Soil moisture estimation models using SIR-C SAR data: a case study in New Hampshire, USA

The technology of using spaceborne synthetic aperture radar (SAR) systems for soil moisture estimation has been refined over the last few years. The potential of microwave sensors to estimate soil moisture is well known, and its continuous monitoring on temporal and spatial bases has been realized recently. Several techniques have been developed for retrieving the surface parameters and soil moisture from the radar backscatter. In order to reduce the confounding effects of surface roughness on soil moisture inversion, the application of multifrequency SAR systems have shown promise. The shuttle imaging radar mission C (SIR-C) had an on board SAR system operating at L-, C-, and X-bands for high-resolution imaging of the Earthʹs surface. Data from SIR-C SAR have been investigated for soil moisture estimation and comparison with in situ data. The models used for soil moisture inversion, viz., (1) the linear regression, (2) the linear statistical inversion, and (3) the neural network models, are presented, and the results of soil moisture estimation using these models are compared. The resulting estimation of soil moisture using the above models is more accurate for the surface soil moisture than subsurface soil moisture estimation, as expected. In general, these models estimate soil moisture within a root mean squared (RMS) error of 3–5%.