Analytical Assessment of GNSS-R Delay Doppler Map Sensitivity to Land Surface Variables Using a Physics-Based Model

Remote sensing using GNSS-reflectometry (GNSS-R) delay-Doppler maps (delay-Doppler maps) over land is an emerging field. Understanding the sensitivity of delay-Doppler maps to geophysical variables of interest, such as soil moisture and vegetation, is needed to understand the performance limitation of their retrievals. This work presents an analytical sensitivity analysis of GNSS-R delay-Doppler maps to land surface variables using the IGOT model and the Cyclone GNSS (CYGNSS) data over three areas with various topographical terrains. The IGOT model is an electromagnetic DDM model for land applications that is valid for topographical terrains with bare-to-intermediate vegetation cover. It divides the surface roughness into three scales. The large scale is governed by a digital elevation model (a DEM), whereas the other two are stochastic. The vegetation effect is modeled as an attenuation layer. Sensitivity to soil moisture, multiple scales of roughness, and vegetation are among the studied land surface variables. Furthermore, numerical results are obtained using realistic scenarios derived from CYGNSS observations. This work finds that the DDM peak reflectivity sensitivity to the large-scale surface roughness parameter and incidence angles varies depending on the geometry of the transmitter, receiver, and specular point. Moreover, the DDM peak reflectivity is inversely proportional to vegetation (due to attenuation effects) and small-scale surface roughness, which is consistent with the literature. Additionally, CYGNSS data over the study are analyzed to study the effect of incidence and azimuth angles. The analysis confirms that the behavior of peak reflectivity varies depending on both angles.