Impact of Sea-Ice Thickness and Permittivity on Polarimetric GNSS-Reflectometry Data Acquired During the MOSAiC Expedition

Abstract

Global navigation satellite system reflectometry (GNSS-R) has long been explored for retrieving sea ice properties, but in situ validation in the central Arctic during the freezing season is rare, limiting its application. The primary objective of this study is to advance the current understanding of multipolarization GNSS-R remote sensing for sea ice application. This article presents observations from the full-polarization GNSS-R (FpolGNSSR) prototype during the multidisciplinary drifting observatory for the study of Arctic climate (MOSAiC) expedition. The FpolGNSSR, with four polarization channels and high antenna gain (11.3 dB), aims to assess the impact of sea-ice thickness and permittivity on GNSS-R data, with observations from October 2019 to January 2020, the onset period of ice growth. First, the reflectivity is simulated by a four-layer model, and the sensitivity of multipolarization GNSS-R to sea ice is qualitatively analyzed. Subsequently, a simplified model reveals a linear relationship between reflectivity and ice thickness, with regression showing a correlation of 0.74 ( P<0.01 ). The retrieval root mean square error (RMSE) of sea-ice thickness retrieval is 0.13 m for first-year ice (0.3-1.0 m thick). Additionally, the imaginary component of sea ice permittivity is estimated, between 0.018 and 0.039. This study provides valuable insights for the GNSS-R community, which could be summarized as: 1) the "H-polarization advantage," which is first proposed in sea ice GNSS-R; 2) the validity of a simplified model, reinforcing the feasibility of satellite-based sea-ice thickness estimations using left-hand-circular, V-, and H-polarizations; and 3) a lower apparent permittivity of GNSS-R than previously reported, corresponding to a deeper penetration depth.