Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #342523

Research Project: Integrating Remote Sensing, Measurements and Modeling for Multi-Scale Assessment of Water Availability, Use, and Quality in Agroecosystems

Location: Hydrology and Remote Sensing Laboratory

Title: Correcting satellite-based precipitation products through SMOS soil moisture data assimilation in two land-surface models of different complexity: API and SURFEX

item ROMAN-CASCON, CARLOS - Universite Grenoble Alpes
item PELLARIN, T. - Universite Grenoble Alpes
item GIBON, FRANCOIS - Universite Grenoble Alpes
item COSME, EMMANUEL - Universite Grenoble Alpes
item KERR, Y. - University Of Toulouse
item BROCCA, LUCA - National Research Council - Italy
item MASSARI, CHRISTIAN - National Research Council - Italy
item Crow, Wade
item FERNANEZ, DIEGO - European Space Agency

Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/15/2017
Publication Date: 10/1/2017
Citation: Roman-Cascon, C., Pellarin, T., Gibon, F., Cosme, E., Kerr, Y., Brocca, L., Massari, C., Crow, W.T., Fernanez, D. 2017. Correcting satellite-based precipitation products through SMOS soil moisture data assimilation in two land-surface models of different complexity: API and SURFEX. Remote Sensing of Environment. 200:295-310. https//

Interpretive Summary: Accurate measurements of daily rainfall accumulations currently form the backbone of most operational agricultural drought monitoring systems. However, a relatively small faction of the global land surface is actually covered by ground-based rain gauges and this fraction has been falling during recent decades. Satellite-based estimates of rainfall accumulations can be used to fill this gap; however, such estimates are prone to large retrieval errors and temporal sampling problems due to gaps in satellite coverage. Recently, investigators have begun to look into the possibility that existing rainfall accumulation estimates (based on direct ground or satellite observations) can be combined with indirect estimates of daily rainfall accumulation inferred from surface soil moisture dynamics (which are readily observable from satellites). These techniques hold great promise for producing the best-possible estimates of rainfall accumulation in areas of the world lacking ground-based instrumentation (including areas of Africa prone to drought). This paper describes new progress in the design of a rainfall estimation system based on this principle. The eventual operational application of this system is expected to significantly improve our ability to monitor rainfall accumulation in critical agricultural regions of the world.

Technical Abstract: Real-time rainfall accumulation estimates at the global scale is useful for many applications. However, the real-time versions of satellite-based rainfall products are known to contain errors relative to real rainfall observed in situ. Recent studies have demonstrated how information about rainfall intrinsically contained in soil moisture data can be used to improve rainfall accumulation estimation. That is, the dynamics of this soil variable are influenced for several days by the amount of rainfall after a rainy event. In this context, soil moisture data from the Soil Moisture Ocean Salinity (SMOS) satellite is used in to correct the rainfall accumulation estimates provided by satellite-based real-time precipitation products such as CMORPH, TRMM-3B42RT or PERSIANN. In particular, an assimilation algorithm based on the data assimilation of SMOS measurements is tested in two land-surface models with different complexities: a simple hydrological model (API) and a more sophisticated state-of-the-art land-surface model (SURFEX). We show how the assimilation technique, based on a particle filter method, leads in significant improvement in the rainfall estimates, with slightly better results for the simpler (and less expensive computationally) API model. This methodology is evaluated for six years in ten sites around the world with different land use and climatological features. Results also show the limitations of the methodology in regions highly-affected by mountainous terrain, forest or by intense radio-frequency interferences (RFI) in the microwave L-band, which notably impacts the quality of soil moisture retrievals from brightness temperatures by SMOS. The satisfactory results shown here demonstrate the feasibility of an operational application in near-real time at the global scale.