Surface reflectivity over Hudson Bay retrieved from TDS-1 mission data

Abstract

In times of a changing climate and the resulting uncertain consequences for nature and society, a special interest is focused on the large-scale recording of sea ice. Among the existing remote sensing methods, reflected Global Navigation Satellite System (GNSS) signals could play an important role in fulfilling this task. Within this project the sensitivity of GNSS reflection data to sea ice properties is evaluated. Estimates of sea ice reflectivity are derived from the ratio of reflected to direct signal power. It is expected that reflectivity of GNSS signals over smooth sea ice is decreasing with increasing sea ice thickness. The surface's reflectivity depends on the sea ice permittivity, i.e. its dielectric property, its roughness and the signal penetration into the ice body. The signals studied here were recorded in the years 2015 and 2016 by the TechDemoSat-1 (TDS-1) satellite. The TDS-1 payload includes a down-looking left-handed circular polarized antenna with high gain peak to acquire the Earth reflected signal. Another hemispherical up-looking right-handed circular polarized antenna was used to receive direct signal. The data is provided by the manufacturer SSTL and the reflection events were further pre-processed by IEEC to derive georeferenced power values. The project focuses on the signals' link budget to apply necessary corrections. The influence of the attitude uncertainty on gain calculation of the nadir antenna was examined and high nadir angles were filtered. Corrections of antenna gain and Free-Space Path Loss (FSPL) have been applied. The reflectivity was calculated from the corrected power using the data of the upand down-looking links. The differences in FSPL requires a correction of the reflected signal up to +6 dB with respect to the direct signal power level, with the loss increasing in magnitude towards higher incidence angles. The antenna gain correction has to account for the difference between a >13 dB peak value of the high-gain nadir antenna and a 4 dB peak value of the hemispherical zenith antenna. The retrieved reflectivity values are compared to model predictions based on Fresnel coefficients. The relation of reflectivity to sea ice thickness is investigated using a sea ice thickness product of the Soil Moisture and Ocean Salinity (SMOS) satellite of the European Space Agency (ESA). First insights into observations over Hudson Bay indicate that retrieved reflectivity decreases as sea ice increases in thickness. These preliminary results show that the developed approach is promising. Further investigations are needed to account for the dependence on surface roughness. Many studies show the potential of spaceborne GNSS-Reflectometry (GNSS-R) to complement existing remote sensing systems cost-effectively with global coverage.