Uncertainties in Modelling Glacier Melt and Mass Balances: the Role of Air Temperature Extrapolation and Type of Melt Models

Abstract:

While glacier responses to climate are understood in general terms and in their main trends, model based projections are affected by the type of model used and uncertainties in the meteorological input data, among others. Recent works have attempted at improving glacio-hydrological models by including neglected processes and investigating uncertainties in their outputs.

In this work, we select two knowledge gaps in current modelling practices and illustrate their importance through modelling with a fully distributed mass balance model that includes some of the state of the art approaches for calculations of glacier ablation, accumulation and glacier geometry changes.

We use an advanced mass balance model applied to glaciers in the Andes of Chile, Swiss Alps and Nepalese Himalaya to investigate two issues that seem of importance for a sound assessment of glacier changes: 1) the use of physically-based models of glacier ablation (energy balance) versus more empirical models (enhanced temperature index approaches); 2) the importance of the correct extrapolation of air temperature forcing on glaciers and the large uncertainty in model outputs associated with it. The ablation models are calibrated with a large amount of data from in-situ campaigns, and distributed observations of air temperature used to calculate lapse rates and calibrate a thermodynamic model of temperature distribution.

We show that no final assessment can be made of what type of melt model is more appropriate or accurate for simulation of glacier ablation at the glacier scale, not even for relatively well studied glaciers. Both models perform in a similar manner at low elevations, but important differences are evident at high elevations, where lack of data prevents a final statement on which model better represent the actual ablation amounts.

Accurate characterization of air temperature is important for correct simulations of glacier mass balance and volume changes. Substantial differences are obtained if we use the common approach of constant in time LRs (even if properly calibrated) or more sophisticated approaches accounting for the different thermal regime off and on-glacier, as distinct thermal conditions exist on and off glacier associated with the presence of the glacier boundary layer where katabatic flow is important.