2026-10-14T03:14:33Z https://www.scidb.cn/oai

10.57760/sciencedb.j00001.01642 2026-08-20T10:32:37Z

2026-08-20 A UAV Multimodal Remote Sensing Dataset for Wind Damage and Flooding Induced by Typhoon Hagas in Coastal Guangdong, China (2025) doi:10.57760/sciencedb.j00001.01642 en The UAV Multimodal Remote Sensing Dataset for Wind Damage and Flooding Induced by Typhoon Hagas in Coastal Guangdong, China (2025) was developed to support fine-scale post-disaster damage assessment, urban micro-topographic hydrological analysis, agricultural and forestry disaster loss estimation, and multimodal remote sensing interpretation. Focusing on post-disaster coastal areas affected by Typhoon Hagas in 2025, UAV-borne LiDAR and visible-light collaborative surveys were conducted from September 26 to 28, 2025, in nine typical post-disaster sample areas on Hailing Island, Yangjiang City, and in Enping City, Jiangmen City, Guangdong Province. The sample areas cover multiple post-typhoon disaster scenarios, including waterlogging in coastal high-rise building areas, roadside tree lodging, agricultural facility damage, reservoir surroundings, and river channel environments.The dataset covers a total valid area of approximately 2.72 km² and has a total data volume of 82.1 GB. It includes true-color point cloud data in LAS format, digital elevation models (DEMs), digital surface models (DSMs), digital orthophoto maps (DOMs), and valid spatial extent vector files (valid_extent.shp) for each sample area. The DEMs and DSMs have a spatial resolution of 0.5 m, while the DOMs have a spatial resolution better than 0.1 m. All spatial data are provided in the WGS 84 / UTM Zone 49N coordinate system (EPSG:32649). Each sample area contains corresponding point cloud, DOM, DEM, DSM, and valid_extent.shp files. The valid_extent.shp files define the common valid coverage of DOMs, DEMs, and DSMs, and serve as mask layers for raster clipping, spatial alignment, and valid extent identification.During data processing and quality control, UAV-borne LiDAR point clouds were processed through trajectory calculation, denoising, classification, and raster product generation. The spatial consistency among DEMs, DSMs, and DOMs was checked using DOM textures, DSM elevation variations, and visual inspection of typical profiles. Abnormal high and low values in DSMs were inspected and locally corrected through three-dimensional visualization, local elevation anomaly detection, and manual verification. After processing, DOMs, DEMs, and DSMs were uniformly clipped and aligned based on the valid_extent.shp file of each sample area, ensuring consistent valid spatial coverage, a unified projected coordinate system, and georeferenced overlay capability among different data modalities within the same sample area.This dataset records representative post-typhoon disaster scenarios, including waterlogging in coastal high-rise building areas, roadside tree lodging, agricultural facility damage, and terrain conditions around reservoirs and river channels. It can provide fundamental data support for post-disaster scene recognition, terrain analysis, three-dimensional structure interpretation, agricultural and forestry loss assessment, urban local waterlogging studies, and multimodal remote sensing algorithm validation. UAV; LiDAR; Typhoon wind damage; High-rise waterlogging; Digital Elevation Model (DEM); Multi-modal dataset xiao xiang Yue Huanyin Li Xuelin PUBLIC https://creativecommons.org/licenses/by/4.0/ dataset Science Data Bank