Accurate identification of soil moisture in Mu Us Sandy Land based on multi-source remote sensing data

Apart from being essential to the development of grasslands and crops, soil moisture has a significant impact on the water cycle and the global climate. Drawing from multispectral Landsat 8 OLI photos, Added to this study, soil moisture was simultaneously measured in the field based on topographic parameters and surface biophysical features. Soil moisture inversion models of multivariable linear regression (MLR) and random forest (RF) were constructed by empirical model method. And compared with inverse distance weighting (IDW). The findings demonstrated the dependence of surface soil moisture on surface biophysical characteristics was high, with the highest inversion accuracy of random forest, in which the R² of RF model was above 0.8 in both months, which was at least 8% higher than that of MLR model, and at least 11% higher than that of IDW model, and the corresponding RMSE was the smallest in both months. In various months, the main variables influencing the soil moisture content were elevation and surface temperature. The investigation's area had low overall soil moisture content, with most of it having less than 20. High, steep slopes in the north and southeast of the area had higher soil moisture content, whereas the region's flat areas in the north-central region had lower soil moisture content.