Sentinels for Science

Climate change scenarios foresee increasing anthropogenic activities and their environmental effects at high latitude regions (especially in the Arctic). In particular, nitrogen oxides (NO_x = NO + NO₂), mainly generated by anthropogenic combustion, play an important role in tropospheric chemistry and have harmful effects on human health. Satellite observations are best suited for atmospheric composition monitoring over large regions where ground-based measurements are only scarcely available or difficult to maintain. However, satellite observations are not yet extensively evaluated and applied for air quality monitoring at high latitudes.

In this context, ILMA-project (supported by the ESA Living Planet fellowship programme) aims at evaluating the existing satellite retrievals at high latitude and increasing the scientific exploitation, with specific focus on NO₂ (and volcanic SO₂) observations in Arctic region. NO₂ and SO₂ columns are currently provided for example by OMI (Ozone Monitoring Instrument), flying onboard NASA’s Aura satellite, which is included in the ESA Third Party mission. The upcoming TROPOMI (TROPOspheric Monitoring Instrument) on Sentinel 5 Precursor will produce observations similar to OMI but with improved spatial resolution (7 × 7 km² instead of OMI’s 16 × 24 km²) and signal-to-noise ratio. The results from ILMA will hopefully be useful in preparation for S5P/TROPOMI data validation and exploitation at high latitudes.

The first results of the project include OMI NO₂ data validation against the ground-based observations from Pandora spectrometer in Helsinki (Finland). The datasets differ on average by -6% and 1% for all skies and clear sky conditions, respectively. The clear sky overpasses mainly correspond to summer days and, thus, to smaller solar zenith angles. Also both datasets show similar seasonal and weekly cycles as surface NO₂ concentration data, with a wintertime peak and a lower signal during the weekend, as compared to the other weekdays.

Also, volcanic SO₂ retrievals from OMI have been evaluated during Holuhraun fissure eruption against Brewer measurements in Sodankylä, Finland. The main differences are related to the unknown vertical profiles and to the algorithm assumptions, which are optimized for middle-low latitudes. These issues cause underestimation of the SO₂ vertical columns. Volcanic emissions were transported at very low altitudes and affected air quality at several stations in Northern Finland. OMI data showed their capability to represent, at least qualitatively, the position of the volcanic plume as compared to ground-based data.

The reconstruction of the ocean surface currents from Sea Surface Temperature (SST) observations can be expressed in terms of a transfer function notation. This transfer function can be theoretically derived using the Surface Quasi-Geostrophic (SQG) equations, assuming that the variability of the density anomaly is dominated by the SST anomaly. Recent study, based on Microwave SST observations, analyzed the validity of SQG at a global scale and revealed that this dynamical model is valid near the major extratropic current systems, which are characterized by an intense mesoscale activity and the presence of strong thermal gradients. Besides, the potential of microwave SST observations to reconstruct ocean surface currents was analyzed using a synergistic approach: an ideal transfer function that combines the phase of SST with the SSH spectra. This synergistic approach revealed that SQG reconstruction can be enhanced. However, despite microwave observations are suitable for global studies, their spatial resolution barely reaches the 60-40 km, scales for which most of the ocean kinetic energy and its dissipation take place. In that sense, simultaneous or nearly simultaneous SST infrared and along-track altimeter observations could be used to further analyzed and characterized the transfer function between them.

In addition, the new operational mission Sentinels-3 will ensure the continuity of simultaneous SST infrared and high resolution altimeter observations that ensures the possibility of retrieving high resolution velocity fields by exploring and exploiting the synergy of both observations. This is one of the main aim of OCEANUS, a research project funded by ESA Living Planet Fellowship. Here we will show some examples of the preparatory activities carried out on the characterization of the transfer function between IR SST and along-track SSH, using existing satellite observations.

The main objective of building an Ocean Virtual Laboratory (OVL) plateform is to allow oceanographers and other scientists to freely discover the existence, visualize and analyse jointly, in a convenient, flexible and intuitive way, the various co-located EO datasets and related model/in-situ datasets over dedicated regions of interest with a multifacet point of view. Synergy tools within the OVL are developed to foster the emergence and prototype new methods and products making use of the complementarity between sensors to study ocean related processes.

The OVL is developed with a special attention to provide the best possible visibility on the upcoming Sentinel1/2/3 datatakes and to help plan and coordinate validation field campaigns. Space agencies are investing in data portals to efficiently distributes specific EO data on one end, and analysis software like the Sentinel toolboxes, IDL/ENVI or Matlab are used for in-depth analysis of a specific dataset. Only a few GIS systems, such as Google Earth or ARCGIS, are able to import several data layers but only very limited data interaction/analysis is possible. The OVL is participating to a long-term effort towards the development of synergy analysis open source tools that fill this gap and provide multisensor data access and analysis capabilities.

We will demo the OVL over the very dynamic Agulhas ocean region.

See more at http://ovl-project.oceandatalab.com/

In the context of global change and population growth, agricultural practices highly impact water availability and quality. Water management strategies in term of quantity and quality have to be analyzed at a regional catchment scales. Yet, crop consumption that account for the main water and nutrient fluxes at the catchment scale needs to be monitored at a high spatial (crop extension) and temporal resolution (crop growth period). This study, funded by an ESA Living Planet Fellowship, aims at demonstrating the improvement brought by synergetic observations of Sentinel-1 and 2 in agro-hydrological modeling. Geo-information time-series such as Leaf Area Index (LAI) with S2 and soil moisture or biomass with S1 are used to re-set soil and agricultural practices parameters of a crop model coupled with a hydrological model.

Two contrasted water management issues are used as demonstrator: stream water nitrate pollution in Gascogne region in south-west of France and groundwater irrigation shortages in Deccan Plateau, in south-India.

In south-west of France, Topography Nitrogen Transfer and Transformation model (TNT2) has been previously used on a small experimental catchment (3.5km², Ferrant et al., 2014a). Leaf Area Index (LAI) derived from S2 type time series (Formosat-2) over 5 years is used to optimize the -1 crop calendars by resetting the seeding dates and 2- within-field crop growth heterogeneity by calibrating soil parameters (depth and porosity) defining the Soil Water Holding Capacity. The synergetic use of S1 time-series (soil moisture and saturated area extent detection) and S2 time series (Leaf Area Index profiles at the pixel size) allows improving soil parameter determination and spatial calibration of hydrological functioning. These spatial calibrations are being evaluated in term of in-stream nitrogen and water fluxes.

In south-India, the Soil Water Assessment Tool model (SWAT) has been previously used on medium scale catchments (around 1000 km²) to simulate the groundwater extraction shortages (Ferrant et al., 2014b).  In this area, farmers irrigate very small plots (<0.1ha), so that high resolution satellite images are needed to accurately identify the extent of irrigated areas. Previous work were limited in term of availability of such data over a period of several years and larger areas. Combining S1 biomass and soil moisture estimation with land cover classification using S2 type time series (from spot5take5 experiment) allows deriving irrigated area extent on several agro-climate contexts of South-India. This seasonal geo-information is used to force the Soil Water Assessment Tool model, an agro-hydrological model adapted to large scale basins, to derive spatial groundwater extraction as a function of climatic variability. An extension of this study to the Sentinel ground coverage will make use of SMOS water fraction and GRACE continental water estimates.

Space Agencies (ESA, NASA, DLR) support users of their satellites with dedicated tools to process and exploit the data from them. These toolboxes comprise generic image processing functionalities, such as image visualisation, geometric image operations, image filters, statistical analysis and format conversions, but also instrument specific capabilities including data ingestion functions for the native formats, sensor data calibration, instrument corrections or sensor specific processing (atmospheric correction, interferometry). As such they complement other opens source and proprietary image processing system. They also provide a simple bridge to GIS systems like ARCGIS or QGIS. Successful examples of such toolboxes are ESA’s BEAM and NEST toolboxes for the optical and SAR sensors on ENVISAT or NASA’s SeaDAS toolbox for SeaWiFS, MODIS, VIIRS and other ocean colour sensors.

Today, after more than two decades of Earth Observation satellites a large archive of heritage data has grown, and the new generation of satellites originating from ESA’s Sentinel or NASA’s PACE mission is adding huge amounts of new data to an already enormous data set of observations. New technological concepts to cope with these data had to be developed and are currently being realised in ESA’s Sentinel Application Platform, SNAP, the Sentinel toolboxes and the Climate Change Initiative Toolbox which aims at exploitation of long term time series combining land, atmosphere and ocean variables.

These specific toolboxes primarily aim at expert users, and in parallel the interest in Earth Observation has grown in the general public, thanks to the increased use of map-type information in public communications (google maps, news). This is currently supported by dedicated tools where a reduced and highly synthesized version of the “raw” EO information is visualised in an attractive way on any kind of devices, including big demonstration screens but also tablets and smart phones. The ESA CCI Visualisation Corner is an example for such a tool. Another set of tools is available for educational purpose, such as ESA / UNESCO's BILKO . This usually includes a reduced sub-set of functionality but is accompanied by tutorials and educational material.

The vastly growing data archives on the one hand, the improvement of processing capabilities especially through flexible cloud processing power, hosted processing and smart end user devices on the other hand require and enable new concepts for user tools. These tools need to work in a highly distributed environment where data streams and processing power are available and “close” to each other in order to provide the requested visualisation capabilities. There is also a new generation of scientists and a knowleable public that grew up with social media and communication tools, and which requires highly interactive tools to share ideas and results. The boundaries between the above classical domains of tools and toolboxes vanish.

In this presentation we will go with the audience through a virtual journey from the past into the future, stop at some milestones of past and current toolbox examples and develop the vision of a Toolbox 2020.