Mapping Glacier Sliding with Machine Learning

Glacial ice rapidly responds to changing climatic conditions, making its processes strongly intermittent in time and heterogeneous in space. In particular, the physical processes controlling glacial basal motion are poorly understood and remain challenging to observe directly.
This lecture presents a detailed study of targeted processes within the highly dynamic cryospheric environment. Using a unique seismic dataset from Glacier d’Argentière (French Alps), a physics-informed feature space, and a gradient boosting decision tree model, glacial sliding can be predicted with high spatio-temporal accuracy. We posit that features of the seismic noise provide direct access to the dominant parameters that drive displacement on the highly variable and unsteady surface of the glacier, making this a new approach to studying glacial basal sliding and uncovering its full complexity.

Further information