Sensitivity Analysis of D-InSAR Soil Moisture Models: AgriROSE-L Campaign

Abstract

Soil moisture models that use repeat-pass differential Synthetic Aperture Radar interferometry (D-InSAR) enable high-resolution estimates of soil volumetric water content. Current and upcoming C-band and L-band missions will enhance global mapping capabilities significantly. These models typically assume the absence of vegetation and deformation. Simple models consider a flat soil surface and an homogeneous soil medium, whereas more advanced models include effects of surface roughness and dielectric profile changes over time. However, it is still uncertain whether adding complexity to the models improves accuracy over bare and vegetated surfaces. This study aims to conduct a sensitivity analysis of recent D-InSAR soil moisture Models at L-Band and VV-polarization. Two models are evaluated with high-resolution airborne data from the German Aerospace Center (DLR) F-SAR system. Special attention is given to model parameterization. The performance of the forward models is evaluated using three reference fields: Potato, Wheat, and Maize.