Evaluating ICESat-2 and support vector regression models for the identification of nearshore bathymetry in Nuuk and Kullorsuaq, Greenland

Estimates of nearshore bathymetry are vital to understanding coastal change processes in Greenland, which is experiencing uplift at rates of 5-23 mm/yr in coastal regions from glacial isostatic adjustment as the Greenland Ice Sheet continues to lose mass. However, current in-situ methods of shallow water depth collection, such as multibeam echo sounding (MBES), cannot efficiently reach the shallowest coastal waters where bathymetric change will be most evident and most impact those who live in coastal settlements. Impacts will include changes to harbor infrastructure, boating routes, habitat of benthic species, and natural hazards. Nearshore MBES collected in Nuuk and Aasiaat, Greenland in June-July 2021 provides a unique opportunity to evaluate the accuracy of local satellitederived bathymetry in Arctic waters. Accurate SDB in Greenland is made more complex by storms, fog, floating ice, and limited sunlight - especially during winter months. Using ICESat-2 altimetry and support vector regression models from Sentinel-2 imagery, we seek to better understand the shallow water (0-20m) bathymetry of Greenland and compare these methods to ground truthed MBES data and local observations from community members.