Remote sensing retrieval of surface soil moisture in Mu Us Sandy Land based on optical trapezoid model

Abstract Maowuosu sandland ecosystem is fragile, and moisture is a key factor restricting the growth and development of sandland plants and ecological construction in semi-arid areas. Therefore, this study constructed the eigenspace based on the data of the 4th period of 2016, fitted the dry and wet edge equations of the eigenspace, calculated the OPTRAM, and utilized the measured in situ soil moisture data from various depths during the corresponding period. The OPTRAM was regressed to construct a model for estimating soil moisture. This model was then subjected to a series of validation procedures to ensure its accuracy. The coefficients for the shortwave infrarouges conversion reflection coefficient (STR) - standardized differential vegetation index (NDVI) feature space dry and wet edge fits increased with the increase of the month, where the dry and the R 2 for the wet side in September was 0.88 and 0.87, respectively. OPTRAM and the measured soil moisture content data in the surface layer were all correlated to some extent, with correlation coefficients above 0.45. The OPTRAM model is more accurate than other models in inverting in situ moisture content at the 0∼10 cm depths within this area.