Assimilation Experiment of LIS Based on AMSR-E Soil Moisture Products

Soil moisture is an important part of the soil, playing an important role in energy and mass exchange between land and atmosphere. It is an important environment factor and the process parameter of hydrology, meteorology and other researches. In order to accurately depict and understand the interaction between land and atmosphere, precise and high resolution soil moisture contour information is needed. But whether using the objective measurement or numerical prediction of land surface process simulation, soil moisture contour line with spatial-temporal continuum and good precision can't be obtained. In order to eliminate the error, data assimilation method (EnKF) is used to process passive microwave inversion products (AMSR-E) and land surface model (NOAH) outputs, and then shallow soil moisture information is transmitted to deeper soil layers by land assimilation system (LIS). These methods are used to calculate soil moisture with high precise, high resolution and low error. Through these methods, 4-layer (10 cm, 30 cm, 60 cm, 100 cm) soil moisture data with spatial resolution of 0.25° by 0.25° for the year of 2003 are obtained.Agro-meteorological-station soil moisture data and ecological-station data are used to verify the 4-layer soil moisture results. The simulation accuracy of the land surface model is improved when data assimilation method is used, the simulation accuracy in different layers is different, and the accuracy in the grassland ecosystem is higher than that in crop and forest ecosystems. The spatial characteristics of assimilation results match with real values, which is high in the southeast and low in the northwest. Assimilation soil moisture data has more detailed information in horizontal scales compared to station data and has more detailed information in vertical scales compared to the passive microwave inversion products.The accuracy of the assimilation process depends on the passive microwave inversion products (AMSR-E). However, AMSR-E has stopped running and inversion errors of passive microwave inversion products will be accumulated in data assimilation run. In order to solve these problems, passive microwave inversion products of FY-3B can be used experimentally. Besides, assimilation experiment can be started by directly assimilate radiation brightness temperature of microwave channels. Simulation accuracy of soil moisture can be further improved through these processes. Effect of assimilation experiment is different in different vegetation, which mainly depends on capabilities of land surface model. Therefore, improving the land surface model is also an effective means to improve soil moisture simulation accuracy.