Three-dimensional numerical modelling of an unstable rock slope in Alaska (US)

• Main Author: Noijons, Benthe Charlie
• Other Authors: Donati, Davide, Borgatti, Lisa
• Format: Master Thesis
• Language: English
• Published: Alma Mater Studiorum - Università di Bologna 2023
• Subjects: slope stability,numerical modelling,glacier retreat,remote sensing; Ingegneria per l'ambiente e il territorio [LM-DM270]; Buttress; Donati; glacier; Alaska
• Online Access: https://amslaurea.unibo.it/id/eprint/29430/; https://amslaurea.unibo.it/id/eprint/29430/1/Noijons_Thesis_30August.pdf

This research project attempts to design a preliminary numerical model of the Fels slide, an unstable slope at the tongue of the Fels glacier in Alaska, United States. Following research on the slope failure mechanism and rupture surface by Donati et al. (2021), a three-dimensional model has been developed to investigate the factors controlling the slope failure using the Finite Element Method in the Rocscience software RS3. As glacier retreat and slope failure have been linked by many studies, this research also aims to assess whether the retreat of the Fels glacier might play a role in failure of the slope. Due to limited data availability in this remote region, remote sensing data was used for the morphology of the study area, as well as the rupture surface proposed by Donati et al., (2021). Material properties were based on field investigation by Newman (2013) and literature. Glacier retreat was estimated from historical imagery. The model resulted in a computed critical strength reduction factor of 0.99, meaning the model accurately represents a failing slope. Not all displacement trends found by remote sensing by Donati et al. (2021) were recognized in the displacement vectors and solid displacement figures of the model output. When adding glacier volumes as elastic material buttress, the SRF increases. This is in accordance with literature, but likely overestimated in this analysis. The preliminary slope model predicts the failure to a certain extent, but is not entirely accurate, particularly with respect to displacement magnitude and distribution. Reassessing the reconstruction of the basal surface, introducing a more complex representation of the unstable slope, including the fast-moving toe, and adding more parameters like groundwater and more specific material properties would make the model more accurate. Nevertheless, it is a step towards prediction of natural hazards from little data, valuable for risk assessment as well as climate change studies.