Effectiveness of Unmanned Aerial Vehicle-Based Remote Sensing for Assessing the Impact of Catastrophic Windstorm Events on Timberland

Abstract

The forests in the southeastern United States are of great importance for their economic, ecological, and cultural values. However, these forests are increasingly threatened by climate-induced windstorms, such as hurricanes and tornadoes, which can cause significant damage to the forest structure and increase the likelihood of secondary ecological disturbances such as wildfires and insect attacks. To assess the impact of these windstorms, advanced remote sensing technology, specifically unmanned aerial vehicles (UAVs) equipped with Light Detection and Ranging (lidar) and RGB camera was used. The study conducted a comparative analysis of three classification techniques, Maximum Likelihood (ML), Decision Tree (DT), and Random Forest (RF), on two datasets, one integrated with lidar-derived Canopy Height Models (CHM) and one without. The results showed that Random Forest consistently outperformed the others, achieving an overall accuracy of 94.52% with CHM and 77.56% without. These findings emphasize the effectiveness of UAV-lidar and RGB imagery as rapid and efficient tools for rapid windstorm damage assessment. These findings have significant implications for landowners, policymakers, and further research in environmental monitoring and disaster management across diverse forested landscapes.