PHASE DIFFERENCES IN LONGER WAVELENGTH POLARIMETRIC SAR - SURFACE OR SUBSURFACE EFFECTS?

Large phase differences between HH and VV polarizations have been observed in vegetation-free areas where doublebounce scattering is unlikely [1]-[2]. Scattering from such terrain can be due to surface or subsurface phenomena, particularly when longer wavelengths are used. In this paper we discuss the causes of such significant phase differences and the conditions under which they can be attributed to either surface or subsurface scattering. Smooth surface and subsurface layers are modeled as lossy dielectric materials. The Small Perturbation Model is used to estimate the relative amplitude and phase of polarimetric radar backscatter returns from each layer, and interactions at the airsurface interface are addressed using the complex amplitude Fresnel transmission coefficients for a lossy dielectric. The model is shown to be in good agreement with NASA/JPL AIRSAR polarimetric SAR data obtained over Death Valley. Anomalous scattering from salt pans on the floor of Death Valley is examined. NASA/JPL AIRSAR polarimetric SAR and SIR-C data obtained over Death Valley exhibit large HH-VV phase differences that cannot be explained by invoking simple surface scattering. These effects are only seen as the incidence angle approaches the Brewster angle for the surface. Field measurements of the dielectric constant indicate that the real part may be quite low, while the imaginary part (and therefore the loss tangent) is quite high, suggesting penetration is not the correct explanation. Anomalous HH-VV phase differences and HH/VV amplitude ratios are consistent with surface scattering from a salt pan, consisting of salt mixed with a highly saline brine solution to form the surface dielectric material. The paper will also discuss the extension of these results to other areas where the soil potentially has a high salt content, and to areas where subsurface scattering from a high salt content layer may be present.