The Impacts of Satellite Data Quality Control and Meteorological Forcings on Snow Data Assimilation over the Sanjiangyuan Region

The effectiveness of snow data assimilation is closely related to the satellite data quality control that affects snow cover data used for assimilation and meteorological forcings that drive land surface model to estimate snow depth, especially over headwater regions where in situ measurements are sparse and land surface simulations are challenging. This study proposes a joint quality control scheme based on precipitation constraints and cloud thresholds, uses the Ensemble Square Root Filter to assimilate the controlled data to improve snow depth estimation from the Conjunctive Surface-Subsurface Process model version 2 (CSSPv2), and explores the impacts of different forcing data on the assimilation. The correlation between the assimilated monthly snow depth data and the in situ measurements averaged over 21 stations during November–February of 2000–2015 is 0.93, and the root mean square error is 0.22 cm. Compared with CSSPv2 model simulation, the correlation increased by 5.6%, and the error decreased by 18.5%. The joint quality control scheme has led to an average accuracy improvement of 47%, while the high-quality forcing data have resulted in an average enhancement of 58%. This study suggests that satellite data quality control and meteorological forcings are important for increasing correlation and decreasing error for snow depth assimilation, respectively.