FROST in the European Alps – Implementing a data assimilation framework for calibration of surface mass balance models.

The demise of glaciers in the European Alps appears inevitable. The remaining question is the rate of ice loss and the timing of complete disappearance. This is particularly important for local water resource management and hazard assessment. Answering this requires glacier models that can robustly project glacier evolution over the coming decades. With the Instructed Glacier Model (IGM), an important tool for such projections has been made available, enabling computationally efficient modeling of glaciers as three-dimensional objects.

The remaining task is to accurately calibrate the glacier model. Fortunately, major advances in remote sensing provide an unprecedented amount of satellite observations that can be used for calibration on regional to global scale, even in remote areas without in-situ observations. The Framework for assimilating Remote-sensing Observations for Surface mass balance Tuning (FROST) utilizes a global elevation change product to infer equilibrium line altitudes and surface mass balance gradients. We tested the framework on 409 individual glaciers in the European Alps that were larger than 1 km² at the beginning of the century and evaluated the results against in situ measurements and end-of-summer snow line altitudes. The results show that the calibrated equilibrium line altitudes agree well with the snow line altitudes, while the surface mass balance gradients differ from glaciological measurements. This discrepancy is partially explained by the presence of small glaciers, which challenge gradient-based surface mass balance models, and by uncertainties of observation that lead to inaccurately modelled glacier flow. The first regional application of FROST shows promising results, provides meaningful insights and reveals challenges  for glacier projections in central Europe.