Theoretical models for microwave snow response and applications to remote sensing

Two theoretical models were developed to characterize snowpacks: (1) a random medium with a variance, a horizontal correlation length, and a vertical correlation length; and (2) a homogeneous dielectric containing discrete scatters. Theoretical results were matched with experimental data collected from snow ice fields. For snow ice fields the horizontal correlation length is no less than the vertical correlation length signifying a more laminar structure. The correspondence between the continuum random medium and the discrete spherical scatter model can be verified when the vertical correlation length is equal to the horizontal correlation length. The vertically polarized backscattering cross section sigma (sub vv) is always greater than the horizontally polarized backscattering cross section sigma (sub hh) for half space scattering media and may become smaller for a two layer model. In order to account for diurnal change exhibited by snow fields in both the active and passive remote sensing cases, a three layer model with a thin top layer caused by solar illumination must be used.