Damage Analysis of the 2018 Jacksonville, Alabama, EF-3 Tornado Using Satellite- and UAV-Based datasets, and Site Survey Observations

Keywords: tornadoes, damage analysis, remote sensing, Sentinel-2, UAV, NDVI, PCA, Survey 123
Session Type: Virtual Paper Abstract
Day: Friday
Session Start / End Time: 2/25/2022 11:20 AM (Eastern Time (US & Canada)) - 2/25/2022 12:40 PM (Eastern Time (US & Canada))
Room: Virtual 17

Abstract

Tornadoes are the most destructive atmospheric phenomena. Current advances in high-resolution multispectral imagery provides a supplementary alternative to site surveys to determine the extent and characteristics of tornado damages. This research uses Sentinel-2 A images to estimate Normalized Difference Vegetation Index (NDVI) change before and after the storm. NDVI-differencing has the potential to capture damages as weak as EF-1 on the Enhanced Fujita (EF) scale. Since NDVI change analysis uses two bands of red and near-infrared, we will apply principal component analysis (PCA) as a second technique to include more bands into the analysis. The results were then correlated with wind speed data collected by NOAA and surveyed building damages to determine the stronger relationships of NDVI and PCs changes before and after storm. The final results were validated and supported by the UAV-based dataset and site survey data that were collected at the time of incidence.