Predicting slope instability driven by rapid depth change along the Canadian Beaufort Margin
Abstract
The Canadian Beaufort Margin (CBM) is a well-studied Arctic continental margin that has been seen an increase in geohazard research interest with numerous multidisciplinary research expeditions in the last decade. Repeat autonomous unmanned vessel and hull mounted multibeam surveys have detected rapid depth change (> 1 m/year) geomorphology. Such geomorphologic features including deepening of thermokarst depressions associated with degradation of subsea permafrost. Shoaling pingo like features widespread on the continental shelf have also been described. Despite the wealth of knowledge gained from small (< 30 km2) repeat high-resolution (4 m lateral resolution, cm-scale vertical resolution) surveys targeting rapid depth change geomorphological features, their activity and significance over the larger CBM is largely unknown.
Here, recently-developed geospatial machine learning methods are used to predict decadal-timescale depth changes across a large region of the CBM spanning the mid-shelf to the upper continental slope. We use observed depth changes, documented in the last decade, to train and validate machine learning models. Predictors used to build statistical models include bathymetry, slope, sedimentation rate, lithology, and Arctic-specific variables such as depth to subsea permafrost, distance from documented pingo-like feature, and annual iceberg scour frequency. Initial model results will be validated using new bathymetry acquired during the 2022 R/V Araon field season, and the presented models are updated using these new observed data. Predicting the location and magnitude of depth change hotspots on the CBM is essential for holistic geohazard risk assessment.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFMOS25D0954O