Suping Nie, National Climate Center, China Meteorological Administration, Beijing, China
Assimilation of FY-3B MWRI soil moisture over China using the EnKF method
Abstract: The Beijing Climate Center (BCC) Land Data Assimilation System (BLDAS) based on the Microwave Radiation Imager (MWRI) surface soil moisture products from the Chinese polar-orbiting Fengyun-3 series satellite (FY-3B) is developed for the operational seasonal climate prediction in BCC. The Ensemble Kalman Filter (EnKF) method is applied to assimilate FY-3B soil moisture retrieval products into the BCC Atmosphere Vegetation Interaction Model (BCC_AVIM) over China. A multi-steps quality control (QC) procedure is incorporated on MWRI soil moisture data to remove outlier records and mismatched values to the simulation of BCC_AVIM. By assimilating the QC-based FY-3B soil moisture during the period from January to December of 2012, the FY-3B products, the open-loop simulations and the assimilation estimates are validated against about 600 in-situ measurements in the China, respectively. The FY-3B products can better capture the variations of surface soil moisture over bare soil surfaces than vegetated surface types (e.g. tree, grass, and crop surfaces). For all land cover types, the FY-3B products have a higher accuracy during the summer period while have significant negative biases in other seasons. Over all sites, assimilation of MWRI observations can improve the mean biases, time correlations, and root mean square errors (RMSE) of soil moisture in both surface and root-zone layers, especially when the system errors were removed prior to the assimilation. These results suggest that the FY-3B soil moisture products have the potentiality to improve the soil moisture status in BCC_AVIM and are suitable for implementing in the BCC next generation operational seasonal forecast system.