A minimal model for reconstructing interannual mass balance variability of glaciers in the European Alps

B Marzeion,M Hofer,T Mölg,A H Jarosch,G Kaser

doi:10.5194/tc-6-71-2012

Abstract

Abstract. We present a minimal model of the glacier surface mass balance. The model relies solely on monthly precipitation and air temperatures as forcing. We first train the model individually for 15 glaciers with existing mass balance measurements. Based on a cross validation, we present a thorough assessment of the model's performance outside of the training period. The cross validation indicates that our model is robust, and our model's performance compares favorably to that from a less parsimonious model based on seasonal sensitivity characteristics. Then, the model is extended for application on glaciers without existing mass balance measurements. We cross validated the model again by withholding the mass balance information from each of the 15 glaciers above during the model training, in order to measure its performance on glaciers not included in the model training. This cross validation indicates that the model retains considerable skill even when applied on glaciers without mass balance measurements. As an exemplary application, the model is then used to reconstruct time series of interannual mass balance variability, covering the past two hundred years, for all glaciers in the European Alps contained in the extended format of the world glacier inventory. Based on this reconstruction, we present a spatially detailed attribution of the glaciers' mass balance variability to temperature and precipitation variability.

Highlights

Glaciers are prominent features of the alpine landscape
We present a minimal model of the glacier surface mass balance
The cross validation indicates that our model is robust, and our model’s performance compares favorably to that from a less parsimonious model based on seasonal sensitivity characteristics

Summary

Introduction

Glaciers are prominent features of the alpine landscape. As they integrate their surface energy and mass fluxes over multi-annual to centennial timescales (e.g. Johannesson et al, 1989; Oerlemans, 2001), the fluctuations of the glaciers’ extent constitutes a naturally low-pass filtered signal of the atmospheric variability. Glaciers are prominent features of the alpine landscape As they integrate their surface energy and mass fluxes over multi-annual to centennial timescales Johannesson et al, 1989; Oerlemans, 2001), the fluctuations of the glaciers’ extent constitutes a naturally low-pass filtered signal of the atmospheric variability. Through this property, glaciers allow people to directly perceive slow changes of the climate system, that otherwise would be overwhelmed in human perception by short-term noise. Changes in glacier extent have been discussed long before climate variability and change received the attention they do today Without the ability to distinguish between different modes of change (e.g. between stochastically forced fluctuations and fluctuations caused by anthropogenic warming), the glacier fluctuations are meaningless to the observer interested in inferring atmospheric variability from them

Results

Discussion

Conclusion