Evaluating the TanDEM-X Digital Elevation Model in Differential Interferometric topographic phase removal process for glacier velocity estimation

摘要

Glacier movement is an important parameter for understanding the climate change, glacier dynamics, ice thickness and mass balance studies. The movement can be measured using field based (Global Positioning System) or remote sensing (Interferometry or offset tracking) based technique. GPS gives spatially limited information and it is very difficult on rugged terrains. Remote sensing techniques give velocity measurements in inaccessible areas and also overcome the spatial limitation problem. Interferometric Synthetic Aperture Radar (InSAR) is a remote sensing technique for measuring the glacier velocity with centimeter accuracy. Topographic phase removal is one of the important steps in the 2-pass Differential InSAR (DInSAR) technique and also it is one of the major error sources in this process. An external Digital Elevation Model (DEM) is required for this step and using an optimizing DEM increases the accuracy level of glacier movement. The time period of SAR image pair is also considerable for the selecting of external DEM due to the changes in elevation (in the process of glacier mass loss/gain). The main objectives of this study are to estimate the Eidembreen glacier movement and evaluating the TanDEM-X 90m DEM with ArcticDEM in DInSAR topographic phase removal step. The ArcticDEM is selected due to the accuracy and resolution. The majority of ArcticDEM data was generated by using optical stereo high-resolution imagery. For this study, Eidembreen glacier is selected and it is located in the Svalbard. Sentinel 1 A/B sensor SAR images selected for the interferometric process with a perpendicular baseline of 12.89 m and the mean coherence value is 0.67. We used both TanDEM-X and ArcticDEM as external DEM to remove topographic phase information and compared the both generated line-of-sight (LOS) velocity maps along the central flow line of Eidembreen glacier. We observed the mean difference value as 0.35 mm/day and it represents approximately 12.7 cm/year. So, this value is almost negligible with the annual glacier velocity range values.