INTERCOMPARISON OF X- AND C-BANDS ACTIVE MICROWAVE SOIL MOISTURE RETRIEVALS OVER DRYLAND WHEAT FIELDS

KOMPSAT-5 is an X-band satellite that has a great potential for high-resolution soil moisture retrieval over agricultural regions. In the present paper, comparison between the soil moisture estimations from the widely used Sentinel-1 (C-band) and KOMPSAT-5 (X-band) is conducted for a dryland wheat field site located in Victoria State of Australia. This study uses a semi-empirical vegetation scattering model to invert radar signal and retrieve soil moisture based on KOMPSAT-5 and Sentinel-1 images and the retrieval model is calibrated by field measurements. In this model, normalized difference water index (NDWI) is applied for the parameterization of vegetation water content and the estimation of volume scattering from vegetation canopy. The results demonstrated that the accuracy of soil moisture retrieved from both satellite instruments was consistent with the field measurements, with the root mean square error (RMSE) reached 0.016 cm(3)/cm(3) at X-band and 0.015 cm(3)/cm(3) at C-band, and the correlation coefficients (R) were 0.726 and 0.517, respectively. In addition, the validating results showed that the X-band provided slightly consistent accurate soil moisture retrieval in comparison with the C-band, with the RMSE of 0.017 cm(3)/cm(3). Consequently, this study indicated the prospect and applicability of estimating soil moisture retrieved from both KOMPSAT-5 and Sentinel-1 remote sensing data and could also provide support and reference for other crop yield estimation effectively.