Distributed surface mass balance of the avalanche-fed Argentière glacier, French Alps

Avalanches are important contributors to the mass balance of glaciers located in mountain ranges with steep topographies. They result in localized mass inputs that are particularly difficult to quantify, due to the difficulty to access these avalanche cones in the field, and the need to account for ice dynamics when analyzing the elevation change signals from digital elevation models. Here, we aim to quantify the avalanche contribution to Argentière Glacier (Mont Blanc massif, French Alps) by inverting its distributed surface mass balance from remote sensing products and modeled ice thicknesses. Ultimately, we run the full-Stokes model Elmer-Ice with and without the additional contribution from avalanches to evaluate the importance of accounting for this process for future simulations of glacier evolution.

We used Pléiades satellite stereo acquisitions, captured at a high temporal resolution (more than 2 acquisitions per year), to generate detailed maps of elevation change and velocity spanning the years 2012 to 2021. We derived the distributed ice thickness of the glacier using three models of varying complexity, constrained by a dense array of ground penetrating radar measurements. To account for the uncertainty in ice thicknesses, we perturbed the modelled thicknesses using sequential gaussian simulations. We then combined ice thickness and velocity to derive the distributed flux divergence and surface mass balance at 20 m resolution across the whole glacier, carefully accounting for the uncertainties following a Monte Carlo approach. We evaluated our results against long-term mass balance measurements from stakes conducted as part of the French glacier monitoring service GLACIOCLIM, and in situ measurements of submergence on one of the main avalanche deposits.

There is a good agreement between our surface mass balance estimates and the stake observations (RMSE < 1.5 m.w.eq) for all ice thickness scenarios, even though ice thickness represents the most important source of uncertainty. Thus, the comparison of our distributed surface mass balance estimate with the mass balance gradient derived from the stake measurements allows us to 1) highlight the ability and potential of such an approach to provide robust estimates of distributed surface mass balance and 2) estimate the contribution of avalanches for Argentière Glacier with a relatively high accuracy. Notably, preliminary results show that the mass balance in avalanche-fed areas of the accumulation zone is approximately 2-10 times larger than in other areas at the same elevation.