Mapping Urban Flash Flood Damage Using High-Resolution Satellite Imagery

​Mapping flood damage in urban areas is challenging due to factors including building density and flash floods exacerbated by impervious surfaces. Often called "Appalachia's Katrina", Hurricane Helene was the deadliest mainland US storm since Hurricane Katrina , with more than 200 deaths and an estimated $78.7 billion in damage. In this study, we used PlanetScope high-spatial- and temporal-resolution satellite data and machine learning to, for the first time, map flood extent across heavily impacted counties in western North Carolina, focusing on Asheville, the region's largest city and previously deemed a "climate haven”. Random forest classification of pre-, peak-, and post-flood imagery achieved high accuracy (98.8–99.9% overall accuracy), though extensive cloud cover limited peak-flood classification, and models struggled to detect narrow waterways. To assess damage, we combined the flood extent maps with FEMA flood zones, FEMA USA Structures, and Black Marble nighttime lights data. Comparison with FEMA's 100-year flood zone in Asheville showed 94.76% overall pixel-level agreement, though only 40.57% agreement among pixels flooded by at least one layer, with more flooding missed within the FEMA zone (4.24% of pixels) than mapped outside it (1.01%). We identified 272 buildings intersecting the mapped flood extent in Asheville, though this likely underestimates total structural damage. Nighttime-lights anomalies were most severe in Asheville and surrounding western counties, consistent with the scale of reported damage there. These findings quantify flood extent and damage across western North Carolina and show how high-resolution satellite imagery can support emergency preparedness and response, while also highlighting key limitations, such as cloud cover, that constrain rapid post-storm imagery-based assessment.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-20
DOI
https://doi.org/10.21203/rs.3.rs-11039001/v1
Primary Topic
Flood Risk Assessment and Management
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Mapping Urban Flash Flood Damage Using High-Resolution Satellite Imagery

Júlio Caineta, Mirela G. Tulbure, Darcy Boast, Mollie Gaines et al.
6 links
Research Square
Flood Risk Assessment and Management
preprint

Mapping Urban Flash Flood Damage Using High-Resolution Satellite Imagery

Júlio Caineta, Mirela G. Tulbure, Darcy Boast, Mollie Gaines, Rebecca Composto
preprint en

Abstract

​Mapping flood damage in urban areas is challenging due to factors including building density and flash floods exacerbated by impervious surfaces. Often called "Appalachia's Katrina", Hurricane Helene was the deadliest mainland US storm since Hurricane Katrina , with more than 200 deaths and an estimated $78.7 billion in damage. In this study, we used PlanetScope high-spatial- and temporal-resolution satellite data and machine learning to, for the first time, map flood extent across heavily impacted counties in western North Carolina, focusing on Asheville, the region's largest city and previously deemed a "climate haven”. Random forest classification of pre-, peak-, and post-flood imagery achieved high accuracy (98.8–99.9% overall accuracy), though extensive cloud cover limited peak-flood classification, and models struggled to detect narrow waterways. To assess damage, we combined the flood extent maps with FEMA flood zones, FEMA USA Structures, and Black Marble nighttime lights data. Comparison with FEMA's 100-year flood zone in Asheville showed 94.76% overall pixel-level agreement, though only 40.57% agreement among pixels flooded by at least one layer, with more flooding missed within the FEMA zone (4.24% of pixels) than mapped outside it (1.01%). We identified 272 buildings intersecting the mapped flood extent in Asheville, though this likely underestimates total structural damage. Nighttime-lights anomalies were most severe in Asheville and surrounding western counties, consistent with the scale of reported damage there. These findings quantify flood extent and damage across western North Carolina and show how high-resolution satellite imagery can support emergency preparedness and response, while also highlighting key limitations, such as cloud cover, that constrain rapid post-storm imagery-based assessment.

Research Square
North Carolina State University (US)
Sustainable cities and communities, Climate action
Flood Risk Assessment and Management
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon
Mapping Urban Flash Flood Damage Using High-Resolution Satellite Imagery — Júlio Caineta, Mirela G. Tulbure, et al. · Research Square (2026) | TGRS Research Map | TGRS