Crevasse density, orientation and temporal variability at Narsap Sermia, Greenland

Abstract

Mass loss from iceberg calving at marine-terminating glaciers is one of the largest and most poorly constrained contributors to sea-level rise. However, our understanding of the processes controlling ice fracturing and crevasse evolution is incomplete. Here, we use Gabor filter banks to automatically map crevasse density and orientation through time on a ~150 km2 terminus region of Narsap Sermia, an outlet glacier of the southwest Greenland ice sheet. We find that Narsap Sermia is dominated by transverse (flow-perpendicular) crevasses near the ice front and longitudinal (flow-aligned) crevasses across its central region. Measured crevasse orientation varies on sub-annual timescales by more than 45 $^\circ$ in response to seasonal velocity changes, and also on multi-annual timescales in response to broader dynamic changes and glacier retreat. Our results show a gradual up-glacier propagation of the zone of flow-transverse crevassing coincident with frontal retreat and acceleration occurring in 2020/21, in addition to sub-annual crevasse changes primarily in transition zones between longitudinal to transverse crevasse orientation. This provides new insight into the dynamics of crevassing at large marine-terminating glaciers and a potential approach for the rapid identification of glacier dynamic change from a single pair of satellite images.

Increased crevassing across accelerating Greenland Ice Sheet margins

“…Third, the optical nature of the source data meant that we cannot extract snow-filled crevasses that may be possible to detect using other methods, such as SAR or GPR 67 . However, the large diameters of crevasses detected here are highly unlikely to fill with snow: in analysis of Sentinel-2 optical imagery with a similar effective resolution for crevasse detection, crevasse density was not observed to change over a seasonal cycle or in an indicative elevation-dependent way that suggested snowfill 28 . The month filtering, ablation zone masking, and median mosaicking we performed during the mosaicking process mean we consider it very unlikely that snowfill can explain any of the large-scale multitemporal change we observe in our study.…”

“…The ability to observe crevasses in 3-D provides a major advance over two-dimensional mapping from imagery alone 27,28 . We have observed significant increases in crevasse volume in pre-existing crevasse fields at low elevations (marine-terminating outlets).…”

“…Here, there exists UAV-derived 15cm-resolution map of crevasses (dated 2018-07-08) classified using object-based machine learning techniques 3 . We compare this against a Sentinel-2-derived map of crevasses using a Gabor filter approach 28 for the date 2018-07-02, and apply our current approach on an ArcticDEM strip dated 2018-06-24. Data is shown in supplementary figures 13 and 14.…”

“…In comparison to the Sentinel-2 approach, our method is sensitive to smaller crevasses, as well as less likely to misclassify the edges of snow/ice boundaries. These advantages are balanced by the much higher temporal resolution of the Sentinel-2 stack, which can detect sub-seasonal changes 28 . Using the UAV data as ground validation, we assess the limit of crevasse width detectable by our method to be approximately 10 m. This matches the previous assessment made using a more rudimentary ArcticDEM segmentation approach 3 .…”

“…Therefore, improved observations are required to develop and validate models of fracture formation and propagation 25 and parameterise their behaviour in models of ice sheet dynamics and hydrology 6,26 . Optical satellite observation methods have progressed from manual delineation 4 to computer vision 27,28 and machine learning 29,30 approaches. However, these are limited to assessing crevasse presence without critical information about crevasse depth, and attempts to map geometry have thus far been limited to profiles 21 .…”