Better use of Remotely Sensed Soil Moisture to Initialize Soil Moisture in Global Climate Models
Wenge Ni-Meister, Jeffrey P. Walker and Paul R. Houser
Recent developments in remote sensing technology provide us a great opportunity to improve our understanding of land surface processes by integrating remote sensing data and land surface processes models. Some assimilation methods have been developed to assimilate the remotely sensed data into Soil-Vegetation-Atmosphere Transfer Scheme (SVATS) for better prediction of land surface processes. This study focuses on how to make full use of remotely sensed soil moisture data to better initialize soil moisture for seasonal and annual weather prediction. We are looking for answers for the issues related to data assimilation methods, such as: 1). Data assimilation method has been used to predict the soil moisture profile using remotely sensed soil moisture at the very thin surface, can we do a good job to predict the soil moisture from surface to root zone using data assimilation? 2). Satellite data and land surface models provide different approaches to estimate soil moisture, is there any system bias between the remotely sensed and model predicted soil moisture? 3). Do the predicted soil moisture profiles using data assimilation show the same seasonal and annual variabilities as the in-situ measurements? Remotely sensed, model predicted and in-situ measured soil moisture data in Eurasia are used for the analysis. The land surface model used in this study is the catchment-based land surface model of Koster et al. (2000). The model is unique in that it models the sub-grid variability of topographic control on soil moisture, surface runoff and evaporation through the hydrologic watershed rather than the conventional atmospheric grid.