Sinkholes detection in the post-mining Olkusz-Bolesław Region (southern Poland) using a drone-based thermal camera

Abstract

Sinkhole activity in the post-mining region of Olkusz and Bolesław (southern Poland) has intensified following recent hydrogeological changes. We introduce a novel approach using Unmanned Aerial Vehicle (UAV)-based thermal imaging to detect and predict sinkholes, focusing on shallow mining-induced voids. Two study areas were analysed: a test area with numerous known sinkholes and a predictive area with high sinkhole formation potential. Thermal orthomosaics, generated from drone flights conducted at sub-zero temperatures, revealed positive thermal anomalies associated with newly formed or reactivated sinkholes, attributed to heat flow from subsurface voids, and negative anomalies in older, stable depressions. This method successfully identified high-risk locations, though some anomalies were influenced by external factors (e.g., buildings, snow cover, stored objects). The results show that UAV-based thermal imaging is effective for making inventories of recent sinkholes and holds promise for predicting their occurrence, particularly in dynamic hydrogeological settings.