, TM, ETM+ archives reprocessing campaigns have been initiated since 3 years. The dissemination to the user of high quality Landsat ESA  products have started. Beside, a  better understanding of the mission behavior, sensor calibration, archive quality has been reached as never in the past. This statement is more than true for the Landsat MSS missions.

The Landsat MSS data have been recorded data from 1972 up to 1990. The MSS instruments have been designed to have four visible/nir spectral bands, large spectral bandwidth, with a spatial resolution of 60 m.

Unlike TM, ETM+, the MSS radiometric processing is based on  in-flight calibration by using internal calibrator data. More over calibration parameters are mainly applicable for mid latitude observations.

As consequences, data observed in bright scene (snow, ice, desert regions), during spring / summer periods, are today strongly affected by image data saturation. This issue is known on the  ESA side and on the USGS side, as well.

Whilst USGS  is not able  improve the radiometric processing because data saturation exist in the  to  Level 0 archive (radiometric calibration applied on Level 0), the situation is somehow different on the ESA side: the Level 0 includes raw data with no calibration applied. 

Although, It has been discovered that even for bright scenes, the Level 0 data is not saturated. It is therefore the Level 1 processing that introduce data saturation.

This presentation shows the proposed change in the  radiometric processing algorithm and also the calibration methods applied to propose a new set of calibration parameters. Snow/ice sites and CEOS endorsed sites have been selected, and inter calibration exercices with TM data have been performed

At the time of writing, it is too early to provide consolidated  results, we are confident, but still leading experiments (please refer to documents attached for more information).

The radio occultation technique based on the refraction of an electromagnetic signal between a GNSS satellite (Global Navigation Satellite System) and a receiver satellite (located in low orbit) provides a way to observe the Earth's atmosphere, especially its temperature, pressure and water vapor, but also of the ionosphere. This technique is now considered a mature concept, the benefits clearly recognized by the communities of weather prediction, climatology and space weather. Since 2006, the GNSS radio occultation measurements have been distributed in near real time in the weather forecasting centers around the world, especially those equipped with global weather models. Currently the measurements from the radio occultation are used operationally in a decade of meteorological centers (for example Météo-France). In addition to the NWP, these measures are used to monitor and track the evolution of some climatic parameters, especially the temperature of the stratosphere. For their climatology use, the GNSS-RO measurements are archived in several centers which can then reprocess and use them either directly in the form of climatological series GNSS-RO, either by assimilating them into a "re-analysis", together with all other observation systems normally assimilated into weather forecasting models. The GNSS signals are "perturbed" not only by atmospheric refraction, but by their interaction with the ionized particles in the ionosphere. In particular, these signals allow us to observe and monitor the ionosphere by measuring its content of electrons (TEC; Total Electron Content). The establishment of a three-dimensional map in near real time of the ionosphere would allow monitoring the phenomena of ionization and scintillation. This monitoring would interest the space platforms located a few hundred kilometers of altitude and all telecommunication systems based on wavelengths which are reflected on the ionosphere.

Radio occultation measurements have the advantage of being very precise, independent of the instrument (no calibration required), little sensitive to weather conditions, to have a very good vertical resolution and global coverage. They perfectly complement the measurements made by other atmospheric sensors (radio probes, sonar infra-red). Increased number of measures need was expressed by such users to improve the performance of their models (16000 occultations per day against currently 3000) for improved weather prediction and better knowledge of the climate. This should be ensured until 2020 thank to the current and future satellite constellations. Beyond 2020, no program is planned this day to ensure such data volume. Consequently, CNES has decided to realize a phase 0 aiming at the elaboration of a concept mission based on a minimalist instrumentation compatible with a constellation of small satellites.

The paper presents the proposed constellation to meet these increased requirements of radio occultation measurements. It describes the instrumental concept chosen for the GNSS receiver in order to be implemented on a small satellite while respecting the objectives of lower costs. It also presents the satellite architecture and the implementation of the system architecture to meet the needs of measurements in near real time expressed by users.

Differential Synthetic Aperture Radar Interferometry (DInSAR) [1],[2]⁠ and Persistent Scatterer Interferometry (PSI) [3]–[5]⁠ are both recognized as well-established radar interferometric methodologies used to measure and monitor land deformation phenomena. Their maturity has been demonstrated in a wide range of applications, including land motion measurements caused by both natural and anthropogenic activities. Considering the limitations that both DInSAR and PSI have when they are used individually, new efforts have been made by incorporating these methods [6],[7]⁠. A key limitation of DInSAR and especially PSI is the poor density of point targets over natural terrains where no or few buildings are present

Hence, we developed a hybrid approach using elements from both DInSAR and PSI methods. The main aim of this synergy was the exploitation of distributed scatterers and therefore the improvement of the spatial coverage. The concept of the methodology includes three main parts: the DInSAR, the PSI and the combined one, ensuring that we use either a single or multi-reference stack configuration. In the DInSAR part, a stack of 2d differential interferograms is generated. The key factor of this part of the analysis is the strong multi-looking in range and azimuth. Strong multi-looking permits phase noise reduction which results potentially in interpretable phases. Multi-looked (ML) differential phases are assigned into point locations in order to make the integration with the PSI approach feasible. In the second, the PSI part of the approach, a stack of 2d point differential interferograms, identical to the first part, is generated. This time, a point candidate list is determined using two different approaches. The first one is based on the spectral characteristics of each point target (PT) and the second on the backscatter or temporal variability. Optionally, a point density reduction can be applied on the point candidate list in order to keep the high quality point targets in the final list [8]. Based on the final PT list, 2d point differential interferograms are also generated. Finally, having the PT candidate list and the initial differential point interferograms from the PSI processing part, we proceed to the combined part of the approach. The two point lists, as well as the corresponding data stacks of the differential interferograms, derived from the PSI and the DInSAR methodologies respectively, are joined. The combined lists and the combined differential interferograms are then used for further processing.

The hybrid approach has been tested successfully in two different types of natural terrain [9], [10]⁠ using L-band data and demonstrated a strong improvement. Now the approach is applied using Sentinel-1 data. This is of particular interest because of the short 12-day repeat interval and the short spatial baselines which both improves the potential of the distributed scatterers, in particular when considering multi-reference stacks with short time intervals.

In our contribution we will present the hybrid time-series methodology using Sentinel-1 IWS data stacks.

References

[1] H. Zebker and R. Goldstein, “Topographic mapping from interferometric synthetic aperture radar observations,” J. Geophys. Res., vol. 91(B5), pp. 4993–4999, 1986.

[2] A. K. Gabriel, R. M. Goldstein, and H. a. Zebker, “Mapping small elevation changes over large areas: Differential radar interferometry,” J. Geophys. Res., vol. 94, no. B7, p. 9183, 1989.

[3] A. Ferretti, C. Prati, and F. Rocca, “Permanent scatterers in SAR interferometry,” Geoscience and Remote Sensing Symposium, 1999. IGARSS ’99 Proceedings. IEEE 1999 International, vol. 3. pp. 1528–1530 vol.3, 1999.

[4] A. Ferretti, C. Prati, and F. Rocca, “Permanent scatterers in SAR interferometry,” IEEE Trans. Geosci. Remote Sens., vol. 39, no. 1, pp. 8–20, 2001.

[5] C. Werner, U. Wegmüller, A. Wiesmann, and T. Strozzi, “Interferometric Point Target Analysis with JERS-1 L-band SAR Data,” no. July, pp. 25–27, 2003.

[6] A. Hooper, “A multi-temporal InSAR method incorporating both persistent scatterer and small baseline approaches,” Geophys. Res. Lett., vol. 35, no. 16, p. L16302, Aug. 2008.

[7] A. Ferretti, A. Fumagalli, F. Novali, C. Prati, F. Rocca, and A. Rucci, “A New Algorithm for Processing Interferometric Data-Stacks: SqueeSAR,” IEEE Trans. Geosci. Remote Sens., vol. 49, no. 9, pp. 3460–3470, 2011.

[8] U. Wegmüller, O. Frey, and C. Werner, “Point Density Reduction in Persistent Scatterer Interferometry density,” in EUSAR. 9th European Conference on Synthetic Aperture Radar, 2012.

[9] P. Kourkouli, U. Wegmüller, A. Wiesmann, and K. Tansey, “An Integration of 2-D Multi-looked Differential Interferograms with Persistent Scatterers Interferometry,” in Living Planet Symposium, 2013.

[10] P. Kourkouli, U. Wegmüller, P. Teatini, L. Tosi, T. Strozzi, A. Wiesmann, and K. Tansey, “Ground Deformation Monitoring Over Venice Lagoon Using Combined DInSAR/PSI Techniques,” in Engineering Geology for Society and Territory - Volume 4 SE  - 35, G. Lollino, A. Manconi, J. Locat, Y. Huang, and M. Canals Artigas, Eds. Springer International Publishing, 2014, pp. 183–186.

The paper addresses the preparatory studies of future ESA mission concepts devoted to improve our understanding of Earth’s mass transport phenomena observable as temporal variations in the gravity field, at different temporal and spatial scales. These are due to ice sheet and glaciers growth or melting, changes in continental hydrology, sea level change and deep ocean mass dynamics, atmospheric dynamics and  olid-earth deformations. The ESA initiatives, started in 2003 with a first study on observation techniques for future solid Earth missions. This was recently continued through several system studies, technology developments and constellation simulations. The activities concentrated for example on propulsion, e.g. tests on miniaturized ion thruster, or distance metrology, e.g. laser interferometry. These activities received precious inputs from the in-flight lesson learnt from ESA’s Gravity field and steady-state Ocean Circulation Explorer (GOCE) mission and the American-German Gravity Recovery and Climate Experiment (GRACE) mission.

Two parallel studies for the “Assessment of a Next Generation Gravity Mission to Monitor the Variations of the Earth’s Gravity Field” (concisely: NGGM) were performed, one led by Thales Alenia Space (Turin, Italy) and another one by Astrium GmbH (Friedrichshafen, Germany). Both consortia included European universities and academic institutions for scientific support and
requirement assessments. Several mission aspects were analyzed, leading to different mission concept features. The preferred mission concepts fitting the at that time defined programmatic boundary conditions have been studied with prioritized science requirements and detailed system designs. In addition to drivers such as tight propulsion requirements and accelerometer calibration issues, resulting in a dominant error source at large scales, technical constraints on power and fuel generally dictate the choice of orbit. Thus, for each considered constellation of
satellite pairs, the different interactions between drag-free and “loose” formation control have been analyzed together with the design of the relative attitude control that is necessary to ensure the inter-satellite laser link all along the mission length.

Since then, several complementary GSP (General Study Programme) and TRP (Basic Technology Research Programme) studies were initiated and some are currently running, notably:

• Next Generation Gravity Mission: AOCS Solutions and Technologies, with the objective to define and evaluate the mission-critical Attitude and Orbit Control System (AOCS) solutions, to identify the critical technologies and to assess their feasibility and the design drivers.

• Miniaturized Gridded Ion Engine (GIE) Breadboarding and Testing: the ongoing activity aims at developing a miniaturized GIE with increased thrust level, with thrust ranging from 50 μN to 2.5 mN.

• Consolidation of the micro-PIM Field Emission Thruster design for NGGM: In particular the study aims to demonstrate that the mN-FEEP thruster is an excellent candidate for the lateral thrusting on NGGM.

• High-Stability Laser with Fibre Amplifier and Laser Stabilisation Unit for Interferometric Earth Gravity Measurements: The objective of this development is the manufacturing of an elegant breadboard (EBB) of a high stability laser (HSL) reaching TRL 6 at subsystem level in 2016.

• Assessment of Satellite Constellations for Monitoring the Variations in Earth's Gravity
Field: aimed at the optimization of “Bender-type” constellations of two pairs of satellites for the retrieval of the time-variable gravity field to monitor mass distribution and transport. The study focused on the reduction of temporal and spatial aliasing by optimizing the design of the constellation and by explorating more advanced methods for the time-variable gravity field retrieval. The latest results of this scientific study will be presented in parallel sessions at the LPS 2016.

At the time of writing this abstract, a new system study – named “ Consolidation of the system concept for the Next Generation Gravity Mission” – is in preparation: the purpose of this technical work is to focus on the design of a single satellite pair of the NGGM constellation that should have the flexibility to operate successfully both in near-polar or inclined orbits.

Meanwhile, the user community, also via a recent resolution of the IUGG (International Union of Geodesy and Geophysics), formally invited the space agencies to extend the gravity field science and application monitoring with a concerted international effort. In the framework of a cooperation on future EO technologies and missions, ESA and NASA therefore established an Interagency Gravity Science Working Group of experts tasked to define the overall objectives and the observation requirements for a future constellation with better performance than that of a concept elaborated by a single agency – in fact, a performance better than “the sum of the parts” when these were not jointly defined. The latest results concerning the preferred satellite architectures and constellations, payload design and estimated science performance will be presented as well as remaining open issues for future concepts. Attention will also be given to the ongoing ESA-NASA inter-agency cooperation to align preparatory work on synergistic mission concepts beyond GRACE Follow-On.

The Radio Occultation (RO) instrument will be part of the core payload of the future MetOp-SG satellites to be flown as part of the EUMETSAT Polar System-Second Generation (EPS-SG) from 2021 onwards. In total there will be 6 Metop-SG satellites, all of them carrying RO instruments. The development of MetOp-SG builds upon the experience of the very successful MetOp-A and –B satellites launched in 2006 and 2012 respectively. RO is nowadays a well-established technique for the retrieval of vertical profiles of temperature, pressure as well as humidity in the lower troposphere and upper stratosphere.

 Initially, the RO instruments will be capable of tracking GPS and Galileo GNSS (Global Navigation Satellite System) signals on the L1 and L5 frequency. The onboard AGGA-4 (tracing back to the initial Advanced GPS/GLONASS ASIC) should also enable the tracking of future GNSS signals from the Chinese COMPASS/BeiDou and the Russian GLONASS constellation at these frequencies. These instruments will contribute to primary mission objectives of the EPS-SG in the areas of operational meteorology and climate monitoring.

One of the primary tasks in the development phase is the preparation of a science plan to detail the scientific work, which is needed to meet the RO related mission objectives. This plan will be prepared by the RO Science Advisory Group (SAG) and external experts supported by ESA and EUMETSAT. The science plan must especially establish the scientific requirements for the RO related component of the EPS-SG ground segment. The plan will be used as reference for scientific activities to be undertaken within and outside the RO SAG in the coming years.

This presentation will start with a short summary of the RO instrument and requirements and then focus on the science plan content and the schedule of the preparation of this plan. The presentation is intended, on the one hand, to inform the wider RO science community about potential future related activities and, on the other hand, to engage them to make recommendations on any issue needed to be tackled to reach the objectives of the RO instrument.

The coast is a unique environment in which atmosphere, hydrosphere and lithosphere contact each other. Coastal zone monitoring is an important task in sustainable development and environmental protection. For coastal zone monitoring, coastline extraction in various times is a fundamental work. A coastline is the area where land meets the sea. Actually, sometimes it is very difficult to determine a precise line that can be called “coastline”, due to the dynamic nature of the sea. Coastal management is becoming vitally important to local authorities faced with problems of erosion and subsidence, which threaten the stability of the land and safety of the people they are responsible for.

In this study, dual-polarimetric C- and X-band SAR data are exploited to extract the continuous coastline. In addition, co-located GPS measurements are exploited to analyze the performance of the C- and X-band coastline extraction methods. The test area is the coastal area of Monasterace, south of Italy, where C-band Sentinel-1 and RadarSAR-2 and X-band Cosmo-SkyMed (CSK) multi-polarization SAR data have been collected together with ground truth GPS measurements and optical images. The area includes sand, dunes and urban areas.

The analysis is undertaken exploiting both single-polarization and dual-pol observables. First the co-polarized normalized radar cross-section (NRCS) is predicted using a two-scale scattering model to analyze the sand/sea contrast in case of wet and dry soil at both C- and X-band. Then dual-polarimetric observables are exploited. The metrics adopted are based on combination of co- and cross-polarized channels. First it is shown that due to reflection symmetry the phase between co- and cross-polarized channels is uninformative and, therefore, only amplitude information is considered. Then, the probability density function (pdf) of the correlation between co- and cross-polarized amplitude channels is analyzed to conceive unsupervised coastline extraction technique able to exploit C- and X-band SAR data.

Experiments confirm the soundness of the approach showing that both C- and X-band SAR data show a remarkable performance with C-band yielding the lowest error with respect to ground truth GPS. Those results confirm the effectiveness of Sentinel-1 for this application.

Theme: Oceanography

Oral presentation

Marina A. Evdoshenko, P.P. Shirshov Institute of Oceanology Russian Academy of Sciences, 36 Nakhimovski Prospect, Moscow, Russia, 117997, E-mail: maarsio@bk.ru

Detecting Imprints of Atmospheric Waves with Ocean Color Scanner Data

Abstract                                                   

Satellite ocean color scanners enable to examine atmospheric gravity waves (AGWs), formed in low troposphere, by their signatures in the sky and on the sea surface. AGWs are known to be generally invisible, but in case of sufficient moisture they display as “cloud streets”. Optical radiation does not penetrate through cloud, so cloud streets are visible on images of level L1b (without the atmospheric correction) calibrated geolocated  radiances L_t(l) at definite wavelengths l, measured by ocean optical scanners. AGWs, due to changing of atmospheric pressure, simultaneously modify a free water level and produce imprints in the form of light and dark stripes on the sea surface. In the absence of cloud the stripes are visible on images of level L2 (corrected to the atmosphere) remote sensing reflectance Rrs(l) measured by the scanners. In many published papers atmospheric gravity waves in different geographical regions were examined on the basis of L_t(l) images by their imprints as cloud stripes.  But we have not found any papers where investigation of AGW signatures on the sea surface was carried out with the help of Rrs data. In the paper data of so called high resolution: 300-m from MERIS-Envisat at all accessible l and 250-m from MODIS-Aqua and -Terra at 645 and 859 nm were used. AGW signatures on the sea surface are most noticeable in waters where dynamical structures or changes of phytoplankton concentrations are slightly expressed. It was found that sea surface signatures in the form of surface waves are totally produced by AGWs propagating at different heights of low troposphere. Using subsequent images of the same water area obtained from MERIS, MODIS-Terra and MODIS-Aqua enable us to trace and examine the dynamics of AGWs. Surface waves induced by AGWs are also visible on images of a downwelling diffuse attenuation coefficient at 490 nm k_d(490). Values z₉₀(490)  reciprocal to k_d(490), represent a 90%-depth of light penetration at 490 nm. Evaluations made on the basis of z₉₀(490)changes showed that a height of surface waves reaches dozens centimeters. Besides the time variability and vertical amplitude, other characteristics of surface wave, reflecting AGWs, were estimated, as a wavelength reaching several kilometers, crest length reaching more than a hundred kilometers, and direction of propagating. Imposition in a difference mode of images obtained simultaneously: L_t(l)with signature of AGWs in cloud and Rrs(l) with signature of induced surface waves give possibility to identify a horizontal phase shift between cloudy and surface waves. It is known that cloud streets in the AGW can be formed between two adjacent counter-rotating atmospheric vortices above the updraft part of the vortices circulations, maximum and minimum amplitudes of the AGW coincide with eddy axes, and cloud streets correspond to areas of near zero vertical displacement of the AGW. From imposition images it follows that a frequently observed situation is when cloud stripes have prolongations in crests of surface waves and are parallel to them. In this case AGWs and surface waves are shifted by minus quarter of the period. Sometimes crests of surface waves are not parallel to cloud stripes and a small horizontal angle between them is observed. This situation can be caused by different orientation of atmospheric waves, being frequently upon influence of the wind, at a height of 1 - 2 km and near the sea surface.

Keywords: Ocean Colour, Coastal Zones

List of satellite: ENVISAT, Other

With the increasing density of operational, high-resolution optical EO data, the need for automatic atmospheric correction of this data for time-series analyses is becoming more pressing. Atmospheric corrections are of course very sensitive, as errors will reduce the comparability of the data sets or even make the determination of absolute values e.g. for plant parameter retrieval impossible.

VISTA has already for many years developed atmospheric correction procedures based on the radiative transfer model MODTRAN. For many years, the correction was done in a semi-automatic way, using user input for determination of the current visibility, which is the most influential factor for an accurate atmospheric correction. For areas, in which a good meteorological station network is available, using vicarious data is not a problem, but for remote areas, these data are often not available.

Hence, VISTA has developed a method to automatically retrieve the visibility from within the EO data based on the radiative transfer model SLC (Soil-Leaf-Canopy). Comparing the results of several runs of the inverse SLC modelling, based on atmospheric corrected images using different visibilities, each pixels’ visibility is defined by the SLC run producing the lowest RMS value.

After the atmospheric correction clouds and cloud shadows are detected. While cumulus cloud and cloud shadow masking works well using the reflective bands in the atmospheric windows, cirrus clouds prove more difficult. Due to their low density, they show up as a mixed signature, including the reflectance of the ground underneath.

Here, Sentinel-2’s cirrus sensitive 1.38µm band helps to support a more differentiated classification of different cloud types and the correction of cirrus influenced areas. Due to the fact, that the 1.38µm radiation is reflected by ice crystals but absorbed by the lower atmospheric water vapor, the surface is completely hidden but the upper cirrus clouds can be detected and classified independently from the ground. Additionally, the information about the cloud density can be used to correct cirrus influence from the reflectance spectra of the other bands, providing that the clouds aren’t opaque.  

For the cirrus correction itself, several algorithms from literature have been tested and adapted for the most successful cirrus removal. Waiting for the availability of Sentinel-2 data, the tests have substitutional been using Landsat 8 data. The results are satisfying but still limited by at least three difficulties:

the parallax of cirrus clouds in different bands caused by the different observation angles
which means the location of the detected cirrus clouds on 1.38µm is not bound to be equal to their location on other bands

the uncertain altitude of cirrus clouds

the irreproducible geometric correction of surface and cirrus clouds
especially in mountainous areas might cause uncorrectable shifts of the cirrus clouds

For minimizing the parallax effect, VISTA’s approach counts on Sentinel-2’s detailed documentation on each detector’s observing angles while estimating the cirrus clouds’ altitude by probability.

In conclusion, with Sentinel-2 an automatic atmospheric correction, deriving current visibility from the image itself, correcting for cirrus clouds and masking both cumulus clouds and cloud shadows with a high degree of accuracy is possible. This allows for an application of the same method running on areas anywhere around the globe, and also for automatic processing of large amounts of data, which is necessary for time-series analyses. For applications not only in the scientific but also in the commercial realm, high quality standardized processing chains like this are necessary for a successful opening of the down-stream market.

Today, utilizing radar imagery in different applications has been more attention by many researchers around the world due to introduction of high resolution SAR satellite sensors and data accessibility in 24 hours for all weather conditions. therefore, using high resolution SAR images in large scale mapping and specially DEM generation have been a practical reality. for example, availability of new high resolution SAR spaceborne sensors as COSMO-SkyMed, TerraSAR-X and RADARSAT-2 offers new interesting potentialities for the acquisition of data useful for the generation of DEMs and other applications. image matching is necessary for these applications. ZNCC is one of the most common methods in area based matching which usually leads to DEM production. ZNCC can be improved by multi size window and expanded window. Among feature based matching method, SIFT and SURF are most common methods whereas census and Ranklet are also used for sparse matching.

in this paper the multi step algorithm originated on the combination of feature and area based images matching. range and doppler are physical equations that commonly as geometric constrain for our proposed image matching

The matching step consists of several main parts. In this algorithm, a SAR image is firstly filtered by a speckle suppression algorithm and then a feature detection algorithm is used to extract feature points. then feature descriptors are extracted and local matching is performed. in continue, using a specific criterion and metric and using geometric and radiometric constraints, the corresponding features are matched. after performing local matching, global matching is done. the aim is to the an optimal multi step image matching algorithm.

In this study, we use a pair of TerraSAR-X Single-Look Slant-range Complex (SSC) images with short and long baseline. The images were acquired over the city of Jam, southern Iran, in spotlight mode and descending orbit

We use a part of these images by 700×700 pixels. the proposed optimization method is able to cope more complications than other methods with a high point correspondence accuracy in short and long base line images. this method match 679 point with subpixel precision for short base line images and 171 point with subpixel precision for long base line images in short time.

According to the results of the experiments, finally a multi-step optimization model for matching were determined. the proposed optimization method is able to cope more complications than other methods with a high point correspondence accuracy in short and long base line images. for example, according to short and long base line image paires the proposed the multi-step approach is 11% and 23% better than feature based matching with primary window description.

The Sentinel-3 Mission Performance Framework has been established by ESA and EUMETSAT to ensure, over the mission lifetime, the required operability and adequacy of the core science products generated by the Payload Data Ground Segment (PDGS) and delivered to the Copernicus Services.

The Sentinel-3 Mission Performance Framework embraces ESA and EUMETSAT ‘in-house’ expertise and experts as well as the Mission Performance Centre (MPC) Team, the Sentinel-3 Validation Team (S3VT), the Copernicus POD service and CNES POD team.

The S-3 MPC provides support to the Sentinel-3 project to ensure:

The calibration and characterization of the S-3 sensors (A and B unit);

The monitoring of the S-3 sensor performance (OLCI, SLSTR, SRAL, MWR);

The products calibration;

The validation of the S-3 core products using external and independent datasets (e.g. in-situ, numerical model or satellite data);

The validation and verification of the S-3 core products using statistic and monitoring tools;

The routine quality control of the S-3 core Land products generated by the Core Ground Station (CGS) and Land Processing Archiving Centre (PAC);

The maintenance and evolution of the calibration, validation and processing algorithms;

The Copernicus Precise Orbit Determination (POD) Service is in charge of the operational provision of the orbit products for the ground processing. On top of this, the POD Service performs long-term performance monitoring of the GNSS sensor, the quality control of the POD orbit products and platform attitude auxiliary data files with the support of POD and GNSS experts. Dedicated POD Quality Working Group meetings are held for the Copernicus Sentinel-1, -2 and -3.

The S-3 Validation Team (S3VT) provides support to the S-3 project in conducting field experiments and campaigns in support of Sentinel-3 Calibration and Validation.

The CNES French Agency performs the long-term performance monitoring of the DORIS sensor on board S-3 satellites as well as the quality assessment of the orbit products they will generate.

Sentinel-3 Quality Working Groups (QWG) regularly convenes for the OLCI, SLST and Altimetry communities with representation of the Mission Performance Framework partners, scientists and operational users. The S-3 QWG meetings, organized and co-chaired by ESA and EUMETSAT, will be the forum for discussing on the S-3 mission quality performance and recommending evolution of the S-3 Optical and Altimetry core products and processing algorithms baseline.

This paper provides an update on the status on the operations of the S-3 Mission Performance Framework during Commissioning Phase, including early results on the S-3 mission PDGS Altimetry and Optical products quality, Cal/Val results as well as early plans related to the upgrade of the Instrument Processing Facilities.

The valiant efforts by the coastal altimetry community to improve the availability and accuracy of the altimeter-derived parameters in the coastal zone are paving the way to a new generation of coastal altimeter products from past and present satellite missions, but research is more needed than ever to improve the products further and enable their full exploitation. Dedicated coastal-oriented retrackers have been developed in the last decade aiming at retrieving accurate information from contaminated radar waveforms. The Adaptive Leading Edge Subwaveform (ALES) retracker focuses on a subwaveform that includes the leading edge, which contains most of the oceanographic information, adapting the width of the fitting window according to a first estimation of the sea state. ALES has been successfully validated in a few locations (Passaro et al., 2014; 2015a; 2015b) and both for Sea Surface Height (SSH) and Significant Wave Height (SWH).

A further improvement to the accuracy of the coastal altimetry measurement is expected from enhancements in the corrections for atmospheric and surface effects. The sea state bias (SSB) is a case in point, being intimately linked to the choice of retracking algorithm, so it makes sense to re-compute it for a new retracker such as ALES. In this work we focus on the re-computation of the sea state bias (SSB) correction along a few Envisat RA-2 track segments (18 Hz posting rate) located around the Spanish coasts during its phase E2 (October 2002 - October 2010). To do this, we used sigma0 from ALES to estimate the wind speed following the methodology described in Abdalla (2012). Then, wind speed and significant wave height (from ALES) were used to estimate the SSB with the look-up table applied to obtain the SSB in the official SGDR product (Labroue, 2007). The performance of the sea level anomaly (SLA) was validated using 9 tide gauges located around the coasts of Spain. We compared time series of SLA using: (1) the recomputed SSB from ALES; and (2) the SSB available in the SGDR. Overall, we observed an abatement of the along-track SLA uncertainty by 18% using our recomputed SSB from ALES. The accuracy (in terms of RMSE) improved by about 15%.

Satellite radar altimetry, initially designed for studying ocean surface topography, demonstrated a strong potential for the continuous monitoring of ice sheets and land surfaces over the last 25 years. If radar altimetry is mostly used for its capacity to determine surface height, the backscattering coefficients provide information on the surface properties.

Spatio-temporal variations of radar altimetry backscattering over land and ice sheets were related to the nature of the surface and its changes against time. This study presents the results of an along-track analysis of radar altimetry echoes over land, Antarctica and Greenland at S, C, Ku and Ka bands using data from Topex/Poseidon, Jason-2, ERS-2, ENVISAT and SARAL on their nominal orbit during the tandem phase of the two missions. Temporal average and deviations are presented at global scale.

Amanda Regan*, Pierluigi Silvestrin†, Diego Fernandez‡, Karl Atkinson#,

Nicolas Leveque#, Rachel Bird^, Johnny A. Johannessen#,

Thomas Nagler, Helmut Rott, Ad Stoffelen&, Martin Wooster°, Neil

Humpage§, John Remedios§

The successful launch of Sentinel-1A and Sentinel-2A signified the beginning of the dedicated space segment for the Copernicus Programme, the result of the partnership between the European Commission (EC) and the European Space Agency (ESA). Sentinel-1A and Sentinel-2A are the first of a long-term operational series of Earth Observation (EO) satellites to be launched by Europe that will complement the already well-established series of meteorological missions.

For the first time, these missions will provide a continuous and longterm European capability for systematic observations of the Earth surface, its oceans and atmosphere to unprecedented accuracies, resolutions, and temporal coverage. If additional cost-effective missions could be flown together with these operational missions (including operational meteorological satellite series such as MetOp (Second Generation - SG) then the possibilities for meeting new Earth science and application objectives could be far-reaching e.g. fulfilling observational gaps, synergistic measurements of Earth system processes, etc. To explore this potential, the ESA initiated three exploratory paper studies (known as the EO-Convoy studies). The aim of these studies is two fold: Firstly, to identify scientific and operational objectives and needs that would benefit from additional in-orbit support. Secondly, to identify and develop a number of cost-effective mission concepts that would meet these objectives and needs. Each EO Convoy study is dedicated to a specific theme, namely: Study 1 - Ocean and Ice Applications,Study 2 - Land Applications and Study 3 - Atmospheric Applications.

This paper will present the results of the EO-Convoy studies including an overview of the user needs and derived convoy concept descriptions. This paper shall focus on the resulting science benefits. Example convoy concepts to be presented include a passive C-band SAR flying with Sentinel-1 and possible free flying thermal infrared payloads flying with Sentinel-2, Sentinel-3 and Landsat-8.

Definition of altimetry vitual stations over inland water bodies requires a very specific processing composed of a refined selection of valid data for estimating water levels and eventually the correction of hooking effect. The Multi-mission Altimetry Processing Software (MAPS) offers the opportunity to end-user to define their own virtual stations using the altimetry data contained in the Geophysical Data Records (GDRs) for the most recent nadir-looking altimetry misisons (i.e., Envisat, Jason-1, Jason-2, and Saral). Once achieved a coarse selection of the altimetry measurements present in the sudy zone using a kml file defined with Google Earth, altimetry heights are computed using the geophysical and environmental corrections for different type of surface (rivers and small lakes, great lakes, ocean, and coastal areas). A graphical user interface allows to visualize not only the altimetry heights derived from the different retracked ranges available for each mission (e.g., Ocean, Ice-1, Ice-2, Sea Ice, MLE-3) inside the study area, but also any corrections used and other parameters (e.g., backscattering coefficients) that can be useful for selecting the valid altimeter heights. Correction of the hooking effects can be performed on the selected or a part of the selected data at any cycle.  The selection of the valid data can be saved at any time and reloaded. Time-series of altimeter heights are then computed based on the user’s selection of the valid data and can be exported. This software, developed as a collaboration between LEGOS-OMP, GET-OMP, and EPOC-OASU will be son available on the CTOH website: htpp://ctoh.legos.obs-mip.fr.

The proposal “Integrated Sentinel-1 PSI and GNSS technical facilities and procedures for determination of 3D surface deformations caused by environmental processes” was accepted for implementation by the ESA. The main objective of the proposal is to combine Sentinel-1 IW images processed by integrated Persistent Scatter Interferometry (PSI) with additional GNSS observations on those important areas, where the use of traditional PSI methods cannot be used or are limited due to the different vegetation cover and extension. This problem can be handled by properly designed artificial backscatters. If these backscatters can be used for both ascending and descending satellite passes the vertical and east components of the displacements can be estimated.  These may be biased by the unknown north component of the displacement. This limitation can be solved by the integrated epoch measurements carried out by systematic GNSS observations.

In this presentation the primary practical results of the proposal are summarized which are based on the first prototype of trimmed twin corner reflector (TCR), which is placed at the Széchenyi István Geophysical Observatory; and on the result of the processing accomplished with Sentinel-1 Toolbox provided by the ESA. The TCR is placed on 1 square meter reinforced concrete block supported by adapters for GNSS, traditional geodetic and gravimetric measurements called as integrated geodetic/geodynamic benchmark. For the test computations the predicted satellite orbits (included in SLC annotation files) and the GNSS derived coordinates of TCR are used, which are given in WGS-84 coordinate system.

The identification of TCR - oriented to the chosen ascending and descending directions - is carried out using the split, deburst and update geo reference modules of the Sentinel-1 Toolbox. The pixels of TCR can be identified very easily on the VV polarised intensity images. The reflectivities of TCR are 59.9 ±0.6 dB and 29.9 ±5.4 dB for VV and VH polarizations in the ascending cases, respectively, while in the case of descending images they are 60.2 ±0.6 dB and 30.0 ±11.7 dB. The VH polarizations are worse; sometimes the complex numbers cannot be estimated. The characteristics of the identification of TCR on the differential interferograms (using coregistration, interferogram formation, deburst, topographic phase removal and update geo reference modules) are similar.

The geometric stability of SLC images are investigated by the closest approach method using the predicted orbits and the GNSS measured coordinates of TCR. The repeatability of the ascending images is: azimuths ±2 arcsec, incidence angles ±10 arcsec and the differences between the radar derived - given in SLC data - and the measured SAT-TCR distances is -5.7 ±1.3 meters. In the case of descending images: azimuths ±2 arcsec, incidence angles ±8 arcsec and SAT-TCR distance difference is -4.7 ±0.6 meters.

These primary results prove that the prototype of the integrated benchmark and the stability of the Sentinel-1A images can fulfil the requirements for the proposed PSI and GNSS data integration. At the moment the missing images, experienced during the test computations, seems to be major limitation, which hopefully will be mitigated when the system will be declared fully operational.

Global Navigation Satellite System (GNSS) stands for worldwide continuously transmitting constellations of satellites dedicated to navigation and positioning purposes. The L-band transmitted signal traveling through the direct path is captured and processed by a receiver located near the Earth’s surface to provide to the user information about Positioning, Velocity and Timing (PVT). However, the receiver also acquires GNSS signals after a scattering off obstacles or surfaces occurs, giving rise to the multipath errors. Within such a context, GNSS-Reflectometry is based on the exploitation of GNSS signals reflected off Earth’s surface to infer geophysical information of the scattering scene. Firstly conceived for altimetry scope, GNSS-R techniques have been developed for several remote sensing applications. The GNSS signals scattered off the observed scene, called Glistening Zone (GZ), is  mapped at the receiver output into the Delay-Doppler (DD) domain, thus the final GNSS-R data is a Delay-Doppler Maps (DDMs) or a 1-D delay waveform. To strongly demonstrate the importance of this innovative technique, dedicated space mission are planned or already operative, i.e. the first UK-Disaster Monitoring Constellation (UK-DMC), the Uk TechDemoSat-1 (TDS-1), the NASA CYGNSS constellation and the International Space Station (ISS) receiver GEROS. Almost all GNSS-R remote sensing applications have been developed by extracting compact parameters from DDMs or waveforms in order to infer geophysical information of the scattering scene. However, the limit of DD domain is that spatial information of the observed scene are lost representing a drawback for applications within the image remote sensing field. Thus, some techniques have been proposed to deconvolve the DDM to reconstruct a radar image of the observed scene. The procedures present in the literature are applied to various simulated marine scenario including no homogeneous wind speed, oil slicks, non homogeneous elements, cyclones. The obtained results encourage the possibility of exploiting GNSS-R tool for this remote sensing topic. Thus, in this paper a radar image of the reflecting scene is reconstructed by processing real DDMs acquired by the UK TDS-1 satellite. The DDMs used in this work refer to a marine scenario including non homogeneous area, i.e. land or ice, where the signal reflection gives rise to peculiar scattering. The 2-D Truncated Singular Value Decomposition technique is applied to reconstruct a radar image of the observed scene and the obtained results demonstrate the soundness of the proposed approach. 

Inspired from Synthetic Aperture Radar (SAR) Technique, a new nadir radar altimeter concept called “Delay/Doppler altimeter” or “SAR Mode altimeter” was proposed in [Raney, 1998] demonstrating theoretically higher precision and resolution capabilities than what is typically seen with conventional pulse-limited altimeters. This new concept has been readily carried on-board the CryoSat-2 satellite [Wingham, 2006] then implemented on Sentinel-3 [Donlon, 2012]. Its principle is quite similar to the SAR imagery processing. The use of the Doppler effect caused by the satellite displacement in the along-track direction (azimuth) allows the radar to a resolution enhancement in this direction (around 300m for Cryosat-2 and Sentinel-3).

Over ocean surfaces, very few studies have been addressed to the development of SAR-mode processing scheme and to the evaluation of the data quality. Although the resolution enhancement is not necessarily a key driver for ocean applications, the improvement in precision that can be observed by the increased number of independent looks makes Doppler altimetry very attractive on these surfaces [Jensen and Raney, 1998]. The use of a new measurement technique raises also the question of the continuity with conventional altimetry missions as Jason-2 that represents today, the reference mission for all scientists. It is thus essential that the processing of Doppler altimeter measurements do not introduce biases with respect to conventional altimetry.

The chief objective of this paper is to give an overview a dedicated retracking algorithm developed by CNES to extract the geophysical parameters from the Doppler echoes (Cryosat-2 Processing Prototype). Subsequently, we assess the proposed processing strategy over three years of Cryosat-2 open-ocean data, using a large set of analysis tools and diagnosis, and rule the benefits and drawbacks of this new technique compared to the conventional altimetry.

The Sentinel-3 Mission (S-3) is part of the Global Monitoring for Environment and Security (GMES/Copernicus) European initiative. For oceanic applications, the S-3 mission will deliver continuity to existing ESA ERS, Envisat and CryoSat missions. The S-3 mission will provide ocean/land colour data, sea/land surface temperatures and sea surface and land ice topography at least at the performance level of corresponding Envisat instruments, the Medium Resolution Imaging Spectrometer (MERIS), the Advanced Along-Track Scanning Radiometer (AATSR) and the Envisat Radar Altimeter (RA). The topography payload consists in a Doppler altimeter SRAL, a microwave radiometer MWR, and 3 instruments for precise orbit computation purposes (GNSS, DORIS and laser reflector). The Sentinel-3 satellites (A and B) will fly on a new orbit with a 27-days cycle.

The SRAL instrument is a nadir Ku/C altimeter, with two operating modes: the low resolution mode (LRM), used on all past oceanographic missions and the high resolution mode (SAR) used for the first time on CryoSat-2. The SAR mode (also called Delay Doppler) provides a better along track resolution (around 320m) with respect to the LRM mode.

The Sentinel-3A satellite will be launched in December 2015, and CNES (with the support of CLS) will provide a strong support to ESTEC project for the topography calibration and validation during the commissioning phase. In this paper, we will present a first data quality assessment performed on CNES side, focusing on inland water. This will include the coverage analysis of the main inland water bodies, the waveforms centering, and the comparison of the water level to in situ data (with the support of LEGOS team), in both open-loop and closed-loop tracking modes. Moreover, the results will be compared to other missions such as Jason-2/3 or SARAL/AltiKa. This analysis will demonstrate the added value of SAR mode measurements for improving the observation of inland water bodies.

The Sentinel-3 Mission (S-3) is part of the Global Monitoring for Environment and Security (GMES/Copernicus) European initiative. For oceanic applications, the S-3 mission will deliver continuity to existing ESA ERS, Envisat and CryoSat missions. The S-3 mission will provide ocean/land colour data, sea/land surface temperatures and sea surface and land ice topography at least at the performance level of corresponding Envisat instruments, the Medium Resolution Imaging Spectrometer (MERIS), the Advanced Along-Track Scanning Radiometer (AATSR) and the Envisat Radar Altimeter (RA). The topography payload consists in a Doppler altimeter SRAL, a microwave radiometer MWR, and 3 instruments for precise orbit computation purposes (GNSS, DORIS and laser reflector). The Sentinel-3 satellites (A and B) will fly on a new orbit with a 27-days cycle.

The Sentinel-3A satellite will be launched in December 2015, and CNES (with the support of CLS) will provide a strong support to ESTEC project for the topography calibration and validation during the commissioning phase. In this paper, we will present a first data quality assessment performed on CNES side, focusing on sea ice. This will include the coverage over sea ice regions, the waveforms centering, and the assessment of the sea ice height accuracy (with the support of LEGOS team). Moreover, the results will be compared to other missions such as Cryosat-2 or SARAL/AltiKa and possibly to ancillary data (in situ data, optical data, SAR data…). This analysis will demonstrate the added value of SAR mode measurements for improving the observation of sea ice.

The Sentinel-3 Mission (S-3) is part of the Global Monitoring for Environment and Security (GMES/Copernicus) European initiative. For oceanic applications, the S-3 mission will deliver continuity to existing ESA ERS, Envisat and CryoSat missions. The S-3 mission will provide ocean/land colour data, sea/land surface temperatures and sea surface and land ice topography at least at the performance level of corresponding Envisat instruments, the Medium Resolution Imaging Spectrometer (MERIS), the Advanced Along-Track Scanning Radiometer (AATSR) and the Envisat Radar Altimeter (RA). The topography payload consists in a Doppler altimeter SRAL, a microwave radiometer MWR, and 3 instruments for precise orbit computation purposes (GNSS, DORIS and laser reflector). The Sentinel-3 satellites (A and B) will fly on a new orbit with a 27-days cycle.

The SRAL instrument is a nadir Ku/C altimeter, with two operating modes: the low resolution mode (LRM), used on all past oceanographic missions and the high resolution mode (SAR) used for the first time on CryoSat. The SAR mode (also called Delay Doppler) provides a better along track resolution (around 320m) with respect to the LRM mode.

The Sentinel-3A satellite will be launched in December 2015, and CNES (with the support of CLS) will provide a strong support to ESTEC project for the topography calibration and validation during the commissioning phase. In this paper, we will present a first data quality assessment performed on CNES side, focusing on land ice. This will include the tracking performances over steep slopes, the waveforms centering, and the assessment of the ice height accuracy (with the support of LEGOS team). Moreover, the results will be compared to other missions such as Cryosat-2 or SARAL/AltiKa. The analysis performed will enable to have a first assessment of the SAR mode performance over land ice.

The problem of variational data assimilation for a nonlinear evolution model is formulated as an optimal control problem to find the unknown parameters of the model. We study the problem of sensitivity of the optimal solution via variational data assimilation with respect to observation errors. On the basis of relations between the error of the optimal solution and the errors of observational data through the Hessian of the cost functional, the algorithms are developed and justified for calculating the coefficients of sensitivity as the norms of the response operators occurring in the equations for errors. A numerical study of the sensitivity of the optimal solution on the example of the problem of variational data assimilation of sea surface temperature to restore the heat flows for the model of thermodynamics is presented. Numerical examples for data assimilation in the Baltic Sea dynamics model are given. This work was carried out within the SAMOVAR project (CNRS-RAS), Russian Science Foundation project 14-11-00609, and the project 15-01-01583 of the Russian Foundation for the Basic Research.

References 

1. Gejadze  I., Le Dimet F.-X., Shutyaev V.P. On analysis error covariances in variational data assimilation.  SIAM J. Sci. Comput., 2008, v.30, no.4, pp. 1847–1874.

2. Gejadze I., Le Dimet F.-X., Shutyaev V.P. On optimal solution error covariances in variational data assimilation problems. J. Comp. Phys., 2010, v.229, pp. 2159–2178.

3. Gejadze I., Shutyaev V.P., Le Dimet F.-X. Analysis error covariance versus posterior covariance in variational data assimilation.  Q. J. R. Meteorol. Soc., 2013, v.139, pp. 1826-1841.

4. Shutyaev V.P., Parmuzin E.I. The study of solution sensitivity for the variational observation data assimilation problem in the Black Sea dynamics model. Russ. J. Numer. Anal. Math. Modelling, 2013, v.28, no.1, pp. 37–52.

5. Gejadze, I.Yu., Shutyaev, V.P. On gauss-verifiability of optimal solutions in variational data assimilation problems with nonlinear dynamics. J. Comp. Phys., 2015, v.280, pp.439-456.

In the frame of the Copernicus programme (the European initiative for the implementation of information services based on observation data received from Earth Observation (EO) satellites and ground based information), ESA has developed the Sentinel-2 optical imaging mission that will deliver data products designed to feed downstream services related to land monitoring, emergency management and security. As part of the Sentinel-2 Mission Performance Centre (S2MPC) activities, the DIMITRI software is used by ARGANS to perform vicarious validation of the Level-1 product. This validation objectives are to assess the quality of the data product at Level-1, and to monitor the products evolution.

DIMITRI (Database of Imaging Multispectral Instrument and Tool for Radiometric Intercomparison) consists of several vicarious calibration methodologies for EO optical sensors: 

1) Rayleigh scattering methodology, applicable over ocean sites where the molecular scattering is the dominant contributor to the Top-Of-Atmosphere (TOA) signal.

2) Sun-Glint Methodology: applicable over ocean sites where the geometrical configurations are close to the specular sun reflectance

3) PICS (Pseudo-Invariant Calibration Sites) methodology applicable over desert sites, enabling the assessment of the temporal stability of Earth Observation sensors.

4) Intercomparison methodology, which enables radiometric intercomparison of TOA reflectances from a number of satellite sensors over fixed validation sites and long time periods

This presentation will provide a brief description of these methodologies in term of advantages, limitations and possible improvements.

DIMITRI has been used in the validation activities during the S2A Commissioning Phase operations and is now ready for the Routine Operations Phase. Cross-comparison with results of the Commissioning team ensure the continuity of the mission performance. The first results from the MSI instrument obtained by these methodologies will be presented. Following this, the PICS methodology outputs from Sentinel-2/MSI, LANDSAT-8/OLI and Aqua/MODIS are analysed. TOA-reflectance from both sensors has been extracted over the Railroad Valley (RadCaTS) site and compared to concomitant ground-based TOA-reflectance. The comparison results show good agreement when the relative difference is within 5%-10%. The uncertainties over the calibration coefficients from Rayleigh methodology are found to be less than 5% and about 3% for the PICS methodology.

This work has support from ESA in the frame of S2MPC and DIMITRI Evolution projects. We gratefully acknowledge the RadCaTs measurements provided by the NASA Landsat Cal/Val Team in supporting this analysis.

The SnowScat instrument was originally designed as a fully-polarimetric scatterometer
for measurements of the radar cross-section of snow at X-band up to Ku-band over a frequency range of 9.2 to 17.8 GHz [1-3].
In 2015, a SnowScat hardware extension, which allows for high-resolution tomographic profiling of a snow pack (and potentially other volumetric targets of interest), was implemented in the frame of the ESA project "Enhancement of SnowScat for tomographic and vertical profiling observation capabilities" (ESA/ESTEC Contract No. 20716/06/NL/EL CCN3). This extension enhances the SnowScat device in order to better respond to the ESAC recommendations which were made on the deselected CoReH2O candidate following the User Consultation meeting in March 2013 for the 7 Earth Explorer mission (see also similar recent work by various authors [4-7]).

The new tomographic profiling capability of the SnowScat setup is obtained by means of moving the antenna along a rail in elevation direction followed by tomographic aperture aperture synthesis [8-9].
See also Fig.1 in the accompanying pdf document which shows the SnowScat device including the new rail mounted on a scaffolding at the test site in Davos, Switzerland.

Last winter (2015/2016) a demonstration campaign was carried out to (1) verify the measurement concept and (2) to verify the processing concept.
First results were presented in [10-11].
When comparing the vertical tomographic profile of a snowpack (acquired with the SnowScat device and processed tomographically) with snow profiles taken in-situ with the snow micro-pen device
it was found that a stratification of the snow pack with layers of melt-freeze crusts was identified correctly, thus confirming the measurement concept and a first-order processing approach
(see also Fig. 2 in the accompanying pdf document).

Since microwave imaging of a snow volume is subject to refraction, an iterative process is required to solve for the tomographic inversion problem.
The concept for a refined SnowScat tomography processing scheme is based on a two-stage approach:
1. As a first approximation the snow volume is considered to be homogeneous with a known refractive index with know depth of the snow layer.
For this scenario the ray paths are calculated and a time-domain back-projection-based tomographic imaging of the snow layer is performed [8-9].
2.  Based on the reconstructed imagery an autofocus scheme is applied in order to refine the focusing and thereby also refining the knowledge of the actual inhomogeneous structure of the snow layer.

While the previous campaign in winter 2014/2015 could successfully be exploited to demonstrate the measurement concept and the potential of the tomographic profiling method,
a limited snow accumulation of only ca. 60cm as well as limited in-situ data hindered a detailed assessment of the tomographic focusing approach required to account for multi-layer refraction within the snow pack.
Therefore the processing concept has to be validated and refined with a more comprehensive set of SnowScat measurements of a substantial snow pack and accompanying in-situ measurements:
Starting from Nov./Dec 2015 a follow-up measurement campaign with the tomographic profiling mode is conducted in the Bernese Alps in Switzerland.

In our final contribution, we will
(1) assess and discuss a refined tomographic processing scheme that better accounts for the refraction in stratified snow packs and validate
    the processing concept with the time series of tomographic snow profiles and the in-situ snow profile measurements.
(2) present an analysis of an extended time series of tomographic profiles of a snowpack that are acquired
    in the coming winter months at a test site in the Bernese alps in Switzerland. The time-series of SnowScat tomographic snow profiles
    is also accompanied by in-situ profiles gained from snow micro-pen measurements and snow-profiles taken from snow pits.

References:

[1]  A. Wiesmann, T. Strozzi, C. L. Werner, U. Wegmuller, and M. Santoro,
     "Microwave remote sensing of alpine snow," in Proc. IEEE Int. Geosci. Remote Sens. Symp.,
     Jul. 2007, pp. 1223–1227.
    
[2]  C. L. Werner, A. Wiesmann, T. Strozzi, M. Schneebeli, and C. Matzler,
     "The SnowScat ground-based polarimetric scatterometer: Calibration and initial measurements
     from Davos Switzerland," in Proc. IEEE Int. Geosci. Remote Sens. Symp., Jul. 2010, pp. 2363–2366.
    
[3]  A. Wiesmann, C. L. Werner, T. Strozzi, C. Matzler, T. Nagler, H. Rott, M. Schneebeli, and U. Wegmuller,
     "SnowScat, X- to Ku-band scatterometer development," in Proc. ESA Living Planet Symposium, Jun. 2010.
    
[4]  K. Morrison, H. Rott, T. Nagler, H. Rebhan, and P. Wursteisen,
     "The SARALPS-2007 measurement campaign on X- and Ku-band backscatter of snow,"
     in Proc. IEEE Int. Geosci. Remote Sens. Symp., Jul. 2007, pp. 1207–1210.
    
[5]  S. Tebaldini and L. Ferro-Famil,
     "High resolution three-dimensional imaging of a snowpack from ground-based SAR data acquired at X and Ku band,"
     in Proc. IEEE Int. Geosci. Remote Sens. Symp., Jul. 2013, pp. 77–80.
    
[6]  L. Ferro-Famil, S. Tebaldini, M. Davy, and F. Boute,
     "3D SAR imaging of the snowpack at X- and Ku-band: Results from the AlpSAR campaign,"
     in Proc. of EUSAR 2014 - 10th European Conference on Synthetic Aperture Radar,
     Jun. 2014, pp. 1–4.
    
[7]  K. Morrison and J. Bennett,
     "Tomographic profiling - a technique for multi-incidence-angle retrieval of the vertical SAR
     backscattering profiles of biogeophysical targets," IEEE Trans. Geosci. Remote Sens.,
     vol. 52, no. 2, pp. 1350–1355, Feb. 2014.

[8]  O. Frey and E. Meier,
     "3-D time-domain SAR imaging of a forest using airborne multibaseline data at L- and P-bands,"
     IEEE Trans. Geosci. Remote Sens., vol. 49, no. 10, pp. 3660–3664, Oct. 2011.
    
[9]  O. Frey, F. Morsdorf, and E. Meier,
     "Tomographic imaging of a forested area by airborne multi-baseline P-band SAR,"
     Sensors, Special Issue on Synthetic Aperture Radar, vol. 8, no. 9,
     D. Riccio, Ed., pp. 5884–5896, Sep. 2008.

[10] O. Frey, C. L. Werner, M. Schneebeli, A. Macfarlane, and A. Wiesmann,
     "Enhancement of SnowScat for tomographic observation capabilities,"
     in Proc. FRINGE 2015, March 2015.

[11] O. Frey, C. L. Werner, and A. Wiesmann,
     "Tomographic profiling of the structure of a snow pack at X-/Ku-band using SnowScat in SAR mode,"
     In Proc. EuRAD 2015 - 12th European Radar Conference, pp. 1-4, Sep. 2015.
    

In SAR radar altimetry, the development strategy of the on-ground processing aimed at reducing the noise speckle of the measurements while retaining the high along-track resolution of the data. The actual Cryosat-2 and Sentinel-3 operating systems allow today to resolve Doppler bins with non-overlapping segments on the surface enabling to mitigate the contamination between adjacent Doppler bins. In addition, the SAR-mode processing allows a better measurement precision compared to the conventional altimetry thanks to the higher number of statistically independent looks contributing to the measurement average.
The principle behind the SAR-mode multi-looking process is to gather a stack of multiple Doppler beams, steering at the same ground location from different look directions, and then incoherently average these collocated Doppler beams to produce a lot less noisy echo (so called Doppler echo). The speckle noise, which de-correlates from different look directions, is reduced by the averaging operation. It is found however that the speckle noise reduction on the altimeter-derived parameters (range and wave height) is not as high as expected. This effect is mostly explained by the fact that the contributing beams to the final averaged waveform have different mean shapes and different amplitude values related to the looking angle of measurement. The off-nadir beams of lowest amplitudes thus contribute very little to the noise reduction, as well as to the geophysical parameter estimates. As computed theoretically by Amarouche in 2013, the effective number of looks that are ultimately involved in the noise mitigation process is lower than the number of beams and depends on the waveform sample and wave height. In order to better exploit the capabilities of the SAR altimeter compared to those obtained with the actual ground processing, it is thus essential to develop alternative methods allowing a better processing that would take maximum advantage from the Doppler processing. This is of major interest for Sentinel-3 and Sentinel-6 missions, embarking both a SAR altimeter.
New SAR-mode processing schemes have been investigated and analyzed in R&D studies funded by CNES. For each new proposed approach, a Cryosat-2 data set of at least one-month duration has been produced over open ocean and their performances assessed through the use of common validation tools and protocols and in comparison with operational-like data products (generated with CNES processor CPP).
Currently, in the frame of the SCOOP ESA project, CLS is investigating new SAR/Doppler processing solutions that would provide potential improvement of the Cryosat-2 SAR altimetric measurements’ precision over ocean. The proposed approach is an alternative methodology to the well-known antenna pattern compensation or stack beam weighting techniques. It consists in processing individual echo beams then averaging the estimates such that all Doppler beams may equally contribute to the noise reduction, with no beams weighting, and thus improve the SAR-mode performance.
This paper will present the principle of these methods, and their benefits and drawbacks will be discussed. The aim of this paper is also to determine whether the new processing schemes have a potential impact in operational use or not.

The aim of this study is to assess the improvement on the three-dimensional water vapor solutions of the atmosphere as a result of the integration of water vapor maps provided by Synthetic Aperture Radar (SAR) interferometry into the GNSS tomography. We will also assess the improvement resulting from the densification of the GNSS network on the three-dimensional water vapor solution. Water vapor distribution plays a crucial role in the atmospheric dynamics being difficult to quantify. Numerical weather prediction models require precise three-dimensional water vapor measurements in space and time, in order to improve the forecast. Space-borne techniques based on GNSS satellites and SAR sensors have been providing in the past years integrated water vapor measurements (IWV) with a similar precision as common meteorological sensors. GNSS atmospheric processing provides nearly continuous IWV measurements at the zenith of the receivers, while SAR interferometry (InSAR) acquisitions allows to produce high resolution maps of IWV measured along the radar line-of-sight, with a temporal sampling depending on the satellite revisiting time. Despite both techniques having electromagnetic signal proprieties in the microwave wavelength, allowing to sense the local atmosphere in all weather conditions, they lack a vertical quantification of the water vapor content along the atmosphere.
The GNSS water vapor tomography technique allows to retrieve the three-dimensional atmospheric state over a region with a network of receivers. Reconstruction of slant path GNSS delay observations from the precise IWV measurements results in several ray paths in the satellite line of view at each instant, allowing to apply the tomography technique throughout the atmosphere. Usually discretization of the local troposphere is performed dividing the space into voxels, forming a three-dimensional grid, with a horizontal resolution of a few kilometers and a vertical resolution of a few hundreds of meters. Assuming a constant value during a short period and inverting the system of equations relating the observations with the grid model, three-dimensional water vapor maps plus time can be obtained from the GNSS tomography. The main drawback of this technique is the sparse coverage of slant rays on the lower levels, due to the lack of low angle observations provided by the GNSS, implying the need for numerical constraints.
Due to the InSAR differential measurements of the water vapor changes occurred between master and slave image acquisition times, the GNSS tomography processing scheme has to be modified to account for the interferogram temporal baseline. The main advantage of including InSAR to constrain the GNSS tomography is the higher spatial pixel resolution compared with a GNSS network composed by a few tens of stations at best. Therefore, a higher spatial resolution of the water vapor is obtained when including InSAR maps, as verified with application of ENVISAT data (Benevides et al., 2015). A tomography experiment of densification of a GNSS network of 9 stations located in a regional area around Lisbon (Portugal), extended the network to 17 receivers during July 2013. One interferogram was generated from two TerraSAR-X images acquired over the region and during the network densification field campaign, with an 11 day temporal baseline. Radiosonde launches every four hours were performed to validate and compare the various possible water vapor tomographic solutions (permanent network, extended network, InSAR constraint). With the continuous development of the Sentinel-1 mission and other possible future SAR missions, the interferograms could be produced within a shorter temporal interval, allowing its inclusion in weather observations techniques like GNSS tomography, in order to obtain water vapor maps to improve numerical weather prediction models.
This study was funded by the Portuguese Science Foundation FCT, under the grants UID/GEO/50019/2013 and SFRH/BD/80288/2011.
Benevides et al., 2015, DOI: 10.1109/TGRS.2015.2463263.

The sea level determination in the Arctic brings challenges with it, due to its high seasonality,
varying surface types and scarce in-situ observations. This has an effect on the performance of
sea surface height determination, ice presence and ice sheet thickness measurements. With its
polar orbit, CryoSat-2 provides frequent observations of this regions while experiencing different
terrains and seasons which is crucial to improve the sea level accuracy.
Since CryoSat-2 is the first satellite to carry a SAR altimeter, it serves as a stepping stone
to improve the performance of SAR observations. Due to the different characteristics of the
returned waveform of SAR measurements compared to pulse-limited waveforms, retrackers that
were used for conventional altimeters cannot be applied. Therefore, the data of CryoSat-2 builds
the foundation to develop and improve SAR retrackers.
Using SAR observations, the sea surface height is currently determined by either empirical
retrackers, such as the primary peak retrackers or physical retrackers like SAMOSA. It should
be noted here that SAMOSA only retracks lead and open ocean waveforms. The primary peak
retrackers on the other hand, are able of retracking also irregular waveforms and therefore
provide a global sea surface determination.
To provide accurate and precise determination of the Arctic surface, a retracker system needs
to be developed that is able to cope with the different terrains and seasons. This goal is reached
by analyzing existing retrackers with respect to their performance and combining their positive
behavior into one retracker system.
The performance of retrackers is evaluated based on their accuracy and precision. Four retrackers
have been evaluated: SAMOSA3, primary peak center of gravity, primary peak threshold and
CryoSat/ESA retracker. During four years of CryoSat-2 SAR observations (2011-2014), the
performance of each retracker is assessed on a monthly basis for given waveform classes (leads,
open ocean and sea-ice). The retrackers with the highest performance on a monthly basis are
then combined within one retracking system.
To remove the bias when switching from one retracker to another, three bias removal strategies
were developed. The first strategy removes a constant offset from one retracker in a given
waveform class to eliminate the relative offset. Despite a constant offset removal in this simplistic
approach, an error is expected to remain due to the different conditions over time.
Another method uses the linear relationship between retracker offset and significant wave height.
It should be noted that the coefficients of this linear relationship vary for different retrackers,
meaning that the coefficients are determined for every measurement condition separately. The
third bias removal approach makes use of a neural network. By training a multilayer neural
network, it will be able to identify irregular waveforms and compare two waveforms of different
retracker with each other. As an output it will provide the offset between two retrackers.
The development of a year round retracker with an implementation of a proper bias removal
strategy will improve the Arctic sea level determination as it is able to cope with varying local
conditions. By analyzing the seasonal performance of retrackers, an insight into the behaviour
of empirical and physical retrackers is gained that can be used for future improvements in the
retracker design for regions with a high seasonality and fast changing surface types.

Along with the navigation technology getting mature, the Global Navigation Satellite System Reflectometry (GNSS-R) concept was put forward. The GNSS-R system is to use the reflected signals received by Low Earth Orbit (LEO) Satellites from Global Navigation Satellite System such as GPS or Galileo to measure the characteristics of the reflection points. Although the power of GNSS reflected signal is weak, this signal source can still be detected and used for altimetry purposes. Water, ice, and snow covered areas have high reflectivity for GNSS signals. The signal hit the ground or ocean surface, and part of it reflected back from the reflecting surface. By measuring the relative delay between the direct signal which received by LEO satellite from GNSS and the reflected signal, the height information of the reflection point can be obtained. This measurement can be used in geoscience to get the quantities describing sea states, such as sea height, roughness and wind speed. Since the GPS coverage is dense and rapid, many measurements can be obtained, providing global sea-surface mapping in times shorter than the traditional altimetry repeat cycles. Also, sea surface heights can be obtained not only along the LEO altimetry ground tracks but also at many (off-nadir) points in between, thus allowing monitoring of phenomena with small spatial- and time-scales (e.g., moving waves and mesoscale eddies). [1] Sensing GNSS signals reflected by the oceans surface from LEO for altimetry or scatterometry applications is meant to be a cheap and powerful approach of obtaining ocean height and roughness with very dense and rapid coverage.

Once the technology is proved to be operational, it will contribute a lot to geosciences. In this work, the geometry of the reflection system and error propagation process are studied to give a better understanding of the topic. The main part of this work is to model and simulate the process of GNSS-R system. The input GNSS data is GNSS orbit, and take GOCE satellite as the receiver to receive the reflected signals. This thesis presents the algorithms of reflection point calculation and analysis of error source effects. The coverage of reflection points getting from different GNSS constellation is also analyzed. The main error sources discussed in this paper includes tropospheric effect, ionosphere effect, and Sagnac effect. First use iteration method to simulate the observation range data. Then add in noise and recalculate the reflection points by calculating minimum distance between two ellipsoids. Using this method, the overall height accuracy is predicted to be 0.5934 meters with white noise being considered.

The temporal and spatial variation of aerosols and its radiative effects is one of the most important uncertainties in the determination of the surface reflectance from remotely sensed data. In Proba-V data the atmospheric (including for aerosol) correction is applied in line with the methods defined for SPOT-VGT, in order to ensure data continuity between the two missions.

In this paper we focus on the validation methods and first results of the retrieved aerosol optical depth and the 1km Proba-V surface reflectances based on the currently applied method wherein aerosol quantities are estimated using an optimization algorithm applied to the blue spectral band (Maisongrande et al. 2000). This method is similar to the so-called dark target technique (Vermote et al. 1997), but instead of using a constant SWIR/Blue relationship for dark targets, the adapted method uses a Normalized Difference Vegetation Index (NDVI) dependent SWIR/blue relationship.

For the validation of the surface reflectance and aerosol OD, one year (2014) of PROBA-V data centered on a series of well-documented targets (more than 50), such as AERONET stations with different types of landcover, will be retrieved and compared to data from AERONET and other satellite missions. 

Additionally, within the recently started PV-LAC project, we want to investigate the added value of another joint surface-aerosol retrieval algorithm to improve both land surface reflectance characterization derived from Proba-V observations. The suggested method will take advantage of previous work (Govaerts et al. 2010) and assess the possibility to apply such approach to Proba-V. The pratical implementation of this method is based on the inversion of physically-based radiative models accounting for radiative coupling, using Optimal Estimation. The method primarily targets to fulfill the requirements of Essential Climate Variables (ECV) generation, which is the most interesting for the processing of the entire SPOT-VGT/Proba-V time series. Given that the characteristics of the Proba-V data are very similar to those of SPOT-VGT, the same method could be applied to that data set in the future.

Two marine altimeter calibration campaigns in the area of Ibiza island have been made in June 2003 and September 2013. The first focused on Jason-1 satellite (Marine Geodesy December 2004) and the second on Jason-2 and Saral satellites (Marine Geodesy in press to appear on January 2016). GPS buoys and a catamaran have been used. A levelling traverse was conducted in the Ibiza Marina de Botafoch harbour in June and September 2013 to link the radar MIROS tide gauge and the GPS permanent station with the levelling network in the area in order to confirm the results of June 2011 from the Spanish Instituto Geografico Nacional IGN and to detect the possibility of settlements.

The technical geodetic infrastructure of Ibiza site is presented. The main objective is the preparation of a new marine calibration campaign in the Ibiza area for the new missions to be launched in the next months, Sentinel-3A of the European Space Agency ESA and Jason-3 of EUMETSAT/CNES/NOAA/NASA. It is expected to make direct altimeter calibration measurements along the satellites tracks and a cross calibration together with an expected methodologies for the processing of the altimeter, GPS and tide gauge data.

The Copernicus POD (Precise Orbit Determination) Service is part of the Copernicus PDGS Ground Segment of the Sentinel missions. A GMV-led consortium is operating the Copernicus POD Service being in charge of generating precise orbital products and auxiliary data files for their use as part of the processing chains of the respective Sentinel PDGS.

The POD core of the CPOD Service is NAPEOS (Navigation Package for Earth Orbiting Satellites) the leading ESA/ESOC software for precise orbit determination, in whose development GMV has participated along the last 15 years. The careful selection of models and inputs is important to achieve the different but very demanding requirements in terms of orbital accuracy and timeliness for the Sentinel -1, -2 & -3 missions. The three missions require orbital products in Near Real Time (NRT), with latencies as low as 30 minutes, in Short Time Critical (STC) , with latencies of 1.5 days and in Non-time Critical (NTC) with latencies of 20-30 days. The accuracy requirements are very challenging, targeting 5 cm in 3D for Sentinel-1 and 2-3 cm in radial direction for Sentinel-3. Although the characteristics and the requirements are different for the three missions the same core POD setup is used to the largest extent possible. This strategy facilitates maintenance of the complex system of the CPOD Service.

Sentinel-1A has been launched on 10 April 2014 being operational since October 2014. Sentinel-2A has been launched on 23 June 2015 planned to be set operational mid October 2015. Sentinel-3A is expected to be launched on 10 December 2015. The launch of Sentinel-1B is planned in spring 2016. Thus the CPOD Service will be operating three to four satellites simultaneously in spring 2016.

The Copernicus POD Quality Working Group (QWG) is supporting the CPOD Service by providing independent orbit solutions. These solutions are used to validate the quality of the orbits. The cross-comparison of orbit solutions from different institutions is essential to improve the orbit accuracy because for Sentinel-1 and -2 this is the only possibility to externally assess the quality of the orbits. The two missions are equipped with GPS receivers for POD. Sentinel-3 does carry instruments for GPS as well as for DORIS tracking. Additionally the satellites are equipped with a Laser Retro Reflector for Satellite Laser Ranging. The three techniques allow therefore for an independent validation but also for combination. The comparisons to orbit solutions from the QWG is needed for Sentinel-3 as well because of the stringent accuracy requirements.  

This paper presents the status of the CPOD Service in terms of operations and orbital accuracy achieved for the different orbit products of the different missions. Results from the main as well as from the redundant GPS receivers are shown. Improvements and updates in the processing schemes of the three missions are reported as well as the work and the results of the Copernicus POD QWG.

The Copernicus POD (Precise Orbit Determination) Service is part of the Copernicus PDGS Ground Segment of the Sentinel missions. A GMV-led consortium is operating the Copernicus POD Service being in charge of generating precise orbital products and auxiliary data files for their use as part of the processing chains of the respective Sentinel PDGS.

Sentinel-3 is the third mission of the Copernicus program. The orbit accuracy requirement is very stringent with 2-3 cm in radial direction because of the radar altimetry observations. In addition to the SAR radar altimeter the Sentinel-3 satellites carry as main payloads an Ocean and Land Colour Instrument (OLCI), a Microwave Radiometer, and a Sea and Land Surface Temperature Radiometer (SLSTR). The Sentinel-3 mission thus assures continuity of ERS, ENVISAT and SPOT vegetation data.

The POD instruments are a GPS receiver, a DORIS receiver and a Laser Retro Reflector for Satellite Laser Ranging (SLR) to the satellite. On the one hand, the three different techniques GPS, SLR and DORIS make POD more complex but, on the other hand, it is very helpful to have independent techniques available for validation of the orbit results. The three techniques are, therefore, equally important to fulfil the orbit accuracy requirements. The Copernicus POD Service will process GPS and SLR data routinely and has the capacity to process DORIS in Non-time Critical (NTC) and reprocessing campaigns; combining the three techniques has been proven very useful to remove biases between the techniques.

Three different orbit products are provided for Sentinel-3. A Near Real Time (NRT) product (latency: 30 minutes; radial orbit accuracy requirement: 8-10 cm), a Short Time Critical (STC) product (1.5 days; 3-4 cm) and an NTC product (28 days, 2-3 cm). The NRT processing has been developed by the CPOD Service but it is running externally at the PDGS Marine Centre (EUMETSAT) and at the Core Ground Station (Svalbard). The STC and NTC processing is running at GMV. The orbit predictions needed for successful SLR tracking are provided by the CPOD Service as well.

The first satellite of the mission Sentinel-3A is expected to be launched on 10 December 2015. The POD processing will be fine-tuned during the five months of commissioning phase. The accumulated experience from the already operational satellites Sentinel-1A and Sentinel-2A and a sophisticated testing before the launch of Sentinel-3A will facilitate the procedures.

This paper presents the status of the POD processes for Sentinel-3A at the CPOD Service. First results from the commissioning phase are shown and comparisons against operational Sentinel-3A orbits provided by CNES (Centre National d´Études Spatiales) are presented. The performance of the three POD instruments is assessed and first quality checks of the different orbital products are done as well.

In order to foster a common policy towards harmonized EO data, procedures and descriptors a common dialogue among stakeholders is key for the Copernicus community.  To date a number of initiatives are paving a way towards this goal. Foremost at European level, progress towards geo-information sharing and integration is led by the guidelines contained within the INSPIRE Directive of the European Union.  

In view of this the Copernicus Coordinated data Quality Control (CQC) service, as part of the Copernicus Space Component Data Access system (CSCDA), has visibility of the whole range of data products provided by the contributing satellite missions to Copernicus users. Through this unique perspective, it has been recognised that there is variability across image providers of orthorectified data. This generates inconsistencies and may cause confusion among users of EO data when working with, and comparing, multiple datasets.

CQC is in a unique position to initiate a dialogue among Copernicus stakeholders  in pursuit of  a common agreement regarding the harmonisation of different ortho-products, and to bring the resulting proposals to the broader INSPIRE platform. This process would result in updated INSPIRE orthoimagery guidelines which will be published, assimilated and enforced. The CQC harmonisation process should require INSPIRE to urge the image providers to issue informative and harmonized metadata of their products: metadata should contain description, rationale and details on processors used, data sources and other processing information deemed relevant for the assessment of data quality by the user. In particular for orthoimagery, the following details should be entered in the metadata: the description of DEMs used for orthoimagery and the positional geo-location accuracy reported in the form of RMS error. 

In addition, through the effort of harmonisation in the Copernicus community, the creation of a new metadata file, INSPIRE.xml, delivered within Copernicus Product Packages is one of the goals already being achieved and is illustrated in this paper. This new metadata is INSPIRE compliant, consistent and harmonized, thus ensuring interoperability across all Copernicus EO products.

UNIFIED GNSS-R SCATTERING MODEL: A STRATOSPHERIC FLIGHT EXPERIMENT DEMONSTRATION

H. CARRENO-LUENGO, and A. CAMPS

Universitat Politècnica de Catalunya – BarcelonaTech Remote Sensing Laboratory and Institut d’Estudies Espacials de Catalunya /UPC, UPC Campus Nord, D3; 08034 Barcelona, Spain  

Tel: +34 93 4017362 Fax: +34 93 4017232

Email: hugo.carreno@tsc.upc.edu; camps@tsc.upc.edu   

Remote Sensing using Global Navigation Satellite Systems (GNSS) signals reflected over the Earth’s surface is a promising new technique. Originally proposed to improve the temporal resolution of mesoscale altimetry, it is being also applied for: Wind speed measurements, ice altimetry and soil moisture determination. At present, NASA’s GyGNSS constellation, ESA’s GNSS rEflectometry, Radio Occultation and Scatterometry experiment on-board the International Space Station (GEROS-ISS), ESA’s PAssive Reflectometry and Interferometry System In-Orbit Demonstrator (PARIS-IoD), and UPC’s ³Cat-2 include as scientific objective the application of GNSS-Reflectometry for scatterometry and altimetry applications. This work presents the results of the TOpography from Reflectometric Measurements: an Experiment from Stratosphere (TORMES) project in the Balloon EXperiments for University Students BEXUS 17 and 19 stratospheric flights using the P(Y) and C/A ReflecOmeter (PYCARO), ³Cat-2 payload. The forward scattering at dual-polarization Left Hand Circular Polarization (LHCP) and Right Hand Circular Polarization (RHCP) using GNSS-R bistatic reflections over boreal forests is studied using dual-frequency GPS (L1, L2), Galileo (E1), and GLONASS (L1, L2) signals from the stratosphere.

The experimental set-up was composed of the PYCARO instrument, a zenith-looking omnidirectional dual-band (L1, L2) and dual-polarization (RHCP, LHCP) antenna patch to collect de direct GNSS signals, a nadir-looking dual-band (L1, L2), and dual-polarization (RCHP, LHCP) 6-elements antenna array to collect the Earth-reflected signals, an On-Board Data Handling (OBDH) subsystem for the experiment management, and an active thermal control. The total gain of the nadir-looking antenna was 13 dB at L1, and 12 dB al L2. The OBDH subsystem was composed of a Commercial Off The Shelf (COTS) microcontroller for housekeeping and scientific data management, communications with the ground station, and data storage. The collected data were registered in two internal SD memories (PYCARO and microcontroller), and simultaneously sent to the ground segment via the E-Link system.

The signatures over boreal forests are used to retrieve information on the correlation process. A second added value is on the use of polarimetric measurements at different elevation angles over forests and lakes so that a full evaluation of the dielectric properties is provided. These data are evaluated against theoretical models using a modified version of the EMISVEG simulator developed in support of ESA/SMOS emissivity studies. The results obtained from this 27,000 m height platform have been evaluated with the final objective to optimize the PYCARO payload parameters for the future ³Cat-2 mission.  

ABSTRACT THEME: GNSS-R, BOREAL FORESTS, NANO-SATELLITE

PRESENTATION REQUEST: ORAL

KEYWORDS: GNSS EARTH OBSERVATION

The detection of clouds and cirrus in multi-spectral satellite imagery is a crucial step in many processing work flows. We present various detection schemes based purely on radiative transfer modeling or manually classified training data and compare their results, complexity, and computational efficiency. All discussed detection schemes are based on state of the art machine learning algorithms and measurements from a single pixel without including any external information such as meteorological or climatological data.

They are validated against a new database of manually classified data from Landsat-7 and 8 as well as Sentinel-2A. Corresponding scenes are distributed throughout the globe and seasons to create a representative dataset. The manual classification was performed by a human expert who relied on imagery created from band math as well as spatially and contextually useful information. The manual classification exercise is based on classes for optically thick and thin clouds and cirrus as well as classes for water and snow/ice.

Our results can be used to assess the additional cost and efforts from more complex detection schemes which additionally use external data and might increase the classification skill at the cost of increased complexity and the need for more computational resources.

One of the most simple case of a detection scheme is a decision tree with only limited depth of not more than four layers. We use this type of classifier as a baseline to compare results from more advanced methods which are in general less easy to decode and are basically used as “black boxes”.

The more one treats this problem as a classical machine learning problem, the more one disconnects from the underlying physical processes and the behavior of various approaches becomes more difficult to grasp while the overall detection skill might, or might not increase substantially. We use some selected classifiers to discuss such possible trade-offs in detail.

Our analysis is based on well known methods such as Decision Trees, RandomForests, Support Vector Machines, as well as linear models learned using stochastic gradient descent with various loss functions. Next to these established methods, we employ our own implementation of a classical Bayesian classification method. We discuss possible additional benefits of an ensemble method like adaptive boosting if a classification method is able to predict not only the class for a certain measurement, but also the probability for that measurement of belonging to each considered class.

We present a specific decision tree for each considered sensor with best classification skill for the presented dataset. Each tree has only few layers, is computationally fast, and it is in general easy to include them into any processing system.

Sediments play an important role in biogeochemical cycling in the aquatic environment since they are responsible for transporting a significant proportion of nutrients and contaminants and also mediating their uptake, storage, release and transfer between environmental compartments. The study of riverine suspended sediments is becoming more important as the need to assess fluxes of  nutrients and contaminants to lakes and oceans increases. In this regards, measurements of ocean color from earth-orbiting satellite sensors have demonstrated to be the most suitable tool available to monitor the transport of key biogeochemical substances such as suspended sediments in river plumes. In theory, algorithms to estimate suspended particulate matter concentration (SPM) from optical remote sensing need to be calibrated locally because the relationship between the mass of aquatic particles and their light scattering/absorption properties depends on the type and size of particles. However, recent results suggest that a single global algorithm can be used for remote sensing of turbidity (T) in different estuarine and coastal waters (Dogliotti et al. 2015). It was also suggested that SPM, the parameter of main interest in sediment transport studies, could be then retrieved by ocean color remote sensing if a region-specific relation between T and SPM is known.

In this context, a new network (TURBINET) is proposed to establish long-term collaboration and data-sharing with the aim of generating an single multi-site dataset to validate satellite ocean colour turbidity products. In the present work we present the first results obtained in the frame of TURBINET project within which joint measurements campaigns have been performed in rivers and inland waters of Belgium and Argentina. Activities performed included harmonization of measurement protocols, setting of quality control criteria, and intercomparison and intercalibration of different instruments for measuring turbidity. The generality of the turbidity algorithm was tested and its application to high resolution satellite images like Sentinel-2, Landsat 8 and Pléiades was evaluated. Long-term expansion of this initial two-partner network is envisaged and encouraged in order to provide ground-based validation data from a wide range of turbidity values and different regions of the world. The potential value of such a network is high since a single measurement from a relatively low cost instrument can be used to validate every cloud-free image of that location from any optical mission (Sentinel-2AB, Sentinel-3AB/OLCI, Landsat-8, VIIRS, MODIS-Aqua, MODIS-Terra, PACE, SABIA/Mar, etc.).

Dogliotti, A.I., K.G. Ruddick, B. Nechad, D. Doxaran, and E. Knaeps. 2015. “A Single Algorithm to Retrieve Turbidity from Remotely-Sensed Data in All Coastal and Estuarine Waters.” Remote Sensing of Environment 156 (0): 157–68. doi:10.1016/j.rse.2014.09.020.

In-water suspended particulate matter concentrations (SPM) is of interest in various oceanographic research fields (e.g. underwater light attenuation and visibility, sediment transport modelling) and can be mapped from satellites. Remote sensing algorithms have been developed to retrieve SPM from marine reflectance (Rrs). However, these algorithms show large uncertainties when applied to waters with specific inherent optical properties (SIOPs) that deviate from those assumed in the bio-optical models. On the other hand, water turbidity (T, a measure of particles side-scattering) has been demonstrated to be not only a good proxy to SPM (Boss et al., 2009), but also to be highly correlated to marine reflectance in the red and near infrared spectral range, e.g. (Nechad et al., 2009). Moreover, (Dogliotti et al. 2015) found that the relationship between T and Rrs is only weakly sensitive to the natural variability of SIOPs and to the particulate scattering phase function.

In the present paper, algorithms for the estimation of turbidity from marine reflectance (Rrs-T) measured by any ocean colour hyperspectral sensor, and at Sentinel-2, Landsat 8 and Pléiades spectral bands are calibrated and validated using a large dataset of in situ measurements of T and Rrs, collected in the southern North Sea, the Ligurian Sea, French Guyana waters, the Rhône, Loire, Gironde, Scheldt and Río de la Plata Rivers, from 2007 to 2015. Next, the relationship between particulate backscattering coefficient (bbp) and side-scattering is investigated using simulated datasets based on Fournier-Forand phase functions, assuming variable backscattering ratio and variable algal and non-algal particles concentrations. Two algorithms for the estimation of bbp at 850 nm and SPM from turbidity are proposed (T-bbp and T-SPM, respectively) which are calibrated and validated using T, bbp and SPM measurements collected in the southern North Sea, the Gironde, Río de la Plata and Scheldt Rivers. Finally, the three algorithms (Rrs-T, T-bbp, T-SPM) are applied to Landsat and Pléiades imagery yielding maps of T, bbp (850 nm) and SPM over the southern North Sea, the Scheldt, Gironde and Río de la Plata Rivers.

References

Dogliotti, A.I., K.G. Ruddick, B. Nechad, D. Doxaran, and E. Knaeps. 2015. “A Single Algorithm to Retrieve Turbidity from Remotely-Sensed Data in All Coastal and Estuarine Waters.” Remote Sensing of Environment 156 (0): 157–68. doi:10.1016/j.rse.2014.09.020.

Boss, E.,  L.Taylor, S. Gilbert, K.Gundersen, N.Hawley, C. Janzen, T. Johengen, H.Purcell, C. Robertson, D. WH Schar, G. J. Smith, M. N. Tamburri. 2009. “Comparison of inherent optical properties as a surrogate for particulate matter concentration in coastal waters.” Vol.7 (11), 803-810.

Nechad, B., K.G. Ruddick, and G. Neukermans. 2009. “Calibration and Validation of a Generic Multisensor Algorithm for Mapping of Turbidity in Coastal Waters.” In SPIE European International Symposium on Remote Sensing. Vol. 7473.

Reflectance anisotropy, caused by viewing and illumination geometry, is a commonly known source of error in remote sensing data that needs to be corrected for. On the contrary, knowledge of reflectance anisotropy can be used as an additional source of information. Multi-angular reflectance measurements can provide an insight in the anisotropic reflectance behavior of a target surface. Laboratory and field goniometers can provide this type of data, however, goniometers are often cumbersome and impractical in their use and in addition, most goniometer setups provide point-based measurements and thus do not provide any spatial explicit information on reflectance anisotropy. For this study, we used a method to capture anisotropic reflectance effects using a unmanned aerial vehicle (UAV). By flying with a snapshot camera (Rikola) mounted on a UAV we were able to extract multi-angular reflectance data based on the position of the UAV relative to the location of the pixels in the individually recorded images during the flight. The UAV flight-lines were programmed to create a lot of overlap in the images, such that pixels were recorded from multiple different positions. In this way, we collected a multi-angular reflectance dataset of a potato field, in which several experimental plots received varying levels of fertilization. The multi-angular reflectance measurements were taken at 16 different bands in the 500 – 900 nm range that partially match the bands of Sentinel-2. In addition to the reflectance measurements, vegetation parameters such as leaf area index (LAI), leaf chlorophyll and leaf water content were collected to provide an insight in the state of the potato plants in the individual plots. We will explore the additional value of the acquired anisotropy information for the extraction of vegetation parameters of the potato plants based on reflectance data. The data is currently being processed.

Although conventional radar altimetry products (Jason1, Jason2, LRM Cryosat-2, etc) have a spatial resolution between 5 and 10 km (due to the size of the radar footprint), the observation of ocean scales in the along track direction smaller than100 kmis limited by the existence of a “spectral bump”, i.e. a geographically coherent error between 10 and 20 km, observed on all conventional altimeters.

Dibarboure et al (2014) have largely discussed the bump artefact which appears to be a mixed effect between the Brown model used for retracking the data (which has been designed for a homogeneous scene and is not fully relevant during backscattering events) and the criteria for selecting the 20 Hz valid data. They showed that dedicated processing could better discard outliers and reduce the energy in the mean PSD of the sea level. More recently, dedicated editing procedure based on the mispointing angle on AltiKa (Poisson et al OSTST 2014) has also enhanced the description of the corrupted data. In parallel, other groups are also working on retracking as Amarouche and Sandwell, also showing some significant improvement.

Another important issue for understanding the oceanic turbulence at short scales is the use of the new delay Doppler technique (SARM) that should yield measurement less altered by the heterogeneities issue since the thin stripe-shaped synthetic footprint of Delay Doppler mode is reduced to 300 m in the along track direction. Preliminary results have shown that the SARM data derived from Cryosat-2 do not show the spectral bump.

As the delay Doppler processing have become more mature in the last couple of years, the content of the shortest spatial scales is further investigated here by analyzing several oceanic regime acquired in SAR mode with Cryosat-2 mission and comparing the results with other LRM processing (AltiKa, Jason-2 and Cryosat-2).

This work allows understanding the outcomes of the SARM processing that we could expect from Sentinel-3 mission. Furthermore, in the frame of the future altimetry missions (SAR for Cryosat -2 and Sentinel-3 missions and interferometry for the SWOT mission), it becomes crucial to investigate again and to better understand the signals obtained at small scales by conventional altimeter missions.

The series of Sentinel satellites mark a major step forward in the collection of Earth Observation data with the commitment to a series of spacecraft and sensors to construct long time series of data suitable for both climate applications and widespread operational use. Each Sentinel mission is based on a constellation of two satellites to fulfil revisit and coverage requirements, providing robust datasets for Copernicus Services.

Sentinel-3 is a multi-instrument mission to measure sea-surface topography, sea- and land-surface temperature, ocean colour and land colour with high-end accuracy and reliability. The mission will support ocean forecasting systems, as well as environmental and climate monitoring.

ESA and EUMETSAT have defined the Mission Performance Framework for qualifying the performance of the Sentinel-3 mission, sensors and products. One important piece of this component is the Mission Performance Centre, which is in charge of the performance of the Optical Mission and of the Surface Topography Mission (STM).

The Sentinel-3 Mission Performance Centre (S3-MPC) has been charged with different main activities for the STM:

Quality Control activities

The calibration, characterisation and performance of the altimeter (SRAL) and the microwave radiometer (MWR) sensors

Validation of the products and ground processing

Assessment of the mission performance

Support for the continuous improvement of the S-3 STM performance

The present paper will present the different results on the STM Sentinel-3 Level 2 product quality and on the mission performance. This assessment is done in synergy with the ESTEC/CNES team who is in charge of the assessment of the sensors performance during the commissioning phase (phase E1).

A focus will be done on the data quality obtained over ocean, considering the whole payload (SRAL, MWR, DORIS, GNSS). We will demonstrate how Sentinel-3 insures the continuity of the ERS-ENVISAT data records and improves the capability of the altimeter observation system thanks to the new SAR mode. Different metrics will be presented for the assessment of the LRM and SAR performances offered by the SRAL altimeter.

In the polar regions, the sea ice and lake ice is typically covered with snow, which plays a major role in development of the underlying ice cover. Currently, snow thickness on sea ice is operationally estimated using microwave radiometer data. Many studies have shown that the retrieved thickness is accurate only over smooth first-year ice. Snow thickness products based on microwave radiometers also suffer from inherent poor resolution of the sensors.

SAR imagery, on the other hand, has the potential of providing the high spatial resolution needed for sea and lake ice applications. Few case studies on C-band SAR based snow thickness estimation have been conducted on sea ice. Lately also ground-based scatterometers at X-band and Ku-band have been used to investigate relationship between backscatter and snow thickness on lake ice. For sea ice at C-band, the results indicate increase in volume or mixed scattering components with increasing snow thickness. For lake ice, the Ku-band is found to be sensitive to the snowpack, whereas X-band is largely insensitive to snow characteristics.

In this study we concentrate on assessing X- and Ku-band capability to estimate snow thickness on ice by using co- and cross -polarized X- and Ku-band SAR backscattering data acquired with the ESA airborne SnowSAR sensor. The SAR data acquisition and co-incident in-situ measurements were conducted in Finland in 2012 within Phase A studies of the ESA CoReH2O (Cold Region Hydrology High Resolution Observatory) mission, a candidate for the ESA Earth Explorer program. A specific study, ESA SILIRIS (Sea Ice, Lake Ice and River Ice Study), was conducted focused on development of sea ice and lake and river ice retrieval algorithms for CoReH2O. The sea ice test area was on landfast ice in the Bay of Bothnia of the Baltic Sea, while lake ice measurements were conducted over Lake Orajärvi in Northern Finland. The ground resolution of the SnowSAR imagery is 2 m. In-situ measurements included ice and snow thickness, snow density, snow water equivalent, temperature and grain size.

Although the direct correlation between backscatter and snow thickness was found to be rather weak, we show that from different frequency and polarization combinations Ku-band VV-polarization (KuVV) and the KuVV/XVV-ratio has the best sensitivity to snow thickness on sea ice. The KuVV backscattering coefficient (sigma0) increases with increasing snow thickness, likely due to increase of snow volume scattering as a combination of snow grain growth and increase in snow mass.

For lake ice, the best sensitivity was found from XVH and KuVH/XVH-ratio. The different results for lake and sea ice might be explained by the difference in dataset sizes, or the snow cover of lake ice being more complex in terms of stratigraphy and snow properties at the ice-snow interface.

The backscattering values for lake ice are shown to be higher than the values for sea ice. It indicates that there is more volume scattering from the snow cover of lake ice due to more complex stratification in the snow cover.

In summary, the relationship between snow thickness on sea and lake ice, and X-band and Ku-band SAR backscattering signatures was rather weak in our dataset. For sea ice, this is partly due to the small variation of snow thickness at the site, which prevented a comprehensive analysis. For lake ice, we assume that poor correlation is mainly due to complex relationships between snow and ice media characteristics and SAR backscatter data. Further studies with larger datasets, more detailed in-situ measurements, and theoretical backscattering models are required to see if snow thickness estimation using SAR data is possible with reasonable accuracy.

In SAR altimetry, three different products are generated before the computation of the range. The L1A contains L0 complex echoes that have been sorted and calibrated. The L1B-S contains geo-located, calibrated azimuth formed complex echoes related to a given location on the ground, after slant range correction (the so called ‘stack’). L1B contains geo-located, calibrated, azimuth formed, slant range corrected and averaged together (i.e multi-looked/stacked) power echoes associated with a fixed point on the ground-track.

In the L1B-S, the samples within the stack containing unwanted information (clutter) such as Doppler ambiguity, land contamination, aliasing, etc., can be removed in order to have cleaner L1B waveforms.

The different options of removing samples within the stack can be classified as a function of the dimension they are applied: the azimuth (or along-track) direction (i.e. selection of a given number of central beams) and the range direction.

However, these corrections may be applied in both directions at the same time. Examples of this are land contamination removal or Doppler ambiguities removal. All the beams containing undesired phenomena can be removed, but we have the option of deleting only the range samples affected by these contamination or ambiguities for each beam of the stack.

All these options can be integrated together in one single mask for each stack. This mask can be built by the user by defining the range bins of the Doppler beams to be removed, either theoretically or by means of an array; by drawing a polygon over the 3D Stack plot and/or by reading a predefined mask from a user file.

This way of cleaning the stack has already been proven to be very powerful, with Sentinel-6 simulated data, significantly improving the performance over ocean. Moreover, it has also been used to clean stacks of CryoSat-2 data in difficult geometric circumstances like coastal areas or lakes.

This algorithm and others will be freely available within the processor developed within the SEOM Study 1 SARAE (beta version planned to be delivered in September 2016).

Some examples are described in this poster, showing the main advantages of this powerful option for removing undesired but expected interferences.

Tandem-L is a proposal for a highly innovative L-band SAR satellite mission for the global observation of dynamic processes on the Earth’s surface with hitherto unparalleled quality and resolution. The Tandem-L mission concept is currently undergoing a phase A study and is based on the use of two SAR satellites operating in L-band (23.6 cm wavelength) with variable formation flight configurations and is distinguished by its high degree of innovation with respect to the methodology and technology [1]-[4]. Examples are the polarimetric SAR interferometry for measuring forest height, multi-pass coherence tomography for determining the vertical structure of vegetation and ice, the utilization of the latest digital beamforming techniques in combination with a large deployable reflector for increasing the swath width and imaging resolution, as well as the formation flying of two cooperative radar satellites with adjustable spacing for single-pass interferometry.

Main mission goals are the global measurement of 3-D forest structure and biomass for a better understanding of ecosystem dynamics and the carbon cycle, systematic recording of deformations of the Earth’s surface with millimeter accuracy for earthquake research and risk analysis, quantification of glacier movements and melting processes in the polar regions for improved predictions of sea level rise, high-resolution measurement of variations in soil moisture close to the surface for advanced water cycle research, systematic observation of coastal zones and sea ice for environmental monitoring and ship routing, mapping of agricultural fields for crop and rice yield forecasts, monitoring of infrastructure and its degradation as well as emergency observations for disaster mitigation, recovery and prevention. The systematic acquisition concept is based on two main imaging modes: 1) 3-D structure mode with a bistatic radar operation and 2) deformation imaging mode with differential SAR interferometry. Thanks to the novel imaging techniques and the vast recording capacity with up to 8 Tbytes/day, it will provide vital information for solving pressing scientific questions in the biosphere, geosphere, cryosphere, and hydrosphere. Above and beyond the primary mission goals, the unique data set acquired by Tandem-L has therefore immense potential for developing new scientific and commercial applications and services. The final paper will provide details and insights into the retrieval techniques, instrument concept with advanced digital beamforming techniques as well as examples of PolinSAR and tomographic applications. An outlook for the future work towards mission implementation will be given.

The optical volume scattering function (VSF), which describes the angular distribution of light scattered from an incident beam, is a fundamental inherent optical property of the aquatic environment. Along with the spectral absorption coefficient, VSF is one of the two inherent optical properties which describe the propagation of light in aquatic environment. Despite its fundamental nature, there’s little known about the range of variability of the VSF in the aquatic environment. One of the main reasons of lack in the measurement data is that instruments, which have been used for measuring VSF, are complicated and there is no commercially available instrument able to take measurements of the function in full angular range.

Tartu Observatory in cooperation with company Interspectrum and MHI of Ukraine developed a prototype of the new instrument for scattering measurement - MVSM. It outperforms in most of the parameters industrial instruments currently used to measure VSF. It has modern programmable multicolour LED light source, measurement angles cover full angular range, interference filters in receiver enable fluorescence measurements at different wavelengths, and closed measurement volume eliminates the background noise and increases sensitivity and dynamic range. Instrument has modular design with 32-bit microcontroller in every module; software in modules executes under certified real-time operating system SafeRTOS allowing safe and robust multitasking as well as fast multiprocessor communication between modules.

In September 2013 the comparison measurements were carried out in Sevastopol, Ukraine. A set of scattering measurements were carried out in laboratory with pure water and with the water mixed up with 5 µm solid particles. On the basis of the information received from this testing, improvements were made and a new test with 5 µm particles was carried out in Tõravere. The third set of laboratory testing was made after the latest prototype release and 1-5 µm solid particles were used.

As the device is very complex, then we also ran a calibration exercise in the Tartu Observatory optics laboratory to find a set-up that would provide sufficient amount of data for calibration documentation.

In the poster the newest design features and software changes are presented together with testing results and results from the calibration exercise. 

The need for operational soil moisture (SM) monitoring and retrieval has been raised by a number of studies and emphasized by the Global Climate Observing System (GCOS). In this respect many SM measuring networks and field campaigns have been initiated with dedication to various applications such as: hydrology, meteorology, agronomy, and satellite data validation. The launch of Sentinel-1 (S-1) satellite constellation, equipped with the C-band (5.4 GHz) SAR systems, opens a new perspective for SM monitoring over unique ecosystems. One of such environments spans over Biebrza Wetlands located in northeastern Poland, which is one of the largest areas in the EU covered with marshes, swamps, and wet meadows. Within the European Space Agency (ESA) project no. 4000112578/14/NL/MP two S-1 SM validation sites covering grassland and marshland are established within the Biebrza wetlands. This region was also included in the SPOT-5 Take5 experiment that provided high resolution satellite optical imagery complementary to S-1 data. Environmental conditions between both sites vary in respect to SM regimes, soil type, and vegetation cover. This diversity allows to evaluate the S-1 SM products across a wide range of conditions. The marshland site spans over 500x500 m area covered by a regular soil measurement grid of 3x3 stations, whereas the grassland site spans over 200x600 m area with grid arranged in two rows with 4 and 5 stations. Each SM station has two probes installed vertically at the 0-5 cm depth and the following probes installed horizontally at the increasing depths of 10, 20, 50 cm. This results in 18 surface SM, temperature and electrical conductivity measurements for a single site, which are collocated with the S-1 and Sentinel-2 (S-2)/SPOT-5 data. Apart from the SM measurements a wide range of other biophysical variables are measured operationally: air temperature, humidity, pressure; precipitation; wind speed and direction; net short- and longwave radiation; photosynthetically active radiation (PAR); leaf wetness; soil heat flux; CO2 and H2O fluxes (eddy covariance method). Once per month a field campaign is conducted over the validation sites to perform additional measurements related to: leaf area index (LAI), biomass, canopy height, chlorophyll content, TDR SM measurement, soil pH, surface temperature. This complementary material expands the SM validation analysis with ancillary information about the variables influencing SAR signal (biomass, vegetation condition) and provides one more reference SM dataset acquired by means of the TDR technique. Moreover, the vegetation state and shortwave radiation measurements are intended to be used for validation of products originating from optical sensors, such as S-2, SPOT-5. The ensemble of all measurements collocated with the spaceborne SAR and optical imagery is used to monitor water, carbon and energy fluxes within the unique wetland ecosystem. The approach to soil moisture derivation from S-1 data is based on the water cloud model (Attema and Ulaby, 1978), that was further modified by Prevot et al. (1993) and Dabrowska-Zielinska et al. (2007). The water cloud model expresses the total SAR backscatter from a canopy as a sum of contributions from vegetation and underlying soil. Since the backscatter is affected by dielectric and geometrical properties of the canopy, it is possible to expand water cloud model by incorporating a robust vegetation descriptors derived from SPOT5 data.

The aim of the proposed presentation is to demonstrate preliminary results of the soil moisture derivation based on S-1 data and to present potential of the Biebrza sites related to SM validation.

SONMICAT - the integrated sea level observation system of Catalonia - aims at providing high-quality continous measurements of sea- and land levels at the Catalan coast from tide gauges and from modern geodetic techniques for studies on long-term sea level trends, but also the calibration of satellite altimeters, for instance. Up to now, the system has started at l'Estartit and Barcelona harbours.

A description of the actual SONMICAT infraestructure and campaigns at Barcelona harbour are presented.

Especially, an airborne LiDAR campaign was made in July 2014, flying along two ICESat/GLAS target tracks over Barcelona area, in order to compare both methodologies. Advantages and disadvantages with respect to various aspects are discussed; a short overview and the major differences between these two technologies are outlined; and results of this comparison are presented. 

Moreover, the comparison between the GOCE gravity field solutions with existing local an regional gravity field models are presented.

The simulation of future images is part of the development phases of most Earth Observation missions. This simulation uses frequently as starting point images acquired from airborne instruments; these instruments provide the required flexibility in acquisition parameters (time, date, illumination and observation geometry…) and high spectral and spatial resolution, well above the target values (as required by simulation tools). However, there are a number of important problems hampering the use of airborne imagery. On one hand, noise and residual errors from atmospheric corrections introduce unrealistic radiometry in the images; on the other hand limited spatial coverage and, in many cases, observation zenith angles (OZA) far from those that the future misisons will use, constrain the usefulness of the images. Some of these problems are not present when synthetic images are used for the simulation, but these types of images fail to represent the full spatial and spectral variability of radiance across a natural scene, and are therefore of limited use.

The remote sensing laboratory at INTA owns and operates two airborne sensors, AHS and CASI. These has been used extensively for the simulation of many Earth Observation missions in the framework of different ESA campaigns. In this work we present, from the expertise acquired during these campaigns and others, our approach for dealing with the problems listed above. In particular, we examine the problem of OZA angles; we present an analysis of the difference in ground reflectance and related parameters (like vegetation indices) at different scales that arises from the use of airborne imagery when variable OZA and/or sun zentih angle. The selected pixel sizes correspond to those of Sentinel 2.

In conclusion, there is no simple solution to the limitations imposed by the very different geometries. A mixed approach, in which the airborne images are combined with other sources of data (field radiometry, satellite images) seems the only way to overcome such limitation.

The autonomous median trackers on-board the satellite altimetry missions have lower performances in the coastal zones and over the continental ice caps and waters (rivers, lakes, reservoirs), mainly because of the land contamination in the radar waveforms and steep topography. As the number of satellite altimetry applications in these regions has significantly grown for a decade, some techniques have been developed in order to answer these new user requirements. In particular, the DIODE/DEM or OLTC (Open Loop Tracking Command) mode has been implemented with the objective to obtain a larger number of exploitable radar waveforms over these areas of interest for the SRAL altimeter on board Sentinel-3.

The principle of the OLTC (Open Loop Tracking Command) mode consists in driving the altimeter with a priori information available on-board: real-time estimates of the satellite orbit with the DIODE navigator and theoretical height of the point located under the satellite, provided by a Digital Elevation Model (DEM) stored in the on-board memory and previously sampled along the satellite track. The sampled DEM is prepared pre-launch by assembling various elevations data: a Mean Sea Surface for the ocean and coastal zones, a global DEM for continental areas, inland water elevations from specific databases and specific DEM over ice caps where available.

The Sentinel-3 mission, to be launched in December 2015, will fly with an on-board DEM. The analysis of the altimeter cycles operated in OLTC mode on previous missions (Jason-2, AltiKa) have shown the great interest of this technique in the areas of interest and some limitations. This study allowed us to identify improvement of new algorithms resulting in a more accurate on-board DEM.

These algorithms show better agreement between the classic mode (autonomous median tracker) and the OLTC mode. Validation tools have been developed to estimate in near real time the OLTC vs. autonomous mode performance. CNES and NOVELTIS will operate these tools during the Sentinel-3 lifetime. This communication will present the first performance results for Sentinel-3.

Since 1986 the Italian Space Agency (ASI) is one of the 55 agencies supporting the Committee on Earth Observation Satellites (CEOS) initiatives.

The CEOS Mission Statement is to “ensure international coordination of civil space-based Earth observation programs and promotes exchange of data to optimize societal benefit and inform decision making for securing a prosperous and sustainable future for humankind”.

The political, economic, and geopolitical importance of the Disaster Risk Management (DRM) continues to rise as the number of casualties and economic losses increase due to the human-induced or natural disasters.

In this context in 2013 CEOS has created the Working Group on Disasters (WGDisasters) in order to foster, and increase the satellite EO contributions to the various DRM phases and to inform politicians, decision-makers and major stakeholders on the benefits of using satellite Earth Observations in each of those phases.

Since 2013 space agencies have initiated a series of concrete actions to support Disaster Risk Management with a focus on Disaster Risk Reduction (DRR) for disaster preparedness and prevention. These actions have been carried out by strengthening, improving and coordinating the use of EO data through single-hazard Pilots projects (currently focusing on floods, volcanoes and seismic hazards) and multi-hazards projects as the Recovery Observatory (RO), the recently started Landslide Pilot Project and the support to the GEO Geohazard Supersites and Natural Laboratories (GSNL) initiative.

ASI supports the Supersites initiative since 2012, participating in the selection process of Supersites and providing COSMO-SkyMed data to the projects, and takes part in the CEOS DRM Pilot projects, coordinating with USGS the Volcano Pilot project.

Currently ASI has signed 13 agreements for COSMO-SkyMed data exploitation to support CEOS initiatives, in particular nine GEO Geohazard Supersites Natural Laboratories (GSNL) projects (Hawaii, Iceland, Etna, Vesuvio-Campi Flegrei, Marmara, Ecuador, New Zealand, Nepal and Sinabung), three Disaster Risk Management (DRM) Pilot projects (Flood, Seismic, Volcano) and the Recovery Observatory (RO).

To support the Supersites initiative and the DRM Pilot projects, ASI has planned to provide 7900 COSMO-SkyMed products until the 2017, while 2794 products have been delivered up until now.

In this work an overview of the projects and the results achieved in various fields of application thanks to the COSMO-SkyMed data exploitation will be shown together with the statistics of the COSMO-SkyMed products, delivered for each project.

In the framework of the CEOS initiatives, since 2014 ASI has joined the CEOS Space Data Coordination Group (SDCG), established in 2011 to support the Global Forest Observations Initiative (GFOI), deciding to provide COSMO-SkyMed data for the GFOI Research and Development (R&D) activities.

The Sentinel-3 Mission Performance Centre (S-3 MPC) is one of the facility of the Payload Data Ground Segment (PDGS). It aims at controlling the quality of all generated products, from L0 to L2. The S-3 MPC is composed of a Coordinating Centre (CC), where the core infrastructure is hosted, which is in charge of the main routine activities (especially the quality control of data) and the overall service management. Expert Support Laboratories (ESLs) are involved in calibration and validation activities and provide specific assessment of the products (e.g., analysis of trends, ad hoc analysis of anomalies, etc.). The S-3 MPC interacts with the Processing Archiving Centers (PACs) and the Marine Centre at EUMETSAT.

The S3MPC covers both optical and topography missions, each of them composed of several instruments. The launch of S3-A is planned in December 2015, with an end of Phase E1 planned in early May 2016.

During that period, the S3-MPC will start its activities, mainly focused on:

Calibration activities, done in close relationship with the satellite commissioning team at ESTEC

Processor verification and update of specifications

Validation of L1 products and assessment of instrument performances

Cross-chekcing of L1 products processed in various processing centres

Progressive implementation of Quality Control activities done at the S3-MPC

Validation of L2 products, in synergy with the Marine Centre for L2 marine products

Lessons learned during that period, main results and key outcomes will be presented as well as the way forward for the ramp-up phase and the routine operations of the S3-MPC.

The Global Change Unit (GCU) at the University of Valencia has been involved in several cal/val activities carried out in dedicated field campaigns organized by ESA and other organisms. However, permanent stations are required in order to ensure a long-term and continous calibration of on-orbit sensors. In the framework of the Spanish National Plan for Scientific and Technical Research and Innovation, the GCU has managed the setting-up and launch of experimental sites in Spain for the calibration of thermal infrared sensors and the validation of Land Surface Temperature (LST) products derived from those data. For this purpose, the spatial heterogeneity of selected sites was analyzed, and permanent stations with thermal radiometers for the continuous measurement of radiometric temperatures were implemented. Currently, three sites have been identified and equiped: the agricultural area of Barrax (39.05N, 2.1W), the marshland area in the National Park of Doñana (36.99N, 6.44W), and the semi-arid area of the National Park of Cabo de Gata (36.83N, 2.25W). The activities of the GCU also included the implementation of an operational processing chain in order to provide in near-real time different remote sensing products, including LST. This work presents the performance of the permanent stations installed over the different test areas, as well as the cal/val results obtained for a number of EO sensors: SEVIRI, MODIS, VIIRS, and Landsat series. We also show the results obtained in the validation of LST products derived from AATSR, with dicussion on the implications for the forthcoming Sentinel-3/SLSTR.

The Earth is a “complex changing planet” on which the climate and ecosystems are changing rapidly, driven by society’s accelerating use of the Earth’s natural resources. Earth observing systems are required to deliver integrated, transparent and permanent monitoring capabilities at the global scale to report on the state and the evolution of the world vegetation, agriculture and water.

The joint Belgian-Chinese Global-V satellite mission concept aims at providing continuous and daily access to data of global land masses, at a resolution of 100m, in 4 VNIR, 1 SWIR and 2 TIR bands. Data derived from this mission will lead to dedicated applications for food security, agriculture, and management of natural resources worldwide.

Global-V will be the first to deliver global acquisitions at 100m resolution on a daily basis, and will fill the clear gap in global monitoring of the TIR spectral range. Users consider this an ideal data stream to complement higher resolution land missions with more constraints on data coverage or availability such as Sentinel-2, the Landsat series and the Chinese Gaofen series. Apart from land masses, the spatial coverage will be extended to include all coastal areas and inland water bodies.

The VNIR/SWIR bands will be designed with the same specifications for radiometric quality and response as SPOT/VGT and PROBA-V, which allows to continue the existing time series since 1998 and extending it beyond 2020. An additional Green band and a new thermal instrument with 2 TIR bands between 10-12 mm are included to develop new capabilities and synergies with high-resolution land missions. In that respect, Global-V will be placed in the same orbit as Sentinel-2 to allow the full exploitation of synergies between the two missions.

The daily coverage and guaranteed data availability of Global-V will make it a valuable resource for long-term monitoring, detection of land cover changes, as well as changes related to climate change or extreme weather events. To this end, the stringent calibration specifications known from PROBA-V will be maintained for Global-V.

The presentation provides an overview of the mission concept, its main performance indicators, envisaged standard data products and data dissemination, and proposes opportunities for new products taking advantage of the daily coverage at 100m spatial resolution, and the combination of optical and thermal bands.

Radar interferometry has been successfully applied to measure ground deformation for over two decades. However, until now the data have been acquired on an ad-hoc basis. The Sentinel-1 mission, operational since October 2014, is the first to acquire global data systematically. Moreover, the default mode of radar acquisition operation for Sentinel-1 is the new Terrain Observation by Progressive Scans (TOPS) mode. Processing these data requires a major shift in methodology in comparison to traditional approaches. We have developed a completely new re-engineered and adapted InSAR time series processing approach, which efficiently processes the data from this new type of SAR constellation, with the goal to deliver ground deformation products with the highest possible precision. In summary, the proposed system approach requires the development of an automatic, almost unsupervised, system that integrates methods to obtain time-dependent surface deformation estimates. The ground velocity maps will ideally meet the desired accuracy of 1 mm/yr/100 km to measure strain-rates (10 nanostrain/yr) at a comparable level to current existing sparse regional GPS measurement networks.

Here, we present in detail the processing steps to achieve stack/time series products (linear velocity and displacement time series), which differ from the traditional stripmap mode interferometry. In this communication, we describe the different steps we have adopted to solve, among other problems, how to coregister TOPS SAR image stacks containing 10s to 100s of images acquired over many years, how to efficiently process newly acquired data, how to filter interferograms optimally in all settings, , and how to select reliable (temporally-stable) pixels for phase unwrapping. We illustrate some of the special features of our processing system with new results for a few selected target regions.

Satellite interferometric SAR is a powerful technique to measure sub-centimeter-scale relative displacements over spans of days to years. However, it is difficult to attribute radar scattering centers to real-world objects, since the geolocalization precision of the measurement points (the scatterers) is still poor. This hampers the proper interpretation of the measurements. Even though for high resolution sensors such as TerraSAR-X, a precision of several centimeters has been reported [2], these results still relate to the azimuth and range position only, not to a terrestrial 3D position. For medium resolution sensors such as ERS1/2, EnviSAT, Sentinel-1 and Radarsat-2, the precision is even worse, at the meter-range, with a highly-inclined cigar-shaped error ellipsoid with typical axis length ratios of 1/3/100 for azimuth, range, and cross-range directions [3].

In this study, we will consider all the error sources for PS positioning, such as the inaccuracies in the orbital state vectors, radar system time errors, atmospheric delay, solid Earth tides, tectonic drift, and the InSAR processing error, to construct the error ellipsoid of every InSAR measurement and interpreting its precision in astochastic manner. Moreover, in order to achieve the 'Object Snap' (referring InSAR measurements to the corresponding actual ground objects) we will locate the PS into a Lidar 3D testing environment by using a Nearest Neighbor search. Here we use the terrestrial laser scanning data to build the 3D surface model with millimeter precision. One of the key points of this work is to find the solution of the collaboration between laser scanning data and SAR data because of different data frames. We propose to use the connection/reference points for coordinate transformation. We demonstrate our method over a number of test sites in the Netherlands by using both high and medium resolution data (TerraSAR-X, Radarsat-2 and Sentinel-1 data). Corner reflectors (CRs) are used to test and validate the connection points, where the geodetic locations of the CRs are measured by GPS. Eventually, we show the precise PS 3D geolocation in the Lidar 3D environment and distinguish the actual ground objects for every PS point.

Reference:

[1] Ferretti A, Prati C, Rocca F. Permanent scatterers in SAR interferometry[J]. Geoscience and Remote Sensing, IEEE Transactions on, 2001, 39(1): 8-20.

[2] Eineder M, Minet C, Steigenberger P, et al. Imaging geodesy—Toward centimeter-level ranging accuracy with

TerraSAR-X[J]. Geoscience and Remote Sensing, IEEE Transactions on, 2011, 49(2): 661-671.

[3] Deenathayalan, P., Small, D., Schubert, A., and Hanssen, R. F. (2015). High precision positioning of radar scatterers. Journal of Geodesy, submitted.

Current space-borne thermal infrared remote sensing capabilities/systems aimed at land surface remote sensing retain some significant deficiencies, in particular in terms of spatial resolution, spectral coverage and number of imaging bands and temperature emissivity separation. Largely building on existing and space-proven technology and the heritage of the German BIRD and FireBID missions, the proposed VISible-to-thermal IR micro-SATellite (VISIR-SAT) mission addresses many of these limitations, providing multi-spectral imaging data with medium-high spatial resolution (80m GSD from 800 km altitude) in the thermal infrared (up to 6 bands, 8 to 11µm) and in the mid infrared (1 or 2 bands, at 4µm). These MIR/TIR bands will be co-registered with simultaneously acquired high spatial resolution (less than 30 m GSP) visible and near infrared multi-spectral imaging data. To enhance the spatial resolution of the MIR/TIR multi-spectral imagery at daytime, data fusion methods will be applied, such as the Multi-sensor Multi-resolution Technique (MMT), already successfully tested over agriculture terrain. Thus, EO products for Land Surface Temperature (LST) and Soil Moisture (SM) with a 30 m spatial resolution can be generated. For high temperature phenomena (e.g. vegetation- and peat-fires), the Fire Disturbance ECV “Active fire location” and “Fire Radiative Power”, as well as effective fire temperature and the spatial extent even for small fire events will be retrieved with less than 100 m spatial resolution from VISIR-SAT data. System characteristics of VISIR-SAT thus go beyond existing and planned IR missions. These comprehensive high-accuracy products from VISIR-SAT (e.g. for fire) may synergistically complement high temperature observations of Sentinel-3 SLSTR. Additionally, VISIR-SAT features a very agile sensor system, which will be able to conduct intelligent and flexible pointing of the sensor’s Line-of-Sight with the aim of providing global coverage of cloud free imagery every 5-10 days with only one satellite (using near real time cloud cover information). It may be flown in convoy with Sentinel-2 and/or Sentinel-3.

Time-series of vegetation index derived from satellite spectral measurements can be used to gain information on vegetation dynamics. This information contributes to ecosystem research, such as phenology and carbon cycles. To obtain this information, TIMESAT provides many smoothing algorithms (Savitzky-Golay filter, double logistic functions, asymmetric Gaussian function, LOESS smoothing, and smoothing spline) to process coarse-resolution data (http://www.nateko.lu.se/timesat). Evaluating the performance of these smoothing algorithms on extracting vegetation seasonality information from satellite data is the aim of our research. For evaluating each smoothing algorithms in TIMESAT, we base the research on comparing the smoothed satellite data with ground references, which contains land-based spectral observations and topography. Land-based spectral observation data provide a destination to smoothing operation, which means the smaller difference between smoothed satellite data and ground observations is the better smoothing operation. The evaluation of smoothing operation focused on the difference of phenology parameter extracted by TIMESAT from satellite-based spectral data and land-based spectral observations. We also navigated the limitation of each smoothing algorithm in reconstructing time-series which contains missing data. Based on the previous evaluations and a design philosophy that is elevation caused temperature variation would cause the difference of vegetation phenology, we finally evaluated performance of the smoothing algorithms in mapping vegetation dynamics in a mountain area. In conclusion, this research provides an alternative evaluation system for selecting and setting smoothing operation on satellite vegetation index time-series. In addition, this evaluation system could be applied on high spatial resolution satellite data, i.e. Sentinel-2 and Landsat 8. The integration of the results from this research can be used for further research, such as vegetation carbon uptake and land use classification.

Land surface temperature and emissivity measurements offer valuable information for many applications such as evapotranspiration, energy balance, mineral mapping, urban studies, and volcano surveillance. In particular, urban studies can take advantage of these data for urban heat island detection, city thermal regime characterization and materials classification, among others.

Computation of land surface temperature and emissivity from radiometric measurements is mathematically unsolvable because one obtains more unknowns than equations. Therefore, different empirical approaches of temperature and emissivity estimation have been developed. One of the most widely used approach is known as the Temperature and Emissivity Separation (TES) algorithm. The TES algorithm was originally developed for the ASTER instrument on board NASA’s Terra platform. Currently, the algorithm is used in data processing from many others spaceborne and airborne sensors. This work introduces an improved version of the TES algorithm. The major improvement is based on the substitution of Normalized Emissivity Method (NEM) module by a new one that takes advantage of relationship between brightness temperature and emissivity to smooth spectral signatures. This newly developed algorithm is further tailored for the TASI sensor, which is airborne thermal hyperspectral scanner developed by Itres Ltd. (Calgary, Canada).

Data set used in this work consists of airborne images and ground measurements. Airborne data were acquired by TASI sensor, which consist of 32 spectral bands situated in the 8 to 11 µm region and having FWHM ≈ 0.11 µm. Data were acquired on 4.7.2015 over city Brno, Czech Republic. Radiometric and geometric corrections were done by sensor specific software provided by the manufacturer. Atmospheric corrections were performed by using MODTRAN 5.3, which used atmospheric temperature and water vapor profiles delivered by MODIS product MOD07_L2. Spectra of different urban surfaces, including asphalt parking lots, concrete pavement, asphalt roof and aluminized foil, were acquired with a portable FT-IR spectrometer Model 102 developed by D&P Instruments (Simsbury, USA). These ground measurements were processed with a spectral smoothing algorithm in order to get surface emissivities. These emissivities were then convolved with TASI response functions.

We then applied our improved version of TES algorithm to the corrected TASI image. The emissivity results were compared with emissivities derived from in-situ measurements. Close agreement between these implies that the improved version of TES algorithm gives promising results. We also conclude that with the suggested procedures airborne thermal hyperspectral data can be effectively used for classification of urban surfaces.

For 16 years, VGT-1 and VGT-2 acquired near-daily global coverage of the Earth’s land mass at 1km spatial resolution in the Blue, Red, NIR and SWIR spectral bands. The time series started in April 1998 with VGT-1 and was continued from February 2003 with VGT-2, until its last image was captured on 31 May 2014. Meanwhile, PROBA-V was launched. One of the main goals of the PROBA-V mission is to provide continuity for the SPOT-VGT time series, allowing the extension of the daily global dataset from 1998 into the present and future. Although the PROBA-V sensor was spectrally defined as similar as possible to SPOT-VGT, there are nevertheless differences to cope with, related to differences in the camera system and geometry, but also associated with spectral characteristics.

The combined SPOT-VGT archive is currently being reprocessed with the objective to improve the time series, to harmonize its content and to obtain a better consistency with the PROBA-V data. The reprocessing includes the application of updated instrument calibration parameters, as determined by the VEGETATION Image Quality Centre located at CNES, to correct for among others observed striping and the smile effect in the blue band of VGT-2, adjustments for incorrect Sun-Earth distance modelling, and a better cloud, snow and ice detection algorithm. This study focuses on the evaluation of the effect of the reprocessing on VGT-1 and VGT-2 surface reflectances and NDVI, and the consistency between SPOT-VGT and PROBA-V.

The analysis, which is performed in parallel with the current reprocessing activities, is based on a relative comparison of maximum value composites between the ‘old’ and reprocessed VGT, between reprocessed VGT-1 and VGT-2, between reprocessed VGT-2 and PROBA-V, and on comparison with external datasets such as from METOP-AVHRR and MODIS. The consistency between SPOT-VGT and PROBA-V is evaluated for 10-daily composites over the entire overlapping time period, and for daily composites for a limited number of days when the viewing geometry of SPOT-VGT and PROBA-V is nearly identical.

The evaluation is done on a spatially subsampled global dataset, with additional sampling schemes defined based on constraints on registration day and view zenith angle. The following aspects were investigated: (1) product completeness: quantification of missing or flagged pixels over land; (2) spatial consistency analysis: spatial distribution of (dis)similarities and analysis of residuals per biome type; (3) global statistical analysis; and (4) spatio-temporal consistency analysis.

Preliminary results indicate higher inter-annual and intra-annual stability of the reprocessed SPOT-VEGETATION surface reflectance and NDVI products. The reprocessing also results in a better consistency between VGT-2 and PROBA-V, with very small systematic bias between products: over 93% (Blue), 87% (Red), 68% (NIR) 74% (SWIR) and 72% (NDVI) of the pixels show a bias within the optimal range (±5% for surface reflectance or ±0.05 for NDVI). Taking into consideration that these observations are still affected by anisotropy effects, this is a good result. Before reprocessing, consistency was considerably lower, with 88% (Blue), 81% (Red), 64% (NIR), 71% (SWIR) and 69% (NDVI) of the pixels with bias within the optimal range. 

Accurate geo-location is one of the fundamental requirements for Earth observation satellite imagery to be suitable for many applications, in particular to generate temporal composites of EO satellite products, and to support change detection and retrieval of bio-geophysical parameters over heterogeneous land surfaces.
In the GEOACCA project, funded by the European Space Agency (ESA) under the Quality Assurance framework for Earth Observation (QA4EO), we developed a tool for assessment of the geolocation accuracy of medium resolution optical satellite data (spatial resolution of 300 m to 1000 m). The tool supports ERS ATSR-2, ENVISAT AATSR and MERIS, and Proba-V VEGETATION, but can be easily extended to other sensors like the upcoming Sentinel-3 SLSTR and OLCI. The tool uses a global reference data set of more than 350 reference points, which was compiled within the project. Reference points are image chips extracted from terrain corrected Landsat Level 1T scenes with 30 m pixel spacing covering lakes and islands with an extension of a few to several km. We used also vector outlines of the SRTM Open Water Bodies data set to check the quality of the Landsat reference image chips. In a first step the medium resolution images of a data take are rectified using a digital elevation model and data windows centered at the estimated location of the reference points are extracted. Then the data windows are oversampled to the resolution of the reference points by bilinear interpolation and the displacements in map projection (Northing, Easting) and in image coordinates (in along track and across track direction) are estimated using an image cross correlation method and are stored in a data base together with image metadata and quality information of the matching process. Image chips affected by clouds, polar night or covered completely or partly by lake ice or snow cannot be used. They are identified by preprocessing steps and labelled in the database. The tool was installed at ESRIN GPOD and is currently operating on the complete ERS ATSR-2 and ENVISAT AATSR and MERIS FR data set. In first tests for a limited time period of 2003 to 2005, ENVISAT AATSR at nadir viewing shows an absolute geolocation error in the years 2003-2005 in the range of 200 to 800 meters, with some outliers above 1 km.
Various options are implemented in a web tool acting as front end to the database to support the visualization and analysis of the geolocation accuracy e.g. for one reference point in time or the spatial distribution for a certain time period. Trends of the geolocation accuracy for the complete life time of a mission or selected periods can be displayed. We show selected results of the geolocation accuracy and the temporal trends for the ESA sensors ATSR-2, AATSR and MERIS.

University College London has a long involvement in the Sentinel-3 SRAL Mission,  through the development of the Sentinel-3 Level-2 Ice Detailed Processing Model (DPM) and Reference Processors, and provision of key land and sea ice auxiliary data to the Sentinel-3 ground segment.  As an Expert Support Laboratory for the Sentinel-3 Mission Performance Centre, here we present preliminary results from the first look characterisation of the SRAL instrument and processor performance over ice sheets, ice margins and ice shelves. During the first few months of the Sentinel-3 commissioning phase it is intended to operate the SRAL instrument in its two different operational modes. For the first 27-day cycle, the SRAL low resolution LRM mode will be selected. Equivalent low rate modes were used in previous altimetry missions such as Envisat over land ice and CryoSat-2 over ice sheets. Subsequently the instrument will be switched to use the high along track resolution SAR mode operating in open-loop tracking for ice margins and closed loop for ice sheets. Prior to Sentinel-3, a SAR mode altimeter has  been operated on CryoSat-2, but predominantly over areas of sea ice. CryoSat-2 uses a separate interferometric mode over ice margins. We present a direct comparison of performance and biases between Sentinel-3’s two main operational SRAL modes over ice sheets and margins and a compare these results from contemporaneous and  previous altimeter missions. We will also compare performance  and biases between the two separate SAR tracking modes - open loop and closed loop. Crossover analysis will be used to show statistics and maps of the spatial patterns of the amplitude of elevation and backscatter residuals over areas of Antarctica and Greenland for the different waveform retrackers used in the level-2 processing. These results provide information on the performance and stability of the SRAL instrument in each mode over different ice surface types and slopes. They also provide information on instrument, orbit and timing errors, polarization and volume scattering issues, the effectiveness and correct implementation of geophysical corrections, slope models and Doppler corrections in the L2 processor. A more detailed analysis of individual tracks, waveforms and the retracker performance over specific test sites will be carried out and where necessary initial recommendations for tuning retracker parameters will be discussed. An analysis of the effectiveness and correct implementation of the DEM derived slope model used to relocate the echo point to the point of closest approach over sloping surfaces in Antarctica and Greenland will be shown as will tests on the initial availability, stability and validity of  L2 geophysical and atmospheric corrections over polar surfaces. Both of these are important components in the calculation of elevation and elevation rates over ice surfaces which are a key  variable in operational climate studies.

The Jason-CS/Sentinel-6 (JCS/S6) mission will consist of 2 spacecraft and will be the latest in a series of ocean surface topography “reference” missions that will span nearly three decades. It is a part of the EC Copernicus initiative, whose objective is to support Europe’s goals regarding sustainable development and global governance of the environment by providing timely and quality data, information, services and knowledge.  The JCS/S6 satellites follow a series of satellite altimeters on-board TOPEX/Posiedon through to the Jason-3 mission (that is awaiting launch re-planning following Falcon-9 CRS-7 failure in June 2015). Jason-CS will continue to fulfil objectives of the reference series whilst introducing a major enhancement in capability providing the operational and science oceanographic community with the state of the art in terms of platform, measurement instrumentation design thus securing optimal operational and science data return.

The design will be based on a platform derived from CryoSat-2 but adjusted to the specific requirements of the higher orbit. Building on the heritage of the ESA CryoSat-2 SIRAL and Sentinel-3 SRAL instruments, the principle payload instrument is a high precision dual-frequency (for high stability ionospheric path delay correction) Ku/C band synthetic aperture radar altimeter (POSIEDON-4).  Retrieval of geophysical parameters (surface elevation, wind speed and SWH) from the altimeter data requires supporting measurements derived from a DORIS receiver (for Precise Orbit Determination (POD)) and the Climate Quality Advanced Microwave Radiometer (AMR-C) provided by JPL (for high stability wet-tropospheric path delay correction). Orbit tracking data are also provided by on-board GPS  and a Laser Retro Reflector. An additional US GPS receiver, GNSS-RO, will be dedicated to radio-occultation measurements.

The JCS/S6 altimeter will be the first Synthetic Aperture Radar (SAR) altimeter used as part of the reference altimeter mission series.  It will also be the first altimeter to operate in a continuous high-rate pulse mode optimizing the RADAR sampling strategy and allowing, for the first time simultaneous production of low-resolution mode (LRM) measurements on-board as well as the processing of SAR echoes on-ground. Both types of measurements will be provided in (separate) Sentinel-6 altimeter data products.

Two satellites are in development with the 1st planned for launch readiness in the 1st half of 2020 with the 2nd satellite 5 years later. Partner Agencies include ESA for development, procurement & early orbit activities/verification for the two satellites; EUMETSAT for mission management, ground segment, flight ops, providing a contribution to the funding of the 1^st satellite and participation in funding for the 2nd satellite; NASA for developing the US payload, launcher procurement and funding US science; EU for funding the operations and participation in funding (with EUMETSAT) for the 2nd satellite; CNES for mission expertise and provision of POD and NOAA whose role is being defined. All the partners are expected to jointly contribute to commissioning and mission validation.

Geosynchronous SAR missions (GEOSAR) are presently being studied to complement present LEOSAR satellites providing unprecented continous SAR Observation of the Earth at continental scale. Multiple applications would benefit from this permanent monitoring approach among them: land stability control, natural risks prevention and accurate numerical weather prediction models from water vapour atmospheric mapping. In the GEOSAR case due to the smaller relative Satellite-Earth velocity and much larger observation range, the synthetic aperture integration time becomes in the order of minutes or even hours when a Geostationary platfom is selected which is the case of GeoSTARe concept under study.  

In this type of missions, focusing the SAR image requires knowing the range history of the satellite with respect to Earth surface with an accuracy better than the radar wavelength which is well beyond the usual orbit determination requirement of satellites in GEO orbits.

To cope with this problem an orbit estimation is proposed based on a group of Active Radar Calibrators (ARC) placed in well-known positions of the observed scene. These devices amplify and transmit back the received radar pulse, which can be used as calibration signal. Hence the ARC can be modelled as persistent scatterers with a selectable Radar Cross Section that depends on the antennas and amplifier chain gains. The range compressed ARC echoes provide the range and range-rate observables, which can be extracted from the radar acquired raw data before SAR processing. The accuracy of these measurements, and hence the accuracy of the orbit estimation, depends on both the achieved Signal to Noise and Signal to Clutter ratios. In the paper we propose a pyramidal subaperture processing of the ARC signals intended to reduce the ARC Radar Cross Section Requirements and the impact of surface Clutter resulting in small ARC antennas and moderate gain.  Using a balanced internal calibration loop the ARC can be operated with a high stability specification both in phase and amplitude. In addition adding appropriate coding in the ARC internal path it is possible to easily discriminate among different ARC and between calibration and surface echoes reducing potential interferences of the ARC on the scenes of interest.  

The range and phase measurements collected during the calibration time constitute the input of an inverse problem. In fact, the approach followed consists in determining an orbit correction with respect to an orbit taken as reference, by the evaluation of the difference between the measured and expected observables. The latter are computed in correspondence of the sensor position along the reference orbit evaluated on the observation epochs. Both batch and sequential method can be adopted for the inversion. The batch Least Square method considers one reference orbit and determines the orbit correction to apply at a given starting epoch considering all the measurements at the same time. The sequential Kalman Filter considers as orbit reference the one determined at the previous epoch and updates the estimation at each epoch considering the new observables measurements. The latter is ideal to follow the orbit variation. However, since this variation is quite slow, the Least Square method can be applied considering calibration intervals for which the orbit error can be assumed constant. The paper will present the overal calibration approach, the proposed configuration of the ARC including preliminary experimental data and simulations of the relative orbit determination problem related to GEOSAR focusing.

In the frame of COSMO-SkyMed di Seconda Generazione (COSMO-SkyMed Second Generation, CSG) programme, extensive research activities have been conducted regarding the SAR sensors data processing. One of the most important, complex and scientifically interesting issue of SAR data processing regards the ability to focus images with extremely high resolution, as well as the adaptive image processing capacity, aimed to equalize the radiometry or reduce noise and artefacts, so as to increase the images quality.

As regards the resolution, it is essential to create a model for the management of all those elements which alter the target phase responses, while it is "integrated" for several seconds, i.e. an order of magnitude greater than the typical integration interval of the currently available systems. The task is very challenging since, while increasing the “integration” time, several phenomena that are usually considered as negligible will significantly affect the image quality such as, for example, the orbital variations respect to the theoretical orbit and the effects due to the altitude and the target position within the image.

Concerning the noise removal, one of the major problems that pertain SAR wide-ranging products is the ability to equalize the image, i.e. compensating all the phenomena that affect the received signal intensity which are due to the fact that the instruments is non-ideal. In particular, research activities are aimed at developing adaptive-iterative techniques for the compensation of inaccuracies on the knowledge of radar antenna pointing, up to achieve compensation of the order of thousandths of degree. The researches have also the goal to develop highly efficient algorithms for processing huge amounts of data (in the order of Giga / Tera-bit) in a very short time. Moreover, the emerging need to generate accurate and reliable DEM (Digital Elevation Model) of the Earth’s surface leads to the necessity to design an Interferometric SAR system based on multi-pass approach, which however is inherently subjected to different types of errors. Thus the research are also aimed at studying and designing  an algorithm which, by exploiting a larger number of input information (such as number of pairs having different geometric characteristics and different types of low performance reference DEM), will be able to obtain products at the required level of accuracy.

ABSTRACT

Sea Surface Height (SSH) estimates from GNSS-Reflectometry can provide valuable data that can complement nadir altimetry, filling the space between altimetric tracks, and providing the opportunity for enhanced observations of mesoscale eddies [1]. Altimetry using GNSS-R was proposed for the first time back in 1993 as the PARIS concept [2], and has undergone numerous improvements since then, becoming recently the main focus of the GEROS-ISS experiment, scheduled for launch in 2019 [3]. The current predicted precision for Sea Surface Height (SSH) estimation using GNSS-R still remains on the order of tens of centimeters, but the multistatic nature of GNSS-R makes it very attractive due to the high space-time sampling, free transmitters and low-cost receivers, that could be launched to form progressively larger constellations, or piggybacked on other satellites. Airborne and tower-based experiments have proven the feasibility of the technique, but with the launch of TechDemoSat-1 (TDS-1) satellite by Surrey Satellite Technology in July 2014, a fairly large dataset of spaceborne oceanic GPS-Reflections has recently become available. The global dataset of GPS-R data is expected to increase dramatically with the launch of the CYGNSS microsatellite constellation in October 2016 [5].

This study presents an analysis of TDS-1 data for SSH estimation [6]. A number of oceanic regions are selected as test areas, and the SSH is extracted from delay waveforms, using the Leading Edge Derivative (LED) algorithm [7]. These represent the first spaceborne observations of SSH using a GNSS-R instrument. The results are compared with the mean SSH, derived using the DTU10 model [8], and a reasonable agreement between SSH derived from TDS-1 and ground truth will be shown. The issues and limitations of the data and of the instrument itself will be discussed, being the GPS-R instrument onboard TDS-1 designed primarily for scatterometric purposes.

References

[1] Zuffada, C., Z. Li, S. Nghiem, S. Lowe, R. Shah, M. P. Clarizia and E. Cardellach (2015), “The Rise of GNSS-Reflectometry for Earth Remote Sensing”, in Proceedings of the International Geoscience and Remote Sensing Symposium (IGARSS) 2015, Milan, Italy.

[2] Martin-Neira, M., (1993), “A Passive Reflectometry and Interferometry System (PARIS):Application to ocean altimetry”. ESA Journal, 17, 331-355.

[3] Wickert J. , O. Andersen, G. Beyerle, E. Cardellach, B. Chapron, C. Gommenginger, P. Høeg, A. Jäggi, N. Jakowski, M. Kern, T. Lee, M. Martin-Neira, N. Pierdicca , C.K. Shum1, C. Zuffada (2014), “GEROS-ISS: Innovative Ocean Remote Sensing using GNSS Reflectometry onboard the International Space Station”, in Proceedings of the Advanced RF Sensors and Remote Sensing Instruments (ARSI) workshop, ESA Estec, Noordwijk, NL.

[4] Ruf, C., R. Atlas, P. Chang, M. Clarizia, J. Garrison, S. Gleason, S. Katzberg, Z. Jelenak, J. Johnson, S. Majumdar, A. O’Brien, D. Posselt, A. Ridley, R. Rose, V. Zavorotny (2015), “ New Ocean Winds Satellite Mission to Probe Hurricanes and Tropical Convection,” Bulletin of American Meteorological Society, doi:10.1175/BAMS-D-14-00218.1.

[5] Cardellach, E., Rius, A., Martín-Neira, M., Fabra, F., Nogués-Correig, O., Ribó, S., Kainulainen, J., Camps, A., D Addio, S. (2013), “Consolidating the Precision of Interferometric GNSS-R Ocean Altimetry using Airborne Experimental Data,” IEEE Transactions on Geoscience and Remote Sensing, 52(8), 4992-5004, doi:10.1109/TGRS.2013.2286257.

[6] Clarizia, M.P., Ruf C., Cipollini P. and Zuffada C. (2015), “First Spaceborne Observations of Sea Surface Height using GPS-Reflectometry”, submitted to Geophysical Research Letters.

[7] Hajj, G.A., and C. Zuffada (2003), “Theoretical description of a bistatic system for ocean altimetry using the GPS signal,” Radio Science, 38(5), doi:10.1029/2002RS002787.

[8] Andersen, O. (2010). “The DTU10 gravity field and mean sea surface,” Second international symposium of the gravity field of the Earth (IGFS2), Univ. of Alaska Fairbanks.

The data acquired by space-born SAR satellites are naturally affected by a Doppler shift, due to the relative motion between the platform and the imaged ground scene. Such Doppler shift is usually called “Geometric”, and assuming a perfect knowledge of the orbit and of the steering law of the sensor, can be fully predicted.

A further Doppler shift can be originated by the motion of the imaged scene: this is the typical scenario for the acquisitions over the ocean, where the wind and the currents introduce an additional Doppler shift on the received signal. Such Doppler shift is usually called “Geophysical” and can be exploited, after subtraction of the geometric predictable geometric contribution, to retrieve meaningful information from SAR data, like ocean currents and winds velocity and direction.

The accuracy of the extracted geophysical information is limited by two main factors:

The intrinsic uncertainty of the Doppler Centroid (DC) estimates from data which can be reduced by increasing the amount of averaged data at the price of reduced spatial resolution.

The accuracy of the available geometric information, introducing both an uncertainty and a bias in the estimated geophysical Doppler contribution.

Modern SAR sensors (like Sentinel-1) are equipped with accurate attitude control systems, allowing good accuracy in the estimation of the geometric Doppler contribution (say tens of Hz). Unfortunately this is not enough for science application (e.g. weather prediction models), possibly requiring errors of few Hz.

The present paper focuses on this issue and proposes a method to perform the calibration of the senor pointing, exploiting the DC estimates of SAR acquisitions over still land areas, where no geophysical Doppler contribution is expected. In particular the method is aimed at identifying recurring biases in the geometric DC estimates possibly originated by non-perfect calibration of the on-board attitude control systems. The estimation and correction of such biases would be fundamental for enabling science applications.

The first part of the paper will introduce the proposed pointing calibration method which, starting from a grid of DC estimates in the imaged swath, performs a Singular Value Decomposition to identify the pointing plane best fitting the available DC estimates. This process, repeated in time, will allow to identify any possible trend in the sensor pointing which, corrected, will allow to increase the accuracy of the geometric Doppler estimates.   

The second part of the paper will provide the results of the proposed pointing calibration strategy applied to Sentinel-1A SAR sensor.

Satellite altimetry missions are intended to estimate some geophysical parameters of the sea surface (topography, wave heights, etc ...) from the backscattered radar signal. It is therefore essential to have a good understanding of the interactions of electromagnetic waves with the surface. With more than twenty years of altimetry missions using Ku, C and S bands, the radar backscattering nadir cross-section is now well documented. However, the recent launch of SARAL/AltiKa in February 2013 and the preparation of the future SWOT (Surface Water Ocean Topography) mission, both working in Ka band, triggers the need for the characterization of the scattering mechanisms at this shorter wavelength.

In this study we make a joint analysis of the Ka, Ku and C bands normalized backscattering cross-section (sigma naught or Sigma0) through the cross-comparison of Jason-2 and AltiKa measurements. A specific processing (new retracker and dedicated editing) is applied on both datasets in order to improve the sigma naught estimates from the GDR products. Systematic comparisons are established for varying values of the main sea state parameters (wind speed, wave height and wave period) using one year of data.

The results show that the more the frequency is high, and the more the backscattering cross-section will vary according to the various couples wind/waves. In addition, the backscattering cross-section in Ka band remains more sensitive to the low winds and waves conditions in comparison with the lower frequency bands. The influence of the wave period on Sigma0 has also been analyzed and allowed to show that, for the three frequency bands, the impact of this parameter was only visible for low wind speeds, Ka band remaining to most sensitive. Finally, the average values of sigma0 showed a decrease with the frequency as expected.

The relationships between Ku and Ka sigma naught measurements have also been reconsidered in the light of a recent analytical scattering model, referred to as GO4, which makes it possible to extrapolate Ka from Ku data with no a priori knowledge on the sea surface spectrum. The theoretical predictions are found consistent with the observations but tend to show a likely calibration bias of about -0.7 dB for the value of the AltiKa sigma naught (which is of the same order of magnitude as the accuracy of the altimeter Sigma0 measurements).

Numerous SAR space-borne sensors are currently operating or will be launched in the near future. In this context, there is a need to reinforce the methodological and thematic developments in relation to future space missions dedicated to major scientific and societal issues: the Sentinel-1 and BIOMASS missions by European Space Agency (ESA). Both missions can be feasible to do tomography. While interferometry SAR technique is well known, the tomography is quite new and state-of-the-art technique which provides the unique information in vertical direction. The rationale of tomography is to employ multiple flight tracks, nearly parallel to each other. The ensemble of all flight lines allows us to form a 2D synthetic aperture, resulting in the possibility to focus the signal in the whole 3D space. In other words, by exploiting tomography, multi-baseline SAR data can be converted into a new multi-layer SAR data stack where each layer represents scattering contributions associated with a certain height [1].

In BIOMASS mission, it features the first ever P-band (wavelength 69cm) and Tomography from space. During the tomography phase, the satellite's orbit is designed to gather multiple acquisitions over the same sites from slightly different orbital positions, so as to image the forest's vertical structure through SAR tomography. The great novelty of the tomography is that, for the first time, a remote sensing method can be used to measure forest biomass up to 500 tons/ha with 10% error [2].

Although Sentinel-1 has not systematic acquisition for tomography, thanks to the 6-day repeat cycle which minimizes the temporal decorrelation effects, the tomographic application in urban, i.e. 3D city reconstruction, can be feasible as it happened with Cosmos SkyMed and TerraSAR data. However, the Terrain Observation with Progressive Scan (TOPS) mode of Sentinel-1 introduces an additional quadratic phase term in the flight (azimuth) direction. In case of a small mis-registration error between a pair of images, this residual term leads to a phase jump the interferometric phase. Due to this limitation, there is no toolbox capable (even with ESA SNAP toolbox with version in October 2015) for TOPS tomography processing.  

In this paper, we present a software platform, which supports the entire processing from SAR, Interferometry SAR, to Tomography (so called TomoSAR). The TomoSAR toolkit is arranged in packages, each dealing with a specific aspect of the processing for ground-based, airborne and space-borne SAR systems. The TomoSAR toolkit is designed to push the services of Irstea GEOSUD and the THEIA Land Data Center. For external, it opens to provide service solutions under collaborations.

The tomographic demonstrations will be shown with forest scenarios (Paracou and Nourages tropical forests) in BIOMASS mission and with urban scenarios (Montpellier City) in TOPS Sentinel-1 data.

Finally, since the TomoSAR toolkit is not limited in tomography, it has full capability of SAR and interferometry processing. We present other thematic applications such as glacier ice velocity monitoring in Greenland, volcano in Réunion and ground subsidence estimating in Ho Chi Minh City with TOPS Sentinel-1 data.

[1] D. Ho Tong Minh, T. Le Toan, F. Rocca, S. Tebaldini, M. Mariotti d’Alessandro, and L.Villard, “Relating P-band SAR tomography to tropical forest biomass”, IEEE Transactions on Geoscience and Remote Sensing, Vol. 52, No. 2, pp. 967-979, Feb. 2014.

[2] D. Ho Tong Minh, S. Tebaldini, F. Rocca, T. Le Toan, L. Villard, and P. Dubois-Fernandez, “Capabilities of BIOMASS Tomography for Investigating Tropical Forests,” Geoscience and Remote Sensing, IEEE Transactions on, vol.53, no.2, pp.965-975, Feb. 2015.

In the present work, we carry out a first regional validation of the Sentinel-3 (S-3) STM L2 SAR (Synthetic Aperture Radar) and PLRM (Pseudo-LRM) data products against an independent implementation of the SAR and PLRM processing chain.

The L2 altimetric geophysical parameters to be validated are the sea surface height above the ellipsoid (SSH), the sea level anomaly (SLA), the significant wave height (SWH) and the wind speed (U10), all estimated at 20 Hz. Two areas are selected to cover different tide dynamics characteristics: German Bight (with large tidal amplitudes) and Mediterranean Sea (with small tidal amplitudes).

The work is structured in two parts.

In the first part, we will process independently the Sentinel-3 SAR data products from L0/L1a until L2 using the same processing baseline as defined in the S-3 STM Payload Ground Segment (PDGS).

The main objective is to reproduce the CAL/VAL results obtained from the nominal S-3 STM PDGS L2 data products and, in case of inconsistencies with these results, to investigate possible anomalies in the S-3 STM PDGS implementation for level L1b & L2 and for mode SAR & PLRM.

In the second part, we will process the Sentinel-3 data products using a different processing baseline than used above, which include a tailoring of the Delay/Doppler Processing and SAR Waveform Retracking. In this enhanced processing baseline, we will introduce new options, both in the L1b and in the L2 SAR chain, as zero-padding and Hamming window in coastal zone and a double extension of the radar range window (for L1b chain) and a Delay-Doppler Stack Masking computed using the exact slant range correction (for L2 chain). We expect in this way to identify the positive impact of these changes in the S-3 data product quality.

TU Darmstadt will provide the PLRM L2 data (built from S-3 STM L1a), the SAR L2 data will be extracted from the ESRIN GPOD Sentinel-3 service. Finally the calibration and validation (CAL/VAL) sites will be the altimeter Calibration Site in Corsica (Ajaccio) and the network of tide gauges and buoys in the German Bight. This work is part of the GB_S3CVAL activities proposed by TU Darmstadt as part of the Sentinel 3 Validation Team (S3VT).

The accuracy of InSAR derived displacements is strongly affected by spatio-temporal variations of tropospheric water vapour, which can cause errors comparable in magnitude to those associated with crustal deformation. Global, real time and reliable estimates of tropospheric delays are thus desirable for monitoring geohazards. Most of previous studies rely on remote sensing water vapor measurements provided by GPS and/or satellite radiometers. However the application of those methods is strongly limited by their data availability (e.g. lack of a dense network for GPS, or the presence of clouds for satellite radiometers). Numerical weather models hence offer potential for overcoming major limitations of traditional methods.

In this work we demonstrate the use of operational NWM output for tropospheric correction of Sentinel-1 data and quantify the skills of different NWMs and different spatial resolution based on three case studies. We compare tropospheric delays estimated from two global models and one regional model which differ both in the spatial resolution and in their dynamics parameterizations. More specifically we analysed outputs from the operational high resolution ECMWF (HRES-ECMWF, ~16 km), available near real-time and the archived reanalysis products from ERA-Interim (~75 km) which shares most of the parameterizations of HRES-ECMWF and mostly used in the geophysical community. In order to test the sensitivity on the spatial resolution, we also ran the Weather Research and Forecasting model (WRF)  at 2 and 5 km resolutions, in a nested configuration with parent and nested domain at 25 km and 5 km resolutions respectively, with meteorological lateral boundary conditions driven from the NCEP Final Analysis.

We quantify model skills based on three case studies representative of different geophysical and meteorological conditions: the Canaries, a region in the Ethiopian rift valley and in west Nepal affected by the 25^th April 2015 Mw 7.8 earthquake. In all cases HRES-ECMWF provides an improvement in the satellite retrieval by reducing tropospheric errors more than the commonly used ERA-Interim. The high resolution WRF simulations provide more accurate results over smaller and geographically complex domains (e.g. over the island of Tenerife), although its operational application is limited by the computational facilities and expertise required.

For the first time, this work demonstrates the value added by using the high resolution ECMWF products not only for reducing tropospheric errors but also for the potential for near real-time processing of Sentinel-1 data. A toolbox for applying tropospheric corrections based on HRES-ECMWF is planned to be released after the 2016 ESA Living Planet Symposium.

Urbanized areas are very complex environments that are continuously evolving and need technological solutions for enabling effective monitoring of resources, especially in case of anthropogenic or natural disasters. The evaluation of time-variable natural and hazard related processes at different spatial scales are fundamental to assess the condition of the whole observed area and its potential impact on built structures, especially when considering buildings located within historical city centres, along steep slope relieves, and more generally within areas susceptible to surface deformations. In order to address urban multi-risk assessments, a holistic approach is suitable. It includes strategies, methodologies and new tools able to work either together or independently from each other with the final aim to provide user-friendly geo-awareness maps and synthetic descriptors of observed areas and structures. In this context, radar sensors have demonstrated their capability to address a broad range of applications, such as surface topography, land monitoring, change detection applications and early warning. These operational benefits have been obtained, in particular, thanks to the radar interferometry techniques, able to estimate deformation processes with space-borne and ground-based instruments. The exploitation of such systems and techniques within a multi-sensor and multi-platform approach could be very interesting for supporting and integrating classical remote sensing analyses as well as for overcoming their constraints, in terms of resolution, coverage, accuracy and repeatability. In fact, some operational benefits are effective, such as the provision of surface displacements along complementary directions at different spatial scales, the multi-temporal monitoring of urban environments and structures, as well as the monitoring of remote areas together with the analysis of both structural and topographic settings for relevant buildings. As a result, the integration of both interferometric remote and proximal sensing, based on radar technologies, represents an interesting topic to be investigated for monitoring built structures and their surrounding environment.

In this framework, a non-invasive radar-based methodology is here proposed, aiming at assessing surface deformation processes and structural vibrations of buildings to support the diagnosis of urban areas for multi-risk assessment and mitigation purposes. The approach combines Synthetic Aperture Radar (SAR) measurements provided by space-borne COSMO-SkyMed satellites, Ground-Based (GB) SAR and GB Real Aperture Radar (RAR) data by means of interferometric processing techniques. It allows providing maps of long-term and regional-scale surface velocity, short-term ground displacement map of local-scale areas, and near real time vibrating monitoring of structures at building scale, respectively.

The study area is the city of Cosenza, located in a seismic prone area of south Italy. In particular, the cultural heritage of S. Augustine Compound, within the historical city centre, is the focus of the local and building scale analyses.

On the one side, the GBRAR measurements allow calculating the power spectral density (PSD) and the cross-power spectral density (CPSD) for some structural elements, identifying relevant peaks of resonating frequencies on the Compound walls. To confirm the results, the frequencies retrieved through interferometric GBRAR analysis have been compared with those obtained from ambient vibration tests performed through velocimeters.

On the other side, the deformation and the velocity maps, obtained from GBSAR and COSMO-SkyMed SAR data respectively, allow detecting possible surface movements of the study area on the almost orthogonal view of ground- and space-based sensors.

The experimental campaigns have been carried out thanks to the activities of the Italian project PON MASSIMO[i] (namely Monitoraggio in Area Sismica di SIstemi MOnumentali) for the monitoring of cultural heritages in seismic area. Some meaningful results are presented in this work.

[i] The present work is supported and funded by Ministry of Education, University and Research (MIUR) under the project PON01-02710 "MASSIMO" - "Monitoraggio in Area Sismica di SIstemi MOnumentali".

With the advent of diverse satellite altimeters and variant measuring techniques, it has become mature, in the scientific community, that an absolute reference Cal/Val site is maintained to regularly define, control and evaluate the responses of any altimetric system. This fiducial reference site for satellite altimeters will consistently and reliably determine (a) absolute altimeter biases and their drifts; (b) relative bias among diverse missions; but also (c) continuously and independently connect different missions, on a common and reliable reference. Results from this fiducial reference site should be based on historic Cal/Val site records, and would be the yardstick for building up capacity for monitoring the climate change records. This ground facility will be capable of defining and assessing any satellite altimeter measurements to known, controlled and absolute reference signals with different techniques, processes and instrumentation.

 The objective of this presentation is to set the ground for the establishment of a Fiducial Reference Site for ESA satellite altimetry in Gavdos and West Crete, Greece. This research infrastructure will aim at monitoring and controlling, in an absolute sense, satellite altimetry measurements and results by (1) continuously keeping track of their quality, biases, errors and drifts; (2) by establishing an absolute reference of altimetry on a common and reliable standard for settling relations among different missions; and (3) by developing reference measures for satellite altimetry, on diverse procedures and instrumentation, as well as on ascending and descending orbits, at the same location and settings.

The INTA Remote Sensing Laboratory has implemented a tool for the direct geolocation of satellite images (i.e., using orbital data and a rigorous sensor model and without the need of ground control points or RPC models). The core of the tool is a C code based on the "Earth Observation Mission CFI SW" from ESA, a set of C libraries that include functions for solving orbital and pointing problems.

The tool accepts different types of inputs for satellite attitude, like euler angles, quaternions or a default attitude model. Satellite position can be provided either in ECEF or ECI coordinates. The sensor model, or line of sight of each individual detector, is imported form an external file or is generated by the tool from camera parameters (focal length, pixel size, principal point coordinates and camera tilt).

The tool has been internally validated by different means. This validation shows that the tool is suitable for georeferencing images from high spatial resolution missions. As part of the validation efforts, a set of routines for simulating orbital info from LEO missions has been produced.

This software is intented as a R+D tool; it is not suited for an operational, production environment, but for offline processing of EO data, for analysis of geolocation parameters (related to orbital info, DEM or sensor model) and for generating simulated geolocated data for future missions.

 The tool has been already tailored for georeferencing images from the forthcoming Spanish Earth Observation mission SEOSat/Ingenio, and for the camera APIS onboard the INTA cubesat OPTOS. The specific issues of these two cases are presented and discussed in this communication. The next step is to configure it for the geolocation of Sentinel 2 L1b images. In this way, we could have an alternative method for generating Sentinel 2 geolocation information, which could be useful for some non-standard  processing procedures.

Light Rare Earth Elements (LREEs) are five chemical elements of the Lanthanoids series, whose physical and chemical properties offer a wide range of applications in the components of modern technical devices. Due to the increased global demand of those devices, the exploration of new deposits and the re-evaluation of existing resources is an ongoing process. This study focusses on the contribution of imaging spectroscopy to the classic REE exploration through a hyperspectral mapping of Rare Earth Oxides. Therefore a recently published REE detection chain (Boesche et al., Remote Sensing, 2015) was adapted and applied to a variety of today’s existing airborne and spaceborne sensors (Aviris, AvirisNG, ISS HICO, EO-1 Hyperion and simulated EnMAP on the basis of Aviris and AvirisNG). The main steps of the modified REE detection are 1) Pre-processing of the delivered image data, 2) Signal enhancement of neodymium- and samarium-related spectral characteristics, and 3) Image classification based on a correlation with reference spectra of a synthetically produced reference material. The results show that  a detection of REE enriched material on the surface using current sensors is possible, and  that a higher spectral resolution and a higher Signal-to-noise ratio of future sensors will provide a higher accuracy of the the mapping of the REE enriched zones. Hence, imaging spectroscopy benefits the classical REE exploration using a direct mapping of REE indicative pixel. In order to highlight enriched zones with certain spots all sensors were suitable, while a full mapping of enriched zone requires a high image signal-to-noise and high spectral resolution.

Image data sharpening is a challenging field of remote sensing science which became more popular after recent emergence of high spatial resolution satellite image sensors. Those satellites usually provide one broad-band panchromatic (PAN) band with a high spatial resolution and multispectral images with lower spatial resolution. The general approach is to use the PAN band to enhance spatial resolution of the multispectral images using sharpening algorithms which are using different  approaches to inject the spatial detail of the PAN image to the higher spectral resolution image data. Besides this approach, it is also possible to sharpen lower spatial-resolution data with any higher spatial-resolution multispectral data, not necessarily with the PAN band.

An optimal sharpening algorithm would preserve both good spatial and spectral information. Both spatial and spectral performance at a time of most of the newly developed methods is still not satisfactory, thus the research in pansharpening algorithms is still a developing field of remote sensing science. There are several traditional algorithms which have been developed and successfully used until present time, such as PCA (Principal component analysis), Gram-Schmidt, Ehlers fusion or Brovey transform, and also other advanced algorithms, such as wavelet transform or contourlet transform. The above mentioned group of algorithms is mostly valued for its easy implementation, broad availability in most popular software and a low computational time. Main observed problems are deviations in the spatial accuracy and in spectral values of the sharpened image (Witharana, et al., 2013). Those are still barriers in classification processes which use the sharpened image. Especially when dealing with geology or mineralogy, precise spectral information is needed.

The main objective of our work was to develop new approaches of the sharpening routines applied on the higher spectral resolution image data which are of a different sensor origin. This is also one of the issues in the development of optimal sharpening algorithm and will be very useful for sharpening a new generation superspectral and even hyperspectral satellite data such as Sentinel-2 and EnMap. From those reasons we adopted for our study the PCA algorithm and Gram-Schmidt principles which we further developed. Besides that we have also tested our own algorithms which were based on the Crisp sharpening algorithm (Winter et al., 2007). Both approaches led to the development of modified sharpening approaches which improved the transfer of the spectral information between the source image and the sharpened image. This is of a special interest of the geological case studies where the precise geological information is demanded. Results were validated by several traditional validation metrics, such as ERGAS, SAM, RASE or RMSE. In consequence, an approach is proposed that provides convincing results for proximal sensing as for sharpening EnMAP with Sentinel-2.

In situ calibration insures regular and long-term control of the altimeter sea surface height (SSH) time series through comparisons with independent records. Usually, in situ calibration of altimeter SSH is undertaken at specific CalVal sites through the direct comparison of the altimeter data with in situ data.

However, Noveltis has developed a regional CalVal technique, which aims at increasing the number and the repeatability of the altimeter bias assessments by determining the altimeter bias both on overflying passes and on satellite passes located far away from the calibration site. In principle this extends the single site approach to a wider regional scale, thus reinforcing the link between the local and the global CalVal analyses. It also provides a means to maintain a calibration time series through periods of data-outage at a specific dedicated calibration site.

The regional method was initially developed at the Corsican calibration sites of Senetosa and Ajaccio. It was then successfully implemented at the Californian site of Harvest and at the Australian site of Bass Strait, in close collaboration with JPL and the University of Tasmania, respectively. The method was used to compute the altimeter biases of Jason-1, Jason-2, Envisat and SARAL/AltiKa at all these sites.

These recent studies gave the first Envisat and SARAL/AltiKa absolute bias estimates at non-dedicated sites using the same method, and showed high consistency with the analyses of the global CalVal teams and the work of the in situ CalVal teams. These results highlight the numerous advantages of this technique for monitoring missions on any orbits, including Sentinel-3 and CryoSat-2.

Spatial and temporal dynamics of turbidity and suspended sediment can provide essential information about the function and water quality of an estuarine system. The Gironde estuary (France) is characterized by severe changes in turbidity concentration along its entire length, going from the extremely turbid waters (> 4000 NTU) in the upstream region (where the Garonne and the Dordogne rivers join) to the low turbidity (> 5 NTU) coastal waters of the Bay of Biscay. This study presents the capabilities of three sensors OLI (Landsat 8), MODIS (Aqua/Terra) and SEVIRI (MSG) to study the suspended particulate material (SPM) and turbidity dynamics in relation to river discharges in a complex estuarine system such as the Gironde. A description of the most suitable atmospheric correction and best performing algorithms to derive SPM concentration and turbidity for the different water bodies existing in the study region are provided for each sensor. OLI images offer a high spatial resolution (30 m) that combined with the moderate spatial and temporal resolution of MODIS (250 m, 2 images per day) and the high temporal resolution of SEVIRI (3.5 km at the Gironde mouth, 1 image every 15 minutes), show the potential of these three sensors to monitor this type of system. In situ spectral measurements acquired during several surveys across the Gironde estuary between 2012 and 2014, were used to validate the atmospheric corrections and to develop regional empirical SPM algorithms. Then, satellite-derived turbidity and SPM maps were validated using automated in situ turbidity measurements provided by the Magest monitoring network stations, and established turbidity-SPM relationships. Additionally, a MODIS algorithm was adapted to select (1) the most suitable atmospheric correction, using near-infrared (NIR) and shortwave infrared (SWIR) bands, and (2) the most suitable SPM empirical model in relation to the turbidity levels present in the estuary, the river mouth and the plume area. Finally, the capacity of SEVIRI to follow the daily plume dynamics coupled with MODIS images is presented.

This multi-sensor approach will be applied to the new ESA Sentinel-2 and 3 missions framed within the FP7-HIGHROC project, offering the possibility to study estuarine and coastal dynamics with both high temporal and high spatial resolution.

Almost continuous worldwide coverage is being achieved from high resolution SAR sensors such as Radarsat and Sentinel-1 and for regions of user interest from very high-resolution radar satellites like TerraSAR-X, COSMO-SkyMed, RADASAT-2. Meanwhile, the most popular available SAR data are C-band (ERS-1/2, ENVISAT, RADARSAT-1/2), S-band(HJ-C), X-band (TerraSAR-X, CosmoSkyMed) and L-band(JERS-1, ALOS) data. In order to map 4 dimensional ground deformation over an extended period of time and particularly for multidimensional time-series analysis, it is desirable to utilise all available InSAR data at the same time, by exploiting the advantages of different space-borne SAR data with varying acquisition geometries, and acquisition parameters (e.g. azimuth and incidence angles, spatial and temporal resolutions, polarization) in order to compensate for the limitations of a particular data set (e.g. low temporal resolution) and in order to provide uninterrupted coverage. Multiple-Aperture SAR Interferometry (MAI) and Offset-Tracking methods have recently been developed to estimate the azimuth surface displacement, which can be integrated with conventional dInSAR to obtain 3D surface displacement. It is well known that high accuracy azimuth displacement measurements can be achieved by MAI when comparing with Offset-Tracking in high coherence area, but MAI cannot get acceptable results in low coherence areas. By contrast, although being less accurate than the MAI interferometric phase measurement, the Offset-Tracking measurement can more robustly overcome temporal decorrelation and does not require any noise-sensitive phase unwrapping process. In this paper, an improved Small Baseline Subset (SBAS) methodology is presented for integration of multiple InSAR data sets, and a variance component estimation (VCE) algorithm is applied to weight the dInSAR, the Offset-Tracking measurements and MAI measurements to yield an optimal 4D(3D and time dimensional) deformation. The accuracy of the dInSAR technique is strongly affected by orbital error, tropospheric and ionospheric phase artifacts, residual topography error, etc. Particularly for long-wave SAR systems such as L- and P-band SAR, ionospheric effects cause severe Ionospheric phase screens, local or global decorrelation, azimuth streaking and long wavelength phase distortion similar to orbital ramp errors. Effective detection and correction of ionospheric phase distortion from dInSAR images are necessary to measure and accurately interpret surface displacement. In this paper, early results from a joint correction of both ionospheric noise and orbital error will be shown. Study sites include landslide areas in the UK & Sichuan, China by using C-band ENVISAT, C-band Sentinel-1A, X-band TerraSAR-X and L-band ALOS JERS-1 Data.This work is partially supported by the CSC and UCL through a PhD studentship at UCL-MSSL.

Global urbanization is one of the great challenges of the 21^st century. A better understanding of urban environments, their spatio-temporal processes and related ecological implications is essential to promote the sustainability of cities. Remote sensing has become a valuable Earth observation tool for urban environmental research and bears the potential to link research, policy, and planning for solving pressing urban and global change issues.

Since the 1970s’ various multispectral sensors have acquired optical images over urban areas. These datasets have been used to produce bio-physical or thematic urban land cover maps, to assess urban growth patterns, or to quantify urban environmental conditions. The emergence of new Earth observation missions appears promising to further enhance urban environmental research by means of remote sensing. On the one hand, data acquired by the novel Sentinel-2 mission complement images from established multispectral sensors such as Landsat. Due the improved spatial resolution between 10-20 m and the increased spectral resolution in the near-infrared region, Sentinel-2 constitutes suitable source for frequent and global urban land cover mapping assessments. On the other hand, images acquired by the upcoming hyperspectral mission EnMAP (Environmental Mapping and Analysis Program) will increase the availability of high quality imaging spectrometer data. Despite the coarser resolution of 30 m and the coverage for specific regions only, the high spectral information content of EnMAP will help to improve mapping and to gain better insights into process understanding.

Based on simulated data along an urban gradient in Berlin, Germany, and empirical regression modeling with support vector machines, this work presents opportunities and challenges of quantitative urban mapping with Sentinel-2 and EnMAP data. Mapping considered the bio-physical composition of urban land cover according to the well-recognized vegetation-impervious-soil (VIS) scheme. The two data sets were simulated from airborne HyMap (Hyperspectral Mapper) data. The trade-off between spatial and spectral detail is mirrored in the results, where VIS components – extended by separation of tree and low vegetation – can be accurately mapped with both data sets (MAE range between 10 and 20%). This shows, that the spectral advantage of the EnMAP sensor is outweighed partly by the increased spatial resolution of Sentinel-2. With original Sentinel-2a data available, these mapping are reproduced. First attempts repeat the positive findings from simulated data. In our contribution, the potential and the challenges related to mapping urban areas with Sentinel-2 data, e.g. for mapping built-up areas from large view angles, will be further elucidated based on selected examples.

We present new products for land surface reflectance and global atmospheric aerosol and for the Copernicus Sentinel-3 mission. The land surface product aims to provide high accuracy surface reflectance intended to facilitate global services such as mapping of biophysical parameters such as albedo, leaf area index and absorbed radiation, and land cover classification. The retrieval has been developed for application to Sentinel-3 to make synergistic use of the two sensors, OLCI and SLSTR, and will provide co-registered surface reflectance spectrum from both, corrected for the effects of atmospheric gases and aerosol scattering. The algorithm has been implemented on the ESA BEAM system and tested on MERIS and AATSR data and results show improved estimation of aerosol properties compared to single-instrument retrievals. The method for global aerosol retrieval using SLSTR is based on the dual angle algorithm developed for the (A)ATSR instrument series, to allow retrieval of aerosol properties and their uncertainties over both land and ocean. This has been developed under the ESA Aerosol Climate Change Initiative for generation of the first 17 year dataset from global aerosol retrieval from ERS-2 and ENVISAT (1995-2012), and evaluated using the global AERONET sun photometer network. First results from Sentinel-3 testing will be presented depending on successful launch and initial processing.

The Fourth (A)ATSR data reprocessing will begin in summer 2016. Its main objective is to generate (A)ATSR Level 1B data products in the same format as SLSTR products from Sentinel-3. In this way, users can easily access the 20-year dataset from the ERS and ENVISAT (A)ATSR missions and carry the analysis forward into the Sentinel/SLSTR era.

This established dataset generated in the new format will now include uncertainty estimates, Bayesian and probabilistic methods for cloud masking, and ECMWF reanalysis fields. Dataset improvements will include enhanced coverage (due to data recovery), the use of orthogeolocation, and improved surface classification via use of the Sentinel-3 land masking dataset.

The ESA projects SST CCI and GlobTemperature will generate the 4th reprocessing Level 2 SST and LST datasets, respectively, making use of up-to-date expert knowledge.

The paper will review the expected improvements to the L1B dataset and itemise the data that has been recovered.

The DUACS system has produced, up to now, as part of the CNES/SALP and the MyOcean projects, high quality multimission altimetry Sea Level products for oceanographic applications, climate forecasting centers, geophysic and biology communities... These products consist in directly usable and easy to manipulate Level 3 (along-track cross-calibrated SLA) and Level 4 products (multiple sensors merged as maps or time series) and are available in global and regional version (Mediterranean Sea, Arctic, European Shelves …).

With the integration of HY-2A data in 2014, the Near Real Time system now merges data from 4 satellites. In parallel of the constellation management, on April 2014, the entire catalogue was significantly upgraded with impacts on scientific content and format, improving the quality and accessibility of the products. Besides, a full reprocessing of the whole altimetry time series has been performed allowing us to make available a set of 21 years of homogeneous along–track and map products.

In 2016, we are now starting a new step. The operational production of the along track and Sea Level maps is now generated as part as the Copernicus Marine Environment and Monitoring Service (CMEMS), a European project launched in May. Besides, new satellites, Jason-3 and Sentinel-3, will be launched, and will contribute to the robustness of this service. Moreover, the processing of the sea level can be improved to fully exploit this increase of real time altimetry observations, combined with the advanced technology of the new sensors. This paper will present the latest development and the main perspectives for the DUACS Sea level products in the coming years.

Russian satellite «Resurs-P» has been operated since 2014. Hyperspectral equipment has next specifications: spectral range from 0.4 to 1.1 nm, 96 spectral bands and spatial resolution of 30 m. At present the complex of researches to assess the information capacities of this satellite is performed.

Next information needed for data acquisition to evaluate the information capacity of satellite hyperspectral data and choose an optimal parameters:

-          field spectral measurements,

-          meteorological observations,

-          synchronous airborne data acquisition,

-          modelling of interval and monochromatic contrasts of scene elements under various conditions of survey and atmosphere.

Were used the objects with known reference spectra that were selected and measured during the satellite data acquisition from validation sites. Adjustment coefficients for radiometric calibration of data were determined from these measurements. Evaluation of the reliability of the spectral calibration was carried out by analyzing changes in the known absorption bands of atmospheric gases and aerosols at certain wavelengths based on the spectral resolution of hyperspectral equipment.

To solve practical tasks on the basis of hyperspectral data the original software is used. And for each task a unique set of features is generated and the most efficient method of detection or classification is selected. Furthermore, to determine the spectrum ranges for the most effective solving of the task, methods of data statistical estimation are used, such as gradient method, spatially scalable filtering method, and others.

As an example of thematic task, we estimated the depth of the water body, the Gulf of Finland, from “Resurs-P” satellite hyperspectral image. The depth from 10 to 20 m was determined based on approaches for identification of the bottoms’s type and assessment of its reflection characteristics, analysis of hydrooptical indicators of water masses, and classification of mineral and organic substances influencing the absorption and scattering of radiation in the water.

Most informative spectral bands of hyperspectral data were used to identify the light scattering and absorption (425, 545 and 605 nm). Semi-analytic models that reflected the relationship between the qualitative characteristics of water bodies, water actinometrical characteristics and performance of hydro-spectral absorption and scattering coefficients, were applied to verify estimated depths. Results of verification performed by in-situ data have shown that the accuracy of "Resurs-P" hyperspectral measurements on depths is at least 90%.

In general, conducted researches confirmed the high information capacities of the "Resurs-P" hyperspectral data and the appropriateness of their wide use in solving tasks of environmental analysis.

Delay/Doppler altimetry is a new technique, inherited from the SAR imaging instruments and adapted to nadir altimetry by Raney 1998. This technique aims to improve the along-track resolution by exploiting the altimetry signal Doppler frequency due to the instrument movement and to reduce the estimation noise by increasing the number of looks available for each surface Doppler resolution cell with respect to conventional altimeters. On CryoSat-2 mission this technique (called the SAR mode) is dedicated to sea ice areas (and ice caps margins with an additional interferometry processing). On Sentinel-3, the same technique is implemented and will be operated full time over the world for the first time.

CLS was in charge of the specifications and scientific validation of the SRAL L1 and L2 algorithms that have been implemented in Sentinel-3 operational Instrument Processing Facility (IPF). The validation of the processing has been performed thanks to the development of the SRAL L1 Ground Prototype Processor (GPP) and the end-to-end SRAL System Performance Simulator (SPS), including a Level 2 prototype processor allowing the reproduction of all the instrument modes using a realistic geophysical scene and real instrument characteristics.

In the frame of Sentinel-3 Mission Performance Center (MPC), CLS is in charge of the Level 1 and Level 2 SRAL products validation and processing parametrisation as soon as Sentinel-3 products are available. This presentation will be mainly dedicated to the SAR mode Level 1 and retracking processing. Indeed, SAR mode will be operated everywhere on the globe from cycle 1 onwards, during the commissioning phase

The first set of configuration parameters, available in the Level 1 and Level 2 operational processing, has been defined using simulated data or inherited from previous missions. Using Sentinel-3 data, these parameters will be checked to review their adequacy and completeness and to adapt them to the in-flight measurement. The PLRM (standard LRM processing) parameters will be of high importance to serve as a reference for the validation of the new SAR mode.

In this presentation, we will provide the main results of the analysis and validation of the first observations of Sentinel-3 data.

INTRODUCTION

The derivation of phase spatial gradients from SAR interferograms has been often exploited in different SAR interferometry applications as a way to bypass the phase unwrapping [1][2][3]. The aim of this work is to analyze how this type of measurement can be used in terms of geophysical modeling. In order to do that we first collect and present the tools to estimate and handle the deformations gradients from the complex interferograms. Then, in order to assess the achievable precisions, a study of principal noise sources is carried out considering the decorrelation of the interferometric signal and the delays coming from the troposphere. This information will further help us to combine the phase gradients meaningfully with surface displacement measurements from the unwrapped interferometric phase.

Provided we have a surface changing process covered from different Line-of-sights (LOS) possible uses of the phase gradients can not only be the analyses of earthquake source models or volcanic deformation sources, but also the reconstruction of strain and rotation tensors to describe, for example, the glaciers flows.

PHASE GRADIENT ERROR ANALYSIS

In order to properly exploit the phase gradient information it is necessary to describe statistically the errors that affect the measurements. The main sources of error for the interferometric signals are the phase decorrelation and the phase delays coming from the propagation in the troposphere.

From the signal processing point of view the estimation of phase gradients from a partially decorrelated  complex interferogram can be seen as the frequency estimation of a noisy complex sinusoid. This is a known problem and estimators and performance varying the different signal-to-noise ratio conditions have been studied in literature. The theoretical precision of the frequency estimation has been derived in [4]. The additive phase delays generated by the propagation of the signal trough the troposphere generate a spatial phase pattern in SAR interferograms often named Atmospheric Phase Screen (APS). Such pattern is very variable according to the season, time of the day and atmospheric conditions. However its behavior can statistically be well described by a covariance function if we assume the stationariety of the process [5]. In general it is possible to derive the covariance functions of the gradients starting from the covariance function of the absolute phase .  As expected the spatial correlation will be much lower compared to the absolute phase, but nevertheless it will be direction-dependent and with no correlation in the differentiation direction and maximum correlation in its orthogonal direction. Examples for two-parameter exponential models (power, correlation length) are given in Fig.1.

EXAMPLE: INVERSION OF EARTHQUAKE FAULT PARAMETERS

The combination of surface displacement and surface displacement gradient modelling has been tested on the interferogram of the 2012 Ahar earthquake doublet (NW Iran), where two magnitude > 6 earthquakes have occurred only 11 minutes apart. For this sequence only the mechanism of the first rupture can be constrained well seismically as the seismic waves of the second earthquake are mixing with late seismic phases of the first.  Coseismic coverage of both earthquakes exists only for an ascending RADARSAT-2 interferogram.

The use of the gradients has been also tested for the fault parameter inversion of L’Aquila Earthquake. Two Envisat ASAR interferometric pair ( Ascending and Descending ) have been used to carry out the inversion.

EXAMPLE: STRAIN AND ROTATION FROM DIFFERENT LINE OF SIGHTS

Another possible application for the interferometric gradients can be related to the estimation of strain and rotation tensors. These types of measurements are useful in glaciology for the study of glaciers flows. In order to give a demonstration a simulated example has been computed, results are shown in Figure 4.

Over the last decade Hyperspectral Imaging (HSI) systems created an enormous outburst in the field of earth observation. Multiple instrument on-board imaging systems are currently available, providing a large amount of hyperspectral imagery for various applications, such as precision agriculture, geology and oceanography. Despite the important advantages hyperspectral imaging systems demonstrate, HSI acquisition and processing stages usually introduce multiple constraints. Slow acquisition time, limited spectral and spatial resolution, low dynamic range, and restricted field of view, are some of the limitations that hyperspectral sensors experience, and require further investigation.

Enhancing the spectral resolution, i.e., number of distinct spectral bands, of the acquired images is critical for both visualization and subsequent analysis, including spectral unmixing, pixel classification and region clustering. High spectral resolution imaging systems are able to capture a huge amount of data, including the 2D spatial and the 1D spectral variations of an input scene over time. Unfortunately, various factors can lead to the introduction of imaging constraints such as the case of Spectrally Resolvable Detector Arrays systems that directly acquire the entire 3D data-cube through a combination of spectral filters and detector elements. Despite the dramatic reduction these systems exhibit with respect to acquisition time, such designs also lead to a reduction of the spectral resolution by associating each pixel with a single spectral band.

In order to overcome the aforementioned limitations, we propose a novel spectral resolution enhancement framework of low spectral resolution imagery, based on the Sparse Representations (SR) framework. Unlike state-of-the-art hyperspectral super-resolution methods that utilize inherent correlations to obtain high spatial resolution images, the proposed algorithm aims at enhancing the spectral content of the imagery. This goal is achieved by introducing the assumption that each high spectral resolution “hyper-pixel” can be estimated from its low resolution version by identifying a sparse representation encoding that directly generates the high-spectral resolution output.

The notion of sparsity has revolutionized modern signal processing and machine learning, and has lead to very impressive results in a variety of imaging problems, including deblurring, denoising, etc. In this work, we enforce the sparsity constraint through learning a joint sparse coding dictionary from multiple low and high spectral resolution training image pairs. To the best of our knowledge, this is the first work that proposes a spectral super-resolution technique for hyperspectral images. A crucial requirement for the recovery process is the proper generation of the two dictionaries, in order to simultaneously sparsify the low and high resolution hyperspectral data. The joint learning of the two dictionaries encodes the assumption that both the high and the low spectral resolution hyperspectral pixels can share the same sparse code, since they adhere to similar statistical characteristics under different spectral-resolution conditions. For this purpose, multiple pixels are sampled from large collections of high and their correspondent low spectral resolution training hyperspectral images.

The proposed inverse spectral resolution enhancement problem recovers high spectral information, capitalizing on the sparse representations framework as prior-knowledge, effectively encoding the relationships between high and low spectral representations. Additionally, the proposed scheme can be extended to handle large ranges of low-to-high resolution enhancements by efficient modifications of the joint dictionary learning process, as well as offering the capability of addressing additional sources of HSI image degradation such as blurring and noise.  

More than six years have been spent since the launch, on November 2, 2009, of the European Space Agency’s (ESA) Soil Moisture and Ocean Salinity (SMOS) satellite carrying a microwave synthetic aperture radiometer working at 1.4 GHz. The aim of the mission is to provide Sea Surface Salinity and Soil Moisture observations, with a spatial resolution of 30-50 km, and an accuracy suited for climate studies.

Startinf from L1brightness temperature (TB) observations, experimental Sea Surface Salinity (SSS) and Soil Moisture (SM) maps are being developed and progressively distributed at the Barcelona Expert Center (BEC). Data are distributed in NetCDF format using THREDDS and maps are served through a Web Map Service (ncWMS), both at the BEC distribution data website (http://cp34-bec.cmima.csic.es/).

BEC ocean products are experiencing a great evolution thanks to the introduction of  new key algorithms at different levels of the processing chain. At level 1, nominal TBs are replaced by the nodal sampled ones.  Nodal sampling is a new image reconstruction technique that reduces tsidelobes and ripples that associated to abrupt changes in TB (see  [1,2]).  At level 2, systematic biases are removed by referring them to SMOS-derived climatologies. As a result, the new maps present a sharp reduction of the land-sea contamination. At level 3, the correlations of the optimal interpolation algorithm have been improved such that the resulting maps show a more coherent spatial structure at all the scales. Finally, at level 4, the spatial and temporal resolutions have been increased by means of multifractal synergy techniques, which allow generating global daily SMOS SSS maps (see [3,4]).  Besides, all these improvements allow the production of a new SSS product fitted to  polar  regions. Surface density products are now also available in our website.

For land applications, high resolution maps (1km at Iberian Peninsula) are produced using reliable non-remote sensing auxiliary parameters so that now these products are  weather-independent and with full spatial coverage. That allows the generation of a new product focused on the fire forest risk prediction.

Finally, new cryospheric sea ice concentration products have been developed following the methodology described in [5].

In summary, the new generation of BEC products, to be distributed soon, represent a real leap forward in quality that will enable a new generation of geophysical studies.

  [1] V. González-Gambau, A. Turiel, J. Martinez, E. Olmedo, and I. Corbella, "A novel reconstruction algorithm for the improvement of SMOS brightness temperatures," in Microwave Radiometry and Remote Sensing of the Environment (MicroRad), 2014 13th Specialist Meeting on , vol., no., pp.124-127, 24-27 March 2014, doi: 10.1109/MicroRad.2014.6878922.

[2] V. González-Gambau, E. Olmedo, A. Turiel, J. Martínez, J. Ballabrera, M. Portabella, and M. Piles,  “Enhancing SMOS brightness temperatures over the ocean using the nodal sampling image reconstruction technique,” Remote Sensing of Environment, August 2015, in review.

[3] M., Umbert, N., Hoareau, A., Turiel,  and J., Ballabrera-Poy, (2014). “New blending algorithm to synergize ocean variables: the case of smos sea surface salinity maps”, Remote Sensing of Enviroment, 146, pp. 188-200.

[4] E. Olmedo, J. Martínez, M. Umbert, N. Hoareau, M. Portabella, J. Ballabrera-Poy, A. Turiel, “Improving time and space resolution of SMOS salinity maps using multifractal fusion”, Remote Sensing of Environment, August 2015, in review.

[5]  C., Gabarro, Q., Pla, P., Elosegui, E. Slominska, A. Turiel, J. Martínez, M., Portabella, V, González-Gambau, 2015. Investigating SMOS data for sea ice concentration determintation. SMOS Science Workshop, ESAC- Madrid.

Sentinel-1A was launched in April 2014 in order to provide radar vision for Europe's Copernicus programme.  Sentinel-1A carries a 12 m-long advanced synthetic aperture radar (SAR), working at C-band. The 'radar interferometry' (InSAR) remote sensing technique combines two or more radar images over the same area to detect changes occurring between acquisitions. 

The performance of the radar interferometry depends strongly on the geometry between the positions of the satellite when the combined acquisitions were taken, namely baselines. For the Sentinel-1A mission small baselines are preferred, therefore the following orbit requirement was defined. “The reference orbit shall be maintained within an Earth-fixed orbital tube of a diameter of 100 meter-rms, at every orbital point, over any repeat cycle, during the nominal mode operation time.” But the orbital tube is just an artefact, i.e., a comparison with an ideal reference orbit, whereas the real performance is given by the resulting interferometric baselines. That is to say, the relative geometry between the actual positions of the satellite over the same area.

An exhaustive analysis of the resulting baselines over one year of data has been performed. The baselines are decomposed into parallel, perpendicular and along-track baselines. Absolute and statistical results will be presented for each case as well as its evolution with time and its dependence with latitude.

Finally the resulting baselines will be linked to the orbital tube control. It will be shown how the tube rather than a tube has an elliptical shape, being each of its axis important for different matters. While the along track axis plays a major role in the baseline results, the radial axis drives also the synchronization drift during acquisitions.

Besides the achieved baselines results of the mission, and important conclusion of the presentation is that they are not only dependent on the orbital tube size, but there are other factors related also to the semi-major axis, inclination and eccentricity control that contribute to the final results. With an orbital tube of a bit less than 155 meters RMS diameter an RMS baseline of less than 100 meters over the whole period is achieved.

The Sentinel satellites will form the majority of the observational capacity of the European Earth monitoring program, Copernicus. This program is designed to support a wide variety of applications by targeting six so-called ‘thematic areas’ – land, marine, atmosphere, climate, emergency management and security. Following on from the recent launches of Sentinel 1 A and Sentinel 2A, the next scheduled addition to this growing family of satellites, Sentinel 3A, will launch in December 2015 and will be focused on providing data for Marine and Land services. OLCI - the Ocean and Land Colour Instrument – is the Sentinel 3 instrument of interest for this project. This is a spectrometer designed to produce wide-swath, multi-channel images of the Earth’s surface.

An understanding of the quality of the data OLCI produces is vital for many applications, including science. For this reason, this project has been working to develop a model to determine the per pixel uncertainty of OLCI images. To properly estimate this a variety of contributors must considered, caused both in the instrument (e.g., noise and stray light) and during ground processing (e.g., radiometric calibration).

Efforts began with a modellisation of the signal processing, from the raw data to the output top of atmosphere (TOA) Level 1b product. From this an analytical description of the total uncertainty on a per-pixel level could be derived in terms of the different contributors encountered. Efforts could then be focused on determining each contributor in turn. This started with a modelisation and on-ground verification of the instrument’s Signal-To-Noise Ratio (SNR), smear and stray light.

In the following we describe a way to retrieve these uncertainties making use of Level 1b data as an input leading to the development of a tool which can be provided to users for the uncertainty retrieval making use of a toolbox, such as S3tbx. The main disadvantage of the approach is the inversion of the Level 1b data back to the unprocessed Level 0 data in order to feed the model. An assessment of this approach is provided making use of the on-ground test results, simulations and a verification with first inflight data.

In order to assess and control the expected Sentinel-3 Altimeter System Performance, a System Performance Simulator (SPS), and a L0/L1b Ground Processor Prototype (GPP), have been designed and developed under a TAS/ESTEC contract for each instrument of the topography payload (SRAL, MWR and GNSS). A SRAL/MWR Level-2 prototype (L2 PAD for Products and Algorithms Definition) has also been developed under an ESRIN/CNES contract. The L1b and L2 processing prototypes have been used as a reference for the operational Instrument Processing Facility (IPF) in terms of:

Algorithms specification

Products definition

Provision      of reference TDS for the PDGS.

Moreover, the SPS and GPPs (including L2 PAD) are powerful and valuable tools allowing:

Generation      of scientifically meaningful data in Sentinel-3 format

Development of new algorithms

Test      of a specific instrument parameter impact on the processing

Test      of specific geophysical effects on the performance (for example, a study      using these tools has been performed for CNES for a theoretical analysis      of the SAR mode SSB)

And many other functions.

In this presentation, the SPS and GPP components will be described along with an overview of the TDS generation capabilities (ocean, coastal and hydrology) providing several examples of simulations showing the behavior of some parameters against the over-flown surface. These TDS have been generated according to ESA-ESRIN TDS User and Science Requirement Document.

The GPP has also been adapted to process CryoSat-2 FBR data. This new tool has been used recently for the generation of new TDS in Sentinel-3 format using CryoSat-2 real data acquired in the SAR mode over ocean, sea ice, ice sheet and inland water areas. Some examples using CryoSat-2 real data will be also presented.

Finally, we will present an assessment of the SRAL L1b SAR processing performances using one month of CryoSat-2 SAR mode real data over ocean.

For many scientists, climate warming induces an increases in evaporation and precipitation leading to an acceleration of the water cycle. Soil moisture is a key parameter to study the water cycle variations.  At global scale it controls the exchange of water and heat between the land surface and the atmosphere through plant evapotranspiration. At local/regional scale is also a relevant parameter. Well knowledge of the soil moisture is of paramount importance for ecological processes, agriculture applications even more considering that is expected an increase of a 45% in the water demands. Therefore, its knowledge will be essential to optimize the use of the fresh water, which in turn will allow to distribute it better, and to improve the quantity and quality of the crop production. In addition, soil moisture, can be useful to predict when excess rainfall can cause water logging and/or flooding instead, it also useful to prevent droughts.

Missions like SMOS (radiometer at L-Band) from ESA, or SMAP (radiometer and radar at L-Band) from NASA, are dedicated missions focused on the retrieval of the soil moisture. Unfortunately, the current spatial resolution offered for that missions (40 and 10 kms respectively) are more suitable for global applications, than for local applications.

Recently the GNSS-R signals have been used for monitoring soil moisture. As other remote sensing techniques (e.g. radiometry), GNSS-R is based on the variability of the soil’s dielectric properties with the ground water content. Therefore, the GNSS reflected signals will present changes as a function of the soil moisture. In the same way, flooded areas can also be detected, measuring in this case 100 percent of water saturation content (very high values of reflectivity). An example of the sensitivity of the GNSS-R signals to soil parameters, can be found in [1], where significant variations in the measured reflection coefficients were related to different soil moisture content, and vegetation conditions. Results obtained on [1] (Leimon project), were confirmed on the GRASS project, were several experimental campaigns were carried out over an agricultural area in the vicinity of Florence.

In this work is proposed to integrate a GNSS-R sensor on a dedicated RPAS platform. One of the main advantages of this technique with respect others techniques (e.g L-band radars), is that it is cheaper and lighter, allowing to embend it onboard of a RPAS, which is of high interest for agriculture applications and river management (flooded area mapping), since it is very flexible and require fewer energetic resources compared with other techniques.

 During summer 2015, two experimental campaigns (Mistrale Project) have been carried out to validate the data processing concept. In those campaigns a complete GNSS-R sensor was installed on an ultralight aircraft, allowing gathering polarimetric GNSS-R data. The flights were done in France over the Camargue area (flooded areas, marshlands and water salinity changes), and Pech Rouge area (agricultural plots). In addition, in-situ measurements using SNR analysis [2] have been acquired as ancillary information, serving as a ground truth.

The estimated reflection coefficients acquired during the flights, have been computed and geo-referenced on ground, showing that the reflection coefficients are sensible to terrain changes. Main results obtained during these experiments, will be presented during the conference. 

The field measurement of HDRF (hemispherical directional reflectance factor) of the Namib Desert, Namibia is reported as a part of the new test site characterisation efforts to produce a calibration of autonomous instrumented reference sites for the radiometric calibration of high- resolution optical Earth Observation sensors.

A new permanently instrumented radiometric calibration site has been set up, in the vicinity of the Gobabeb Research and Training Centre, Namibia.  The site will become part of the RadCalNet (Radiometric Calibration Network using automated instruments) initiative established by the CEOS (Committee on Earth Observation Satellites), which is a globally distributed network of autonomous instrumented reference sites for the radiometric calibration of high-resolution optical Earth Observation sensors. The principal aim of the Gobabeb site is to provide daytime nadir top of atmosphere spectral reflectance every thirty minutes to facilitate satellite L1 radiance comparison and calibration. In addition to nadir measurements, the site provides multi-angular reflectance measurements to account for variations in the viewing angles of overpassing satellite sensors. On site, a permanently monitoring CIMEL sun photometer provides information on atmospheric conditions and surface reflectance at multiple angles.

As part of a project to establish the ESA (European Space Agency) / CNES (Centre National d’Etudes Spatiales) RadCalNet test site, NPL (National Physical Laboratory), working with the CNES performed an initial surface radiometric characterisation of the site in November 2015. It included nadir and multi-angular spectral reflectance measurements, to assess the spatial variability of surface properties. Ground-based measurements of HDRF of the site were carried out with GRASS (Gonio RAdiometric Spectrometer System). GRASS is designed to record quasi-simultaneous, multi-angle, hyperspectral measurements of the Earth’s surface reflectance. Signal collectors attached on a hemispherical frame and aiming at the same target area are connected with fibre optic cables to a V-SWIR spectroradiometer, operating over a wavelength range from 400 nm to 1700 nm.

During the campaign, measurements were made with clear sky conditions over the viewing zenith angles 0º-50º with a 10º interval and azimuth angles 0º-360º with a 30º interval. Changes in illumination were monitored with an upward facing integrating sphere mounted on the instrument and a reference measurement was recorded at nadir over a Spectralon panel. An area of 300 m x 300 m was selected in the first stage of the project based on pre-determined objective criteria, and random points within the area were sampled. Repeated measurements at different sampling points allowed the collection of HDRF datasets over a range of solar zenith angles. Furthermore, measurements were timed to correspond to key satellite overpasses (Sentinel 2A MSI and Landsat 8 OLI). The GRASS measurements will be used for the comparison of HDRF with the CIMEL instrument, contributing to an operational RadCalNet site.

The mainstream ocean colour remote sensing sensors (MODIS, VIIRS and the future OLCI) require “matchup” data from simultaneous sea-based instruments for validation. In particular validation of the water-leaving radiance reflectance is crucial as this parameter is the basis of all derived  marine parameters, including total suspended matter (TSM) and chlorophyll a (CHL-a). Belgian waters have been a key site for validation of satellite ocean colour radiometry since the launch of MERIS in 2002. Ship-based measurements have been regularly compared with satellite-derived marine reflectance to indicate any performance weaknesses. However, the ship-based measurements provide only a limited number of matchup measurements per year. The Belgian strategy for ocean colour validation has consequently shifted focus to the setup of continuously measuring systems in order to provide early validation feedback to missions such as Sentinel-3/OLCI. Moreover, in addition to the expanding range of mainstream ocean colour sensors, there is a growing interest in exploiting data from high-resolution sensors on polar-orbiters originally designed for land applications, such as Landsat-8/OLI giving 30m spatial resolution, Sentinel-2/MSI (10-60m) and Pleiades (2m). A single validation site can thus provide information for validation of a large number of spaceborne missions.

 The radiometric validation of ocean colour data is presented here in a multi sensor perspective with a focus on Belgian waters and the use of two AERONET-OC sites, one in turbid nearshore waters, one further offshore in clearer waters. The systems are based on a standard CIMEL SeaPRISM instrument but are equipped with a redesigned data acquisition and transmission system using a low-power embedded computer suitable for harsh environments with an on/off time switch, router and GPRS antenna. This approach has multiple advantages compared to the traditionally used satellite data transmission system: (1) dramatically reduced physical footprint  (+/- 35%) of the system on the offshore platforms, (2) remote access to pc system and (3) added level of data security as data is locally stored in case of data transmission malfunction.   The MOW1-Zeebrugge station has been actively collecting ocean color data between February 2014 and February 2015 when it was dismounted for maintenance and calibration. In this one year period the CIMEL/SeaPrism has collected a total of 1865 water leaving radiance measurements. After a first quality control by NASA/JRC 222 level 1.5 water leaving radiance measurements were retained for use in calibration/validation efforts of satellite data. The Thornton-Cpower site is installed on the Offshore Transformation Station (OTS) of the C-power windpark located 26 km offshore. The Thornton-CPower station has been actively collecting ocean color data since April 2015. Up to Oct 16^th 2015 it has collected a total of 1319 water leaving radiance spectra of which 370 spectra have passed the level 1.5 quality control of NASA/JRC.

 This poster will show the validation results of water leaving radiance reflectance from multiple sensors (i.e. MODIS-AQUA, VIIRS, Landsat-8/OLI, Sentinel-2/MSI and Pleiades) in Belgian waters using the AERONET-OC sites.

In the frame of the ESA Copernicus Sentinel-2 mission the Sen2Cor Level 2A processor has been developed in order to provide end users with a tool for generating L2A products from a single date L1C product. Sen2Cor is currently integrated in the Sentinel-2 toolbox available through SNAP (Sentinel Application Platform).

Sen2Cor Level 2A processing is applied to Top-Of-Atmosphere (TOA) Level 1C ortho-image reflectance products. Sen2Cor L2A outputs are a scene classification image, aerosol and water vapour maps and the ortho-image Bottom-Of-Atmosphere (BOA) corrected reflectance product.

The module of atmospheric correction in Sen2Cor is based on ATCOR (DLR) algorithm and therefore relies on a set of Look-Up-Tables (LUTs) that give for each sensor band pre-calculated values of 6 fundamental radiative transfer functions for many different atmospheric and viewing acquisition conditions.

The first version of Sen2Cor was delivered with a single Look-up-Table covering different contents of water vapor and a large scale of solar and viewing angles, visibilities and elevations, however limited to mid-summer conditions, with a hypothesis of rural aerosol and a default value of ozone concentration, besides initial computation was performed using out-of-date Sentinel-2A spectral response functions. It has been established that 24 LUTs would be necessary to cover most of atmospheric conditions on Earth for the Sentinel-2 mission: rural & maritime aerosols, winter & summer atmospheres with six different Ozone concentrations.

Telespazio France was selected by ESA to perform the generation and the validation of these additional LUTs for Sentinel-2A.

Originally in ATCOR these LUTs are based on MODTRAN5 radiative transfer calculation. Within the frame of Sen2Cor development, ESA requested the use of libRadtran, a library for radiative transfer developed by LMU Münich. It was proposed by Telespazio France to perform an inter-comparison with the same LUTs generated by MODTRAN5 to ensure libRadtran and Modtran were giving comparable results over a large range of conditions.

This poster describes the process of LUTs generation with these two Radiative Transfer tools and presents the LUTs inter-comparison methods and results. An original visual method for comparing the LUTs was proposed to identify quickly the areas of discrepancies between the LUTs.

Additionally we present an assessment of the impact of libRadtran and Modtran generated LUTs on the atmospheric correction performed by Sen2Cor, in terms of aerosol optical thickness, water vapour retrieval and surface reflectance. 

Driven by Copernicus and GGOS (Global Geodetic Observing System) initiatives the user community has a strong demand for high-quality Earth observation products. In order to derive such high-quality products precise orbits for the Sentinel satellites are needed. Earth observation missions meanwhile span over more then two decades, in which our understanding of the Earth has increased significantly. As also the models used for orbit determination have improved, the satellite orbits are not always available in an uniform reference system. Homogeneously determined orbits referring to the same global reference system are, however, needed to improve our understanding of the Earth system.

The Navigation Support Office at ESA/ESOC has been providing precise orbits for ESA Earth observation satellites since 2001. The Navigation Support Office has with its NAPEOS software package the capability to process all three satellite geodetic tracking techniques (SLR, DORIS and GNSS). Therefore, we are in the unique position to do orbit determination by combining different types of data, and by using one single software system for different satellites, which matches the most recent improvements in orbit and observation modeling and IERS conventions. Thus we are able to generate a homogeneous set of precise orbits referring to the same reference frame for the different Sentinel  missions. Furthermore we are able to quickly re-process all solution allowing us to continuously upgrade the various solutions for all satellites.

This presentation focuses on the latest results from the efforts carried out by ESA/ESOC for the generation of precise and homogeneous orbits for the Sentinel missions. We will present the orbit determination results and evaluate the orbit accuracy by comparing our orbits with external orbits generated by other centers and will highlight some of the improvements obtained from our most recent upgrades.

In the frame of the German spaceborne hyperspectral EnMAP mission a lot of different methodological research is conducted before launch that aims on having ready-to-use applications when the satellite transmits the first data after commissioning phase. In addition to the application development a persistent monitoring of the instrument performance is required during the operational phase. This relates to geometric, radiometric and spectral parameters and their temporal alteration. Most of the parameters will be continuously monitored by the DLR ground segment and are used for frequent calibration/re-calibration of the system. Besides this, independent validations of all EnMAP products will take place for maintenance of a high data quality. .

In this work we propose a framework for the accuracy or performance assessment of the current geometric validation chain. Contrary to EnMAP’s capacity for radiometric and spectral self-calibration, there exists no on-board calibration device for geometry. Therefore, a validation of foregoing parametric modelling of the sensor trajectory in time and space by the EnMAP ground segment and product related data transformations need to be validated from time to time. Since image based approaches are limited regarding the quantity of parameters that can be assessed, a set of geometric performance related indicators have been selected as key indicators:

The level of spatial coincidence between the VNIR and the SWIR for the radiance product L1c

The relative intra-band keystone for the L1b product

The absolute across-track detector Modulation Transfer Function for the L1b product

The absolute accuracy of the overall pixel Line-Of-Sight for the L1c product

The relative geometric accuracy to the references Landsat-8 OLI and Sentinel-2

The accuracy assessment of these key indicators and the validation chain is investigated based on the simulation of EnMAP data products from three test sites combining the GFZ EnMAP end-to-end simulator with the DLR geometric simulator and L1C-processor. Airborne mosaics having an average ground resolution of better than 1 m served as basis for hyperspectral simulations. The mosaics originated from the region covering the bridge of Lake Pontchartrain in Louisiana, U.S., parts of the city of Phoenix, Arizona, U.S. and some regions around Mullewa, Western Australia.

Key parameters are calculated using iterative approaches that integrate automatic tie point detection, Fourier space local affine distortion and MTF modelling and polynomial registration modelling from micropixel precise local phase correlations.

Foregoing methodological research showed that the achieved micropixel accuracy for validation fosters the validation of future high accurate EnMAP products and provides enough place for all potential applications. It is assumed that parts of the approaches can be utilized for other validation campaigns such as Landsat-8 and Sentinel-2 as well as planned future sensors.

The aim of this contribution is to study and to record the occurring geophysical phenomena as well as their temporal behavior in the Lower-Rhine-Embayment in the southwest of North Rhine-Westphalia, Germany. Within this region there is one of the largest brown coal occurrence in Europe. The related groundwater changes lead to deformations of the Earth's surface in the range of a few centimeter per year.
To detect ground deformation time series in the Lower-Rhine-Embayment we use differential radar interferometry (D-InSAR) data. In more detail we use a multitemporal D-InSAR stack of 191 interferograms from the European Remote Sensing satellites ERS 1/2. These D-InSAR images are estimated using the Small Baselines Subsets (SBAS) method.
The SBAS method assumes small spatial and temporal baselines between the SAR image pairs to reduce decorrelation effects. Except of a few ERS 1/2 tandem pairs with one day seperation, the repeat cycle is a small multiple of 35 days for each satellite. Especially in areas of very rapid surface changes or deformations the 35-day repeat cycle is too long and leads to decorrelated and thus noisy data. The Lower-Rhine-Embayment is a very rural area characterized by just a few numbers of coherent pixels which further increase the measurement noise.
All these effects make it a challenging task to unwrap the phase in the Lower-Rhine-Embayment. A rather popular technique to reconstruct the phase is the Minimum Cost Flow (MCF) approach. The problem is recast into a network with nodes and arcs searching for the minimum cost flow.  The basic algorithm only works within one single interferogram. In order to simultaneously unwrap the whole D-InSAR stack, extended versions are often discussed in literature. These approaches exploit both the spatial as well as the temporal information. One of these approaches is the extended MCF approach, which is based on two Delaunay triangulations in the spatial as well as in the temporal plane. The algorithm involves two main steps. First phase unwrapping is performed in the spatial plane applying the basic MCF approach. These estimated phases are used in a second step as starting points to spatially unwrap the phase in each single interferogram via the basic MCF approach again. The aim is to improve the correctness of the unwrapped phase in the Lower-Rhine-Embayment with help of such a three-dimensional approach.

The Jason-3 altimeter mission is expected to be launched by the end of 2015 / beginning of 2016.

The objective of Jason-3 is to provide continuity to the unique accuracy and coverage of the TOPEX/Poseidon, Jason-1 and OSTM/Jason-2 missions in support of climate change monitoring, research and forecasting, as well as operational applications related to extreme weather events and operational oceanography. In order to ensure this continuity, the first couple of months after the launch will be dedicated to verification activities, on engineering as well as on geophysical level. Therefore the unique opportunity of the formation flight phase (Jason-3 and Jason-2 on the same orbit, following one the other with a couple of minutes time lap) will be used, as both systems will observe the same environment. Data coverage and quality will be analysed. Altimeter and radiometer parameters, as well as sea level anomaly will be compared between Jason-3 and Jason-2 and cross-calibrated, in order to detect possible offsets and ensure that there are no drifts between the two satellite systems. The parameters analysed will include significant wave height as well as altimeter wind speed. This cross-calibration can be done on global level (comparison of daily global statistics), but owing to the formation flight phase also at a very fine level: measurement to measurement comparison is possible. Cross-calibration with other altimeter satellites, such as SARAL/AltiKa and Sentinel-3A (once launched), as well as in-situ data are also planned.

First results of this Jason-3 data quality assessment will be presented.

To cover the Earth’s surface in 10 days, Sentinel-2A acquires images at a rate greater than  100 million pixels per second, due to its swath of 290km and its 13 spectral bands with a resolution up to 10 meters. It is a considerable volume of image data to process each second, including image decompression, radiometric corrections, resampling to cartographic tiles, generation of quality masks, metadata computation and compression to JPEG-2000. The Instrument Processing Facility (S2-IPF) is in charge of giving to each pixel its correct reflectance value and location on ground, producing images up to level 1C.

With its 38 linked software components, the image processing chain is configurable to allow the fine tuning of algorithms such as cloud detection, mask vectorization, compression and geolocation. It also adapts to the available hardware resources such as RAM and number of cpu cores.

In order to meet the mission quality and timeline requirements, the S2-IPF implements state-of-the-art algorithms, while near real-time processing is supported by the massively parallel architecture of the payload data ground segment. From the inside, this challenge is overcome by a careful design of data circulation, stream processing of the large images and optimized implementation of algorithms.

The development of the IPF has gone one step further than initially expected, giving birth to the new geolocation library Rugged. This open source library implements new algorithms for direct/inverse location on the Earth’s rugged terrain, intersecting digital elevation models (DEM) with very high numerical and time performance. The open source Orfeo Toolbox was also the basis for pixel processing, providing the building blocks of most of the radiometric corrections and image resampling functions. It has required specific optimizations and a new configurable pipeline design to be used in operations.

Our presentation will highlight technical choices that have made our software a successful IPF, including trade-offs between quality and performance, or clever design decisions in our solution. We will also present what we have learned from this project and what we will do better next time.

C-S is the prime contractor for the development of the S2-IPF, inside the industrial consortium also composed by GMV and GAEL. Taking in charge the development and integration of the main part of the processing chain, the team of dedicated engineers at C-S has done its best to deliver an operational IPF, contributing to the success of Sentinel-2. The IPF was at the rendezvous last June with the launch of Sentinel-2A, supporting the exceptional images of our colorful planet delivered by the mission.

Earth Observation Satellite data products are key assets to support the study in the Earth, its environmental systems and variations in global and local environmental parameters. Over time the quality of the data products can be improved by different means, better algorithms, improved auxiliary information etc.  as such can be subject to data reprocessing campaigns creating newer versions of the processed data. Understanding data validity, accuracy and the traceability of data management is therefore essential. As part of the ESA DSI (Data Service Initiative) service contract managed by ESA’s Ground Segment Operation division, the X-PReSS (eXpert Product Reprocessing Scalable Service - http://xpress.sp.serco.eu/ ) consortium led by Serco, has developed a Data Information System (DIS), to manage the configuration information relating to data products.  This allows each product’s characteristics as well as its relationships with other items (such as software processor versions and auxiliary files used), to be captured, therefore improving their usefulness overtime.

As the Data Information System becomes more established as a supporting tool in the operating procedures, more uses emerge and the tool continues to receive new requirements and evolve to allow a broader use. This paper outlines the evolution of the data information system, the increased scope and the importance of a proper data management for on-going projects in a broader context. 

Sen2Three is a level 3 processor for the spatio-temporal synthesis of bottom atmospheric corrected level 2A (or alternatively) uncorrected top of atmosphere level 1C products. The key functionality of the level 3 spatio-temporal synthesis algorithm is, generating synthetic output images of a certain geographical area of interest using time series of input images of the area and replacing step by step all "bad" pixels of the output with "good" ones. Input images are a collection of level 1C or level 2A products, acquired sequentially during a certain time period, as determined by the user. Only tiles matching a certain time window will be accounted for the processing.

Level 2A images can either be prepared from level 1C images using the Sen2Cor level 2A processor from the Sentinel 2 Toolbox, which can act in this way together with this processor, forming a processing chain. Alternatively, the user can also select atmospheric uncorrected level 1C images as input. In the latter case, Sen2Three processes only the necessary scene classification as a pre-processing step and performs a spatio-temporal synthesis on the level 1C images afterwards.

 Input products are expected to be placed within one single directory of user's choice. If the processor detects a level 1C product without an equivalent level 2A counterpart, it automatically performs a scene classification first, before performing the level 3 synthesis in a second step.

 Pixel Criteria

 "Good" pixels are primarily identified to be of one of the three alternative pixel types: vegetation, soils, or water as can be identified on clear sky images using the Sen2Cor Scene Classification algorithm.

Clouds are by default classified as being "bad" pixels. Other pixel types like: saturated pixels, dark areas, cloud shadows, thin cirrus and snow can be optionally configured as belonging to the "bad" pixel group as well.

Algorithms

The Sen2Three algorithms work on tile base. Four different algorithms are implemented, which determine the method how the final output product will be synthesized. For all images in all bands the bad pixels will be replaced by good ones, according to following rules:

Most recent: bad pixels of the output scene will always be replaced by good pixels of the current scene, where the actualization criterion is the time information of the tile.

Temporal Homogeneity: all pixels of the output scene will be replaced by the equivalent good pixels of the current scene, if the sum of good pixels of the current scene is higher than the sum of good pixels of any scene in the past. This algorithm will always prioritize the "best" scenes in the course of time.

Radiometric Quality: bad pixels of the output scene will be replaced by good pixels of the current scene, if either:

the average of the Aerosol Optical Thickness is less, or

the average of the Solar Zenith Angle is higher,

than the  equivalent parameter of the best scene in the past. The criteria are configurable. Again, this algorithm will always prioritize the "best" scenes in the course    of time.

Average: the output scene is an average of the good pixels of all processed scenes. Averaging can be useful in situations, when only a collection of very noisy input images are available, in order to homogenize the output product. 

Classification Map 

The Scene Classification map will be updated during each sequential step. In an ideal synthesized scene only the static classes: vegetation, soils, water, urban areas and snow/ice (if permafrost occurrs) would remain.

Mosaic Map

A new level-3 map type will be generated and updated during the synthesis: the Mosaic Map is an index map allowing conclusions on the history of the sequential processing.

For the average algorithm it keeps the per pixel sum of all good pixels of the past scenes and is used for calculating the most recent average. For the other algorithms the index number is an unsigned integer value referring to one specific tile ID within a sequential processing. The meaning of this index depends on the chosen algorithm and is further specified in the metadata.

Railroad Valley, in Nevada, is a homogenous test site which has been used since the 1990s for the radiometric calibration and verification of satellite-borne instruments. The Railroad Valley site contains a collection of instruments monitoring both the ground radiance and the atmospheric properties. Together, these instruments, along with a library of surface reflectance spectra and a unique processing algorithm, form the automated Radiometric Calibration Test Site (RadCaTS) system operated by the University of Arizona. This system provides a unique opportunity for the ongoing calibration and verification of satellite-borne instruments through the provision of Top-of-Atmosphere (TOA) radiances.

In addition, the site is also part of the Radiometric Calibration Network (RadCalNet) initiative. For this initiative, ground radiances of the site will be provided to NASA who will then propagate these to TOA. These TOA radiances will be held in a central database (by NASA) which will be accessible to all interested parties. There are currently four sites which are due to become part of RadCalNet: Railroad Valley (USA), La Crau (France), Baotou (China) and Gobabeb (Namibia).

Both the RadCaTS and RadCalNet systems represent a key advance in the methods used for vicarious calibration.  However, further benefit can be gained from providing a complete, robust and fully metrologically traceable uncertainty budget for each of the sites. This will allow the un-biased comparison of ground-to-satellite measurements and work towards fully traceable in-flight calibration of satellite-borne instruments where no methods other than vicarious calibration are available.

The current work provides such an uncertainty budget for the RadCaTS system at Railroad Valley. This site (and the specific processing method) is chosen for this work due to its advanced stage of operational set up, the many years’ worth of data available and due to the prominence of this site with the RadCalNet initiative.

The derivation of a comprehensive uncertainty budget includes consideration of many factors. For example, the various data sources included in the processing system (i.e. the exo-atmospheric solar irradiance model) need to be considered both in terms of data provenance and any additional uncertainty they may introduce. In addition, all of the assumptions made within the system, for example the assumption that the ground measurements are representative of the entire site, need to be carefully considered and any additional uncertainties characterised. Another element to be considered is the impact of modelling on the uncertainties, for example the propagation of the ground data to TOA using the radiative transfer code, MODTRAN.

Uncertainties associated with each of these elements are derived using a combination of analytical and Monte-Carlo approaches.  The estimates are then brought together to provide a robust end-to-end uncertainty budget for TOA radiances derived from ground measurements at the Railroad Valley site.

The upcoming fleet of Sentinel satellites will provide so far unique opportunities for global environmental monitoring. However, the capability to effectively and efficiently access, manage, process, and analyze the mass data streams from the Sentinels, but also from other “big data” missions such as the Landsat program, still poses major conceptual and technical challenges. DLR works towards bridging the gap between the immense data volumes collected by modern Earth Observation missions and their application-driven, on-demand exploration through geo-information services. Here, a key element is the testing of new concepts for automated processing and analysis chains, with customized data access, composed of generic modules, and embedded in powerful hard- and software environments for distributed and operational processing of mass data sets. This conceptual framework builds upon DLR’s heritage in geometric correction and orthorectification (ORTHO) and atmospheric correction (ATCOR) modules and integrates previously established automated processing chains (CATENA). One of the key elements is the integrated system engineering approach focusing on the demand driven set-up of tailored software/hardware solutions allowing for an efficient production of information products. After a short introduction of the background and objectives, first technical solutions for optimized data access and pre-processing will be presented along with exemplary outcomes of the thematic processing functionalities implemented.

An important part of the image processing for satellite water quality products is a good and reliable cloud screening (i.e. cloud and cloud shadow detection and cloud classification). Imperfections in cloud screening, present even in the standard processing of mature ocean colour missions, can lead to completely erroneous data being supplied to users.

The FP7/HIGHROC (“HIGH spatial and temporal Resolution Ocean Colour”) Project is developing the next generation of optical products for coastal water services, combining mainstream ocean colour sensors (Sentinel-3/OLCI, VIIRS) with higher spatial resolution sensors (Landsat 8, Sentinel 2 (S2)) and higher temporal resolution sensors (SEVIRI). Within this project, common pre-processing tools have been evaluated and implemented for the different sensors.

An important investigation during that time was developing and expanding the cloud detection and the cloud classification for the spatial high resolution sensors, aiming at Sentinel 2 and using Landsat 8 as a precursor for S2. Three different algorithms have been tested, based on several (different) methods: the ACOLITE/PCL algorithm developed by RBINS; the ACCAm algorithm created by VITO, and Idepix developed at BC. We present here a short description of each algorithm, and their individual results.

Developments are currently underway for S2 and S3; the algorithms were developed for MERIS and L8, and are now being adapted in order to exploit the new bands, and the different resolutions. Cloud screening development for OLCI will be heavily based on the experience gathered with MERIS, especially during the 4^th reprocessing. We present here some results for these two sensors.

We show the results of the validation experiment using a pixel identification database (PixBox) gathered over many scenes for each sensor (e.g. 21 for Landsat 8), representing several 10 of thousands of pixels manually classified (e.g. 34020 pixels for Landsat 8), within 8 different categories representing different grades of cloudiness, ranging from totally cloudy over turbid atmospheres to clear sky over different surfaces.

This reference is used to produce confusion matrices between manually selected pixels and pixel identification within the pre-processing steps as well as statistics showing the agreement of both data sets.

Results are currently available for Landsat 8 and we expect to complement this with results from Sentinel 2 as well as Sentinel 3 for the Living Planet Symposium.

PROBA-V has been developed for monitoring vegetation and its changes in a changing climate as a gap-filler between the SPOT-VEGETATION and Sentinel eras. The instrument performs observations of reflected solar radiation in four spectral bands between 0.45 and 1.60 µm, with the chosen wavelengths and spectral responses being largely similar to those of SPOT-VGT. PROBA-V is currently in its third operational year, with the mission expected to continue until mid-2018.

Accurate vegetation monitoring from satellite observations is highly dependent on a robust cloud screening. The PROBA-V cloud detection algorithm is currently being modified using GlobAlbedo surface albedo information and data containing the updated cloud screening will become available to users by mid-2016. First results of the modified algorithm indicate significant improvements compared to the current operational cloud detection, mainly for ice cloud cases and over bright surfaces, such as desert areas. However, up till now the modified algorithm has not been evaluated against an independent satellite reference dataset.

This paper presents the evaluation of the modified PROBA-V 300 m cloud detection algorithm with the 1 km Meteosat-SEVIRI cloud mask (Bley and Deneke, 2013), which synergistically uses data of the geosynchronous Meteosat-SEVIRI’s nominal (3 × 3 km² at nadir) and High Resolution Visible (HRV, 1 × 1 km² at nadir) spectral channels. The 15-minute SEVIRI temporal resolution guarantees a sufficiently small time difference with the PROBA-V observations to perform a proper collocation of the two cloud datasets. The evaluation is performed over various climatological regions in Europe and Africa and spans the entire seasonal cycle of 2014. Additional to this evaluation, we will prove that improved PROBA-V cloud detection leads to an improved quality of vegetation parameters derived from the observed cloud screened reflectances by showing various examples using the existing and modified cloud detection.

As the crowning achievement of almost ten years of mission and system design and development, the first satellite of the Sentinel-3 series (Sentinel-3A) is slated for launch at the end of 2015. Carrying a suite of state-of-the-art instruments, Sentinel-3 is set to play a key role in the Copernicus environmental monitoring programme. Its multi-instrument payload, covering both optical and microwave measurements, provides systematic observations of Earth’s oceans, land, ice and atmosphere to monitor and understand large-scale global dynamics and provide critical information for ocean and weather forecasting.

Sentinel-3 payload suite encompasses the Ocean and Land Colour Instrument (OLCI) covering 21 spectral bands with a swath width of 1270 km, the Sea and Land Surface Temperature Radiometer (SLSTR) covering 9 spectral bands and yielding a dual-view scan with swath widths of 1420 km (nadir) and 750 km (oblique view), the Synthetic Aperture Radar Altimeter (SRAL) working in Ku-band and C-band, and the Microwave Radiometer (MWR) operating in dual-frequency at 23.8 GHz and 36.5 GHz.

In the early stages of mission and system design, one of the key issues to be addressed consisted of the Sentinel-3 reference orbit selection, with a flow-down impact on the space segment architecture, the operations concept and the mission return. The main driver for the orbit selection was the requirement to achieve a revisit time of two days or less globally over ocean areas with two satellites (i.e. 4-day global coverage with one satellite). The orbit selection was seamlessly coupled with the OLCI instrument design in terms of field of view (FoV) definition driven by the observation zenith angle (OZA) and sun-glint constraints applied to ocean observations. Thus, an iterative and coupled orbit-instrument design process was carried out, with the overarching objective to fulfil the coverage requirement, one of the paramount figures of merit to assess the Sentinel-3 mission return.

The criticality of the global coverage requirement for ocean monitoring derives from the sun-glint phenomenon, i.e. the impact on visible channels of the solar ray reflection on the water surface. This constraint was finally overcome thanks to the concurrent optimisation of the orbit parameters, notably the Local Time at Descending Node (LTDN), and the OLCI instrument FoV definition.

The approach for the selection of the orbit altitude and repeat cycle was divided into three main phases:

Identification of orbits with short repeat cycle (2-4 days) that would be optimal orbits for a mission with only optical coverage requirements. The rationale was firstly to minimise the time required to achieve global coverage with existing constraints, and then to minimise the swath required to obtain global coverage, in order to minimise the maximum required OZA. This step yielded the selection of a 4-day repeat cycle orbit, thus allowing 2-day coverage with two adequately spaced satellites. The selected sun-synchronous orbit (SSO) was: 14+1/4, reference altitude ~803 km, LTDN=10h00.

Identification and selection of suitable candidate orbits with higher repeat cycles in the proximity of the selected altitudes. The rationale was to keep the swath for global coverage as close as possible to the previous optimum case, but to tailor the repeat cycle length (i.e. the ground track grid) in order to optimise the performances of the topography mission.

Selection of the reference orbit based on the results of topography performance analysis. The final choice converged on the SSO 14+7/27, reference altitude ~800 Km, LTDN=10h00.

At each step of the design process, extensive coverage analyses were carried out to exhaustively characterise the mission performance and the fulfilment of the requirements, encompassing revisit time, number of acquisitions, observation viewing geometry and swath properties.

Global Navigation Satellite System Reflectometry (GNSS-R) is a recently developed technique for global remote sensing of the ocean and land surface. After the first experimental demonstration of satellite reflectometry from the UK Disaster Monitoring Constellation, new satellite missions such as TechDemosat, CYGNSS and GEROS-ISS [1,2,3] will establish new challenging perspectives about the capability of measuring the ocean roughness in extreme rain and wind conditions and validating the retrieval algorithms.

The main purpose of this study is to define and validate a new algorithm for wind speed estimation over the ocean. Up to now, the approaches for wind speed retrieval using GNSS-R are:

1. The 2-D least squares (LS) fitting of averaged delay-Doppler maps (DDM) with the theoretical Zavorotny-Voronovich model.

2. The retrieval from scattering coefficient maps after deconvolution.

3. The retrieval based on geophysical model function.

The rationale of this work moves along the line defined by the first and second approach. Starting from the expression of the averaged delay Doppler map as a 2-D convolution, it is shown that the volume of the scattered power is given by the ratio between the volume of the delay-Doppler map and the volume of the ambiguity function of the GNSS signal. After an inverse mapping from delay-Doppler to spatial coordinates, the scattering function turns out to depend upon the geometry of the observation system and on the wind speed and direction. The wind parameters, specifically, are embedded into the joint probability density function (pdf) of the sea surface slopes [4].

A Least Squares fitting is used to retrieve the wind speed by adaptation of the estimated pdf and the model (Gaussian) pdf, after the compensation of all the geometric terms that include the space-varying area of the resolution cells, the scattering vector, the distance from platforms to the single resolution cells and the antenna radiation pattern. However, the retrieved wind speed is affected by several error sources: the thermal noise, the speckle noise, the limited size of the DDMs and the data calibration inaccuracies.

The performance of the algorithm is assessed by comparing the wind speeds computed from real data received by TechDemosat-1 (TDS-1) platform with those obtained from the Advanced Scatterometer (ASCAT) on Metop-A that are assumed as ground truth. Precisely, the validation is carried out by representing a large number of TDS-1 and ASCAT retrievals on a scatter plot and by computing appropriate statistics for quantifying the discrepancy. Corrected data are obtained through a calibration polynomial that shows the best fit between retrieved TDS-1 and ASCAT data.

REFERENCES

[1] M. Unwin, P. Jales, P. Blunt, and S. Duncan, “Preparation for the first flight of SSTL’s next generation space GNSS receivers,” 2012 6th ESA Workshop on Satellite Navigation Technologies, pp. 1–6, Dec 2012.

[2]  C.Ruf, A.Lyons, M.Unwin, J.Dickinson, R.Rose, D.Rose, and M.Vincent, “CYGNSS:Enabling the Future of Hurricane Prediction,” IEEE Geoscience and Remote Sensing Magazine, vol. 1, no. 2, pp. 52–67, June 2013.

[3]  J. Wickert et al., “GEROS-ISS: Innovative Ocean Remote Sensing using GNSS Reflectometry onboard the International Space Station,” European Geosciences Union General Assembly 2014.

[4]  V.U. Zavorotny, A.G. Voronovich,“Scattering of GPS Signals from the Ocean with Wind Remote Sensing Application,” IEEE Transactions on Geoscience and Remote Sensing, vol. 38, pp. 951–964, March 2000.

With the launch of the Sentinel-3 satellite data from new sensors will become available: the main instruments being  the Ocean and Land Colour Instrument (OLCI);  the Sea and Land Surface Temperature Radiometer (SLSTR);  and the SAR Radar Altimeter (SRAL).  Calibration and Validation activities during the commissioning and operational phases come under the Mission Performance Framework established by the European Space Agency (ESA) and the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT).  This framework brings together the agency experts; the Sentinel 3 Validation Teams (S3VT); and the Sentinel 3 Mission Performance Centre (MPC).  One of the activities of the Oceanography Group at EUMETSAT is to implement marine calibration and validation tasks as part of  the Mission Performance Framework.

In the context of Ocean Colour, the Oceanography Group, has concentrated, in the first instance, on the development of a level 3 processor.  By analysing the resulting level 3 products the system performance can be assessed and monitored.  The overall aim of these initial investigations is to provide a preliminary assessment of the quality and stability of the various level 1 and marine level 2products: to evaluate their consistency in time and space and, through these and associated analyses, to improve the data quality.

Specifically, Level 3 data will be generated globally at various time periods and further analysed over specific regions. These datasets will be used, often in collaboration with other teams, to: provide radiometric validation; provide verification of cross track artefacts; provide support to vicarious calibration activities; monitor the processor and its implementation; assess OLCI consistency with other missions; monitor the performance of algorithms over time and space; monitor and assess impact of algorithm developments and alterations; and to assess aerosol algorithm performance, to name a few.  This presentation will provide an overview of these activities.  

The experimental site “LakeLab” is a freshwater enclosure facility installed and operated in Lake Stechlin by the Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB). It is a floating construction that consists of one large enclosure 30 m in diameter and 24 smaller enclosures, each 9 m in diameter. Each enclosure is isolated from the lake by walls made of plastic sheets that reach down into the sediment of the lake at a depth of about 20 m. This set of different enclosures offers the unique opportunity to simulate various optical conditions occurring in natural waters under equal external environmental conditions. In the framework of the MARS project (Managing Aquatic ecosystems and water Resources under multiple Stress; www.mars-project.eu) two different treatments were performed: (1) Phosphorus fertilization and (2) browning, an addition of humic substances. For the Phosphorus treatment a seven step gradient and for the browning a three step gradient was applied. This led to a set of 21 enclosures in close proximity with optically different properties, according to the amount of added material. To expand our knowledge about optically active water constituents, validate satellite remote sensing data and improve existing models, there is a strong necessity for datasets coupling remote sensing with in-situ data. During outstanding weather conditions in July 2015 various in-situ and on-site measurements were performed at the LakeLab. Inherent optical properties (absorption and scattering) and apparent optical properties (radiances) were recorded using different instruments for 9 different enclosures. This data will be coupled with the data from a HySpex overflight during the measurements which provides a very good opportunity for validation and modelling.

In recent years numerous federal projects in Europe have engaged LiDAR technology in mapping land cover across whole countries. Such large-scale initiatives often compromise on data resolution standards, leading to lower scanning densities. This may cause difficulties in reconstructing urban environments, especially when complicated structures are present. In current study we attempt to improve the recognition of man-made structures and develop a more efficient method of automatic point cloud classification by implementing additional PolSAR data. A consequent challenge then is developing a method for automatic detection of specific types of buildings.

PolSAR data used in our work have been collected by RADARSAT-2 satellite and are quad-pol type, carrying information measured in multiple polarization planes. A couple of new classification algorithms have been tested in order to optimize building extraction.

During the whole process we strive to achieve satisfactory results while using mainly free and/or open-source software, like LAStools by rapidlasso for LiDAR point cloud analysis and ESA’s PolSARpro for handling quad-pol data. The reason is to make the outcome of our work available for use and testing to all interested parties, resulting in further progress in this particular field of study.

Sentinel-3 Mission Performance Centre has developed a set of tools to allow validation of OLCI products. MERMAID and ODESA processing and analyzing tools have notably been updated to support OLCI land and water product validation. A major issue of OLCI product validation in the early stages of the mission is the access to quality controlled in situ data. Foreseen reference in situ dataset for OLCI CalVal include: FluxNet and NEON for land products, BOUSSOLE, MOBY and AERONET-OC for water products. In the absence of in situ data, representative Cal/Val sites covering a large diversity of land cover and water types are monitored to assess the coherence of OLCI products. These locations includes core validation sites (New forest, Barrax, Valencia Alicante and Nebraska sites for land products and a selection of case 1 waters going from oligotrophic (South Pacific Gyre, South Indian Ocean … ) to eutrophic environment (Benguela and Peru upwellings …).

In the time frame of commissioning phase E1 (five months), quantitative validation will mainly be possible over water thanks to in situ data collected on permanent mooring or stations like BOUSSOLE or AERONET-OC. Owing to the short term delivery of these in situ data, it is acknowledged that they will not be fully consolidated and quality controlled. They will nonetheless be of great interest for a first assessment of OLCI performances. Additional OLCI product quality control will be performed on bio-Argo data which provide chlorophyll measurements over a larger diversity of oceanic regions.

In addition to level-2 products validation, inter-comparison of OLCI products with contemporaneous (MODIS, VIIRS) and historical missions (MERIS, SeaWiFs) will be performed at level-3. In a first case, weekly products will be used to assess the coherence of OLCI, MODIS and VIIRS level-3 products. In the second case, OLCI level-3 products will be compared to climatology derived from contemporaneous and historical missions.

Oscillator phase noise has negligible impacts in a monostatic Synthetic Aperture Radar (SAR) but can severely worsen the performance of a bistatic SAR. In bistatic SAR systems, there is no cancellation of low frequency phase errors as happens in monostatic SARs, where the same oscillator signal is used for both modulation and demodulation. In bistatic systems, uncompensated phase noise may cause degradation of image quality (e.g. geometric resolution widening and localization errors) and degradation of phase based applications (e.g. interferometry) performance.

The current paper investigates the impacts of the limited oscillator stability. In particular, we will address the specific case of SAOCOM – CS mission, taking into consideration the constraints coming from the hardware (e.g. Ultra Stable Oscillator, USO, phase noise power spectral density) and the mission requirements. Our analysis focuses on the impacts of mis-synchronization on SAR images: we will briefly recall the mathematical models used to quantify the impact of phase error on SLC data [1] and interferometric phase in order to later apply them to the case of our interest. According to our experience, we expect that interferometric applications shall be considerably impacted in by this additional phase noise.

Once assessed the limitations on performances caused by non compensated oscillator phase noise in SAOCOM – CS mission, we consider, as a further step, the possibility to mitigate this error. According to the constraints of the SAOCOM – CS mission, we investigate some possible phase correction options.

In this paper, we address the problem of synchronization presenting  a mathematical model to predict the residual phase error after synchronization. This budget shall account also for the error contribution of the link itself, which may suffer of e.g.: thermal noise, phase noise interpolation error and aliasing [2]. Then, we will apply the model to the aforementioned context of interest for SAOCOM – CS mission

We also consider an alternative technique of interferometric phase calibration achieved using a set of Ground Control Points (e.g. from SRTM) as external reference. This alternative approach is useful to allow comparison with other methods and to provide a solution to be applied during data processing.

The actual benefits of each synchronization method are measured evaluating the residual phase noise and the IRF worsening. A comparison with the non-compensated case shall pave the way for discussions and remarks.

[1] G. Krieger et al. “Impact of oscillator noise in bistatic and multistatic SAR,” IEEE Geoscience and remote sensing letters, Vol. 3, No3, July 2006.

[2] M.Younis et al., “Perfomance prediction of a phase synchronization link for bistatic SAR,” IEEE Geoscience and remote sensing letters, Vol. 3, No3, July 2006.

Copernicus Sentinel-1 Constellation: Spacecraft Architecture

Francesca Spataro⁽¹⁾, Giacomo Taini⁽¹⁾,

Mathias von Alberti⁽²⁾,

Ramon Torres⁽³⁾, Svein Lokas⁽³⁾, David Bibby⁽³⁾

⁽¹⁾Thales Alenia Space Italia Via Saccomuro 24, 00131 Roma, Italy

francesca.spataro@thalesaleniaspace.com, giacomo.taini@thalesaleniaspace.com,

⁽²⁾AIRBUS Defence and Space, D-88039 Friedrichshafen, Germany

mathias.alberti@astrium.eads.net

⁽³⁾ESA/ESTEC – Keplerlaan 1 Postbus 299 AG 2200 Noordwijk, The Netherlands

ramon.torres@esa.int, svein.lokas@esa.int, david.bibby@esa.int

The Copernicus Sentinel-1 Mission is based on a Constellation of two C-band SAR satellites to fulfill revisit and coverage requirements, providing robust datasets for main following services areas:

Monitoring sea ice zones and the arctic environment;

Surveillance of marine environment (wind speed, oil spill and ship detection);

Monitoring land surface motion risks;

Mapping of land surfaces: forest (climate change, management, fire), water and soil, agriculture (food security, crop monitoring);

Mapping in support of humanitarian aid in crisis situations.

The Copernicus Sentinel-1 Earth Radar Observatory, a project funded by the European Union and developed by ESA, is a constellation of two C-band radar satellites.

The Sentinel-1 satellites are built by an industrial consortium headed by Thales Alenia Space Italy as Prime Contractor and AIRBUS Defence & Space as SAR Instrument contractor. Thales Alenia Space Italy, acting as Prime Contractor has designed the Satellite System providing innovative solutions to comply with the challenging requirements and to perform the sub-systems integration and testing at Spacecraft before the launch and the flight operations.

The paper will include a description of the Spacecraft general architecture and a description of all functional subsystems, including main features and challenging developments devoted to this C-band Synthetic Aperture Radar Satellite for Earth Observation.

Finally the Spacecraft configuration and performances will be presented giving particular emphasis on Satellite peculiar and innovative design features.

Sentinel-3 (S3) Mission Performance Framework (MPF) Activities span contribution from European Space Agency (ESA), European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT), Sentinel-3 Validation Team (S3VT) and from Sentinel-3 Mission Performance Centre (MPC). During both Commissioning (E1) and Routine Operations (E2) phases EUMETSAT is contributing as defined in EUMETSAT Mission Performance Implementation Plan (EMPIP).

We will give an overview of the cal/val activities planned and implemented at EUMETSAT, with the focus on activities covering optical Sea and Land Surface Temperature Radiometer (SLSTR) and Ocean and Land Colour Instrument (OLCI) instruments. For selected critical activities we plan to show preliminary results depending on the progress of the overall S3 mission.

Many cal/val activities will be implemented through Mission Performance Monitoring Facility (MPMF) within Payload Data Ground Segment (PDGS), and certain activities will be implemented using open source tools (e.g. Felyx, Naiad) or in-house developed tools (e.g. L3 processor).

Considering SLSTR during the Commissioning phase, the main focus will be on Level 1 radiometric inter-comparisons of SLSTR infrared channels with Infrared Atmospheric Sounding Interferometer (IASI) which is adopted as an inter-calibration reference instrument by Global Space-based Inter-Calibration System (GSICS).  Upon successful validation of Level 1 product, the focus will be on sea surface temperature (SST) validation using SST data from in situ and shipborne radiometers as well as from other satellite missions.

OLCI cal/val activities during the Commissioning phase will be focused on radiometric validation using Level 3 water-leaving radiances to verify OLCI absolute calibration and temporal degradation model. Next phase will cover vicarious calibration activities, both for visible and near-infrared channels, validation of the key ocean colour parameter (Chlorophyll concentration) using in situ measurements and Level-3 data. 

Sentinel 3 includes two optical instruments providing simultaneous measurements of the same geographical area: OLCI (Ocean and Land Colour Instrument) and SLSTR (Sea and Land Surface Temperature Radiometer). OLCI is a push-broom imaging spectrometer with five cameras and a total swath width of 1270 km. It acquires data with a spatial sampling of 300m in 21 spectral bands in the [400, 1020] nm spectral range. SLSTR is a dual view conical scanner with a swath width of 1420 km for the near-nadir view and of 750 km for the inclined view. It acquires data for 9 channels in the [0.55, 12] micrometres spectral range with a spatial sampling of 500m for VIS-SWIR channels and 1km for the MWIR-TIR channels.

The combination and co-registration of these two data sets allow the generation of high-quality surface reflectance at all VIS-SWIR channels (except OLCI channels dedicated to atmospheric absorption) and aerosol data over land masses, thanks to the wider spectral range and dual-view, as the SY_2_SYN product. In addition, the continuity with the SPOT-VGT products is ensured by dedicated products: the SY_2_VGP (TOA reflectance) SY_2_VG1 and SY_2_V10 (1-day and 10-days synthesis of surface reflectance and NDVI). These products will be presented and explained as well as an overview of the instruments data co-registration and of the geophysical retrieval algorithms.

From the first months of Sentinel 3 acquisition, the results of the SYN algorithms and the quality of L1/L2 products, as evaluated by the S3-MPC, will be presented and discussed. Finally, potential improvements defined by the SYN ESLs of the S3-MPC will be proposed.

The next generation, post-Sentinel-1, ESA’s C-band synthetic aperture radar (SAR) system is conceived to provide simultaneously high azimuth resolution and wide swath width (HRWS).

There are different ways in which the imaging capabilities of the HRWS SAR system can be exploited, which translate to different operation modes. The more attractive are the wide swath modes, operating in ScanSAR, with 400 km swath width and a resolution of 5m x 5m for single/dual-polarization and, maybe even more noteworthy, the fully-polarimetric 280 km swath width at 5m x 5m single-look resolution. These modes represent a factor four improvement in terms of azimuth resolution with respect to Sentinel-1. Considering also the extended swath or the quad-pol capabilities, the information rate will increase by close to and order of magnitude.

Indeed, wide unambiguous swath coverage and high azimuth resolution pose contradicting requirements on the design of spaceborne SAR systems. Nevertheless, recent studies have shown that by applying Digital Beam Forming (DBF) techniques, such as Scan-on-Receive (SCORE), and Multiple Azimuth Phase centers (MAPS), it becomes possible to overcome these fundamental limitations of conventional SAR systems. The use of MAPS in azimuth enables the decoupling of the high azimuth resolution and wide-swath SAR coverage. It employs a multichannel receiver in combination with mutually displaced multiple aperture elements and the azimuth resolution results determined by the length of the individual sub-aperture elements. At the same time, employing multiple channels in elevation, according to the SCORE technique, allows to collect radar echoes from a wide image swath despite using a receiver aperture with large vertical extension. The trade-off between antenna gain and swath width can thus be relaxed.

In this framework, DLR has reviewed the capabilities of the HRWS SAR system in light of the associated requirements provided by ESA and of the science requirements associated to operational GMES applications. Indeed, many and potentially new applications can benefit from the HRWS SAR operational modes.

Moreover, a HRWS application performance toolkit has been designed and implemented to compare product-level performance for different operating modes and mission scenarios.

The established applications defined within the GMES services and selected for the HRWS performance study are:

Deformation monitoring

Regional land cover

Ocean applications (wind and currents retrieval and oil spill monitoring)

Land ice (wet snow mapping and ice drift)

Sea ice (iceberg detection and ice motion)

Iceberg detection

Security (vessel detection )

Thus, th resulting HRWS toolkit includes for every application analytical expressions or numerical models and, if these are not available, real SAR images as well as numerical algorithms and some explicit simulations of the data and of the inversion process are employed. The tool uses as input the HRWS SAR instrument performance for the different applicable modes and produces as output results comparable with the existing C-band SAR missions.

Due to its wider swath, high resolution and multipolarimetric capabilities, the performance for the HRWS SAR system show a substantial improvement when compared to those of Sentinel-1A, for most of the applications and operational scenarios.

In the final paper a short description of the employed product-level performance models together with the main results will be provided. Furthermore, an analysis based on the different applications performance and on their relative relevance will give a single operational mode for the best compromise.

THEIA is a French national multi-agency organization which promotes the use of satellite data by scientific community and public policy actors. This consortium aims at helping monitoring human and climate impacts on ecosystems and territories by delivering a large panel of products and mutualised services allowing the user community to get the largest benefit of data and products from space missions.

As Data and Services Infrastructure of THEIA, MUSCATE is designed to acquire, process and distribute high resolution satellite images from SPOT 1 to 5 (up to Level-1C), LANDSAT 5-7 and 8, and SENTINEL-2 (up to Level-2A) and covering France territories and worldwide areas of interest.

MUSCATE Data and Services Infrastructure integrates existing CNES components whose quality and efficiency have been proven on other satellite projects, in order to minimize development cost as well as to master computation time: PHOEBUS, SIGMA and MACCS. PHOEBUS is in charge of the processing orchestration: it sends the jobs over distributed resources, manages their priority and, via its interface, operators can monitor processes progress and act when required. It enables to define a specific workflow for each type of acquired product. SIGMA integrates complex algorithms which allows to correct acquisition geometrical model and to orthorectify satellite products. This orthorectification step ensures a correct multi-temporal co-registration of images.

The paper zooms then in particular on Level-2A generation using Multi-Mission Atmospheric Correction and Clouds Screening (MACCS) software. The algorithm implemented in MACCS is briefly presented: conceived by CESBIO, it allows to process temporal series of images at high resolution, high revisit and under constant viewing angles. LANDSAT and SENTINEL-2 data cope perfectly with these pre-requisites. The recursive and multi-temporal algorithm is implemented in a core that is the same for all the sensors and that combines several processing steps: estimation of cloud cover, cloud shadow, water, snow and shadows masks, water vapor content, aerosol optical thickness, atmospheric correction. The algorithm implemented gets its robustness from the use of the temporal dimension to improve the knowledge of the area that has been imaged and make the distinction between what is slowly changing, i.e. the landscape itself, and elements quickly varying such as clouds, clouds shadows and aerosols. This robustness makes this algorithm particularly suitable for Level-2A production in operational conditions, since a same set of Ground Image Processing Parameters can be used worldwide with no need of customization according to the area produced.

After a presentation of MACCS architecture and functionalities and the rationale that triggers automatically Level-2A production in MUSCATE, the paper will give an overview of the production results.

The report presents some features of the development of international university small satellite “Condor”. The proposed international project has to provide the development of spacecraft for remote sensing and integrated observation of ionosphere, which will provide new fundamental results in the field of geophysics and space physics. Subsequently, the obtained data can use for the prediction of earthquakes.

The scientific complex that is used for remote sensing of the Earth's surface and monitoring of space weather and near-Earth space environment on board the satellite is carried out as the part of the project represented.

The main components of the scientific complex are:

-        Package of instruments for space weather measurement.

-        Remote sensing camera.

-        Onboard digital computer, which payload control, collection and processing of scientific data, generation of output packets transmitted to ground control for further analysis.

The package of instruments for space weather measurement consists of:

-        Electron temperature and density probes:

measure temperature and density of the thermal electrons in front of the sensing elements in ionosphere;

study ionosphere variation at the satellite altitude and calibrate tomograph.

-        Gyro-accelerometer for measurement of the variations of density in neutral atmosphere, estimation of the variations of recombination frequency and chemical composition in ionosphere and detection of the satellite braking in the atmosphere expanded due to its local heating above the pre-Earthquake regions.

-        Embedded magnetometer for monitoring of plasma convection in the polar caps, and in particular, the longitudinal currents as the main factors of the magnetosphere-ionosphere relations.

-        External magnetometer. It is located at the distance of the satellite body and measures unperturbed magnetic field around the electron temperature and density probes for data cross calibration;

-        Tomograph. Its main features are:

It is based on photo-multiplier-tubes (PMT): 3 couples of tubes with FOV of ~ 24° for NADIR, 45° for forward and backward directions.

Provides simultaneous observation of the electromagnetic emission of the ionosphere for 630.0 and 135.6nm wavelengths.

Electromagnetic emissions of the ionosphere for 630.0 and 135.6nm wavelengths are typical emissions of the atomic oxygen that is the main component of the F ionosphere layer. Three optical axes with certain orientation allow studying the vertical distribution of ionization in the main ionosphere F-layer and defining the location of the auroral oval and its variations connected with the electron and proton emissions. Luminosity in anomalous line of 630.0nm plays an important role in pre-earthquake areas detection. Variations of the oxygen ion concentration and excitation levels allow controlling such perturbations as solar eclipses, tropical cyclones and typhoons, motion of the equatorial anomaly and auroral oval and so on.

One of the features of the project presented is the co-operation scientific instruments of the flight model of the satellite and its ground-based instruments. The ground-based instruments use the simulators that forms the ionospheric environment on the base of data received from the satellite instruments.

The ideas of Space Plug-and-Play technology were used for the development of the scientific payload complex, that provides the interchangeability of payload components and allow increasing its reliability as well as accelerating the satellite development, assembly and test process.

The Sea and Land Surface Temperature Radiometer, SLSTR, is a development of the Along Track Scanning Radiometers (AATSR) to continue the time series of accurate SST and LST measurements. SLSTR is designed as a dual conical scanning imaging radiometer with a highly curved swath and a gap of up to 900 km between the nadir and oblique views acquired simultaneously. In order to provide L1 and L2 products with a geographical consistency between the two views, as well as to cope with operational constraints, data from different time windows for each view are used to cover the same area on ground. This implies to define for L1 and L2 products a locally quasi-Cartesian regular grid aligned with the sub-satellite track onto which the curved instrument grid pixels are projected. Consequently, some variables and references had to be slightly adapted from their acquisition definition, mainly driven by the scan index.

In the frame of the S3-MPC, efforts are made to help the ESLs and users to get familiar with these new products. After a brief description of the products and their structure, the authors will present keys and methods to go from image grid/definition to acquisition grid/definition.

Examples of products acquired during the first months of acquisitions will be shown.

Aperture synthesis is an interferometric technique in which the signals received by pairs of small antennas are processed in a way to synthesize a single large antenna [1]. This concept has been adapted from radioastronomy to Earth remote sensing. Thanks to this technique, limitations on antenna size in microwave passive remote sensing through satellites have been overcome [2].

A new concept based on a passive spatio-temporal interferometry has been proposed as the newest generation of the well-known SMOS mission successfully operating since November 2, 2009 [3]. The aim of the proposed concept is an enhancement of the currently achieved geometric resolution to meet the stringent user’s requirements in local scale hydrological applications where sub-kilometric resolutions are needed [4]. This interferometric concept is based on the idea of integrating the displacement of the observer (satellite’s antenna), and hence the time variable, in the calculation of the correlation, or visibility, function, which yields the creation of virtual baselines between the positions of antennas at different instants, in addition to the physical ones formed between the instantaneous antennas’ spatial positions.

Unfortunately, the additional information due to the virtual baseline is exactly cancelled by the Doppler shift [5]. But we show that when combining the multi-time correlations in the aforementioned spatio-temporal interferometric system by a Fourier correlation imaging procedure, consisting in cross-correlating the Fourier components at different frequencies of the fluctuating electric fields received by a pair of antennas, the 2D position dependent brightness temperature on the surface of the Earth can be reconstructed. The analytical derivation of the correlation function gives rise to a relationship linking the measured correlations to the position-dependent brightness temperatures by means of a highly oscillatory integral (HOI) kernel. The HOI kernel leads to the remarkable property that a correlation still exists between the signals at slightly different frequencies observed by two different antennas. While existing systems had, until now, only considered the 1D information contained in the correlation at the same frequency, the two–frequency correlation function bears 2D information. Based on this, one is capable, in principle, of reconstructing 2D brightness temperatures starting from a simple 1D geometry, i.e. two antennas arranged perpendicular to the direction of flight.

We test this method of Fourier Correlation Imaging by reconstructing the image of a single point-like source. Based on this, we arrive, at a first estimation, to a spatial resolution of the order of one km.

[1] A. R. Thompson, J. M. Moran, and G. W. Swenson, "Interferometry and Synthesis in Radio Astronomy," New York: Wiley, 1986.
[2] David M. Le Vine, "The sensitivity of synthetic aperture radiometers for remote sensing applications from space," Radio Sci., vol. 25, no. 4, pp. 441-453, 1990.
[3] Y.H. Kerr, P. Waldteufel, J.-P. Wigneron, S. Delwart, F. Cabot, J. Boutin, M. J. Escorihuela, J. Font, N. Reul, C. Gruhier, and al. "The SMOS mission: new tool for monitoring key elements of the global water cycle," IEEE, Institute of Electrical and Electronics Engineers, 2010, 98 (5), pp. 666-687.
[4] Marc F. P. Blerkens, "Global hydrology 2015: State, trends, and directions," AGU Publications: Water Resources Research, 17 Jul, 2015.
[5] D. Braun, Y. Monjid, B. Rougé, and Y. Kerr, "Generalization of the Van Cittert-Zernike theorem: observers moving with respect to sources," submitted to Measurement and Science Technology, 2015.

The Ocean Colour Climate Change Initiative (OC-CCI) provides a long-term time series of ocean colour data and investigates the detectable climate impact. A reliable and stable atmospheric correction (AC) procedure is the basis for ocean colour products of the necessary high quality.

The selection procedure of the atmospheric correction processors is repeated on a regular basis in a round robin exercise, at the latest when a revised production and release of the OC-CCI merged product is scheduled. Most of the AC processors are under constant development and changes are implemented improving in the quality of satellite-derived retrievals of remote sensing reflectances.

The changes between versions of the inter-comparison are not restricted to the implementation of AC processors. There are activities to improve the quality flagging for some processors, and the system vicarious calibration for AC algorithms in their sensor specific behaviour are widely studied. Each inter-comparison also starts with an updated in-situ database, as more spectra are included in order to broaden the range of possible satellite match-ups. While the OC-CCI has been focused on case-1 waters in the past, it has now expanded its goal to provide good case-2 products as well. In light of this goal, new normalisation procedures for the remote sensing spectra have been included in the study.

As in-situ measurements are not always available at the satellite sensor specific central wavelengths, a band-shift algorithm has to be applied to the dataset.

In order to guarantee an objective selection from a set of four atmospheric correction processors, the common validation strategy of comparisons between in-situ and satellite-derived water leaving reflectance spectra, is extended by a ranking system.

In principal, the statistical parameters such as root mean square error, bias, etc. and measures of goodness of fit, are transformed into relative scores, which evaluate the relationship of quality dependent on the algorithms under study.

The sensitivity of these scores to the selected database has been assessed by a bootstrapping exercise, which allows identification of the uncertainty in the scoring results.

A comparison of round robin results for the OC-CCI version 2 and the current version 3 is presented.

The Sentinel-3 Mission Performance Centre provides an environment to perform the calibration and validation of the Sea and Land Surface Temperature Radiometer data.  In this paper the authors present the initial results of the evaluation of the first SLSTR level-1 and 2 products generated during the early mission phase.   

Comparisons against in-situ buoy measurements; ship borne radiometers and other satellite datasets will give an early evaluation of the accuracy of the SST and LST measurements. 

The validation work will also investigate the response of the SLSTR channels over  active fires, gas flares and volcanoes.

Here we describe the compilation of bio-optical in-situ data used for the validation of the second version of ocean-colour products from the ESA Ocean Colour Climate Change Initiative (OC-CCI), as well as results of the validation exercise. The in-situ dataset was compiled from several sources (MOBY, BOUSSOLE, AERONET-OC, SeaBASS, NOMAD, MERMAID, AMT, ICES, HOT, GEPCO) and comprises observations of remote-sensing reflectance ("rrs"), chlorophyll-a concentration ("chla"), algal pigment absorption coefficient ("aph"), detrital and coloured dissolved organic matter absorption coefficient ("adg"), particle backscattering coefficient ("bbp") and diffuse attenuation coefficient for downward irradiance ("kd"). The compiled set of data spans between 1997 and 2012, and has a global distribution. Methodologies are briefly described regarding homogenisation between different sources of data, quality control, removal of duplicated data and merging of all data into one unique table. The final merged table consists of 44191 observations of “rrs”, 39860 observations of “chla”, 1276 of “aph”, 1123 of “adg” and 638 of “bbp” and 2454 of “kd”. The final dataset is discussed in terms of distribution of each variable in time and space and bio-optical relationships. The dataset has been utilised for the validation and uncertainty characterisation of the satellite-derived OC-CCI products. Key findings of the validation of the OC-CCI v2.0 product are given, demonstrating the good performance of the products at the global scale. We have also performed a quantitative assessment of the product performance relative to the GCOS requirements for the Essential Climate Variables (ECVs) of chlorophyll-a and remote-sensing reflectance. The OC-CCI chlorophyll-a product appears to meet the GCOS accuracy requirements and the remote-sensing reflectance is close to meeting the GCOS requirements at multiple wavelengths.

Sentinel-3 onboard sensors include a Sea and Land Surface Temperature Radiometer (SLSTR) with a spatial resolution of 1 km in the thermal infrared channels, providing sea surface temperature (SST) to an accuracy of better than 0.3 K. Both Sentinel-3 A & B SLSTR´s shorten the revisit time to less than a day (www.esa.int).

Accurate and precise knowledge of SST variation plays a crucial role in the climate change perception. Understanding the ocean-atmosphere heat transfers drives scientific and operational communities to comprehend our weather and climate, which ultimately impact all strands of human life. Moreover, the effectiveness of the marine environment protection across Europe, as established in the Marine Strategy Framework Directive, is highly dependent on a good assessment of the environmental status of national marine waters (European Commission, 2008).

As part of the Eastern Atlantic border, the Iberian Peninsula coast is characterized by upwelling regimes and colder water seaward filaments strengthening with northerly summer winds followed by relaxation and flow inversion periods (Relvas et al., 2007, Fiuza et al., 1982). In contrast, the Madeira archipelago waters are considered more stable and warmer along the year and the Azores is characterized by a complex current system that causes strong horizontal temperature gradients (Bashmachnikov et al., 2004).

This study beholds these depicted phenomena through a comparison of satellite-derived SST against coastal in situ moored buoys. Starting in 1996, temperature measurements from 13 buoys belonging to the national network are available with some sparse gaps, nonetheless, providing ground truth to validate satellite data. Undoubtedly, satellite-derived SST, acquired by high temporal resolution remote radiometers, provide the means to continuously assess wide oceanic areas. As technology improves and ground pixels get to less than 1 km of spatial resolution, we push the validation of SST data towards coastal regions.

Satellite-based SST estimates from thermal infrared sensors are validated against a set of coastal moored buoys across the Portuguese mainland and the Madeira and Azores archipelagos. Although thermal IR frequencies are more affected by clouds, aerosols and water vapour, they provide better spatial resolution and are less sensitive to coastal proximity than microwave ones. Thus, we use data from the Advanced Along-Track Scanning Radiometer (AATSR) and its predecessors Along-Track Scanning Radiometers 1 and 2 (ATSR 1-2), as well as the Moderate Resolution Imaging Spectroradiometer (MODIS) complemented with the Advanced Very High Resolution Radiometer (AVHRR) to ensure temporal coverage.

In order to avoid solar heating and minimize the differences between the first thin layer of SST (SST_skin) retrieved from IR satellite radiometers and the subsurface depth SST (SST_bulk) measured with buoy thermistors at 0.7 to 1m depth, we use the nighttime data sets, following the postulated in the work of A. Stuart-Menteth, that temperatures from the surface to 5~10 meters all collapse to the same value before local sunrise (The GHRSST-PP International Project Office, 2008).

Targeting the validation of SST data from satellite against buoy measurements and result discussion, statistical analysis is performed through the RMSE, linear regression and correlations. Ultimately, this research aims to set the ground for validation of the Sentinel-3 SLSTR and its derived L2 products.

Bashmachnikov, I., V. Lafon & A. Martins, 2004. Sea surface temperature distribution in the Azores region. Part II: space-time variability and underlying mechanisms. Arquipélago. Life and Marine Sciences 21A: 19-32.

European Commission, 2008. Directive 2008/56/EC of the European Parliament and of the Council of 17 June 2008 establishing a framework for community action in the field of marine environmental policy (Marine Strategy Framework Directive). Off. J. Eur. Union L164,19–40.

Fiuza, A., Macedo, M., Guerreiro, M., 1982. Climatological space and time variation of the Portuguese coastal upwelling. Oceanologica. Acta 5, 31–40.

Relvas P., E.D. Barton, J. Dubert, P.B. Oliveira, A. Peliz, J.C.B. da Silva e A.M.P. Santos (2007) Physical oceanography of the western Iberia ecosystem: latest views and challenges. Progress in Oceanography, 74, 149-173.

The GHRSST-PP International Project Office, 2008. GHRSST-PP Science & Applications. (https://www.ghrsst.org/science-and-applications/sst-definitions/)

The Sea State Bias affects the range measurement of the altimeters. It is a bias between the true mean sea surface and the mean surface “seen” by the altimeter. It is related to the sea state and is due to several physical effects as the non-linear interactions between large scale waves (skewness bias) and also to hydrodynamic modulation of the short scales by the longer ones (electromagnetic bias). The Sea State Bias is also affected by the processing methodology.

Delay/Doppler altimetry is a new technique that is implemented in Sentinel-3 and that will be operated full time over the world for the first time. The measurement principle and the processing of Doppler altimeters differ from conventional altimeters and consequently bring some questions about the Sea State Bias:

The SSB is related to the difference between the      mean of the electromagnetic surface and the mean of the geometric surface.      Doppler processing technique uses Doppler beams acquired with different looking      angles. Although these angles are very small, does the mean electromagnetic      surface depend on these angles?

The received signal of these altimeters is      Doppler processed to improve the along track resolution. Is the mean sea      surface elevation affected by this new processing?

For an altimetry mission, any anticipated knowledge of the SSB behavior, before launch, would allow to anticipate the activities and efforts needed in the data processing. Unfortunately, despite numerous efforts, no theoretical reliable evaluation of the SSB is operationally used and the SSB corrections are still performed using real data with crossover or along-track differences techniques using the altimeter estimates of Significant Wave Height (SWH) and wind speed as input parameters.

The study presented in this talk is based on two different methods allowing independent analyses and then comparison between them. The first method consists in a theoretical analysis using a mathematical development based on ocean surface spectra and electromagnetic models. The second method is based on simulation tools. Indeed, in the last years, several simulation activities have been performed in CLS with CNES collaboration for the analysis of the SSB. The simulators are based on a realistic modeling of the ocean surface, of the instrument behavior and of the on-ground processing. Thus, it is possible to introduce several physical phenomena on the simulated surface and assess the corresponding SSB by applying the instrument on-board and on-ground processing and compare the results to theory.

These investigations allowed us to deduce that the Doppler beams looking angles are small enough to not add a complementary bias to the nadir looking one and that the Sea State Bias for the Delay/Doppler altimetry is equivalent to the one observed in conventional altimetry. In the end, a comparison has been performed against an independent empirical SSB assessment on CryoSat-2 SAR data over ocean and showed a consistency between the different methods.

This presentation will give an overview of the results obtained using theoretical analysis of the SSB and those obtained by the simulation. We will also show the comparison results with the estimates of the SAR mode deduced from the Pseudo-LRM mode. Considering that Sentinel-3 data will be available very soon, we will be able to conclude with a comparison to the first SSB empirical estimates using real data.

The Radiative Transfer Model (RTM) SCIATRAN provides calculations of radiation between 175 to 2400 nm. Recently the model was extended for a coupled atmosphere-ocean system to provide calculations of the light field in the ocean body including water constituents (Rozanov et al. 2014). Additionally, it accounts for all optical relevant parameters (absorption, elastic and inelastic scattering). Three different forms of inelastic scattering as transpectral processes are included in the model: Vibrational-Raman-Scattering, fluorescence of chlorophyll-a and CDOM (colored dissolved organic matter). To study the influence of absorption, elastic and inelastic scattering processes of oceanic water on satellite measurements for different light field conditions, the Top-Of-Atmosphere (TOA) radiance for 23 different chlorophyll-a concentraions and 14 solar zenith angles have been modeled. In addition, in-water radiative flux calculations have been performed for the same scenarios. The relationship between in-water flux and TOA radiance calculations are investigated and comparisons with in-situ ship and SCIAMACHY satellite measurements are shown in this study.

References:

Rozanov, V.V., Rozanov, A.V., Kokhanovsky, A.A. and Burrows, J.P.: Radiative transfer through terrestrial atmosphere and ocean: Software package SCIATRAN, J. Quant. Spectrosc. Ra., 133, 13–71, doi:10.1016/j.jqsrt.2013.07.004, 2014

The need for high-frequency, high-quality optical measurements at sea level to support and validate satellite measurements of ocean colour has long been recognised (Zibordi, Berthon, et al. 2009). With the aims of creating new satellite products for Sentinels 2 and 3, and greatly increasing the availability of marine in situ data for validation, the HIGHROC proposal was submitted to Framework 7, and subsequently received funding for four years. In work package 5 of the project, a large dataset of in situ measurements of water quality will be generated using automated SMARTBUOY and FERRYBOX systems to generate high quality measurements of parameters such as suspended sediment load, chlorophyll concentration and underwater light penetration. These are the types of data which are typically used in national assessments of water quality, and in Environmental Statements by maritime industry. Utilisation of automated in situ measurements can greatly increase the number of match-ups with satellite data (Neukermans et al. 2012).

In addition to the in-water measurements, readings of the radiance leaving the water surface are very important for satellite validation (Zibordi, Mélin, et al. 2009, Zibordi et al. 2015). The water-leaving radiance (Lw) is equivalent to the radiance measured by a satellite after correction for attenuation of light in the atmosphere. There are at present no suitable operational measurements of Lw in UK waters, indeed there were none in the whole of the North Sea prior to HIGHROC. To improve the coverage of sea-level radiance in the North Sea, Cefas and the University of Hull are working together with the Offshore Renewable Energy Catapult centre in the development of an AERONET ocean-colour measuring station to be located at an experimental meteorological tower in the northern North Sea. The presentation will describe the steps involved in the design and installation of the new AERONET-OC site, and show initial results of the optical characterisation of the case-2 waters in the northern North Sea.   

Neukermans G, Ruddick KG, Greenwood N (2012) Diurnal variability of turbidity and light attenuation in the southern North Sea from the SEVIRI geostationary sensor. Remote Sens Environ 124:564–580

Zibordi G, Berthon JF, Mélin F, D’Alimonte D, Kaitala S (2009) Validation of satellite ocean color primary products at optically complex coastal sites: Northern Adriatic Sea, Northern Baltic Proper and Gulf of Finland. Remote Sens Environ 113:2574–2591

Zibordi G, Mélin F, Berthon J-F, Holben B, Slutsker I, Giles D, D’Alimonte D, Vandemark D, Feng H, Schuster G, Fabbri BE, Kaitala S, Seppälä J (2009) AERONET-OC: A Network for the Validation of Ocean Color Primary Products. J Atmos Ocean Technol 26:1634–1651

Zibordi G, Mélin F, Berthon J-F, Talone M (2015) In situ autonomous optical radiometry measurements for satellite ocean color validation in the Western Black Sea. Ocean Sci 11:275–286

SAR mode altimetry over the ocean was demonstrated for the first time in orbit with the ESA Cryosat-2 mission. Scientific studies since then have convincingly established the improved performance of SAR altimetry in terms of reduced altimetric noise, finer along-track spatial resolution and improved performance near land. Currently, no solution is available for sea state bias in SAR mode, and there remain uncertainties about the sensitivity of SAR mode altimetry to long ocean surface waves (swell) and their direction of propagation. We present an analysis of our investigation into these issues using available SAR altimetry, Envisat ASAR and Wavewatch 3 data, as well as a compilation of the arguments in support of these results.

High-accuracy analysis of data from Global Navigation Positioning System (GNSS) is able to provide, as a by-product, precise tropospheric corrections for the effect of neutral atmosphere delaying signals observed by ground receivers.

There are several ways in which meteorology can benefit from GNSS tropospheric products. First, these products, when made available in (near) real-time, are suitable for supporting numerical or non-numerical weather forecasting.

Second, final GNSS tropospheric products generated for the highest product accuracy, i.e. without requesting a strict short latency, are valuable for an independent assessment or inter-comparison of numerical weather models. They are stable in long-term and provide high accuracy and good temporal and spatial resolution.

The EUMETNET EIG GNSS water vapour programme, E-GVAP, provides near real-time GNSS tropospheric products used already by several meteorological services in Europe. We illustrate both mentioned usages of the products on the case of an ensemble forecast system based on the GEFS global ensemble (NCEP, USA). The ensemble has 20 members with horizontal resolution of cca half-degree and can be used to produce probabilistic forecasts of different meteorological variables.

We investigate the impact of assimilating GNSS delays taken from the E-GVAP database on an ensemble forecast for Central Europe, based on downscaled members of the GEFS ensemble. The downscaling is performed with the mesoscale numeric model WRF-ARW (NCAR). Conventional data and GNSS delays have been assimilated with the help of different techniques like 3DVar and Ensemble Filtering. Assimilation methods have been tested for the downscaled ensemble. Their performance and the influence of GNSS delay data will be illustrated on selected episodes.

Independently, the mesoscale model characteristics, in terms of GNSS tropospheric delay parameters, are assessed with the final GNSS product and compared with the global model used for the downscaling.

The combination of topographic and hyperspectral information is usually realized by using the distance measuring capabilities of LIDAR and the reflectance measuring capabilities of the hyperspectral sensors. Usually both data entities are generated in separated workflows except for the orthorectification and atmospheric correction procedure of the hyperspectral data. The combination of data entities is limited to the gridded digital elevation model of the LIDAR scanner and to the gridded hyperspectral data cube. This separation results in discretization and substantial information loss. The sensors’ potential, especially in combination, has not been exhausted yet. The presented fundamental fusion of the passive and active sensor characteristics is aimed at improving and developing the complete, sensor inherent data density. All synergies in terms of geometric and spectral data alignment as well as process refinement are acquired.

The data fusion approach is separated into two major processing steps, the geometric and the spectral alignment. The optimal geometric alignment based on rigorous parametric co-registration is the key aspect of the proper sensor fusion. Subsequently the spectral alignment based on integrated radiometric cross-calibration of the LIDAR intensities and hyperspectral intensities is realized.

The main linkage between both sensors is the detection of intensities reflected by a surface object in the overlapping wavelength range. In order to make both resulting intensity images comparable, the characteristic of the respective sensor has to be adapted concerning spatial and spectral properties. The time of flight capability of the LIDAR sensor allows orthorectified intensity images which serve as geometric reference. Based on this adopted LIDAR intensity image a rigorous parametric co-registration of the hyperspectral data is performed. This iterative geometric alignment optimizes the flight attitude and position parameters based on the collinearity principles. It does not require additional resampling steps and it is realized by an automated and adjustable tie point detection algorithm incorporating all adjacent flight stripes and intensity information overlaps. The parametric co-registration procedure incorporates ray tracing technics to overcome the grid inherent discretization and to ensure the highest geometric accuracy as well as the highest inherent resolution.

The ray tracing approach also establishes capabilities for further advanced radiometric and spectral sensor fusion.For the spectral alignment, the active measurement principle of the LIDAR sensor is exhausted to compensate the radiometric deficiencies of the hyperspectral sensor, the unknown illumination intensities and unknown morphology of the reflecting targets.The subsequent spectral alignment is based on the iterative adaptation of the hyperspectral intensity data to the radiometric calibrated LIDAR intensities. This is realized by difference modeling of the overlapping intensities of the LIDAR intensities and hyperspectral intensities.

This characteristic advantages are used for compensating geometric, resolution and radiometric deficiencies of the hyperspectral sensor. The active sensor characteristic can also improve geometric sharpening, BRDF correction, spectral un-mixing, and de-shadowing of the hyperspectral data. On the other hand, the spectral resolution of the hyperspectral sensor can be used to assign a single spectral signature to every LIDAR point. In addition, this significantly reduces impacts of central projection on the geometric accuracy. However, this work is embedded in the framework of the spaceborne hyperspectral EnMAP mission to provide both the optimal base for preliminary EnMAP and Seintinel-2 product simulations and an automatic approach for airborne CAL/VAL campaigns of spaceborne missions.

ESA and CNES have initiated preliminary studies to investigate the interest and feasibility of an ocean colour geostationary mission in the last decade. The scientific community is strongly supporting this initiative as a geostationary mission would open new perspectives in terms of scientific breakthrough, algorithm development and environmental monitoring. A geostationary sensor would notably offer a unique opportunity to investigate diurnal cycles of phytoplankton growth from oligotrophic to eutrophic regions. In the dynamic coastal areas, this would offer a potential to analyse fast changing suspended sediment load, due to tidal effect, river discharge or coastal drift. In cloudy regions, a geostationary mission would maximize the number of useful scenes therefore improving the potential monitoring applications.

An ocean colour mission (GOCI on-board COMS) was launched in 2010 by the Korean Space Agency (KARI/KORDI) demonstrating the potential of an Ocean Colour geostationary mission. Early 2015, the CNES has launched a phase A study for a mission covering the entire Earth disk, called OCAPI. The current mission specification recommends a mission centred over Europe (0°E) with a revisit time of one hour and a spatial resolution of 250m. With a location at 0°e, most of European waters, western North Atlantic, the entire South Atlantic and Africa’s western coasts would be sampled.

The mission science objectives and technical specifications will be presented. They are deliberately ambitious and they represent scientific and technological challenges. These scientific and technological breakthroughs represent an advanced level for EU Earth Observations programs.

In 2012, ENVISAT mission was interrupted, after 10 years of altimetric measurements over ocean. But, since the end of the mission, Envisat data are still ingested in MyOcean off line data products and the quality of its data keeps on improving.
After the first official reprocessing of 2011, teams are now getting ready for the second whole mission reprocessing. The development and pre-validation of all the selected evolutions is now finished and most of the expected results can already be anticipated.
Part of the evolutions are Envisat specific instrumental improvements proposed by the Expert Support Laboratories (CLS, isardSAT, UoP…). The other evolutions aim at keeping the mission as compliant as possible with the new missions (Sentinel 3…) best standards: for instance Tides, Mean Sea Surface, Orbits….
This paper presents the expected improvements of Envisat data quality over ocean (new orbit, tide models, sea state bias, wet tropospheric corrections…). It relies on monomission calval diagnosis as well as on comparison to other flying precise altimetric missions and in situ external reference.

For all nadir altimetry missions, the quality of the orbit ephemerides is crucial for the computation of the Sea Surface Height (SSH). Impacting mostly large scales, spatially and temporally, the errors attributed to the orbit are regularly quantified and analyzed precisely for all the altimetric missions (Envisat, Jason-s, Cryosat, Sentinel 3…) in the frame of CNES and ESA data performance assesment at CLS.
Thanks to the recent progress of Precise Orbit Determination teams, the quality of short term MOE orbits (Preliminary Orbit Estimation) computed within a 3 days latency is getting closer and closer to the POE (Precise Orbit Estimation) computed within a month latency.
Combining several calval tools and skills, we propose an analysis of the discrepancies between those solutions in terms of Sea Surface Height performance. Using mono-mission and multi-missions diagnosis as well as in situ database comparisons, different spatial and temporal scales are analyzed and discussed to address the main applications domains.

A new retracker has been developed recently for the LRM mode altimetry products to reduce the bump effects impacting the Sea Surface Height spectra. This bump is due to inhomogeneities of the ocean surface (blooms, rains cells or any other phenomenon) that affect the altimeter waveforms shape so that the retracking Brown model is no longer appropriate to process them. This bump degrades the observation of ocean scales smaller than 100 km.

This retracker (DCORE for parameters DeCOrrelation REtracker) is based on a modified waveform analytical model that decorrelates the surface estimated parameters so that the deformations of the trailing edge have the least impact as possible on the estimates. First, this retracker has been tested using simulated and few cycles of real Jason-2 data. Then, a whole year of Jason-2 data has been compared to SAR CNES CPP (CryoSat-2 Processor Prototype) CryoSat-2 SAR mode data over the Agulhas Current. It has been shown that this retracking significantly reduces the spectral bump but also the noise on the estimated parameters.

More recently, this retracker has been applied to CryoSat-2 CPP PLRM data in order to assess its performances on altimetry waveforms affected by a higher speckle noise due to the low number of independent pulses averaged in the PLRM mode. The bump and noise reduction are consistent with what has been shown on the LRM mode. The objective of this analysis is to demonstrate the high advantages of this retracking for the SAR mode because it enhances the PLRM mode products quality and then makes available higher quality reference data for the SAR mode validation.

This presentation will first provide a description of the DCORE retracker and the main analysis results on LRM and PLRM data and then will mainly focus on the high advantages of this retracking for Sentinel-3 SAR mode validation to conclude on our recommendations to implement it on Sentinel-3 processing.

For the preparation and performance monitoring of the future generation of hyperspectral InfraRed sounders dedicated to the precise vertical profiling of the atmospheric state, such as the Meteosat Third Generation hyperspectral InfraRed Sounder, a reliable assessment of the instrument radiometric error covariance matrix is needed.

Ideally, an inflight estimation of the radiometrric noise is recommended as certain sources of noise can be driven by the spectral signature of the observed Earth/ atmosphere radiance. Also, unknown correlated noise sources, generally related to incomplete knowledge of the instrument state, can be present, so a caracterisation of the noise spectral correlation is also neeed.

A methodology, relying on the analysis of post-retreival spectral residuals, has been designed and implemented to derive in-flight the covariance matrix on the basis of Earth scenes measurements.This methodology was  successfully demonstrated using IASI observations as MTG-IRS proxy data and made it possible to highlight anticipated correlation structures explained by apodization and micro-vibration effects (ghost). This analysis was corroborated by a parallel estimation based on an IASI black body measurement dataset and the results of an independent micro-vibration model.  

Recently we showed that it is possible to compensate the amplitude variations of the delay-Doppler map that are not related to the sea state and likewise also compensate for a dilation of range that causes a variation along track in the power. With both amplitude compensation (AC) and dilation compensation (DC) the resultant AC/DC delay-Doppler map is much simpler (Ray et al., 2015).

Here is presented an application of that simplified delay-Doppler map (or stack).
It is shown that because of the simple shape of the AC/DC delay-Doppler map the antenna pointing can be computed algebraically from the AC/DC delay-Doppler map. 
This technique has been validated using simulated data, and then it is used to estimate the bias of the CryoSat star-tracker and compared to other previously presented methods.

This work has been carried out under the ESTEC/ESA contract to develop the Sentinel-6 Poseidon-4 Level 1 Ground Processor Prototype.

Ref.: Chris Ray, Mònica Roca, Cristina Martin-Puig, Roger Escolà, Albert Garcia; "Amplitude and Dilation Compensation of the SAR Altimeter Backscattered Power", IEEE Geoscience and Remote Sensing Letters(2015); DOI   10.1109/LGRS.2015.2485119

The semi-arid region of Copiapó, Chile, is dominated by olive orchards and vineyards and it is affected by a sever water scarcity and low ground water supply. In this zone, an assessment of water use efficiency based on evapotranspiration retrievals is key for sustainable water use. To address this point, disaggretation methods are useful in order to improve the surface energy flux retrievals at considering a better estimation of fraction vegetation cover in the semi-arid regions. To this end, this work presents a method to disaggregate the Land Surface Temperature (LST) for evapotranspiration retrievals. Data set from Landsat-8 and MODIS (MOD 11-LST) was used in the work from 2013 to 2015. The LST retrieved from Landsat-8 and MODIS data was assimilated in order to provide a 8-day composite product of LST at 100 m resolution. The disaggregation method was developed by taking into consideration the direct relationships between Landsat-8 and MODIS LST and between Landsat-8 LST and Normalized Difference Vegetation Index (NDVI). After then, both dissagregate LST (including LST and NDVI) were weighted by the temporal parameters which relates both LST at 1 x 1 km and 100 x 100 m. In-situ data made in vineyards and olive orchards were used to estimate the LST and the actual evapotranspiration (ETr) accumulated for 8 days. The ETr was also estimated by the SEB model by using completary meteorological station. The disaggregation method was calibrated during 2013 – 2014 and validated during 2015 in order to provide LST and ETr images at 8 days.

Results showed that disaggregated LST presents a RMSE lower than 3 K when comparing to in-situ measurements. The retrieved ETr presents good results in terms of seasonal variations and comparing to in-situ measurements. The ETr difference between 1 x 1 km² and 100 x 100 m² over growing seasonal crops such as vineyards are, in average, 10 mm (8 day)^-1 for the summer season, although in the case of sparse vegetation such as Olives no statistical significant differences were achieved (p<0.05) since the soil and vegetation fractions are almost similar during the whole season. The RMSE between disaggregated and in-situ ETr for both olives and vineyards orchards was less than 7 mm (8 day)^-1. The method presented here can be also adapted for future satellite missions such as the Sentinel-3 which will include thermal sensors capable to provide LST in order to derive ETr products.

An update of the data on the bio-optical characterisation of Portuguese waters at the Sagres site is presented  as a contribution to the upgrade and validation of  algorithms for the MERIS  4^th Reprocessing . This  involves the reprocessing the data set collected between 2008 and 2012 [1] ; much of this data has now been published in peer refereed journals [2,3,4], but there will now be a reassessment  of  this data and  then measurements protocol will be revised and updated based on the decisions taken . In  addition there has been further sampling since 2012 to continue with the goal of validating the bio-optical models that will be used for the 4^th reprocessing  and, in particular, to focus on the applicability of the OC4Me algorithm for  Sagres waters based on the documentation associated with MERIS processing and validation [5].

In addition, the Sagres data has been used to produce a  regional total chlorophyll a product (Chla)  using  an inversion scheme based on Multi-Layer Perceptron (MLP) neural nets, where the regional results are expressed as MLP(Rrs^SITU) and MLP(Rrs^MER), and where the quantities in parentheses indicate MLP inputs. The range, for which the regional algorithm is valid, has been evaluated through a novel detection scheme [6]. Different sets of wavelengths have been tested to derive the regional Chla concentrations from space-born Remote Sensing Reflectance ( Rrs). Results indicate that applications based on Rrs at 490, 510 and 560 nm allow for improvements in the quality of data products within the applicability range for the regional algorithm [7].

Finally an intercomparison has been implemented  for  the determination of algal pigments by High Pressure Liquid Chromatography (HPLC) to guarantee consistency and comparability of Chl a databases for validation of satellite products for the Portuguese coast at two laboratories involved in ocean colour validation activities in Portugal. The overall aim has been to :  quantify single laboratory uncertainties;  quantify differences between the two laboratories considering both HPLC methods and extraction procedures.

[1]  Icely, J., Cristina, S., Goela, P., Moore, G., Danchenko, S., Zacarias, M., Newton, A., 2013. Summary of the project outputs between 2008-2012 for Technical Assistance for the Validation of MERIS Marine Products at Portuguese Oceanic and Coastal Site. Final Report,   Annex A 81 pp., Data - Tables 102 pp., Annex B - Dissemination and Education 6pp. and copies of posters & publications.

[2] Cristina,S.,  Moore, G., Goela,P., Icely,,J., Newton,A.,  2014. In situ validation of MERIS marine reflectance off the southwest Iberian Peninsula: assessment of vicarious adjustment and corrections for near-land adjacency. International Journal of Remote Sensing, 35(6): 2347–2377.

[3] Goela, P.C., Icely, J., Cristina, S., Newton, A., Moore, G., Cordeiro, C., 2013. Specific absorption coefficient of phytoplankton off the Southwest coast of the Iberian Peninsula: A contribution to algorithm development for ocean colour remote sensing. Continental Shelf Research, v. 52, pp. 119-132. DOI: 10.1016/j.csr.2012.11.009   ISSN: 0278-4343

[4] Goela, P.C., Danchenko,S. Icely, J., Lubian, L.M., Cristina, S., Newton, A., 2014. Using CHEMTAX to evaluate seasonal and interannual dynamics of the phytoplankton community off the South-west coast of Portugal. Estuarine, Coastal and Shelf Science, v. 151, pp. 112-123. DOI: 10.1016/j.ecss.2014.10.001  ISSN: 0272-7714

[5] https://earth.eo.esa.int/ocs/envisat/meris/documentation

[6]. D’Alimonte,] D MelinF. , . ZibordiG,  Berthon, J.-F. ,  2003., Use of the novelty detection technique to identify the range of applicability of empirical ocean color algorithms. Geoscience and Remote Sensing, IEEE Transactions on, vol. 41, no. 12, pp. 2833–2843.

[7]     Cristina, S., Davide D’Alimonte,D.,  Priscila Costa Goela,P., , Tamito Kajiyama;T. , Icely,J. G Moore, G.,  Fragoso;B.,  Newton,A. Standard and regional bio-optical algorithms for chlorophyll a estimates in the Atlantic off the Southwestern Iberian Peninsula. (under submission).

Synthetic Aperture Radar (SAR) data are of key importance in Earth resource management and natural hazards monitoring. However, as demonstrated by the recent Huygens-Cassini mission to Saturn and its moons and the ESA planned JUICE (Jupiter Icy moons Explorer) mission, SAR sensors and imagery are an essential tool also for analysis of other celestial bodies [1]. In the latter context, one of the most important issue is the possibility to obtain a first (rough) estimation of the topography, named Digital Elevation Model (DEM). This problem can be arguably considered no more interesting on Earth, where the SRTM mission provided a rough DEM of the entire terrestrial globe,  and where LIDAR data, providing much more accurate DEMs, are already available, at least in the most developed countries.

In the last decades, several techniques aimed at estimating the local topography of the sensed surface have been developed, as stereoscopy [2], [3], shape from shading (or radarclinometry) [4]-[6], polarimetry [7], interferometry [8]. At variance with other methods, the shape-from-shading approach presents the very interesting feature of requiring only a single SAR image, that translates in a much simpler system architecture and image acquisition process, key features in planning space missions. However, a vaguefeeling of skepticism accompanies application of Shape-from-Shading (SfS) techniques on SAR images. SfS is nicely valued if optical images are in order; but within the radar community (which also terms it, radarclinometry) SfS is commonly considered impracticable. This impracticability is partially due to the use of inadequate models describing electromagnetic scattering from a natural scenario, as the Lambertian model commonly used in the radarclinometry community.

In addition, the complexity of the SfS problem, can be significantly reduced considering the problem of the local incidence angle estimation instead of the local height estimation. An accurate inversion procedure aimed at retrieving the local incidence angle map is a necessary step to a reliable DEM estimation via the SfS approach. It is worthy to note that an adequate modelling of the SAR image is required. In this paper we provide an inversion approach to the problem of local incidence angle estimation from a single SAR image of natural (terrestrial or not) landscapes. In particular, the proposed method is based on fractal electromagnetic and surface models, adequate to describe natural surfaces and corresponding electromagnetic scattering processes.

Besides the height retrieval problem, local incidence angle estimation has other somehow surprising applications, in primis despeckling, as recently shown by the authors [9], [10].

References

[1] O. Grasset et al., "JUpiter ICy moons Explorer (JUICE): An ESA mission to orbit Ganymede and to characterise the Jupiter system," Planetary and Space Science, vol. 78, pp. 1-21, April 2013.

[2] U. R. Dhond and J. K. Aggarwal, "Structure from stereo – a review," IEEE Transactions on Systems, Man and Cybernetics, vol. 19, no. 6, pp. 1489-1510, Nov. 1989.

[3] I. Dowman, C. Pu-Huai, O. Clochez, and G. Saundercock, "Heighting from stereoscopic ERS-1 data," Proceedings Second ERS-1 Symposium, pp. 609-614, 1993.

[4] D. N. Ostrov, "Boundary Conditions and Fast Algorithms for Surface Reconstructions from Synthetic Aperture Radar Data," IEEE Trans. Geosci. Remote Sens., vol. 37, no. 1, pp. 335-346, Jan. 1999.

[5] R. L. Wildey, "Radarclinometry for the Venus radar mapper," Photogramm. Eng. Remote Sensing, vol. 52, no. 1, pp. 41-50, 1986.

[6] M. J. Brooks and B. K. P. Horn, Shape From Shading. Cambridge: MA: MIT Press, 1989.

[7] D. L. Schuler, J. S. Lee, and G. De Grandi, "Measurement of topography using polarimetric SAR images," IEEE Trans. Geosci. Remote Sens., vol. 34, no. 5, pp. 1266-1277, Sep. 1996.

[8] D. Massonnet and T. Rabaute, "Radar interferometry: limits and potential," IEEE Trans. Geosci. Remote Sens., vol. 31, no. 2, pp. 455-464, 1993.

[9] G. Di Martino, A. Di Simone, A. Iodice, D. Riccio and G. Ruello, "Non-Local Means SAR Despeckling Based on Scattering," in IEEE Int. Geosci. Remote Sens. Symp., Milan, Jul. 2015.

[10] G. Di Martino, A. Di Simone, A. Iodice and D. Riccio, "Scattering-Based Non-Local Means SAR Despeckling," submitted to IEEE Trans. Geosci. Remote Sens.

The ionosphere is the portion of the upper atmosphere, where ions and electrons are present in quantities sufficient to affect the propagation of electromagnetic waves. Radio waves passing through ionosphere experience transmission delay proportional to the Total Electron Content (TEC) that is the total number of electrons present along a path between a satellite and a receiver. TEC is a highly variable quantity influenced by different helio-geophysical parameters such as solar activity, season, time of the day. Among the others effects, fluctuations of TEC can produce streaks in SAR images (Gray et al., 2000, Geophysical  Research Letters).

Such study aims at catching qualitative correspondences between spatial and temporal gradients of the TEC derived from GNSS and the occurrence of streaks in ALOS-PALSAR images. To carry out this study, an analysis will be applied over Italy, using GPS data from the RING (Rete Integrata Nazionale GPS) network managed by INGV (Istituto Nazionale di Geofisica e Vulcanologia). The area covered by this study has been chosen for a twofold reason: (a) RING is a dense network, composed by more than 100 receivers and offers a fine representation of the ionospheric features as derived from GNSS signals; (b) ionosphere over Italy, being located at middle latitude, allows the characterization of the impact on SAR when SAR is less affected by the erratic behavior of the electron density in the ionosphere. To study the TEC fluctuation effect on ALOS-PALSAR images it is assumed that the ionosphere is a thin layer located at an altitude of 350 km above the Earth’s surface. This is a suitable assumption for the mid latitude ionosphere. Moreover ALOS ascending passages over the area of interest occur during night and a nightly, mid-latitude ionosphere can be assumed to be “frozen” during a time interval of minutes. So it is possible calculate the TEC fluctuation in a interval of few minutes around the two passages of ALOS increasing the number of observations available. In order to minimize the biases affecting the GNSS measurement (Ciraolo et al., 2007, Journal of Geodesy), TEC values, calculated along each ray path (slant TEC), are processed with a calibration technique to obtain “calibrated” slant TEC. Calibrated slant TEC values are then projected to vertical in order to have TEC values which are almost geometry free, i.e. not dependent on the position of the GPS receivers at ground. In order to characterize ionospheric conditions affecting two different passages of ALOS, TEC values are mapped over a fine regular grid by using the natural neighbour interpolation (Cesaroni, 2015, PhD thesis). This procedure allows to calculate a differential TEC map to be compared with the streaks on the ALOS-PALSAR image. The preliminary results are shown.

In the ESA-DUE GlobAlbedo project (http://www.GlobAlbedo.org), a 15 year record of land surface albedo was generated from the European VEGETATION & MERIS sensors using optimal estimation. This was based on 3 broadbands (0.4-0.7, 0.7-3, 0.4-3µm) and fused data at level-2  after converting from spectral narrowband to these 3 broadbands with surface BRFs. A 10 year long record of land surface albedo was generated from Collection 5 of the MODIS BRDF product for these same broadbands. This was employed as an a priori estimate for an optimal estimation based retrieval of land surface albedo when there were insufficient samples from the European sensors. This so-called MODIS prior was derived at 1km from the 500m MOD43A1,2 BRDF inputs every 8 days using the QA bits and the method described in the GlobAlbedo ATBD which is available from the website (http://www.globalbedo.org/docs/GlobAlbedo_Albedo_ATBD_V4.12.pdf).  In the ESA-STSE WACMOS-ET project, FastOpt generated fapar & LAI based on this GlobAlbedo BRDF with associated per pixel uncertainty using the TIP framework.

In the successor EU-FP7-QA4ECV* project, we are working on the generation of a 35 year record of Earth surface spectral and broadband albedo (i.e. including the ocean and sea-ice) using optimal estimation for the land and where available, relevant sensors for “instantaneous” retrievals over the poles (Kharbouche & Muller, this conference). This requires the longest possible land surface spectral and broadband BRDF record that can only be supplied by a 15 year of MODIS Collection 6 BRDFs at 500m but produced on a daily basis. Due to delays in C6 processing, Collection 5 data have been re-processed over a 15 year time period from 3/2000-3/2015 using the CEMS Big Data computer at RAL to generate 7 spectral bands and 3 broadband BRDF with and without snow and snow_only. This product will be presented and the broadband new prior compared against the previous MODIS “prior” and the spectral prior against the ADAM (http://adam.noveltis.com/) 3-year climatology. The impact of the new MODIS “prior” will also be evaluated on the resulting GlobAlbedo broadband products and future prospects for the 35 year time series will be discussed.

We will discuss the progress made since the start of the QA4ECV project on the production of a new fused land surface BRDF/albedo spectral and broadband CDR product based on four European sensors: MERIS, (A)ATSR(2), VEGETATION, PROBA-V and two US sensors: MISR & MODIS. For the European sensors, an uniform atmospheric correction scheme has been employed to generate spectral BRF products and these have all been mapped into MODIS spectral bands whilst the US sensors have employed their own level-2 BRF retrieval schemes with associated uncertainty information. Progress is also demonstrated on the use of TIP for fapar/LAI retrieval from the broadband BRDFs as well as fapar from AVHRR based on retrievals from MERIS and OLCI. In parallel, work has taken place at two of our partners on the production of a new geostationary broadband BRF and associated albedo and this will also be reported on.

* QA4ECV has received funding from the European Union’s Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 607405

The NASA Cloud Absorption Radiometer (CAR) is an unique airborne multi-wavelength scanning radiometer (iFOV = 1 deg.; field of regard= 190 deg.) that can perform several functions including (1) measuring the angular distribution of scattered radiation of various surfaces types, (2) determining the single scattering albedo of clouds at selected wavelengths in the UV, visible, and near-infrared, and (3) acquiring imagery of clouds and the Earth’s surface from multiple angles. It has acquired an enormous amount of multi-angular data over sites in the US and Africa since 1998 in near-annual flight campaigns.

CAR was designed to operate from a position mounted on an aircraft, primarily in the nose-cone, so that its scan is unimpeded as it scans from zenith to nadir. CAR is capable of measuring scattered light in fourteen spectral bands. Eight data channels in the range 0.34 – 1.27 µm are sampled simultaneously and continuously, while the ninth data channel is selected from among six spectral channels (1.55 – 2.30 µm) on a filter wheel. The filter wheel can either cycle through all six spectral bands at a prescribed interval (usually changing filter every fifth scan line), or lock onto any one of the six spectral bands and sample it continuously.  CAR calibration is traceable to the National Institute of Standards and Technology (NIST). The estimated errors associated with radiometric calibration vary from ±1% to ±3% for all spectral channels.

CAR operates ether in full spherical mode as the aircraft proceeds along a flight-path or in stare mode for 2 minute loops at different altitudes looking at a fixed zenith and a complete set of azimuths. Its spectral channels are chosen to match MODIS but they also match a variety of European sensors such as MERIS and VEGETATION. Recently, the CAR data (http://car.gsfc.nasa.gov/) have been re-processed and have been compared against a variety of spectral BRF derived from different spaceborne EO sensors. This represents an unique capability to validate spectral BRF, particularly in the Sentinel era and the opportunity for future validation of multi-spectral BRF will be discussed in the context of existing efforts for validation of Sentinel 2 & 3 products.

The correction and/or masking of optical satellite imagery for atmospheric effects such as aerosols and cloud is an important first step in the use of such data. The correction of imagery typically involves running at atmospheric model, such as 6S or Modtran, to estimate the path radiance for the scene and therefore induce the surface reflectance. Parameterising these models can be difficult and time consuming as parameters such as aerosol optical thickness (AOT) vary over short temporal and spatial baselines and ground measures are either spatial or temporally limited for the correction of imagery. Alternative sources such as MODIS are also limited in spatial resolution and have an offset in acquisition time. Therefore, it is ideal to estimate these parameters from the input image itself.

ARCSI is an open source software package to automatically process satellite data, including the derivation of atmospheric parameters through an inversion of the 6S model. To date all Landsat sensors, Rapideye and WorldView-2 are supported with others being added, such as Sentinal-2, and functionality includes: a) parsing of sensor specific image header files, b) conversion to radiance, c) calculation of top of atmosphere reflectance, d) automatic no data region masking (e.g., LS7 bands), e) automatic cloud masking, e) dark object subtraction, f) estimation of AOT surfaces, g) correction to surface reflectance using a single parameterisation of 6S and h) full correction to surface reflectance using a look up table (LUT) for elevation and AOT.

The ARCSI software is simple to parameterise with global support and allows for rapid batch processing of satellite imagery and support for high performance computing (HPC) environments to process large spatial and temporal sequences of imagery (e.g., the Landsat archive). To date, results have demonstrated a high degree of consistence between scenes and good correlation to field measurements of reflectance.

Small BAseline Subset (SBAS) technique is being increasingly used as a geodetic method of choice for time series deformation analysis related to a variety of natural and anthropogenic processes. The method relies on an appropriate interferometric combination of SAR images with small temporal and spatial baseline. In this study we modify several aspects within the chain of the standard SBAS processing and develop a method called Enhanced SBAS (ESBAS) to improve deformation monitoring using SAR data, in particular for mountainous regions. Our modification includes filtering prior to phase unwrapping, topographic correction within 3-dimensional phase unwrapping, and reducing the effect of atmospheric noise either by external data like GPS or based on the band-pass decomposition of both topography and interferometric phase.

We evaluate the effectiveness of our enhanced method (ESBAS) for several case studies including deformation related to tectonic, volcano and mass movement processes. We show that the use of modified processing makes the long-term monitoring of fault or landslide creep more feasible as compared to the standard SBAS processing, where either atmospheric artifacts or DEM error effects disturb the creep signal. The results are compared with field measurements to better evaluate the performance of ESBAS algorithm. 

Abstract

The dynamics of surface and subsurface water events can lead to slope instability resulting in slough slides or other anomalies on earthen levees [1]. These slough slides are the primary cause for creating levee areas which are vulnerable to seepage and failure during high water events [2]. The loss of lives and property associated with the catastrophic failure of levees can be extremely high. In order to ensure the desired level of flood protection system performance, identification and repair of anomalies on earthen levees are critical to maintain long-term system reliability. Improved knowledge of the condition of these levees would significantly improve the allocation of precious resources to inspect, test, and repair the ones most in need. Early detection of these anomalies by a remote sensing approach could save time versus direct assessment. The roughness and related textural characteristics of the soil in a slough slide area affect the amount and pattern of radar backscatter. The type of vegetation that grows in a slough slide area differs from the surrounding levee vegetation, which can also be used in detecting slough slides [3]. In this paper, we implemented a supervised classification algorithm-- the minimum distance classifier-- with a majority filter and morphology filter for the identification of anomalies on levees using polarimetricSynthetic Aperture Radar (polSAR) data. PolSAR data encompasses information on scattering mechanisms by diverse target structures and materials. Synthetic Aperture Radar (SAR) technology, due to its high spatial resolution and soil penetration capability, is a good choice to identify problematic areas on earthen levees. The characteristics of the airborne SAR instrument used here are: 1.25 GHz carrier frequency, bandwidth of 80 MHz, range resolution of 1.8 m, and full quad polarization. Although the raw ground sample distance is 1.6 by 0.6 meters, the multi-look 5 by 7 meter data is used to minimize speckle effects [4]. The identification of slough slides is good with the classification results, and greatly improved with the majority filter and morphology filter [5-7]. This study employed remote sensing data from the NASA Jet Propulsion Laboratory’s (JPL’s) Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) instrument, using its fully quad-polarimetric L-band polSAR data to characterize levee segments for variability and anomalies that could lead to problems with the levees. UAVSAR is capable of penetrating dry soil to a few centimeters depth. Thus, it is valuable in detecting changes in levees that are key inputs to a classification system [8-9]. The study area is a section of the lower Mississippi River valley in the southern USA, where earthen flood control levees are maintained by the US Army Corps of Engineers.

Keywords:Synthetic Aperture Radar (SAR), UAVSAR, Levee Classification, Radar Polarimetry, Supervised Classification

References

J.V. Aanstoos, K. Hasan, C. G. O'Hara, S. Prasad, L. Dabbiru, M. Mahrooghy, R. Nobrega, M.L. Lee, and B. Shrestha, “Use of Remote Sensing to Screen Earthen Levees,” 39^th Applied Imagery Pattern Recognition Workshop(AIPR), pp. 1-6,  2010.

J. Dunbar, “USACE’s lower Mississippi valley engineering geology and geomorphology mapping program for levees,” presentation at Vicksburg, MS, 2009.

A. K. M. A. Hossain, G. Easson, and K. Hasan, “Detection of Levee Slides Using Commercially Available Remotely Sensed Data,” Environmental and Engineering Geosci., vol.12, no. 3,  pp. 235-246, 2006.

J.V. Aanstoos, K. Hasan, C. O\'Hara, L. Dabbiru, M. Mahrooghy, R. A. A. Nobrega, and M. Lee, “Detection of Slump Slides on Earthen Levees Using Polarimetric SAR Imagery,” IEEE Applied Imagery Pattern Recognition Workshop, 2012.

I. I. Chang and S.S. Chiang, “Anomaly detection and classification for hyperspectral imagery,” IEEE Transactions on Geoscience and Remote Sensing, vol. 40, no. 6, pp. 1314-1325, 2002.

M. Pesaresi and J.A. Benediktsson, “A new approach for the morphological segmentation of high-resolution satellite imagery,” IEEE Transactions on Geoscience and Remote Sensing, vol. 39, no. 2, pp. 309-320, 2001.

ENVI version 5.1, “Exelis visual information solutions user guides and tutorials,” http://www.exelisvis.com/Learn/Resources/Tutorials.aspx

J. V. Aanstoos, “Levee Assessment via Remote Sensing,” Phase 2, SERRI Project Report: 80023-02, 2012.

J.V. Aanstoos, “Levee Assessment via Remote Sensing,” Phase 1, SERRI Project Report: 80023-01, 2012.

The future altimetry missions planned for the coming months (Jason-3, Sentinel-3) or for the coming years (SWOT) aim to deliver a measurement of the topography at a finer spatial resolution, a higher temporal rate and over heterogeneous surfaces, open ocean but also coastal regions, hydrological targets and over ice and sea-ice.

In this perspective, the role played by the wet tropospheric correction (WTC) is critical, due to its large temporal and spatial variability and its crucial weight in the final budget error.

Current algorithms are based on empirical approaches parameterized using measured (radiosonde) or modeled (numerical weather prediction analysis) atmospheric profiles; in both cases, a radiative transfer model relates the integrated content of WTC to the top of the atmosphere brightness temperatures and a relation is fitted between the two datasets.

This method is valid over open ocean only, where a model of the emissivity is available. The performances are then degraded wherever the instrumental measurements are contaminated by other surfaces (land, ice, sea-ice); solutions exist to correct for this contamination but will not be able to satisfy the future constraints on the retrieval errors over coastal regions or hydrological surfaces.

In this context, a one-dimensional variational approach (1D-VAR) for wet tropospheric correction retrieval is a good candidate to provide a unique method well adapted to all surfaces. Where current algorithms directly provide an integrated value of WTC, this latter aims to estimate the atmospheric profiles that best explain the TOA TB measurements. The WTC is then computed from integration of the retrieved profiles. Depending on the surface, the emissivity is provided by a model (open ocean) or emissivity atlas (other surfaces) estimated from TOA measured brightness temperatures. Previous work (Desportes et al. 2010) has already shown the potential of 1D-VAR over coastal regions.

In this presentation, we will present the first assessment of the performances. Comparing a 1D-var approach to the reference GDR wet tropospheric correction of Jason-2 over open ocean, a sensitivity study on the impact of the covariance matrix, being defined either globaly or regionally, is performed and conclusions are drawn for an operational implementation of the 1D-Var.

Space-borne active instruments like radar altimeters (nadir looking) and scatterometers (side-looking) are invaluable instruments for meteorological and oceanic applications. Both are able to provide estimates for the surface wind speed. Both wind speed products are compared against each other wherever possible. Further verification is carried out against the ECMWF atmospheric model fields and in-situ (buoy) measurements. Wind speed products from Jason-2, Cryosat-2 and SARAL/Altika radar altimeters and ASCAT-A and ASCAT-B scatterometers are used for this study.

Error estimation is carried out using the triple collocation technique which is a powerful tool for estimating random errors as far as the errors of various data sources are not correlated. The issue of correlation between the scatterometers and the atmospheric model assimilating them as well as the correlation between scatterometers and altimeters leads to wrong results. This will be shown. Correction measures are used to alleviate this adverse impact.

Several European operational missions are being developed for launch in the post-2013 time frame.  The aim of these missions is to provide reliable and stable long-term EO data streams, e.g. the Copernicus Sentinels and MetOp (2nd Generation) satellite series. Several European national space initiatives also contribute to this monitoring effort.

In the frame of the EO-Convoy studies, ESA has investigated the feasibility and the expected added-value of cost-effective missions that could be flown together with these operational missions and would provide synergetic observations. 

In September 2013, NASA released a request for information for their Sustainable Land Imaging Architecture Study regarding the definition and programmatics estimate of the next generation land imaging system. Airbus Defence and Space submitted a proposal in response to this RFI including some architectural component ideas focusing on possible American and European cooperation. To support NASA during the Sustainable Land Imaging Architecture Study, ESA triggered a short activity to further develop the ideas exposed in the proposal and to elaborate possible scenarios. This study was carried out by Airbus Defence and Space with the support of CESBIO (Centre d’Etudes Spatiales de la BIOsphère).  

The activities included a user’s needs survey, the identification of possible scenarios representing a European contribution to a post Landsat-8 in-orbit architecture, and the analysis of the most promising option candidate consisting in a fast track/low cost thermal infrared imaging satellite flying in formation with either Landsat-8 or Sentinel-2. 

The proposed paper will present the outcomes of the study, from the users’ needs to a mission concept featuring a TIR instrument in the [8-12µm] range based on microbolometer technology offering a 100m GSD with systematic observation with up to 3 bands.

There is a growing need for using surface reflectance products as input for down-stream products. Space agencies, as ESA and NASA, are planning to generate (or plan to generate in the short term) surface reflectance products at global scale for Landsat-8 and Sentinel-2 missions, but there is a need to establish traceability and inter-operability across missions and products. The uncertainty of surface reflectance products is still not well established at global scale and needs to be further investigated. Thus, an atmospheric correction inter-comparison exercise has been decided between NASA and ESA in the frame of CEOS/Working Group on Calibration & Validation (WGCV).

This inter-comparison exercise aims at getting a better understanding of the BOA (Bottom of the Atmosphere) reflectance uncertainties by setting a comparison of the outputs generated by different AC schemes. The understanding of the differences through this exercise should help in: (1) Better understand the different uncertainty contributors, and (2) Improving the different AC schemes.

The project consists on inter-comparing a set of atmospheric correction schemes for optical multi-spectral imaging sensors developed and used within ESA or NASA frame. It consists by comparing (1) the surface reflectance, and (2) cloud and shadow masks. As inputs, we use Landsat-8 L1T and Sentinel-2 L1C TOA reflectance products. Several test sites has been defined by the COES/WGCV partners and include, for example, coastal, desert, forest, agricultural and snow areas.

We present here the frame of this inter-comparison exercise, the defined protocol, the used models and the preliminary results.

OLCI on-board Sentinel-3 is a programmable, medium-spectral resolution, imaging spectrometer operating in the reflective solar spectral range between 390 nm to 1040 nm. Its 21 spectral bands are programmable by ground command both in width and in position by steps of 1.25 nm. The field of view of about 1265 km is shared between five identical cameras arranged in a fan shape configuration. The image is constructed using the push-broom principle, where a narrow strip of the earth is imaged into the entrance slit of the spectrometer defining the across track dimension while the motion of the satellite provides the along track dimension. The spectral dimension is achieved by imaging the entrance slit of the spectrometer via a dispersing grating onto a 2-D CCD array, similar to the MERIS instrumental design. The CCD covers the spectral range with a nominal 1.25 nm spectral sampling interval, with a Full-Width Half-Maximum (FWHM) between 1.5 nm and 2.0 nm, corresponding to the ground characterization measurements. Due to the imaging principle the band centre and the bandwidth of each viewing OLCI pixel vary across the field of view. However, the precise knowledge of the centre wavelength of the OLCI bands for each pixel within 1 nm is necessary for most of the ocean and land applications and further an accuracy of 0.1 nm is necessary for all algorithms using the O2 and H2O absorption bands.

We present the procedures, as they will be applied during commissioning phase and lifetime of S-3, to infer the spectral characteristics of OLCI by means of central wavelengths and bandwidth. For the spectral calibration within the entire spectral domain of the OLCI instrument we use a number of Fraunhofer lines as well as the spectral absorption features of the atmospheric oxygen. In particular the Fraunhofer lines of the Solar spectrum around the Ca II (393.4 nm + 396.8 nm) , Hss (486.1 nm), H 656.3 nm), Ca II (854 nm) and H (1005 nm) as well as the atmospheric absorption bands of Oxygen around 760 nm and 685 nm are used. We assume a Gaussian relative spectral response functions for the individual micro bands, however, the variability of FWHM is taken into account.

Synthetic Aperture Radar Interferometry (InSAR) has proven to be a highly valuable tool for deformation monitoring applications, and with the emergence of the operational Sentinel-1 mission, the expectations have further increased. However, the technique is also notorious for being challenging, both in implementation and in interpretation. As such, it has mainly been applied in an opportunistic manner - stacks of archived data were processed in batch over specific areas of interest, results would be somehow validated, and reported to the end-user.

Due to lack of a stable and reliable SAR data source, it has so far been very hard to build an operational large-scale system for nationwide InSAR deformation monitoring.

The Sentinel-1 constellation will bring a paradigm shift to the field with its operational characteristics: mission configuration, acquisition planning, and data distribution policy. However, operational workflows are still to be designed and deployed. State-of-the-art processing algorithms might be considered ready, but other components of the operational chain are still to be profiled if not designed from scratch. Moreover, an important factor for a nationwide mapping system is dissemination of the results to a broader audience than the one consisting of InSAR domain specialists.

The objective of this contribution is to, as far as possible at this early stage, address the following questions and/or identify the bottlenecks, from the national-scale perspective:

How to integrate algorithmic state-of-the-art in a single processing system?

How to operationally deploy the re-defined state-of-the-art?

How to validate results?

How to communicate results to non-InSAR communities?

These questions will be addressed through a case study for the country of Denmark. Specifically, a strategy for development of a nationwide Sentinel-1 InSAR deformation mapping system will be discussed. Denmark is a good case for a first assessment, both due to its size, and due to the expected signal characteristics. Moreover, the proposed map product will be periodically updated and will be of a different resolution for urban and non-urban areas. The availability of reliable in-situ measurements on both local and national scale therefore will over time allow for reliable validation.

Initial results will be presented and will serve as basis for a discussion on how to communicate and streamline a portfolio of subsidence products to end users, which is a challenge in itself. We will conclude with a discussion on remaining open questions regarding how to address these issues as a community.

This contribution will focus mainly on the overall system level, rather than on algorithmic details.

The ESA Sentinel 3 OLCI series will provide continuity of satellite-based ocean colour measurements with a consistent quality, high accuracy and reliability and in a sustained operational manner for end-users. There is, however, a need for fiducial reference measurements to test the quality of Earth Observation data. The Plymouth Marine Laboratory runs the annual UK Atlantic Meridional Transect (AMT) between the UK and South Atlantic. AMT traverses the poorly sampled oligotrophic gyres along with mesotrophic waters in the North Sea, English Channel and Celtic Sea. On AMT high-quality in situ biogeochemical and apparent & inherent optical property (IOP) data are collected both on-station with depth profiles and near surface along-track with novel autonomous systems using established satellite validation measurement protocols. The along-track flow-through absorption-based systems, in particular, can provide many hundreds of match-ups with satellite data. In this talk we will present results of ocean colour sensor and algorithm validation on recent AMT cruises. The performance of standard and non-standard level 2 and 3 products using algorithms that will be available for OLCI in both coastal and open ocean regions will be presented. The results will be discussed in the context of previous experience from ESA and NASA missions and future plans for validation of Sentinel-3 OLCI and Sentinel 2 MSI.

Regularly and automatically updating land cover mapping has been an aim of remote sensing scientists for many years. With the recently launched Landsat-8 and Sentinel 1 and 2 sensors data has become available for such systems to be implemented and applied as operational systems. This paper presents a new method for detecting change and updating a land cover classification that can be applied within a near real time monitoring system for land cover.

Creating a land cover map is a significant investment in both time and resources in many cases requiring imagery from different modalities (e.g., optical, SAR and DEMs) and from different seasons (e.g., spring, summer and autumn). The map-to-image method proposed in this paper therefore uses the existing classification as the base map and aims to identify the regions of change on a per-class basis. This ensures that change slithers are not present and boundaries in the existing map are maintained. The change method is based on a simple assumption that within a class the response (spectral and/or backscatter) is homogeneous and where changes occur they have a distinct signature from the class, outside of the response distribution of the original class. It is also assumed that the amount of change within a class is small compared to the overall area of the class within the scene and that the original land cover map has a high degree of accuracy in representing the land covers for the area. Using this method we have demonstrated that in excess of 90% of the change can be identified within two case studies. Firstly, change detection of mangrove forests using the ALOS PALSAR and JERS-1 data from a 2010 baseline with new mangrove maps defined for 16 sites globally for 1996, 2007, 2008, 2009 and 2010. Secondly, changes in forest extent have been classified for a region of Tanzania using Rapideye (2012, 2014) and Landsat (2014, 1998) mapping both forest loss and gain.

The method has been shown to highly adaptable and applicably to a number of datasets, including both optical and SAR along with a number of land covers. Current work is on using this technique to build land cover monitoring systems using Landsat and Sentinel 1 and 2 data, for example the Global Mangrove Watch (GMW). The components of the algorithm have been made freely available with the open source Remote Sensing and GIS Software Library (RSGISLib; rsgislib.org).

Surface soil moisture content is one of the most important variables in the hydrological cycle and is very variable in time and spatial. The southwest of Iran is the most important agricultural area and the largest storage dams and numerous modern irrigation networks are located here.Due to the limited water resources in Iran and frequent droughts, simulation and retrieval of soil moisture is necessary to management and planning of water resources. In addition to, Meteorology and hydrology models, weather forecasting and climate change monitoring need to these data. The in-situ measurements of soil moisture are not available for longer periods in this case study. Therefore the main objective of this research is to retrieve and validate of soil moisture from European Space Agency’s Soil Moisture and Ocean Salinity (ESA’s SMOS mission ) satellite products ) Level 1C and Level 2 Soil Moisture data over the south west of Iran. Validation of SMOS L1C will be done using ground based measurements and radiative forward model (L-band Microwave Emission of the Biosphere- L-MEB). Validation and Calibration of Level 2 soil moisture will be done with measurements of soil moisture in meteorological stations in the southwest of Iran. The result of this research gives valuable information on the errors and uncertainties in SMOS Products in this region.

The 3D phase unwrapping is a key step to extract the deformation signal from the temporal differentiel interferograms. It consists on determining the ambiguous phase values across all the 3D_differential interferograms, where the third dimension is the time.

The Generalization and the temporal continuity of 2D / 3D phase unwrapping involve methods of contextual voxels unwrapping. In that sense, methods of 2D phase unwrapping can be applied to each 2D image in the 3D phase volume independently of the entire volume unwrapping. However, these methods cannot detect the phase shift necessary to ensure temporal continuity of the unwrapped phase by the third dimension. In contrast, the 3D phase unwrapping which operates in the volume of the phase can avoid this problem and has the additional information provided by the third dimension. In this context, the algorithm we propose is based on the phase unwrapping process that is guided by the reliability of the edges to prevent the errors spread. The basic method of this process, inspired by the algorithm of Herraez et al [Herráez, 2002] based on a 3D  quality map and edges of the latter on the three planes (x,y, z). In fact, an edge being the intersection of two pixels of the quality map that are connected in the x, y or z, the unwrapping follow a discrete path, going from the most reliable edge at first and the lowest quality edges at last. However, two main criteria determine the phase unwrapping results: i) the choice of the quality map, ii) how to determine the path that minimizes error propagation the best?

 We have, for this purpose, quality maps generated by the first and second derivatives 3D and 3D pseudo-correlation. We tested the process on interferograms simulated to verify the algorithmic side then applied to real interferograms obtained from SAR images acquired on seismic regions.

The availability of Earth Observation missions, such as DEIMOS-1 and UK-DMC2, providing high-resolution images (22m/pixel) over a very wide swath (650 km) on a high-frequency revisit basis has facilitated the unprecedented generation of frequent cloud-free coverages at continental scales. This type of coverages are useful inputs for spatially detailed applications targeting land surface dynamics, such as land cover change or monitoring of agricultural fields or forested patches. The Sentinel-2 constellation, with its 5-day global revisit, will deliver when fully deployed high-frequency, high-resolution time series that will greatly benefit agriculture worldwide.
In the meantime, the Copernicus program has made available for its users an unprecedented series of monthly cloud-free European datasets in the frame of the Data Warehouse 2 initiative. Deimos Imaging’s DEIMOS-1 satellite, in cooperation with its twin satellite, DMCii’s UK-DMC2, has provided an unprecedented series of seven monthly wall-to-wall continental coverages, with 22-m multispectral imagery, fulfilling stringent cloud coverage limits below 10% and 20%. The datasets have been collected from April to October 2015 by the two CCME, achieving the minimum possible cloud coverage over the different regions in order to assure a consistent monthly temporal resolution.
This presentation describes the monthly monitoring campaign over Europe carried out by DEIMOS-1 and UK-DMC2 in 2015, detailing its main results and achievements, and the multitemporal monitoring service and its main products. The main purpose of this campaign was the acquisition and delivery of orthorectified imagery, to be used as an input for the control and monitoring of environmental practices in the study area. All products were delivered, ortho, on a monthly basis after a selection of the products contributing to the best available coverage was performed in close cooperation with ESA.

The wet tropospheric correction (WTC) is a major source of uncertainty in altimetry budget error, due to its large spatial and temporal variability: this is why the main altimetry missions include a microwave radiometer (MR) The commonly agreed requirement on WTC for current missions is to retrieve WTC with an error better than 1cm rms.

JPL for NASA/CNES (Jason-2) missions on one hand and CLS/IPSL for CNES (AltiKa) and ESA (Sentinel-3) missions on the other hand based their retrievals on similar approaches with still identified differences.

Both are based on an empirical relation relating the top of atmosphere brightness temperatures (TB) to the WTC.

With CLS approach, this relation is established using ECMWF analysis for both the reference WTC and the inputs of the radiative transfer model used to compute simulated TB. A single neural network (NN) is used to invert the relation at a global scale.

With JPL approach, radiosondes are used to compute the reference WTC. A Log-linear parametric model is used to invert the relation in a stratified scheme: a set of coefficients is established for different ranges of wind speed and WTC.

In the frame of CNES project PEACHI-J3, we defined a combined approach using the ECMWF analysis and the neural network in a stratified scheme. A sensitivity study is performed on the range definition of wind speed and WTC.

Stratified-NN WTC are computed for tri-channels (Jason-2 and Jason-3) and bi-channels (AltiKa and Sentinel-3) radiometers and compared to the WTC in operational products. Performances are discussed and conclusions are drawn for future operational retrievals.

Measuring subsidence map time series in urban areas with Spaceborne Interferometric Synthetic Aperture Radar (InSAR) is already established with commercial offerings available through a number InSAR service providers. Nevertheless, engineering end users still routinely question that InSAR meets the accuracy of traditional quantitative survey methods such as GPS or leveling. This controversy hampers the widespread acceptance of InSAR as an equivalent method to map instable ground and infrastructure integrity in urban areas. The controversy centers on absolute referencing and bias problems, and whether all spatio-temporal scales are captured properly in the InSAR displacement map series.

From an InSAR processing complexity stand point, such doubts cannot be dismissed outright as there are definitely complications that can bias final results. The most important of these are: (i) Removal of the atmospheric phase contribution leading to positive or negative long-range displacement bias due to over-filtering or incomplete atmospheric removal, respectively; this is particularly severe for urban areas where often subtle motion exists with a-priori unknown spatial distribution and at similar scales than atmospheric phase. (ii) Adaptive phase filtering, both in space and time biasing the recovery of temporally non-linear displacement (due to meteorological factors such as temperature, soil moisture, and precipitation) as traditional implementation cannot guarantee that temporal non-linearity is not removed as statistical outliers. (iii) A-priori (wrapped) phase component modeling and phase unwrapping producing errors and corresponding bias when sharp spatial discontinuities of thermal displacement (and un-modeled topography) exist as is common for urban areas with tall buildings and infrastructure. (iv) Traditional network processing of InSAR stacks introducing both bias and noise into displacement series of only intermittently coherent (temporary) targets; such temporary targets inevitably occur for most longer image stacks due to construction activities and additionally also due to seasonal processes such as snow or vegetation cover. Lastly, (v) simplistic projection of InSAR displacement results to vertical introducing bias; this is because many urban “subsidence” phenomena have significant horizontal movements associated, which combines with the vertical displacement in the slanted InSAR look direction; this problem contaminates also the simple combination of displacement series from two interleaving ascending and descending image stacks as the true 3D movement is not unambiguously constrained and results are thus model dependent.

We present several important features of the 3vGeomatics advanced InSAR processing chain that address the aforementioned complications and that are crucial to traditional surveying quality. They are: (1) Superior APS determination consisting of two steps: initial a-priori prediction of water vapor spatial statistics using meteorological modeling, which in turn is used to identify areas that are safely non-moving at the accuracy limit of (~1 mm/yr). Statistics is then updated over these areas and is inverted to construct optimum per scene APS. This step also includes the iterative statistical recovery of long-scale displacement from the APS estimates. (2) Optimized adaptive phase filtering that preserves temporal non-linearity by also considering the expected atmospheric error statistics as weights. (3) Combined bias-optimized wrapped phase modeling and novel 3D unwrapping using redundant arcs. This approach dramatically reduces unwrapping errors and their associated biases by effectively “bridging” over and then resolving thermally active and multi-bounce areas on buildings and infrastructure with PCA being used to model unknown sources of temporally non-linear motion. (4) Optimized temporary target inversion (in velocity instead of displacement space) using minimal assumptions to constrain target velocity. Finally, (5). Appropriate treatment of 2D and 3D displacement derivation using simultaneous processing of interleaving ascending and descending image stacks and realistic assumptions for constraining the movement in a (spatially variable) plane.

The quantitative effect of these enhanced features on displacement series accuracy is analyzed individually and in combination. We use results from the processing of several high resolution inSAR stacks from RADARSAT-2, TerraSAR-X, and Cosmo-Skymed acquired over urban areas. Comparison with available GPS and leveling data suggests “traditional surveying quality” standards are generally met by the InSAR-derived displacement time series.

On June 19, 2014 the DEIMOS-2 very-high-resolution Earth Observation satellite was successfully launched, and CAL/VAL activities began shortly after. Owned and operated by Deimos Imaging from its control center in Spain, DEIMOS-2 is the highest-resolution fully private satellite in Europe.
DEIMOS-2 is an agile platform, based on Satrec-I SI-300 architecture, with accurate and agile three-axis attitude control supports precise imaging operations. The payload, HiRAIS, is a push-broom type camera (TDI linear array) with 1 m Ground Sampling Distance (GSD) for a panchromatic band and 4 m GSD for four multi-spectral bands (NIR, red, green and blue), which produces 75-cm pan-sharpened imagery after ground processing. The 12-km swath can be increased to 24 km in the wide-area acquisition mode.
The radiometric calibration and validation (CAL/VAL) is of paramount importance for product quality since it enables the data to be used together with other data sources including, but not solely, additional remote sensing sources. This data compatibility creates synergies for the development of applications which face societal, science and economy challenges such as urban growth, climate change and crop monitoring.
CAL/VAL activities were performed by Deimos Imaging in cooperation with the Satrec-I team. The post-launch calibration plan first addressed the task of validating the sensor’s radiometric model, which was developed using pre-launch data, by performing specific acquisitions to ensure that the sensor behavior was consistent with the laboratory characterization.
Once the radiometric consistency was assured, the sensor absolute calibration was then addressed. This process leveraged on the experience of DEIMOS-1 calibration, in which data coming from confidence sources, like Landsat-7/ETM+ was used since 2009 to cross-calibrate DEIMOS-1, and to monitor sensor trending, complemented with vicarious data to validate the cross-calibration. The DEIMOS-2/HiRAIS sensor characteristics, especially its spectral response, require state of the art calibration procedures, due to the lack of a confidence reference sensor to cross-calibrate.
The methodology we used relies on pseudo-invariant calibration sites (PICS) to perform the absolute calibration, being Libya-4 the main one. To do so we had to thoroughly characterize the sites beforehand, using EO-1/Hyperion hyperspectral data and other data sources, and comparing our results with similar characterizations found in the literature.
Once developed, we checked the new methodology and found good consistency using Landsat-8/OLI and DEIMOS-1/SLIM6 data, so we considered the procedure ready to be used for the absolute calibration of the DEIMOS-2/HiRAIS. The methodology is also valid to monitor the sensor trending. Vicarious measurements are still needed to provide a strong confidence level to the absolute calibration based on PICS.
In this presentation we will show how we validated the DEIMOS-2/HiRAIS radiometric model, the method to characterize PICS with EO-1/Hyperion and other data sources, the comparison with similar characterizations found in the literature, the methodology validation using Landsat-8/OLI and DEIMOS-1/SLIM6 data, the results of the methodology applied to DEIMOS-2/HiRAIS and the scheduled work for the future.

ECMWF has been involved in global monitoring, validation and assimilation of satellite wind and wave products since late 1980’s. This includes products from Altimeters, SAR’s and Scatterometers. Operational monitoring, validation and assimilation of altimeter products realised in the early 1990’s with ERS-1 followed by ERS-2, ENVISAT, Jason-1/2, Cryosat-2 and SARAL/AltiKa. This is expected to continue by including all current and future altimetry missions that offer relevant products in near real time (NRT), i.e. within 1-5 hours, including Sentinel-3 radar altimeter (SRAL).

ECMWF runs the Integrated Forecasting System (IFS), which is one of the most comprehensive global weather forecasting systems, operationally twice a day. Model output will be used mainly for the global monitoring and validation of Sentinel-3 wind and wave products from Level 2 STM Marine Product (SR_2_WAT___) and possibly from the equivalent land product SR_2_LAN___ to cover the inland water bodies like the Caspian Sea. Although this will be done mainly in NRT, monthly and annual assessments will be carried out in a delayed mode.

The validation of significant wave height (SWH), ocean surface wind speed, wet tropospheric correction and water vapour content will be done by collocating Sentinel-3 products with IFS model fields and available in-situ data. Direct comparison (i.e. plots and statistics) as well as triple collocation (Sentinel-3, model, in-situ) analysis will be carried out to assess the quality of Sentinel-3 products.
The results will be presented and discussed.

Earth Monitoring with GNSS signals is a promissing novel area for Earth Observations systems. It has recently received much attention, mainly due to the increasing number of GNSS constellations foreseen. The increasing number of sources, together with the low level of requirements (power, size, budget, etc.) and possibility of using COTS hardware, makes these systems potentially very attractive for future space missions, making this a very attractive complementary technology to traditional active radar systems. Since the use of reflected GNSS signals was proposed in 1993, the number of research activity and scientific publications has been steadily increasing, especially in recent years, where the attention devoted to GNSS-R has increased exponentially. Recently, NASA has approved the first operational use of GNSS-R from space, the eight-satellite CYGNSS constellation.

The European GNSS-R Earth Monitoring project (E-GEM, www.e-gem.eu), started in 2104, is an FP7 funded project which joins many of the European experts in the field of GNSS-R. It involves ten of the top European institutions in this field, and is currently the largest single project ongoing in Europe dedicated to GNSS-R. The goal of this project is quite ambitions, covering both instrument development and development of algorithms for a number of applications. An overview of the project's accomplishments so far is presented, together with discussion of planned activities on instrument development, data analysis and supporting studies.

The project comprises three distinct instruments, following distinct priorities and approaches: 1) a satellite-borne instrument with launch set for 2016, which carries the payload instrument PYCARO aboard the 3Cat-2 satellite. Thjis instrument was was recently tested in the BEXUS stratospheric balloon campaign, sponsored by ESA Educational Office. During this flight the first ever multi-constellation (GPS, GLONASS and Galileo), dual-band (L1 and L2), and dual-polarization (LHCP and RHCP) reflected signals were collected, showing a good performance of the instrument; 2) an airborne instrument using Cryowing MKII platform. The development of the SPIR-UAV payload has now been completed, with all the electronics and raw measurements interfaces, enabling the acquisition of extensive amounts of raw data, able to test different techniques, from conventional or interferometric. The flight campaign will be held in Svalbard, Norway and will overfly Ocean, Sea-Ice, Glaciers and Land in the Spring of 2016; 3) the ground-based instrument SARGO first batch of development culminated with in-situ tests by the Tagus river (Lisbon, PT), which derived vertical distance to the water with Galileo E1 and E5a signals. This instrument is now deployed on a long term basis on the Tagus river, providing valuable data on water surface monitoring, side by side with concurrent measurements for validation purposes.

On the algorithms, applications and data processing side, the E-GEM project focuses mainly wave height and surface winds, but also monitoring of ice layers, soil moisture, and biomass. Several studies from the project members have been devoted to the development of new processing methodologies, mitigation of interferences, error analysis, instrument simulation, studies of the impact of integration of GNSS-R data in operational Earth monitoring systems, such as assimilation in oceanographic and weather models. The application areas for these studies cover not only high precision SSH, SWH and surface winds retrieval, but also soil moisture studies, biomass retrieval and other fields. The development and consolidation of algorithms for the interpretation of GNSS-R signals is also a central piece work performed, representing a significant contribution for the operationalisation of GNSS-R as remote sensing technique in areas targeted by E-GEM and to the processing of the data to be generated by the three instruments.

 In this paper the formulation of the hybrid decomposition has been tested to PolInSAR data sets. As it has already been in study with freman-durden decomposition.The procedure assumes that the interferometric cross-correlation for the linear basis can be decomposed into The three mechanisms (i.e.direct, double-bounce and volume scattering mechanisms) . The retrieved parameters are defined with magnitude and phase, and hence the power contribution is estimated jointly with the phase center of the scattering mechanism. Therefore, the re­trieved magnitudes are also associated with their corresponding interferometric phases which are related to the vertical locations and, actually,

The main objective of this study is to use this feature besides to the sensitivity of the interferometric coherence to the polarization state, to enhance different medium characteristics. Also the result are compared with Freman-durden decomposition over PolInSAR data.

Lidar characterization of canopy structure has exploded as lidar data have become more widely available to the biomass and biodiversity research communities.  We will discuss the rapid advance of lidar measurements of canopy structure and the applications that have evolved most rapidly in recent years in terms of 3D habitat characterization, species-specific habitat utilization, and the potential of new space-based lidar measurements designed specifically for terrestrial vegetation mapping and monitoring.  We will focus on advances thus far but emphasize the potential of future missions, particularly the Global Ecosystem Dynamics Investigation (GEDI), designed to install and operate a vegetation canopy lidar on the international space station.  GEDI will provide unprecedented lidar canopy structure information with high density sampling across the tropical and temperate regions beginning in 2018.  GEDi will thereby also provide long-desired information on essential biodiversity variables (EBVs) from space.

The effective processing of optical remotely sensed data requires the detection of clouds and correction of atmospheric effects. This correction step is essential mainly for time series analysis, as different atmospheric conditions over time affect surface reflectance and therefore the change detection outputs. To date, many different atmospheric correction processors have been developed, with the aim of applying them to datasets acquired by various sensors. In this study, the model developed by German Aerospace Center (DLR), i.e. Atmospheric & Topographic Correction for Small FOV Satellite Images (ATCOR), is used for correcting Landsat-8 and Sentinel-2 imagery. In order to validate the algorithm performance, different study areas, representative of main surface and atmosphere types, are considered, e.g. coastal, agricultural and deserts.

The validation of the implemented atmospheric correction is accomplished by the assessment of the surface reflectance products in each test site. To this end, the Aerosol Optical Thickness (AOT) and Water Vapour (WV) estimates are compared with the in-situ AERONET Level 2.0 data for the test sites. In particular, the AERONET AOT averaged within 30 min before and after the data acquisition. The study area is a 10km x 10km spatial subset of the imagery centred on AERONET station. The data concern a six-month period of observations, to provide a range of atmospheric conditions and thus to enable reliable and accurate statistic results. For the AERONET Water Vapour (WV) as well, the average of the two AERONET water vapour retrievals is calculated associated with these two closest AOT retrievals. The ATCOR surface reflectance products for each of Landsat-8 and Sentinel-2 spectral band were compared pixel-by-pixel to the corresponding AERONET surface reflectance data. Only pixels that were not flagged out by the L1 processor and not labelled as cloudy are considered for a pixel-by-pixel comparison of the ATCOR corrected product and surface reflection reference.

Introduction 

Image coregistration aims at stacking two or multiple images in a way such that, for each image, the same pixel corresponds to the same point of the target scene (often with sub-pixel accuracy). Typical remote sensing applications in which image principles of coregistration are used are: multi-sensor classification, change detection, mosaicking, synthetic aperture radar (SAR) interferometry, radargrammetry, photogrammetry, multi-sensor data fusion and so on.

We can distinguish two families of image coregistration problems, basically depending on if the images to be coregistered are taken by sensors of the same or different type (e.g. sensing different wavelenghts) and with similar or different illumination and acquisition geometries (e.g. different sun illumination conditions and/or different acquisition incidence angles).

Whilst the first type of image coregistration is well established, multimodal coregistration is not yet well founded and ready for operational use due to difficulty of finding correspondences between the images (tie points) in a robust way, and the available approaches often recur to manual assistance.

The work presented here focuses on automatic multimodal image registration of very high resolution SAR and optical images.

Coregistration of SAR and optical images

SAR and optical sensors have very different radiometries. SAR is an active imaging sensor emitting coherent microwave radiation and receiving the backscatter from the Earth surface, while optical instruments basically acquire the incoherent Sun radiation reflected by the Earth surface and the atmosphere, or their thermal emission.

Moreover, SAR and optical sensors have also very different geometric acquisition mechanisms, which makes SAR and optical images very different also from the geometric point of view.

For these reasons, SAR and optical images over the same scene appear very dissimilar, and it is sometimes difficult even for an expert photo-interpreter to find correspondences between images.

Multimodal coregistration method

The proposed coregistration technique overcomes the problems due to differences in radiometries and in geometries by exploiting two main concepts:

The mutual geometry information between SAR and optical images is determined and exploited in different steps of the coregistration process in order to: 1) minimize the disparities between the images; 2) reduce false matchings; 3) optimize the processing; 4) increase drastically the number of control points.

The mutual information is used as similarity metric in order to deal with the strong radiometric differences between SAR and optical images.

The matching between the two images is performed using the SAR geometry as common working geometry and it is based on two main logical steps. First global image displacements across the epipolar curves are removed by a bundle adjustment in which orbit and acquisition parameters are corrected. Then the local residual disparities along epipolar curves are estimated through a matching step based on mutual information.

The disparities along the epipolar curves are mainly due to inaccuracy of the elevation model knowledge. Therefore the knowledge of these disparities at the control points are used to determine the correct elevation which finally is used to orthorectify both images and to achieve the final co-registration.

Validation on real data 

The proposed approach was validated on a large set of very high resolution SAR and optical images from Cosmo-SkyMed and GeoEye missions, selected to cover rural areas, mountains, and urban scenarios. In the performed tests coregistration is effectively achieved. Moreover, the method appears general and robust enough to be successfully applied to SAR and optical images also from other sensors.

Satellite SAR interferometry is a technique widely used to detect and monitor slow terrain movements with millimetric accuracy. Among the applications, monitoring of ground deformations due to mining activities is of high relevance but particularly difficult because the scene changes with time, which causes loss of interferometric coherence, and also makes typically impossible to have updated digital elevation models (DEMs).

In this work, we show the successful monitoring of several mining structures (including open pits, waste dumps, conveyor belts, water and tailings dams) in Minas Gerais state, Brasil, performed with COSMO-SkyMed SAR data processed by means of the persistent scatterer pair (PSP) technique.

The PSP technique [1], thanks to the property of exploiting only the relative properties of neighboring pairs of points for detection and analysis of persistent scatterers (PSs), is less affected by atmospheric delays, orbit or DEM inaccuracies, or terrain displacement strong accelerations. Moreover, by exploiting a redundant set of pair-of-point connections, the PSP approach guarantees extremely dense and accurate displacement and elevation measurements also on natural terrains.

The COSMO-SkyMed satellite constellation provided frequent acquisitions (48 per year), and we could conceive a processing approach based on a sliding window selecting acquisitions taken in yearly periods, in order to detect the PS that remain coherent only for short periods.

The obtained results made possible to observe many important phenomena, and to have a consistent global view of the whole mining area, which would be impossible by traditional on-the-field measurements. Among the observed ground deformation phenomena, we mention compaction on waste dumps (see Figure 1 of annex) and tailings dam massif (see Figure 2 of annex), showing the increase of rate deformation from the base to the top, in function of constructive method. Inside open pits, phenomena associated with toppling blocks on mine slopes interfering with ore conveyor belt were observed (see Figure 1 of annex). Additional results, related to other phenomena, will be reported in the final paper.

References

[1] Costantini, M., Falco, S., Malvarosa, F., Minati, F., Trillo, F., Vecchioli, F., “Persistent Scatterer Pair Interferometry: Approach and Application to COSMO-SkyMed SAR Data,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 7, no. 7, pp 2869 – 2879, July 2014.

Current Earth observation systems are not able to adequately observe sub-mesoscale ocean features w.r.t. surface currents. This means that scientific requirements in order to understand the dynamics of ocean processes and ocean atmosphere interactions are still not met. Higher resolution data and better sampling at a global scale is required, especially w.r.t. the accurate measurement of ocean surface currents. In order to close this observational gap, the European Space Agency (ESA) has initiated a number of studies to investigate possible future mission concepts with the goal of making progress in the retrieval of the Total (two-dimensional) ocean Surface Currents Vector (TSCV). In addition to the development of the Wavemill-concept [1], there have been two Ocean Surface Current Mission Studies (OSCMS) based on a Ku-band dual-beam along-track SAR interferometer [2]. This paper presents the outcome of one of those studies and compares it with a Ka-band variant of the concept. Specifically, the paper discusses pros and cons of the two possible frequency bands, and which impacts on the system are expected.

The driving mission requirements – besides the revisit-time and global coverage - are a large swath coverage of 200 km and a TSCV accuracy of 5 cm/s in combination with the (sub-mesoscale) product resolution of 4 km. In order to deal adequately with the mission requirements a performance model was derived which yields corresponding requirements for the radar instrument.

The basic ATI measurement principle is the generation of a pair of SAR images of a surface with a short time-lag but with almost identical geometry. The observed interferometric phase provides basically a Doppler measurement which represents a motion on ground, i.e. the motion of the sea surface. Unfortunately a large portion of the mean Doppler frequency is not caused by the surface current velocity. Instead, it reflects the NRCS weighted average of the radial velocities resulting from surface waves, which in turn are mainly driven by the surface winds. Strong correlations between Doppler centroid anomalies and surface winds were evaluated e.g. in ENVISAT’s ASAR observations [4]. The compensation of these geophysical biases has to be kept in mind while talking about surface current measurement with ATI. Additional scatterometry measurements might be necessary in order to deal with this challenge. However, the objective of the mission study at hand is primarily accurate dual-beam ATI measurements.

The dual-beam measurement concept requires two squinted antennas each looking simultaneously in fore and aft direction, which ensures the observation of the two-dimensional current vector [3]. The instrument concept is based on a javelin, single side-looking geometry using the ScanSAR imaging mode in order to achieve the requirements. A Digital beamforming (DBF) approach, i.e. SCan-On-Receive (SCORE), was chosen to fulfill the strong sensitivity requirements. Additionally, polarization diversity, namely hybrid polarization, will support the geophysical inversion process [5].

References:

[1] C. Buck, M. Aguirre, C. Donlon, D. Petrolati, and S. D’Addio, “Steps towards the preparation of a wavemill mission,” in Geoscience and Remote Sensing Symposium (IGARSS), 2011 IEEE International, Jul. 2011, pp. 3959–3962.

[2] S. Wollstadt, P. Lopez-Dekker, F. D. Zan, M. Younis, R. E. Danielson, V. Tesmer, and L. M. Camelo, “A Ku-Band SAR Mission Concept for Ocean Surface Current Measurement using Dual Beam ATI and Hybrid Polarization,” in Proc. IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Milan, Italy, 2015, to be published.

[3] S. Frasier and A. Camps, “Dual-beam interferometry for ocean surface current vector mapping,” IEEE Transactions on Geoscience and Remote Sensing, vol. 39, no. 2, pp. 401–414, Feb. 2001.

[4] B. Chapron, F. Collard, and F. Ardhuin, “Direct measurements of ocean surface velocity from space: Interpretation and validation,” Journal of Geophysical Research: Oceans, vol. 110, no. C7, 2005.

[5] V.N. Kudryavtsev, B. Chapron, A.G.Myasoedov, F. Collard, and J.A. Johannessen, “On dual co-polarized SAR measurements of the ocean surface,” Geosci. Remote Sens. Lett. IEEE, vol. 10, no. 4, pp. 761–765, July 2013.

In this study we investigate stability of Urmia Lake bridge (also known as Shahid Kalantari’s highway bridge) in northwest of Iran using high-resolution satellite radar imagery. The UL bridge crosses lake Urmia and together with two embankments on its eastern and western side makes Lake Urmia Causeway (LUC). The causeway connects East Azerbaijan and West Azerbaijan together reducing the driving distance between them by 135 kilometers. The bridge has a total length of about 1.7 km and consists of 4 different parts with symmetric geometrical structures including approach, transition, viaduct and arch bridge in its eastern and western parts. The bridge includes 18 caissons filled with concrete; two of them in the embankments and others in the lake. The steel deck of the bridge has three parts; the central part intended for single-track railway while two outer parts are for double lane vehicle traffic.

Preliminary analysis using Envisat, ALOS and TerraSAR-X data show that two embankments on the eastern and western side of the bridge are subject to subsidence due to consolidation caused by dissipation of excess pore pressure in embankments. Here we complement the previous study and assess stability of the bridge using high resolution X-band radar imagery acquired in spotlight mode by TerraSAR-X satellite from February 2014 to February 2015.

We show that thermal expansion of the bridge clearly appears in the interferograms causing a correlation by up to 96% between the phase difference over the bridge and the difference in temperature of the images used for interferogram generation. Such nonlinear effect influences long-term stability analysis of the structure using standard InSAR time-series processing and therefore should be considered and modeled as an additional parameter in the chain of processing before estimating long-term velocity of the object.

The Sentinel-1 satellites allow for near real time monitoring of movements on Planet Earth. With large parts of the globe covered several times per month, InSAR deformation measurements are about to enter a new phase. We now go from case driven stand-alone projects, isolated in time and space, towards continuous global monitoring.

At SkyGeo, we make available wide area products at the scale of subcontinents, based on routinely updated Sentinel-1 based InSAR deformation measurements. In this contribution, we show examples of the current status of this large-scale map, and elaborate the processes and decisions involved in making, maintaining and presenting it.

The processing of these complex data involves interaction with the Sentinel Scihub, stack formation, coregistration, initial analysis and subsequent updating of the deformation maps. We discuss turnaround times, reliability of the updated maps, quality control and quality reporting, as well as user uptake of the data.

The Ka-band frequency is obtaining a lot of attention from various areas and especially from spaceborne remote sensing for Earth Observation applications. It has been already several years that the interests moved to the 35 GHz frequency band, investigating the advantages and disadvantages of such band for the SAR technology. The 35 GHz band has been used already for certain applications, for example telecommunication from satellite and cloud and precipitation observations from the ground. However, it has not been used so far for Earth Observation from the space.

The Ka-band frequency at 35 GHz needs to be exploited for the use in the Earth Observation, especially from space. At this frequency, high resolution with small and light payload can be obtained for a large variety of applications. This advantage creates a very important opportunity which needs to be investigated further. In fact, the Ka-band SAR sensor can be used for a large variety of applications in different areas:

High resolution imaging and mapping

Reconnaissance and Surveillance

Oceanography

Glaciology

Atmosphere

Agriculture

Transportation

Climate studies

Hydrology and land use

DEM’s and cartography

Search and Rescue and disaster managements

To date, only very few datasets of airborne SAR at Ka-band are available for the investigation and scientific studies due to a lack of airborne instruments.

MetaSensing has internally developed an airborne Interferometric SAR sensor prototype at Ka-band for the acquisition of datasets and to develop and implement a dedicated data processor.

The radar instrument is based on Frequency Modulated Continuous Wave (FMCW) technology with 400 MHz bandwidth, allowing for 37.5 cm range resolution, and adjustable parameters (PRF, Transmitted power, etc.) configurable for every acquisitions. The transmitting antenna and the two receiving antennas are mounted on a pan and tilt unit which compensate in real time the attitude of the aircraft by using the attitude information from a high precision Inertial Movement Unit (IMU). The stabilization of the antennas is needed in order to achieve very accurate generation of the interferometric images. The two receiving channels with two antennas can be used for Along-Track Interferometry (ATI) as well as Across-Track Interferometry (XTI) depending on the application requirements.

The acquired data are processed off-line with a recently internally developed processor in order to obtain the SAR images and the interferograms. The novel processor is implemented to run on GPU optimizing the processing time to significantly reduce the time to generate an image after the flight.

In this paper, the system including the sensor and the processor, along with the results of different flight campaigns are described and the analysis on the interferometric images shown. 

Due the continuous growth of Earth Observation (EO) image data collections acquired from a great variety of sensors, we can observe an increasing need for methods and techniques of querying remote sensing images, not only by their annotations but also by their semantic content. Various methods of content based image retrieval (CBIR) have been proposed in the remote sensing domain, but no general approaches are available. Regardless of the used method, the CBIR systems have the same function, to identify the most similar images with the query image. Some authors developed powerful CBIR tools such as GeoIRIS system of Shyu C. et al. which is a content-based multimodal Geospatial Information Retrieval and Indexing System and includes automatic feature extraction, visual content mining from large-scale image databases, and high-dimensional database indexing for fast retrieval, KIM - knowledge-driven information mining proposed by Datcu M. et al. which is based on human-centered concepts and implements new features and functions allowing improved feature extraction, search on a semantic level, the availability of collected knowledge and interactive knowledge discovery, SemQuery,developed by Sheikholeslami G. et al., which is a semantics-based clustering and indexing approach, used to support visual queries on heterogeneous features of images. Regarding the idea of finding a common ground between synthetic aperture radar (SAR), optical data and even data fusion products, the goal is to develop an application capable to join feature extraction algorithms and classification algorithms.  Therefore, this paper is presenting a framework of feature extraction methods for SAR, Multispectral and Data fusion image products that can be used in automatic or semi-automatic classification of urban areas. Our results demonstrate the usability of patch based image classification techniques that can be applied on Sentinel-1 and Sentinel-2 public data sets.  Also, another goal is to demonstrate how data fusion products perform in the context of patch based image classification and automatic annotation of urban areas. In order to do so, the selected scene is grouped in a few generic classes like buildings, vegetation, forest, water, streets etc. Then we use feature extraction methods such as Gabor filter banks and Weber Local Descriptors in combination with Support Vector Machine (SVM) and k-Nearest Neighbours (k-NN) to define an application to be tested on SAR, optical data and data fusion products. The result of the study is intended to establish the optimum number of classes that can be found in the Sentinel-1 and Sentinel-2 images when using patch based image classification techniques. Also another important objective of this paper is to determine the best patch sizes suitable for this type of classification and that can be used to obtain the best results for Sentinel-1 and Sentinel-2 EO images.

Interferometric coherence is a well-defined parameter used as a measure of the quality of the interferometric phase in SAR related applications. It is understood that interferometric coherence decreases with time between SAR acquisitions as a result of changes in surface reflectivity, reducing the quality of SAR interferometric (InSAR) measurements. This is exactly the reason why coherence contains intrinsic information on the land surface characteristics and can be considered as an expression of temporal decorrelation.
Multiple studies have shown the applicability of InSAR coherence for various applications related to land cover/use classification. However, the multi-temporal dimension given in studies dealing with coherence change detection is limited to a small number of well-selected SAR acquisitions suited temporally to the application in hand. Such approaches are partly driven by the lack of sufficient amount of SAR data to extract robust statistical metrics for coherence temporal behaviour and in turns for temporal decorrelation.
The availability of significant number of archived data from previous SAR missions of the European Space Agency (ESA) consist a wealth of information to be exploited. The emergence of SAR systems able to provide systematic acquisitions, such as the Copernicus Sentinel-1 missions, are introducing a major change in the concept of monitoring InSAR coherence.
It has been already demonstrated, in previous work presented by the research team, that tailored analysis of interferometric coherence exploiting the large time series of SAR data enables the accurate quantification of temporal decorrelation. The methodology to translate the observed rate of coherence loss into decorrelation times is further discussed herein and applied and tested for land cover mapping application. The advantages and disadvantages of using temporal decorrelation estimates versus simple temporal coherence average are addressed.
Though the dependence of decorrelation on various land cover types is already been documented, the provision of additional information regarding the expected time of decorrelation is of practical use especially when Earth Observation (EO) data are utilized in operational activities. The performed analysis is viewed within the improved capacity of current and future SAR systems, while underlining the necessity for exploitation of archive data. The objective is to demonstrate a potential usage of SAR data from any available SAR mission presenting an approach that builds on these large data stacks in a more systematic matter.
The pilot site selected for the demonstration of the proposed technique is the broader area of Athens, Greece. Processing of SAR data was performed using SNAP/Sentinel-1 Toolbox and GAMMA software packages. The entire processing was implemented on ESA’s Grid Processing on Demand (G-POD) facilities, a Grid and Cloud-based operational environment, where EO scientific algorithms can be integrated and executed.
Validation of obtained results using existing land cover/use datasets, such as the Urban Atlas, was considered. Taking into account differences in the reference time of these products, several advantages can be seen both in terms of driving the classification procedure and provision of an important source of information to systematically update existing land cover maps.

The need for accessing historical Earth Observation (EO) data series strongly increased in the last ten years, particularly for long-term science and environmental monitoring applications. This trend will increase even more in the future, as it enables global change monitoring and policy makers decisions on the observed changes particularly in the climate system (atmosphere, ocean, cryosphere, carbon and other biogeochemical cycles, sea levels).
The content of EO data archives is extending from a few years to decades and therefore, their value as a scientific time-series is continuously increasing. Hence there is a strong need to preserve the EO space data without time constraints and to keep them accessible and exploitable, as they constitute a humankind asset. This paper describes the ESA Long Term Data Preservation (LTDP) programme activities aimed at the generation of long time series of coherent data through valorisation and curation of heritage data and for providing continuity with current and future missions.Data Preservation, Long Time Data Series, Data Curation, Data Valorisation, Data Management, Data Stewardship and Multi-disciplinary Applications

After the launch of the Sentinel‐3A satellite, the commissioning phase (Phase E1) will start including the functional and performance verification of the satellite, its instruments and the associated ground segments. In order to verify the production and quality of the optical products, a specific processing facility was set-up at ESTEC to allow the verification and the initial validation of Level 1 products. For this verification also some first Level 2 product will be used (e.g., instrument straylight). The paper will report of the early findings during the commissioning phase and will provide an outlook on activities still to be planned and/or to be continued during phase E2.

Consortium led by THALES ALENIA SPACE as prime contractor for Sentinel-3
satellite will contribute to the phase E1 activities until the In Orbit Commissioning
Review which shall declare the spacecraft and system ready for operational phase.
As part of these activities, Thales Alenia Space France which have built the Altimeter
instrument (SRAL) and Airbus Defence and Space Spain which have built the
Microwave Radiometer (MWR) will be in charge of System & Instrument In Orbit
Validation activities (SIOV). Those activities shall allow to verify the adequate in orbit
behavior of the instrument and finally validate the processing models and parameters
which have been used and provided to ESA to build the operational processing. After
SIOV is successfully completed the CALVAL phase of the commissioning will be
officially authorized to start.
The paper presents the industry activities and main results regarding in-flight
verification and validation of instrument level performances.