Mapping and identification of ice-sheet and glacier features using optical and SAR data in parts of central Dronning Maud Land (cDML), East Antarctica

Abstract

High-resolution satellite data is used to extract ice-sheet features in Princess-Astrid Coast of Queen Maud Land, in Droning Maud Land between 70° and 72°S and 5°-20°E, East Antarctica. Their mapping and monitoring is cumbersome and time-consuming using traditional survey methods. Therefore, to achieve desirable results, RS based methods were employed. The main ice-sheet and glacial features mapped are BlueIce-Area (BIA), melt water channels, glacier radar zones, crevasse, nunataks, moraine, supraglacial lakes, and wind scoop. Support-vector-machine (SVM) and supervised classification algorithm were applied over Landsat-8, RISAT-1 and Sentinel-1 data to classify these features. Surface patterns of snow and clouds were found in the continental ice-sheet, which becomes an obstacle to clearly extract sub-surface icesheet features in optical data. Therefore, to improve mapping of snow and ice features, high-resolution microwave data was also used. By comparing optical and microwave data, it was found that more sub-surface features could be mapped using microwave data. Further, RS was used to map the temporal change in the glacial features using two-year images for the austral summer of 2016 and 2017. It was found that diurnal as-well-as seasonal changes in surface ice features occur due to changes in local weather and can be mapped using temporal RS data.