Assimilating Remote Sensing based Soil Moisture in an Ecosystem Model (BEPS) for Agricultural Drought Assessment

Process-based terrestrial ecosystem models inevitably need model initialization and parameters specification. In this study, remotely sensed surface soil moisture derived from near infrared and shortwave infrared bands was assimilated in BEPS (Boreal Ecosystem Production Simulator) to initialize soil moisture in BEPS and fine-tune BEPS key parameters which are closely related to soil moisture estimation including maximum stomotal conductance, leaf area index (LAI) and root density. An Ensemble Kalman Filter is used to perform data assimilation and parameter adjustment. The result shows that using the optimized parameters, the performance of model predictions of 0-10 cm soil moisture was greatly improved compared with the surface soil moisture fields derived from remote sensing data. It is demonstrated that the method of assimilating remotely sensed soil moisture in the BEPS model can help improve the soil moisture results of the BEPS model in the arid and semiarid area and provide a feasible way to monitor drought and to assess its influence on agriculture.