Estimating volumetric surface moisture content for cropped soils using a soil wetness index based on surface temperature and NDVI

Surface soil wetness determines moisture availability that controls the response and feedback mechanisms between land surface and atmospheric processes. A study was carried out to estimate volumetric surface soil moisture content (θ v ) in cropped areas at field (<10² m) to landscape (<=10³ m) scales. Triangular scatters from land surface temperature (LST) and normalized difference vegetation index (NDVI) space were utilized to obtain a soil wetness index (SWI), from which θ v was derived, with the combination of dry and wet edges using data from ASTER (Advanced Space borne Thermal Emission and Reflection Radiometer) for field scale and MODIS (MODerate resolution Imaging Spectroradiometer) AQUA for landscape scale studies. The root mean square error (RMSE) of field scale θ v estimates was higher (0.039m³ m⁻³) than that of the landscape scale (0.033m³ m⁻³). The narrow swath (~60km) of finer resolution sensors (e.g. ASTER) often fails to capture the surface heterogeneity required in the triangle method for deriving SWI and could be one of the main reasons leading to relatively high error in θ v estimates. At both the scales, the lowest error of θ v estimates was found to be restricted within the NDVI range of 0.35-0.65. A geostatistical technique was applied to assess the influence of sub-pixel heterogeneity as an additional source of error for cross-scale comparison of θ v estimates obtained from LST-NDVI scatters. The overall errors of θ v estimates from LST-NDVI space were comparable with other globally available test results. The comparison of landscape scale θ v from MODIS AQUA with large-area global estimates from a passive microwave sensor (e.g. AMSR-E) with longer microwave frequency (e.g. C-band) yielded 75% correlation and 0.027m³ m⁻³ root mean square deviation (RMSD) for fractional vegetation cover less than 0.5. The study recommends the synergistic use of shorter microwave frequency (e.g. L-band) and optical-thermal infrared bands as the best way of satellite based passive soil moisture sensing for vegetated surfaces.