Comparison Between Surface Melt Estimation From Sentinel-1 Synthetic Aperture Radar and a Regional Climate Model. Case Study Over the Roi Baudouin Ice Shelf, East Antarctica

Abstract

Global warming, melting ice sheet, sea levels rise, natural disasters - these are words and sentences that we have heard a lot in recent years. In a changing world, studying the poles and their conservation is very important. However, it is often hard to obtain high-resolution data on this vast and distant continent. Thanks to the Copernicus space program providing free and open-access to high-quality data, this master thesis aims to show the complementarity between Sentinel-1 images and Modèle Atmosphérique régional (MAR) data over Antarctica. This study is conducted over Roi Baudouin Ice Shelf. The complementarity between the two datasets is established by a quantitative, temporal and spatial comparison of the amplitude information of the radar signal and several variables modelled by MAR. To archive it, mosaics of Sentinel images had to be created to cover a larger area and to allow a comparison of same-day interpolated data. Obtained results are quite promising. Comparisons show very strong spatial correlations between MAR variables representing melt and the backscatter coefficient recorded by the satellite. Although temporal and quantitative analyses also give great results, there are still questions in some areas of the platform that do not behave in the same way as the rest. This thesis, therefore, opens up many possibilities in research by showing that most of the time, the results match but in some cases, it is more subtle. Certain questions thus remain open, arousing curiosity but also prospects of complimentary use of radar satellites and climate models.