Thermal diffusivity of mountain permafrost derived from borehole temperature data in the Swiss Alps

Abstract

Mountain permafrost is warming and thawing worldwide due to climate change, with ground temperature being a key control of its mechanical stability. Heat conduction is the dominant mode of heat transfer in frozen ground, and thermal diffusivity governs the rate at which temperature changes propagate through the subsurface. Despite its relevance, there are few field-based estimates of thermal diffusivity. In this study, we develop three mathematically independent formulations of the heat conduction equation to derive thermal diffusivity from borehole temperature data: (i) simple linear regression model (sLRM, statistical estimation), (ii) numerical inversion (semi-implicit finite-difference model), and (iii) analytical solution (explicit calculation). We first evaluate the validity of these three different approaches using a synthetically derived temperature dataset with known thermal diffusivities and then apply them to the 29 borehole temperature time series of the Swiss Permafrost Monitoring Network (PERMOS), enabling us to constrain thermal diffusivity in mountain permafrost with depth without prescribing any additional material properties. We obtain lumped thermal diffusivity values of the permafrost body for various permafrost landforms, with significant differences (median ± median absolute deviation): 1.5 ± 0.6 mm2 s−1 for bedrock, 1.1 ± 0.2 mm2 s−1 for talus slopes, and 1.3 ± 0.3 mm2 s−1 for ice bearing terrain like rock glaciers. This first compilation and analysis of empirically derived thermal properties of mountain permafrost advances the understanding of its thermal regime and provides key constraints for ground temperature and energy balance modeling. The application to real-world data further enables the identification of short-term non-conductive heat fluxes and the estimation of ground composition.