Abstract:
The Gravity and Recovery Climate Experiment (GRACE) has revolutionized large-scale remote sensing of the Earth’s terrestrial hydrologic cycle and has provided an unprecedented observational constraint for global land surface models. However, the coarse-scale (in space and time), vertically-integrated measure of terrestrial water storage (TWS) limits GRACE’s applicability to smaller scale hydrologic applications. In order to enhance model-based estimates of TWS while effectively adding resolution (in space and time) to the coarse-scale TWS retrievals, a multi-variate, multi-sensor data assimilation framework is presented here that simultaneously assimilates gravimetric retrievals of TWS in conjunction with passive microwave (PMW) brightness temperature (Tb) observations over snow-covered terrain. The framework uses the NASA Catchment Land Surface Model (Catchment) and an ensemble Kalman filter (EnKF). A synthetic assimilation experiment is presented for the Volga river basin in Russia. The skill of the output from the assimilation of synthetic observations is compared with that of model estimates generated without the benefit of assimilating the synthetic observations. It is shown that the EnKF framework improves modeled estimates of TWS, snow depth, and snow mass (a.k.a. snow water equivalent). The data assimilation routine produces a conditioned (updated) estimate that is more accurate and contains less uncertainty during both the snow accumulation phase of the snow season as well as during the snow ablation season.