Merging Study Of Multiple Active And Passive Microwave Remotely Sensed Soil Moisture Products In China

In situ soil moisture data from Automatic Soil Moisture (ASM) of CMA which was seriously quality controlled was used to determine the reliability of one modeled soil moisture product, three passive microwave remotely sensed soil moisture products and one active microwave remotely sensed soil moisture product over China in year of2012, namely CLDAS, WINDSAT, SMOS, FY3B and ASCAT. The CLDAS soil moisture product was used as a reference data set for merging multiple active and passive microwave satellite soil moisture products over China in year of2012, by the cumulative distribution function (CDF) matching approach. The merged soil moisture product (MSMP) was compared with the American soil moisture operational product (SMOPS).The main conclusions were as follows.(1) It could be known from multi-source soil moisture products comparison that the CLDAS soil moisture product was of the best quality performance over China in year of2012, the ASCAT active microwave remotely sensed soil moisture product rowed second. Among the three passive microwave satellite soil moisture products, the WINDSAT product was of the minimum deviation with the observed values although drying serious, and the FY3B product backward, while the SMOS product was hardly affected by radio frequency interference, resulting in bad quality.(2) Systematic bias of the microwave remote sensing soil moisture products could effectively be reduced by CDF matching method along with the original change information of each product preserved. The correlation coefficient between each microwave remotely sensed soil moisture product after CDF matching and the ASM product had increased, but not obviously. After CDF matching process, the soil moisture became moist with deviation from negative to positive, and the root mean square deviation reduced as well. (3) Fusing multiple active and passive microwave remotely sensed soil moisture products could obtain improved soil moisture products and also could increase the usage of remotely sensed soil moisture products. Compared MSMP with SMOPS could found that two products were similar on the spatial distribution and time series trend over China. Both merged products performed wetter in the whole year of2012when compared with ASM product. It could be known from the error characteristics that both MSMP and SMOPS were of high correlation which were mainly concentrated in between0.3-0.9, and MSMP was slightly higher than that of SMOPS.