Towards slope-scale assessment of avalanche formation: Exploring UAV-borne GPR for unveiling spatial snowpack variability

This study explores the potential of unoccupied aerial vehicle- (UAV-) borne ground penetrating radar (GPR) for high-resolution, slope-scale snow depth monitoring in challenging alpine environments to advance current methods of assessing the avalanche formation probability. To ensure accuracy under varying flight conditions, we systematically examined how UAV altitude and orientation influenced radar backscatter, assessing the necessity for altitude-based power calibration. These experiments revealed the impact of flight dynamics on signal return and helped refine data processing in complex terrain. Field campaigns were conducted across various sites in the Austrian Alps over two winter seasons, comparing GPR data with traditional reference measurements. Results revealed a strong correlation between GPR-derived and probe-measured snow depth ( R 2 = 0 . 94 , r = 0 . 97 ), demonstrating the reliability of the UAV-borne GPR method. Additionally, our findings highlight substantial small-scale variability in snow depth, even over short distances, underscoring the limitations of conventional point-scale observations. The UAV-borne GPR system used in this study features minimal deployment complexity, relying on off-the-shelf components, making it accessible to both scientists and practitioners. By providing high-resolution snow depth mapping, it complements fixed weather stations and significantly enhances traditional in situ observations as well as air- and spaceborne snow depth products. This method offers a safer, more efficient, and more detailed approach to data acquisition, with the potential to enhance avalanche monitoring and forecasting while contributing to improved safety measures in alpine environments. • Enhancing avalanche monitoring with detailed, safer slope-scale snowpack data. • The effects of UAV altitude and orientation on radar backscatter were investigated. • GPR depth matches probe measurements, confirming method reliability. • Reveals small-scale snow depth variability missed by conventional point data. • Utilizes off-the-shelf parts for easy and accessible deployment by scientists and practitioners.