Nonlinear sensitivity of glacier mass balance to future climate change unveiled by deep learning

Jordi Bolibar,Harry Zekollari,Antoine Rabatel,Isabelle Gouttevin,Clovis Galiez

doi:10.1038/s41467-022-28033-0

Abstract

Glaciers and ice caps are experiencing strong mass losses worldwide, challenging water availability, hydropower generation, and ecosystems. Here, we perform the first-ever glacier evolution projections based on deep learning by modelling the 21st century glacier evolution in the French Alps. By the end of the century, we predict a glacier volume loss between 75 and 88%. Deep learning captures a nonlinear response of glaciers to air temperature and precipitation, improving the representation of extreme mass balance rates compared to linear statistical and temperature-index models. Our results confirm an over-sensitivity of temperature-index models, often used by large-scale studies, to future warming. We argue that such models can be suitable for steep mountain glaciers. However, glacier projections under low-emission scenarios and the behaviour of flatter glaciers and ice caps are likely to be biased by mass balance models with linear sensitivities, introducing long-term biases in sea-level rise and water resources projections.

Highlights

Glaciers and ice caps are experiencing strong mass losses worldwide, challenging water availability, hydropower generation, and ecosystems
mass balance (MB) rates only begin to approach equilibrium towards the end of the century under Representative Concentration Pathway (RCP) 2.6, for which glaciers could potentially stabilize with the climate in the first decades of the 22nd century depending on their response time (Fig. S1a)
This oversensitivity directly results from the fact that temperatureindex models rely on linear relationships between positive degree-days (PDDs) and melt and that these models are calibrated with past MB and climate data

Summary

Introduction

Glaciers and ice caps are experiencing strong mass losses worldwide, challenging water availability, hydropower generation, and ecosystems. Various global glacier evolution models have been used to provide estimates on the future sea-level contribution from glaciers[7,8] All these glacier models, independently from their approach, need to resolve the two main processes that determine glacier evolution: (1) glacier mass balance, as the difference between the mass gained via accumulation (e.g. snowfall, avalanches and refreezing) and the mass lost via different processes of ablation (e.g. melt and sublimation of ice, firn and snow; or calving)[9]; and (2) ice flow dynamics, characterized by the downward movement of ice due to the effects of gravity in the form of deformation of ice and basal sliding. The use of ANNs remains largely unexplored in glaciology for regression problems, with only a few studies using shallow ANNs for predicting the ice thickness[14] or mass balance[13] of a single glacier

Methods

Results

Conclusion