Soil Moisture And Vegetation Optical Depth Retrievals From FY-3B Satellite Microwave Brightness Over China And The Comparative Analysis

Surface parameters such as soil moisture and vegetation water content are important parameters of the hydrological model, climate model and land surface process model. So using microwave remote sensing to obtain the soil moisture and vegetation optical depth has the important application value and the practical significance. Based on the microwave data of FY-3B satellite, we have retrieved the land surface soil moisture over China (LPRM soil moisture) and the daily vegetation optical depth (VOD) over the eastern China using a radiative transfer model. The soil moisture retrievals are then compared with in-situ soil moisture observations from agrometeorological stations and reanalyzed soil moisture dataset from ERA-Interim, NCEP databases and the official MWRI soil moisture daily product of the National Satellite Meteorological Center. The spatio-temporal variations in the derived VOD data is then analyzed and compared with Normalized Difference Vegetation Index (NDVI) data, evaluating the specific performance of LPRM soil moisture and VOD data in China. The main conclusions are as follows:(1) It appears that, the FY-3B soil moisture agrees well with in-situ and reanalyzed soil moisture with respect to their spatial variations over China. The systematic errors of FY-3B retrievals are closely related to the vegetation density. In terms of the seasonal variations, LPRM soil moisture agrees very well with the in-situ observations of agrometeorological sites. However, LPRM soil moisture shows a strong and negative correlation with ERA-Interim and NCEP reanalyzed soil moisture where soil is relatively wet, especially in south China and part of northeast China; and this negative correlation is mainly caused by their differences in seasonal variations. It is suggested that the reanalyzed soil moisture in such regions is of large uncertainty and must be used with great caution.(2) Generally, the spatial distribution of LPRM soil moisture over China agrees well with the official MWRI soil moisture daily products. However, the seasonal variations of LPRM soil moisture and the official MWRI soil moisture daily products show obvious differences. The two datasets show significantly positive correlations over northern China where it is drier, but negative correlations over southern China and the northeastern forest region where it is wetter. The temporal variations of soil moisture with time of the LPRM soil moisture agrees better with the in-situ observations of agrometeorological sites than that of the official MWRI soil moisture daily product. Therefore, compared with the official MWRI soil moisture daily product, the LPRM soil moisture more accurately captures the seasonal variation of soil moisture.(3) It is revealed that, generally the VOD data has higher values in southern China and lower ones in northern China, which is consistent with the NDVI data. The VOD data, prior to the NDVI data, can however better represent the spatial heterogeneity of vegetation at regional scales. With respect to the temporal variations. Theoretically, the VOD data represents the variations in the total above-ground biomass, which compensate for the deficiency of NDVI. Therefore, although NDVI shows significant seasonal variations, the amplitudes of seasonal variations in VOD are rather reduced. This is particularly true in densely vegetated areas, where VOD is concluded to have better capacity for describing seasonal variations of vegetation.