Statistical Analysis of Land-Based GNSS-IR/R Over Bare and Vegetation Surfaces

This study provides the statistical characteristics of the power of the reflected global navigation satellite system (GNSS) signals and establishes associations between the probability density function (PDF) characteristics and different land cover situations. The feasibility of utilizing the PDF of signal power for determining the presence of vegetation cover and inverting associated parameters has been demonstrated. The simulation demonstrates that the signal-to-noise ratio (SNR) of land surface-reflected signals from various reflectors follows a Weibull distribution for the dual-antenna model, and an F-distribution for the single-antenna model. The moment-generating function (MGF) is used for the calculation of 1st–4th-order moments to study the characteristics of the PDF. Furthermore, the distribution parameters and moments are influenced by both the land-cover type and the reflector’s physical parameters. Experiments were conducted on mud flats and farmland for four months to validate the simulation model utilized in this study. Moreover, the simulation and experimental results demonstrate a mathematical correlation between the moments of the PDF and land surface parameters, such as soil moisture content (SMC), soil roughness, and vegetation density.