Photogrammetrically UAV based terrain data generation and automatic extraction of torrential properties

Debris flows are a severe hazard in mountainous regions. However, cost-effective long-term studies of debris seldom, which leads to substantial uncertainties mitigation methods. This paper investigates whether remote sensing techniques can be applied assess the hazards mountain torrents and gather accurate information on development watershed. Torrents that prone often devoid vegetation thus well surveyed using photogrammetric methods based uncrewed aerial vehicle (UAV) surveys. The possibility extracting automatically torrent parameters from high-resolution terrain models, such as cross-sectional area or slope, is explored. presented methodology yields continuous derived along torrent, major advantage over pointwise field Cross-validation with measurements reveals strong agreement. These very highly incised sections, while they severely limited sections steep adjacent hillslopes and/or dense vegetation. We show these kinds assessments greatly gain UAV data followed by automatic parameter extraction. extracted provide insights so key weak points identified accurately assessed field. find contribute comprehensive, reproducible objective assessment processes predispositions. ground-based fieldwork still essential further research sensing-based crucial.