Arctic communities on permafrost face infrastructure instability from climate-driven thawing, a process exacerbated by the insulating effect of snow accumulation around 1 buildings and roads. To monitor this threat, we assess the feasibility of using a lowcost, consumer-grade UAV (DJI Mini 2) with Structure-from-Motion photogrammetry to map snow depth and derive thermal insulation metrics in Aklavik, NWT, Canada.By differencing high-resolution snow-covered (March 2024) and snow-free (September 2022) datasets, we found a mean snow depth of 0.65 m across the 0.98 km² study area.Critically, snow was deeper adjacent to roads (mean: 0.65 m) than buildings (0.59 m), creating stronger thermal insulation along transportation corridors. This translated into a substantial warming bias (∆T s > 2.0 °C) for over 45% of road-adjacent and 35% of building-adjacent areas. Validation revealed a slight overestimation on bare ground (+8.24 cm) but a systematic underestimation relative to probe measurements (bias: -17.32 cm), primarily because the snow-free survey captured the top of shrub canopies rather than the true ground surface. This study demonstrates that carefully processed data from accessible UAV technology can provide valuable, high-resolution insights into hazardous snow distribution, offering a cost-effective and scalable tool to support permafrost stability and infrastructure management.
Mueller et al. (Thu,) studied this question.