The Significance of Convection in Supraglacial Debris Revealed Through Novel Analysis of Thermistor Profiles

Melt from debris-covered glaciers represents a regionally important freshwater source, especially in high-relief settings as found in central Asia, Alaska, and South America. Sub-debris melt is traditionally predicted from surface energy balance models that determine heat conduction through the supraglacial debris layer. Convection is rarely addressed, despite the porous nature of debris. Here we provide the first constraints on convection in supraglacial debris, through the development of a novel method to calculate individual conductive and nonconductive heat flux components from debris temperature profile data. This method was applied to data from Kennicott Glacier, Alaska, spanning two weeks in the summer of 2011 and two months in the summer of 2020. Both heat flux components exhibit diurnal cycles, the amplitude of which is coupled to atmospheric conditions. Mean diurnal nonconductive heat flux peaks at up to 43% the value of conductive heat flux, indicating that failure to account for it may lead to an incorrect representation of melt rates and their drivers. We interpret this heat flux to be dominated by latent heat as debris moisture content changes on a diurnal cycle. A sharp afternoon drop-off in nonconductive heat flux is observed at shallow depths as debris dries. We expect these processes to be relevant for other debris-covered glaciers. Debris properties such as porosity and tortuosity may play a large role in modulating it. Based on the present analysis, we recommend further study of convection in supraglacial debris for glaciers across the globe with different debris properties. Plain Language Summary Glacier melt under a surface layer of rocky debris is typically predicted by determining the energy available at the debris surface, then calculating heat transferred via conduction through the debris layer. However, heat can also be transferred in the debris layer through other processes such as air movement and the evaporation of melt water. We analyzed temperature data in the debris to show that these processes are significant, accounting for up to 43% of heat transferred via conduction. These must thus be investigated further and incorporated into models of glacier melt.