Assessment of Post-Typhoon Vegetation Damage in a City Park by a UAV equipped with LiDAR and Multi-spectral Camera

Typhoon Faxai (also known as the Reiwa 1 Bōsō Peninsula Typhoon) made landfall in Japan’s Kanto region on September 9, 2019. The typhoon brought a record-breaking rainstorm; the maximum instantaneous wind speed was recorded at 10 meteorological stations in Chiba Prefecture, Japan. In such a case, the local government seeks to conduct an investigation of the damage caused by the typhoon. An unmanned aerial vehicle (UAV) is an effective tool for investigating such damage when the flight conditions are appropriate. Here, we evaluated a detection method of damaged vegetation at a City Park in Chiba Prefecture before and after Typhoon Faxai obtained by a UAV equipped with a light detection and ranging (LiDAR) and a multi-spectral camera. The badly damaged vegetation in the study area was sampled from the spatial model subtracted a digital elevation model published by Geospatial Information Authority of Japan from a digital surface model generated by the LiDAR data. We proposed the use of a spatio-temporal difference ratio (SDR) based on two observations to assess the level of damage to vegetation derived from multi-spectral characteristics. These characteristics were based on five band reflectance values obtained by the UAV, and they were analyzed in the region of interest from the LiDAR data. The results demonstrated that the SDR can be used for the quantitative assessments of the damage level of vegetation in the range of 0 to 1.0. Our method can thus be used for detecting and evaluating the level of damage to vegetation in urban parks when natural disasters occur.