Abstract
Brittle failure by crevasses and rift formation is the most directly observable expression of stress state in glacier ice. Both satellite imagery and satellite altimetry are effective tools for characterizing them. However, the ice shelf surface exhibits nonfracture-related features that can appear similar to crevasses and rifts in images or elevation profiles, making it challenging to distinguish them using a single observational perspective. Here, we combine satellite altimetry and satellite imagery to map these fracture features and provide their 3-D structure. During this process, we develop a novel algorithm for detecting and characterizing depressions on the ice surface using Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) L2A Global Geolocated Photon Data, Version 6 dataset (ATL03). The strengths of the algorithm include estimation of surface elevation in 10 m along-track segments, broad applicability to different crevasse types, and robustness against surface irregularities. We apply this workflow to the Thwaites Eastern Ice Shelf (TEIS) as a case study and estimate the general depth and width of various crevasse types. Integrating altimetry with imagery, we find that V-shaped depressions, occurring at the early stage of fracturing, could serve as an indicator of the development of large-scale surface crevasses. Within the shear zone upstream of the pinning points, apart from shear failure, the ice surface also deforms due to strong compression, creating peaks a few meters high. This study highlights that including 3-D morphology in the detection and analysis of ice shelf crevasses and rifts improves the ability to distinguish between crevasse types and provides insights into ice shelf deformation.