Snow Density and Ground Permittivity Retrieved From L-Band Radiometry: A Retrieval Sensitivity Analysis

Abstract

Aboveground dry snow influences the L-band ground–snow system emissivity as the result of impedance matching and refraction effects. Against this background, a retrieval scheme was proposed to estimate dry-snow density and ground permittivity from passive L-band measurements. In this study, the sensitivity of the recently proposed retrieval scheme with respect to surface roughness and snowpack density profile heterogeneities is investigated using synthetic brightness temperatures $T_{\rm B}^{p}$ . While the original retrieval algorithm proposed to use $T_{\rm B}^{p}(\theta)$ at observation angles θ and both polarizations p = “H” and p = “V”, the present analysis involves three polarization retrieval modes: “HV”, “H”, and “V” to identify the most robust one. The analyses based on synthetic $T_{\rm B}^{p}(\theta)$ suggest the use of exclusively vertical polarization (“V” retrieval mode) in case of low “instrumental noise” of $\sigma T_{\rm B}\,{ , as it yields most accurate results in the presence of heterogeneities in profiles and imperfect parametrization of ground surface roughness. The applied retrieval scheme still yields accurate results even in the presence of “instrumental noise” ( $\sigma T_{\rm B}\,\geq \,\text{0.5}$ K) in retrieval mode “HV”. Furthermore, it is shown that ground-permittivity retrievals are less affected than snow density retrievals by ground roughness variations or heterogeneities in snow density profiles. Altogether, our sensitivity investigations indicate the robustness of the retrieval scheme applied thereto with respect to snow density profile heterogeneities, which is expedient for its use with spaceborne radiometry data from, for example, “Soil Moisture and Ocean Salinity” or “Soil Moisture Active Passive” satellites.