Earth Observation for vegetation monitoring and water management
Napoli, 10-11 November 2005
Improvement of modeled soil moisture under vegetated terrain using SAR C-band retrieved bare soil moisture and SPOT derived LAI
Valerie E. Janssens 1, Vincent Guissard 2, Valentijn R. N. Pauwels 1, Niko E. C. Verhoest 1,
Pierre Defourny 2
1. Laboratory of Hydrology and Water Management Ghent University, Coupure links 653, B-9000 Gent, Belgium tel: +32 9 2646137, fax: + 32 9 2646236, e-mail:
2. Department of Environmental Sciences, Université Catholique de Louvain, Place Croix du Sud 2 (bte 16), B-1348 Louvain-la-Neuve, Belgium tel: +32 10 472613, fax: +32 10 478898, e-mail:
Objective of the study
Monitoring soil moisture under vegetated terrain is of great importance for crop growth modeling, and irrigation scheduling. Radar remote sensing has demonstrated its large potential as a monitoring technique for surface soil moisture. However, when using a C-band SAR, its applicability during the growing season is restricted due to the attenuation of the soil surface backscattering within the canopy.
Another way to obtain the water content under a vegetated surface is through hydrological modeling. However, soil moisture estimations from hydrological models, such as soil-vegetation-atmosphere transfer schemes (SVATS), are prone to errors due to a variety of reasons. In order to reduce these errors in the modeled soil moisture, and thus improve the modeled water and energy balance fluxes and states, modeled soil moisture values can be updated with observed soil moisture data through data assimilation. During autumn, winter and early spring, the catchment consists of a patchwork of vegetated and bare soils and SAR C-band remotely sensed (i.e. ERS-2 and ENVISAT) soil moisture over bare soils can be assimilated into the hydrological model. Yet, during the major part of the growing season, generally no bare soil fields are present in agricultural areas. In this case, LAI values derived from optical data (i.e. SPOT and MERIS) can be inputted in the model in order to better describe the vegetation, yielding to improved soil moisture estimations by the hydrological model.
In this study we investigate to which degree we can improve TOPLATS (i.e. TOPmodel [Beven and Kirkby, 1979] based land-atmosphere transfer scheme [Famiglietti and Wood, 1994]) modeled soil moisture under vegetation through the assimilation of soil moisture from adjacent bare soils and through the update of the model LAI with LAI maps derived from optical remote sensing. The results are validated against a dataset from intensive field measurements during 2003 over the Zwalm catchment in Belgium.
1