All-weather Soil Moisture Retrieval Based On Geostationary And Polar-orbiting Satellite Data

Soil moisture is a key factor involved in surface energy balance and atmospheric circulation.It is an important indicator variable for monitoring land drought and soil degradation.At the same time,its spatial dynamics characteristics are of great significance to research in the fields of climate,hydrology,and agriculture.Remote sensing technology can observe the changes of soil moisture in a wide range in real time,meeting the current needs of various spatial scales in different research fields.Optical remote sensing has been widely used to obtain soil moisture with high spatial resolution.However,optical remote sensing observations are easily affected by clouds,which makes it impossible to obtain effective observation data of cloudy pixels.Therefore,exploring how to get spatially and temporally continuous soil moisture within a wide range has become one of the mainstream directions of quantitative remote sensing research.Inner Mongolia is an important agricultural and pastoral area in China,and timely acquisition of information on the temporal and spatial changes of soil moisture in this area is crucial to the protection of grassland vegetation and the sustainable development of agriculture and animal husbandry.Although it is currently possible to obtain soil moisture data under clouds through the relationship between aerodynamic resistance and surface resistance based on meteorological elements,it is found by comparison that the accuracy of cloud pixels is lower than that of clear sky pixels,resulting in an overall decrease in the accuracy of spatially complete soil moisture retrieval.Therefore,the paper is based on the"LST-FVC"trapezoidal feature space method composed of Land Surface Temperature(LST)and Fractional Vegetation Cover(FVC),and cooperates with FY geostationary meteorological satellites and polar orbiting satellites to perform pixel-to-pixel soil moisture inversion,obtained spatially continuous semimonthly soil moisture data in the study area.This paper obtained the soil moisture in the study area for a single day,5 days,10 days and 15 days,firstly.Then,based in the common microwave soil moisture products,two main types of land use(grassland and farmland)are selected for accuracy analysis in the study area.It is shown that:(1)Compared with MODIS land surface temperature,FY-4A is significantly underestimated(bias is about-5.185?),resulting in an overestimation of soil moisture(compared to SMAP soil moisture products,bias is about 0.112m~3/m~3).After correction of FY-4A's surface temperature,it can significantly reduce the deviation of soil moisture inversion to about 0.043 m~3/m~3.(2)As the number of synthetic days increases,the spatial coverage of soil moisture becomes larger,the average spatial coverage of soil moisture synthesized in 5 and 15 days can reach 88%and97%,respectively,in addition,compared with SMAP soil moisture products,the RMSE is0.130 m~3/m~3 and 0.118 m~3/m~3,respectively.