ASSIMILATION OF SOIL-MOISTURE INFERRED FROM INFRARED REMOTE-SENSING IN A HYDROLOGICAL MODEL OVER THE HAPEX-MOBILHY REGION

Hydrological models are generally unable to simulate correctly water e xchanges at the soil-atmosphere interface and the time evolution of su rface soil humidity. This drawback often leads to poor simulation of w ater flows after periods of low flows because of the wrong estimation of the surface soil water content. In this paper, we describe how remo te sensing can be used to account for the vegetation in the estimation of the actual evapotranspiration, and to estimate soil moisture regul arly throughout the year and use it to correct the model simulation. T his work has been done on the Adour river basin, in the framework of t he HAPEX-MOBILHY experiment. The results presented show the improvemen ts that result from the use of remote sensing data in hydrological mod elling: better simulation of the soil moisture and of water flows at t he outlets, and more realistic calculation of evaporation.