The Mountain Glacier forecast framework – MAGIC

We will focus on two main sources of uncertainty
(i) we will improve and develop data assimilation methods to better constrain the basal topography and model parameters for a better representation of the current state and dynamics of the flow
(ii) we will implement advanced surface mass balance models that show stable relationships on multi-decadal time scales

Finally, based on ensemble modelling, the model will allow for a better quantification of the propagation of the different sources of uncertainty.

Methodological developments for data assimilation are being developed and tested on synthetic applications.
Different surface mass balance models have been calibrated and validated for the Schiaparelli glacier in Patagonia.
We have already carried out the first 3D simulations on the scale of the Mont Blanc massif.

The model will be validated over the recent period with available satellite observations. This will enable the forecasts made for 2100 to be refined.

Software development:
- Développement du logiciel libre Elmer/Ice

Communications :
- Cook, S. and Gillet-Chaulet, F.: 2D sequential data assimilation in Elmer/Ice with Stokes, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-105, doi.org/10.5194/egusphere-egu21-105, 2021.

General glacier retreat around the globe has become iconic to illustrate the effects of global warming. Model results show that this demise will continue and significantly add to global sea- level rise. Ice loss can also induce shifts in regional water resources, putting some places at high risk of year-around freshwater availability. Better projections of glacier retreat are therefore essential in terms of coastal impact, water management and hydro-glaciological risks.
The overall aim of the MAGIC project is to provide a framework to improve past and future glacier volume projections. The process-based framework will be based on the open-source ice-flow model Elmer/Ice and will allow to reduce three potentially important sources of uncertainty. Anticipating the continuously growing body of information from satellite remote sensing, we will improve and develop robust data assimilation methods to infer the basal topography beneath the ice and calibrate ice-flow parameters for a better representation of the initial state and ice dynamics. The flow model will be coupled to an enhanced surface mass balance model that exhibits a stable melt relation over relevant multi-decadal timescales. Based on ensemble simulations, the framework will allow to better quantify the interaction and propagation of these uncertainties.
As a proof of concept, the framework will be applied to hindcast and forecast glacier evolutions in the French Alps and in Cordillera Darwin (Chile). These sites have been chosen because for both glacier sites ice-flow is considered important and because of the availability of in-situ and remotely sensed observations. The two sites will further serve as challenging test setups to scrutinise the applicability, performance and robustness of the modelling framework under different climatic conditions and dynamic settings.
The complementary expertise is covered by the two groups at Institut des Ge´osciences de l’Environnement (IGE) in Grenoble and at the Institute of Geography at the University of Erlangen-Nuremberg (FAU) to provide major developments to the Elmer/Ice model and run the proposed applications.