Satellite Imagery-based Hurricane Debris Assessment: Model Development and Application in Rural Florida

Abstract

Abstract The Southeastern United States is hit by hurricanes frequently. To provide a quick response to these natural disasters and expedited recovery, state and federal agencies should be provided with information concerning the severity and extent of hurricane damage, specifically in terms of hurricane debris volume and locations. In this paper, we focused on the post-hurricane debris in Calhoun County, Florida, which was hit by Hurricane Michael in 2018 and provided a macro-level assessment using satellite image collections obtained from the Sentinel-2 satellite in Google Earth Engine (GEE) and Geographic Information Systems (GIS). The specific focus was the assessment of hurricane debris impact on roadways and communities, which were performed in twofold. Firstly, satellite data was used to assess post-hurricane debris impact in the entire county and its population. Secondly, satellite data was used to assess post-hurricane debris with a focus on roadways and its impact on accessibility. Spectral indices from satellite images were used to derive assessment parameters. Normalized Difference Vegetation Index (NDVI) values were obtained for both pre-hurricane and post-hurricane images of the study area and NDVI scores of roadway segments were analyzed. Findings indicate that high volumes of vegetative debris were present in the north-eastern parts of the county, which is occupied by the highest population of county residents. The roadway segments in this region also recorded highest debris volumes, which is a critical information for state and local agencies that need to know highly impacted locations.