This work presents a comparison of the meteorology and the surface energy and mass fluxes of the clean ice and debris-covered ice surfaces of the Djankuat Glacier, a partly debris-covered valley glacier situated in the Caucasus. A 2D spatially distributed and physically-based energy and mass balance model at high spatial and temporal resolution is used, driven by meteorological data from two automatic weather stations and ERA5-Land reanalysis data. Our model is the first that attempts to assesses the spatial variability of meteorological variables, energy fluxes, mass fluxes, and the melt-altering effects of supraglacial debris over the entire surface of a (partly) debris-covered glacier during one complete measurement year. The results show that the meteorological variables and the surface energy and mass balance components are significantly modified due to the supraglacial debris. As such, changing surface characteristics and different surface temperature/moisture and near-surface wind regimes persist over debris-covered ice, consequently altering the pattern of the energy and mass fluxes when compared to clean ice areas. The eventual effect of the supraglacial debris on the energy and mass balance and the surface-atmosphere interaction is found to highly depend upon the debris thickness and area: for thin and patchy debris, sub-debris ice melt is enhanced when compared to clean ice, whereas for thicker and continuous debris, the melt is increasingly suppressed. Our results highlight the importance of the effect of supraglacial debris on glacier-atmosphere interactions and the corresponding implications for the changing melting patterns and the climate change response of (partly) debris-covered glaciers.

We used a spatially distributed and physically based energy and mass balance model to derive the Østrem curve, that is the supraglacial debris-related relative melt alteration versus the debris thickness, for the Djankuat Glacier, Caucasus, Russian Federation. The model is driven by meteorological input data from two on-glacier automatic weather stations and ERA-5 reanalysis data. A direct pixel-by-pixel comparison of the melt rates obtained from both a clean ice and debris-covered ice mass balance model results in the quantification of debris-related relative melt-modification ratios, capturing the degree of melt enhancement or suppression as a function of the debris thickness. In doing so, our model is the first attempt to derive the glacier-specific Østrem curve through spatially distributed energy and mass balance modelling. The main results show that a maximum relative melt enhancement occurs on the Djankuat Glacier for thin and patchy debris with a thickness of 3 cm. However, insulating effects suppress sub-debris melt under debris layers thicker than a critical debris thickness of 9 cm. Sensitivity experiments show that especially within-debris properties, such as the thermal conductivity, the vertical porosity gradient and the moisture content of the debris pack, highly impact the magnitude of the sub-debris melt rates. The Østrem curve is also shaped by the local climate. Our results highlight the need to account for site-specific debris properties and their variation with depth, as well as for the effects of changing local climatic conditions in order to accurately assess (partly) debris-covered glacier behavior and its climate change response.