Improving Landslide Susceptibility Mapping with Drone and Helicopter Surveys

Landslides are recurrent hazards in Brazil’s Serra do Mar, intensified by steep slopes, shallow soils, and heavy rainfall. This study evaluates how aerial surveys using drones (UAVs) and helicopters can improve probabilistic landslide susceptibility mapping in the Rio Sagrado watershed (Morretes, Paraná). A baseline model based on the Monte Carlo method and GIS inputs underestimated susceptibility, with nearly 60% of mapped scars located in “no-susceptibility” zones. Helicopter imagery enabled broad detection of scars in inaccessible terrain, while UAV surveys provided high-resolution mapping of scarps and tension cracks. Integrating these inventories into the model allowed back-analysis of shear strength parameters and raised scar capture accuracy to 88%. The comparison shows helicopters are best for rapid, large-area assessments, while UAVs excel in detailed monitoring. By combining remote sensing with probabilistic modeling, this approach strengthens susceptibility predictions, supporting risk management, disaster prevention, and more reliable land-use planning.