GNSS Earth Observation

The presentation will cover ongoing EUMETSAT (European Organisation for the Exploitation of Meteorological Satellites) RO (Radio Occultation) activities, including the currently operational EPS program, the future EPS-SG and Jason-CS/Sentinel-6 programs, and recent RO related reprocessing and research at EUMETSAT.

EUMETSAT runs the operational EPS (EUMETSAT Polar System) program with its Metop satellites. Currently 2 Metop satellites (Metop-A and -B) are operational in the mid-morning sun-synchronous orbit (Equator crossing time at 09:30), both carrying the GRAS RO receiver. A third Metop (-C) satellite is scheduled for launch around 2018, thus RO observations within the EPS program should be provided by the GRAS instruments into the mid-2020. An overview of the performance of the GRAS instrument is given, along with possible future enhancements regarding the signal tracking and an extension of the covered altitude range to allow tracking up to 120km instead of only up to 80km altitudes (which is the current limit).

The follow-on program to EPS is the EPS-SG (-Second Generation) one, which will host on all its satellites, among other instruments, also RO receivers capable to initially track GPS and Galileo. EPS-SG includes in total 6 satellites which will continue the EPS orbit coverage, providing observations into the 2040s. The RO instrument is planned to include also future Global Navigation Satellite System (GNSS) signals from GLONASS and COMPASS/BeiDou, bringing the total number of occultations per day to > 2000. All these GNSS constellations will be observed at the L1 and L5 GPS frequency. A short overview of the instrument capabilities and some open points is given.

A further program with a RO instrument and EUMETSAT involvement is the Jason-CS/Sentinel-6 one. The RO instrument will be provided by the US and allows observing GPS and current GLONASS signals. The RO instrument design is based on the one used for COSMIC-2. An overview of the expected performance is given in this talk.

Finally, reprocessing and research related activities are presented. Reprocessing of RO data at EUMETSAT is ongoing, the complete Metop-A and -B data set has already been reprocessed up to end of 2014 and data quality evaluation and potential improvements are investigated. Reprocessing of the US/Taiwan COSMIC RO data set is underway and first validations of these consistently reprocessed data streams will be shown. Furthermore, RO measurements have been co-located to GRUAN radio sonde data. This will include also an assessment of the latest COSMIC reprocessing stream provided by the University Corporation for Atmospheric Research (UCAR, Boulder, US).

In recent years the possibility to apply the GNSS-R (GNSS-Reflectometry) or PARIS (Passive Reflectometry and Interferometry System) techniques for investigating on geo-bio physical parameters of the Earth has been in depth investigated. Despite of being initially conceived as a means toward sea surface altimetry, GNSS reflections offer many other potential applications which have been proved by numerous experimental activities such as ocean wind speed, soil moisture changes , sea surface state determination, and sea ice detection and classification. The possibility to extend this technique for the monitoring of other components of the Cryosphere has been also investigated. In Austral summer 2009 a pilot experiment for deriving the physical properties of the ice sheet using this technique has been conducted. The experiment, having duration of one month, has been carried out at Dome-C in the middle of the East Antarctica Plateau, where the Italian-French base of Concordia is located. The data collected by the GPS Open Loop Differential Real-Time Receiver (GOLD-RTR) have been used to sense the sub-surface structure of the snow by means of a new holographic observable. The obtained reflected signals present interferometric patterns consistent with the coherent superposition of signals reflected in different layers of the snow sub-structure, down to a few hundred meters depth. A new radio-holographic observable is also defined, to help identify the layer at which the internal reflections occur. In order to interpret these data an electromagnetic  forward model which consider the multiple reflections occurred in the numerous layers composing the ice-sheet has been developed to compare the amplitude and phase behavior of the interferences with the data. The model propagates and traces the signals down into the snow layers; accounts for attenuation, circular co-polar interface-transmission and circular cross-polar interface-reflection losses; accounts for the total phase delay of each internal reflection ray; sums up all the contributions coherently (phase and amplitude); and convolves with the GPS C/A code. Whereas simulated and measured data show similar patterns the lack of detailed information on ice sheet structure limits the capability to correctly reproduce the data. In order to overcome this point a new experiment is planned starting from Austral summer 2015. In this experiment GNSS-R and PARIS data will be acquired, again at Concordia base, in a pristine area for around one month. The same area will be overviewed by a multi-frequency GPR system, in the 500-900 MHz range, installed in a sledge. Data collected by this system will be able to accurately reproduce the snow structure as observed by the GPS reflected signal down to around 1000 m.  Moreover the snow structure of the first two meters below the surface (i.e. where the capability of GPR is limited) will be investigated by using Near Infrared Radiometry pictures collected in several trenches manually excavated in the same area.  Acquired data allows us to correctly model the ice sheet structure and then to better interpreting  GNSS-R data. At the end of this experiment the system will be devoted to the observation of the ice sheet for a period of one year in order to investigate its capability in the monitoring of surface and subsurface modifications of the first layers of the ice-sheet. These latter has been observed by radiometric measurements at L-band which were collected in the same area but, up to now, not been completely explained because of the lack of temporal data of the snow structure. Results obtained in this project will be fundamental for applying the same technique to air-borne or space-borne platform for the monitoring of the whole Antarctic continent.

Knowledge of the current snow situation, especially on the water stored as snow, is of major interest to regional hydrological operators everywhere in the world. Especially in Northern regions, such as Canada or Scandinavia, the awareness of snow cover dynamics is essential for the daily operations of renewable energy production and early warning / flood forecast.

Due to the low density of reliable in-situ information, only available from costly stations or field campaigns, there is a large need for improvements. Especially in remote areas with a low density of technical infrastructure, like communication or winter accessible road networks.

Within SnowSense, an ESA IAP Demo Project (2015-2018), the team consisting of science, application and hardware specialists, will apply up-to-date earth observation, environmental modeling and novel in-situ hardware components. These methods are applied on the background ofdetailed user requirements, developed during the early stages of the project, (e.g. Fast Track Feasibility Study).

Within the presentation, the specific situation of the user and their requirements in Newfoundland, Canada along with the transferability to large parts of the cyrosphere (e.g. Europe and Scandinavia) will be provided. Based on the regional extension of the watersheds the application and results of Sentinel-1 (EW and IW mode), near-real-time processing and wet snow monitoring results (already investigated in 2014/2015) will be presented. To obtain consistent background information on the snow cover (esp. spatial SWE distribution) a detailed physically based surface process model (PROMET) is applied. Driven from global / national numerical weather prediction resources (e.g. GFS / GEM), an additional information layer on snow probability and dynamics can be integrated. Another key element of SnowSense is the application of a novel in-situ system for snow parameter monitoring, based on GNSS signal interpretation. This method, enhanced by sat com capabilities (planned), will allow to retrieve distributed real time information on the snow parameters.

Based on the components of local measurements, the regional NRT EO and model results a reliable information on snow cover, its dynamics, and the prediction for run-off formation for the next days will be enabled.

For the symposium, the preliminary results of the activities in 2015 in the Alps and in Newfoundland will be presented.

GNSS-Reflectometry (GNSS-R) is already recognized as a well suited technology for several Earth Observation applications. Applications to sea state monitoring and altimetry can be considered quite mature and spaceborne missions are already planned (e.g., CYGNSS, ISS-GEROS) or even already operating (TechDemoSat-1) with the main focus on ocean. Land applications have been also investigated, showing that significant contributions can be gathered in hydrology, agriculture and forestry exploiting ground based receiver networks and the interference pattern between the direct signal and the one reflected at grazing angles, as well by combining direct and reflected signals collected on ground by separated uplooking and downlooking antennas. GNSS-R receivers can have also potential land applications from airborne or spaceborne platforms since bistatic radar measurements at L-band around the specular direction show to be sensitive to soil moisture and vegetation parameters. As for soil moisture, the relevant mechanisms is the sensitivity of soil permittivity to moisture, which on its turn affects the surface reflectivity [Zavorotny,et al., 2003]. However, the received signal is also affected by soil roughness and is a combination of coherent and incoherent contributions that make the retrieving problem quite challenging. Regarding biomass, the signal reflected coherently by the terrain is attenuated by the vegetation dependending on the biomass amount and/or vegetation height, but again the phenomenon is dependent on vegetation type and other parameters, like soil conditions [Ferrazzoli et al., 2011].

This perspective has been investigated in two experimental activities funded by the European Space Agency, based on a GNSS-R sensor developed by Starlab (Barcellona): the LEiMON [Egido et al., 2012] and GRASS [Egido et al., 2014] campaigns. The two projects have provided data suitable for the validation of a simulator (SAVERS) of the data collected by a GNSS receiver looking down to the surface as a function of the main surface parameters, namely the soil roughness and moisture and the superimposed vegetation biomass (both crop vegetation and forest) [Pierdicca et al., 2014].

In this paper we shortly review the simulation package. The overall validation activity and results performed using the data from the two mentioned campaigns (ground and airborne based) is presented, showing the fairly good performances in predicting the measured down-looking antenna signal magnitude normalized to the up-looking one, at least when considering the LHCP polarized signal. The simulator is then used to investigate the capability of a satellite based GNSS receiver in different configurations to retrieve moisture and vegetation parameter. In fact, in order to cope with the difficulty associated to the combined effects of vegetation, moisture and roughness on the signal, it is necessary to acquire a suitable set of independent measurements. The potential of multipolarization measurements (LHCP and RHCP) as well as multistatic (specular and backscattering) is investigated. When possible, an analysis of the data made available in the frame of the TechDemoSat-1 mission is preliminary carried out.    

Egido A., M. Caparrini, G. Ruffini, S. Paloscia, E. Santi, L. Guerriero, N. Pierdicca and N. Floury, “Global Navigation Satellite Systems Reflectometry as a Remote Sensing Tool for Agriculture”, Remote Sensing, 4, pp. 2356-2372, 2012.

Egido A., S. Paloscia, L. Guerriero, N. Pierdicca, E. Motte, M. Caparrini and N. Floury, “Airborne GNSS-R Soil Moisture an Above Ground Biomass”, submitted to IEEE Journal of Selected Topics in Applied Remote Sensing, 2014

Pierdicca, N., Guerriero, L. , Brogioni, M.,  Egido, A., “SAVERS: a simulator of GNSS reflections from bare and vegetated soils”, in print on IEEE Transactions on Geoscience and Remote Sensing, 2014

Ferrazzoli P., L.Guerriero, N. Pierdicca, R. Rahmoune, ― Forest biomass monitoring with GNSS-R: theoretical simulations”,  Advances in Space Research, 47, pp. 1823-1832, 2011.

Zavorotny, V. U., D. Masters, A. Gasiewski, B. Bartram, S. Katzberg, P. Axelrad, and R. Zamora, ”Seasonal polarimetric measurements of soil moisture using tower مbased GPS bistatic radar”, in Proceedings of IEEE 2003 International Geoscience and Remote Sensing Symposium, IGARSS 2003, vol. 2, pp. 781–783, 2003.

Monitoring the atmosphere to gain accurate and long-term stable records of essential climate variables (ECVs) such as temperature is the backbone of atmospheric and climate science. Earth observation from space is the key to obtain such data globally. Currently, however, not any atmospheric ECV record can serve as authoritative reference from weekly to decadal scales so that climate variability and change is not yet reliably monitored, despite of satellite data since the 1970s.

We aim to solve this decades-long problem for the fundamental state of the atmosphere, the thermodynamic state of the gas as expressed by air density, pressure, temperature, and tropospheric water vapor, which are the fundamental ECVs for tracking climate change and in fact fundamental to all weather and climate processes. We base the solution on the unique SI-traceable data of the GNSS radio occultation (RO) observing system, available since 2001 and scheduled long-term into the future. We introduce a new system modeling and data analysis approach which, in contrast to current RO retrieval chains using classical data inversion, enables us to exploit the traceability to universal time (SI second) and to realize SI-traced ECV profiles, accounting also for relevant side influences, with unprecedented utility for climate monitoring and science.

We work to establish such a trace first-time in form of the Reference Occultation Processing System rOPS, providing reference RO data for cal/val and climate applications. This rOPS development is a current cornerstone endeavor at the WEGC Graz over 2013 to 2016, supported also by colleagues from EUMETSAT, ECMWF, DMI Copenhagen, UCAR Boulder, JPL Pasadena, and others. The rOPS approach demands to process the full chain from the SI-tied raw data to the ECVs with integrated uncertainty propagation.

We first briefly summarize the RO promise along the above lines and where we currently stand in quantifying RO accuracy and long-term stability. We then introduce the new concept and design and discuss the development status and early results from the rOPS, with emphasis on its value to provide RO data capable to serve as an SI-traceable reference standard for cal/val and climate. We close with an outlook to next steps of work, in particular regarding the integrated uncertainty estimation, and to the first rOPS climate records 2016.