Water and energy budgets simulation over the AMMA-Niger super-site spatially constrained with remote sensing data

The SEtHyS_Savannah model [Saux-Picart et al., submitted for publication. SEtHyS_Savannah: a multiple source land surface model applied to sahelian landscapes. Agricultural and Forest Meteorology] was developed as an extension of the SEtHyS land surface model to simulate the water and energy fluxes over dry savannah landscapes. The vegetation cover is represented by a two layer model and a mulch approach is used for the soil description. The SEtHyS_Savannah model was regionalized over the AMMA-Niger super-site (about 50 km by 40 km), with the help of remote sensing data. The model uses a regular 1 km grid and each cell is divided in sub-grid patches in order to represent land cover and soil heterogeneities (the approach). The vegetation cover parameters were prescribed according to the land cover map and the seasonal evolution of the Leaf Area Index (LAI), both derived from SPOT-HRV (Satellite Pour l'Observation de la Terre - High Resolution Visible) data imagery. The atmospheric forcing was assumed homogeneous over the area and provided by a meteorological station installed at the Fakara experimental site. The surface water and energy budgets were simulated over a one-year period (2005) at a 5-min time step and validated against MSG-SEVIRI (Meteosat Second Generation - Spinning Enhanced Visible and Infra-red Imager) land surface temperature and ENVISAT-ASAR (ENVIronnement SATellite - Advanced Synthetic Aperture Radar) soil humidity products. The results show realistic surface fluxes and good agreement with the MSG-SEVIRI temperature observations. The soil moisture comparison presents significant correlation but large root mean square errors. These discrepancies are the consequence of both the use of a non-spatialized atmospheric forcing and to residual vegetation effects on the radar signal. Despite these uncertainties. the results increase confidence in the model representation of Sahelian soil-vegetation processes and open new perspectives to quantify the effects of vegetation changes on evapotranspiration and runoff over the region