Wetland hydrological conditions play a key role in the structure and function of the wetland ecosystem. In this study, we use three drought monitoring methods, Temperature Vegetation Dryness Index (TVDI), Perpendicular Drought Index (PDI), and Normalized PDI (NPDI) as well as soil moisture observation data in the field to study marsh wetlands of an arid area in the middle and lower reaches of the Shule River using a remote sensing technology based on the Landsat8 Operational Land Imager (OLI). We obtained the following results in this study:1) TVDI, PDI, and NPDI were significantly negatively correlated with soil moisture in different soil layers (P < 0.01), the goodness of fit among the three drought indexes was in the following order of magnitude NPDI > TVDI > PDI with soil moisture in different soil layers, and the fitting results using the NPDI index were the best with topsoil moisture; 2) there were numerous significant differences in soil moisture among different soil layers of the same marsh type (P < 0.05), and an increasing trend in the soil moisture of different marsh types was observed with increasing soil depth; 3) the soil moisture was significantly different between different marsh types of the same soil layer (P < 0.05), and the comparison of soil moisture in the different soil layers showed the following decreasing order of magnitude:marsh meadow > herbaceous marsh > seasonal saline marsh > inland salt marsh.