Soil moisture estimation using land surface temperature and soil temperature at 5 cm depth

Soil Temperature (ST) data, obtained from either field works or satellite imagery, has frequently been studied for Soil Moisture (SM) estimation. However, a combination of ST data at different depths and soil surface temperature, i.e., Surface Radiometric Temperature (SRT) or Land Surface Temperature (LST), has not yet been well investigated for accurate SM prediction. In this study, an empirical model was first developed to estimate SM at 5 cm Depth (SM5D) over areas with no or sparse vegetation cover using the field SRT and field ST data at 5 cm Depth (ST5D). A Root Mean Square Error (RMSE) and a correlation coefficient (r) of 0.037 m3 m−3 and 0.8 were obtained using this model, respectively. Then, the SRT was substituted by the LST obtained from Landsat thermal bands and ST5D was estimated using the ST data collected at the nearest weather station to the study area by developing a regression equation. The second model demonstrated an RMSE and r of 0.035 m3 m−3 and 0.71, respectively. Overall, it was concluded that the proposed models had high potential for SM estimation using the ST data at different depths collected in the field or acquired by optical satellites.