Analyzing the Influence of Wet Biomass Changes in Polarimetric Differential SAR Interferometry at L-Band

Abstract

The displacement estimated with differential SAR interferometry (DInSAR) might not be unique when more than one polarization channel is available. For the case of agricultural areas, these ambiguities have been mainly related to complex vegetation dynamics, i.e., vegetation growth. This study intends to explore the potential of a synergistic use of DInSAR with SAR Polarimetry (PolDInSAR) in tracking changes within agricultural vegetation covers. The connection between the PolDInSAR observables (i.e., herein, the DInSAR phases at various polarization channels and/or their linear combinations) with wet biomass and soil water content changes is empirically investigated with linear regression techniques. This is done in the frame of an L-band airborne DInSAR dataset. The impact of vegetation vigor differs depending on the type of crop analyzed. For those crops exhibiting a birefringent electromagnetic propagation (i.e., barley, wheat, and rapeseed), the influence of wet biomass is particularly pronounced in the VV DInSAR phase but also in the HH–VV phase difference. Contrarily to the former, the latter shows also a scarce sensitivity to changes in soil water content. Therefore, this PolDInSAR observable is used to generate biomass maps of the analyzed test site. The predicted biomass variations are in good agreement with the collected in situ measurements, i.e., the coefficient of determination varies between 0.8 and 0.9.