Abstract:  Soil moisture is one of the most important components of land-atmosphere coupling system, and soil moisture monitoring plays a significant part in climate, hydrology, and agriculture. Active microwave and passive microwave are two basic microwave approaches which are used to monitor soil moisture. As of now, the passive microwave method is widely used due to its longer wavelengths and stronger penetrating power. It was considered that the passive microwave retrieved method could work well in effectively monitoring spatial and temporal changes of soil moisture in large-scale areas. However, the data retrieved by satellites needs further evaluation and verification. At present, various microwave methods have been proposed for soil moisture retrieve, and a number of corresponding soil moisture products have also been published. Compared to station-based data, remote sensing data can better reveal the dynamic change of soil moisture in a certain region at grid points. Based on the observed data of station-based soil moisture at the upper soil layer (0-10cm) during the growing season (May-October) in 2018, this paper collected and examined the remote senescing datasets from FY-3B, FY-3C, ASMR2 and SMOS which were consistent with the station-based data in time and space. Furthermore, the applicability of FY-3B/3C fusion in different regions of Inner Mongolia was evaluated, which may provide a reliable scientific basis for the application of soil moisture products based on Fengyun Satellites and other related researches. The ascending and descending data of FY-3B and FY-3C were fused respectively by employing weighted average method. In order to evaluate and compare the applicability of remote senescing datasets from AMSR2, FY-3B/3C and SMOS in Inner Mongolia, FY-3B/3C datasets were then formed by random forest method. The results showed that daytime data were of better quality than night data of FY-3B ascending/descending and FY-3C ascending/descending. The data quality of fused FY-3B and FY-3C processed by weighted average method exhibited no significantly improved. And the data quality of FY-3B/3C products formed by random forest models was significantly enhanced. In the rainy season of high vegetation coverage area (NE), the quality of FY-3B/3C data products were better than those of SMOS and AMSR2. Overall, in Inner Mongolia，SMOS is more applicable in Middle (M) and Southeast (SE) regions, AMSR2 has poor applicability in the whole region, while FY-3B/3C performs the best.

Key words:  FY-3B/3C, Soil moisture, Data fusion, Remote sensing monitoring, Applicability