Variational Assimilation of the SMAP Surface Soil Moisture Retrievals into an Integrated Urban Land Model

Abstract

Soil moisture is a key parameter in land surface modeling. In this study, a variational data assimilation algorithm was applied to an integrated urban land model (IUM) to assimilate the NASA Soil Moisture Active Passive (SMAP) mission Level-4 Soil Moisture product. The 10 cm soil moisture data observed in situ at eight sites was used for validation. A very simple analytical algorithm was developed to characterize the error weighting matrix in the cost function. The results indicated that with assimilation, the simulation results of the surface volumetric soil moisture improved in almost the whole research region as compared with the SMAP data. In most of the time periods, accuracy of simulated surface volumetric soil moisture increased. With assimilation, as compared with the observations at eight sites, the 10 cm volumetric soil moisture improved over the whole research time period.