Applicability of Four Soil Moisture Reanalysis Datasets over Eastern Northwest China, North China, and Jianghuai Region

Based on the observational soil moisture data collected at 778 agrometeorological stations in China from 1992 to 2010, ERA-Interim reanalysis soil moisture data, JRA55 reanalysis soil moisture data, NCEP-DOE R2 soil moisture data, and Twentieth Century Reanalysis (20CR) data, four statistical quantities, i.e., mean bias, correlation coefficient, standard deviation of differences, and ratio of standard deviations were calculated first. The applicability of these four reanalysis soil moisture datasets over eastern Northwest China, North China, and Jianghuai region were then investigated based on the four quantities by using the Brunke ranking method and the empirical orthogonal function analysis (EOF). Major conclusions are as follows. In the spring and summer, the JRA55 data is drier in eastern Northwest China with the seasonal average deviations at most stations are between -0.08 m³ m^-3 to 0.08 m³ m^-3. The soil moisture content in ERA-Interim, NCEP-DOE R2, 20CR is larger than observations; the average deviations in southern North China and Jianghuai region are less than the average deviations in northern North China and eastern Northwest China. For the interannual variability, the ERA-Interim reanalysis data agrees best with the observational data; it also best reproduces the variation tendency of the observed soil moisture in eastern Northwest China, North China and Jianghuai region. Overall, the ERA-Interim reanalysis data shows the best relationship with the observed soil moisture data, followed by the JRA55, NCEP-DOE R2 data, and the 20CR data is the worst.