Improvement of land surface model simulations over India via data assimilation of satellite-based soil moisture products

Realistic representation of surface states using the land surface model (LSM) is extremely challenging owing to human-induced changes and uncertainty in forcing data. In this study, we focus on two crucial objectives pertaining to hydrology namely (a) to understand the ability of different soil moisture (SM) products to improve simulation of unmodeled irrigation processes through data assimilation process, and (b) to learn the feasibility of these SM products to correct the spatial surface soil moisture artifacts caused due to error in precipitation forcing. The utility of SM products evaluated in the present study are retrieved from different satellite sensors and algorithms such as the active satellite-based the Advanced Scatterometer (ASCAT), merged SM from European Space Agency Climate Change Initiative (ESA CCI V4.2) and the latest passive microwave-based SMOS INRA-CESBIO (SMOS-IC) SM product. The results presented for three years (2010, 2011 and 2012) suggest that assimilation of ASCAT and CCI based products effectively captures the SM changes due to irrigation. Similarly, it corrects the spatial artifacts caused due to precipitation errors. However, the single sensor based products have a limitation in spatial samples per day which is critical to capture dynamic SM products over a larger area. Hence, blended products are more effective on a larger area to capture the dynamics in a more effective manner at daily temporal resolutions.