Urban Flood Extent Mapping by Integrating SAR Intensity and Coherence within a Gaussian Mixture Model Framework
Urban areas are highly vulnerable to floods due to impermeable surfaces, inadequate drainage, and high population density, which intensify flood impacts. Accurate flood maps are crucial for stakeholders in disaster mitigation. In this regard, Earth observation data from active sensors like Synthetic...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2025-07-01
|
| Series: | ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| Online Access: | https://isprs-annals.copernicus.org/articles/X-G-2025/847/2025/isprs-annals-X-G-2025-847-2025.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Urban areas are highly vulnerable to floods due to impermeable surfaces, inadequate drainage, and high population density, which intensify flood impacts. Accurate flood maps are crucial for stakeholders in disaster mitigation. In this regard, Earth observation data from active sensors like Synthetic Aperture Radar (SAR) have made significant contributions to flood delineation due to their ability to acquire data both day and night and penetrate cloud cover. However, flood mapping using only SAR backscatter in urban environments is challenging due to radar ambiguity introduced by the double bounce effect, commonly observed in inundated urban areas. Inundated regions typically appear as areas of low backscatter in SAR images, whereas flooded urban areas show a significant increase in backscatter due to the double bounce effect. This problem of under-detection caused by double bounce can be addressed by incorporating interferometric coherence (InSAR) as an additional input. Urban areas generally have high coherence, but it decreases in flooded areas. This decrease in coherence can be utilized to distinguish flooded urban areas. We present an unsupervised framework based on a Gaussian Mixture Model (GMM) that integrates intensity and coherence from Sentinel-1 SAR data to map the 2023 Greece floods. Upon validation against high-resolution optical imagery, our framework demonstrates that coherence significantly enhances urban flood mapping. |
|---|---|
| ISSN: | 2194-9042 2194-9050 |