Fast Flood Mapping with Synthetic Aperture Radar Data Using Google Earth Engine
Abstract
This study provides a fast flood extent delineation method in the coastal regions using the Synthetic Aperture Radar (SAR) data on Google Earth Engine (GEE) cloud-computing platform. We use the multi-temporal SAR imagery change detection technique and map the flood of the 2021 Hurricane Ida and the 2017 Hurricane Harvey in the Gulf Coast of the United States. We adopt a sensitivity analysis to calibrate SAR indices (Normalized Difference Flood Index, Ratio Image, and Difference Image Index) thresholding values to extract flooded areas. This analysis demonstrates that constant thresholding values fail to produce reliable results across the geographies. Comparison of the flood extent results with available Landsat 8 and Sentinel-2 multi-spectral imagery flood extent verifies the accuracy of this methodology (over 77%) to produce rapid flood extent map and highlights the capability of SAR data to delineate floods under the cloud cover. The obtained flood extent information helps decision-makers and emergency responders during the extreme weather conditions.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFM.H55M0739H