Numerical study of co-polarized InSAR phase bias in remote sensing of layered media

Abstract

We numerically explore, for a three-layered dielectric medium, Interferometric Synthetic Aperture Radar (InSAR) coherence phase bias arising from co-polarized interferometric observations of electromagnetic (EM) interrogation of, and scattering from, penetrable subsurface media which can be approximated (at least locally, at the SAR pixel level) as planar-layered. A recently-developed incoherent scattering model now allows prediction of InSAR phase bias arising from the radar wave undergoing an (if neglecting radar time-gating) unending succession of subsurface specular reflections (“multi-bounce”), which is crucial for more comprehensively understanding interferometric observations (both terrestrial and extraterrestrial) of many low-loss layered structures. Our paper's results are as follows. First, for increasing subsurface wave attenuation the phase bias approaches zero (backscattering top interface) or the thickness of the subsurface slab (backscatter-free top interface). Second, increasing dielectric contrast between the central and outer two layers elevates (reduces) phase bias for a top interface weakly (strongly) backscattering power relative to the bottom interface. We conclude that subsurface scatter-enhanced phase bias should become significant primarily for geological structures characterized by a weakly-backscattering (i.e., very smooth) top interface and low-attenuating subsurface, which are attributes that may reasonably be used to describe the EM scattering properties of many manifestations of ice, snow, dry soil, and hyper-arid sand or regolith-mantled bedrock structures.