GPS interferometric reflectometry: Forward and inverse modeling of GPS signal strength data applied to remote sensing of snow

GPS interferometric reflectometry (GPS-IR) is a method that exploits multipath for ground-based remote sensing. It has been demonstrated to be capable of retrieving a number of environmental parameters of importance to the study of the water and carbon cycle, including surface soil moisture, snow depth, and vegetation. GPS-IR could be used to provide validation of spaceborne sensors. It also provides an intermediate-scale footprint that augments other measurement systems. Initial results for GPS-IR adopted a mostly empirical data processing approach: changes in observation frequencies and amplitudes have been correlated with changes in environmental parameters. In parallel, a theoretical model based on the physics of electromagnetic scattering was proposed, but has not been used directly in aiding the environmental retrievals. We will present results seeking to bridge these two efforts. We have utilized the physically-based model as a forward step, in conjunction with a statistical inverse model; the former is based on geometrical optics, while the latter is based on non-linear least squares. We will examine time series of inverted parameters as well as post-fit residuals to illustrate and discuss model fitting issues in GPS-IR, such as biases, uncertainty quantification, and quality control. We will compare our model-based results to in situ observations of snow depth and snow water equivalent.