While rock temperature is a key factor for many geomorphic phenomena in high mountains, assessing the small-scale temperature distribution remains challenging. We adapted the program package WUFI ® , originally designed to calculate energy and moisture fluxes in building walls, for modeling natural rock faces. We determined physical rock properties and local meteorological parameters as input data for the simulations at rock walls in the Dachstein Mountains, Austria, and calculated annual temperature fluctuations at the surface and at 3, 10, 30, 50 and 100 cm depth, including different elevations, aspects and slope angles. Thirty-seven temperature sensors inserted into drillholes of different depths at ten locations (altitudes from 2300 to 2700 m) were used to validate the simulation. Additionally, surface temperatures were determined from infrared images; these were calibrated for surface emissivity using further temperature sensors. ). However, the transferability of this approach to natural settings by means of temperature sensor data has not yet been validated.