Multi-annual Analysis of the Greenland Ice Sheet between 2010 and 2017 based on TanDEM-X DEM Data

Kurzfassung

The Greenland ice sheet represents the world’s largest ice mass outside Antarctica covering an area of approximately 1.7 million km2. During the last decades, observations reveal dramatic changes, which can be attributed to anthropogenic induced global warming with considerable implications for environment and human society. The melting of the entire Greenland ice sheet would raise global mean sea level by 7.36 m. In view of this, it is crucial to understand spatial and temporal glacial dynamics for determining the contributions to sea level rise and predicting responses to climate change. Interferometric SAR based Digital Elevation Models (DEMs) are powerful techniques to collect information on glaciers and ice sheets worldwide, particularly in polar regions located in high latitudes. To date, multi-date DEMs at high spatial resolution is lacking for entire Greenland. Existing DEM products are either outdated or show low spatial resolution, often mono-temporal. The German TanDEM-X satellite mission offers great potential for monitoring of the Earth’s polar regions at unprecedented spatial resolution due to the global availability of consistent and precise interferometric data. However, so far TanDEM-X data have been mainly used on local scales. The first goal of this study is to provide a novel change detection analysis for the entire Greenland ice sheet at 12 m spatial resolution using TanDEM-X interferometric acquisitions between 2010 and 2017. One major issue with X-band SAR data is related to penetration of the signal into the snow and ice surface, which influences significantly the height accuracies of the delineated DEM. Therefore, we defined a second goal, which aims to investigate the X-band penetration over different snow and ice characteristics to correct possible elevation bias. As a reference, we used IceBridge ATM L2 Elevation data from spring 2012 to analyze the penetration depth over different snow zones, which are determined by specific physical parameters (grain size, snow density, stratigraphy, surface roughness, and water content), which influence the SAR backscatter. Our results reveal penetration depths up to ten meters and a high correlation between penetration depth and backscatter intensity as well as interferometric coherence and height of ambiguities. This information can help to improve vertical accuracy of TanDEM-X data over the Greenland ice sheet to obtain more reliable elevation change from different time spans.