Inferring Debris Properties on Debris-Covered Glaciers: Implications for Glacier Modelling

Abstract

Debris, ranging from thin surface dust to medial moraines and thick, continuous layers in ablation zones, partially covers glaciers all around the world. By modifying energy transfer from the atmosphere to the ice, the supraglacial debris layer fundamentally controls sub-debris melt rates. Debris physical properties such as surface roughness (z0) and thermal conductivity (k) have only been derived from local measurements at a few sites, and modelling studies of debris-covered glaciers have often relied on literature values. The correct representation of these properties in energy-balance models is crucial for understanding the climate-glacier dynamics and how debris-covered glaciers will behave in the future. There are several established methods to derive these properties from field measurements, yet relatively few studies undertake to measure properties for their sites, or to evaluate the resulting property values.We undertook an observational campaign to investigate supraglacial debris properties at Pirámide Glacier, in the central Chilean Andes. First, we used established approaches, as well as some variations on those approaches, to derive z0 from wind-temperature tower data and k from thermistor strings in the debris at three glacier locations. Second, we determined locally-optimal k and z0 values to reproduce observed ice melt: we optimised k by simulating energy conduction through the debris with the surface temperature as an input, then optimised z0 by running a complete energy-balance model using the observed surface meteorology. We then conducted point-scale energy-balance modelling using the z0 and k values obtained i) with the derivations from field measurements; ii) through optimisation, or; iii) from the typical values found in literature. This allowed us to evaluate how the different methods perform by comparing the modelled and measured ice melt. Our results show that deriving local debris properties from measurements is challenging and that measured values can differ significantly from common literature values. The values derived from measured data can vary significantly depending on the method employed. It is important to note that these values can also differ significantly from the values required by an energy-balance model to accurately represent sub-debris ice melt. Furthermore, energy-balance models typically assume a representation of heat transfer within the supraglacial debris layer based solely on conduction and require a bulk thermal conductivity value. This highlights the necessity of efforts to reevaluate measurements in the field and reconsider our definition of debris properties in melt modelling.