Robust Flood Monitoring Using Sentinel-1 SAR Time Series
Abstract
The 2017 hurricane season in North and Central America has resulted in unprecedented levels of flooding that have affected millions of people and continue to impact communities across the region. The extent of casualties and damage to property incurred by these floods underscores the need for reliable systems to track flood location, timing and duration to aid response and recovery efforts. While a diverse range of data sources provide vital information on flood status in near real-time, only spaceborne Synthetic Aperture Radar (SAR) sensors can ensure wall-to-wall coverage over large areas, mostly independently of weather conditions or site accessibility. The European Space Agency's Sentinel-1 constellation represents the only SAR mission currently providing open access and systematic global coverage, allowing for a consistent stream of observations over flood-prone regions. Importantly, both the data and pre-processing software are freely available, enabling the development of improved methods, tools and data products to monitor floods in near real-time. We tracked flood onset and progression in Southeastern Texas, Southern Florida, and Puerto Rico using a novel approach based on temporal backscatter anomalies derived from times series of Sentinel-1 observations and historic baselines defined for each of the three sites. This approach was shown to provide a more objective measure of flood occurrence than the simple backscatter thresholds often employed in operational flood monitoring systems. Additionally, the use of temporal anomaly measures allowed us to partially overcome biases introduced by varying sensor view angles and image acquisition modes, allowing increased temporal resolution in areas where additional targeted observations are available. Our results demonstrate the distinct advantages offered by data from operational SAR missions such as Sentinel-1 and NASA's planned NISAR mission, and call attention to the continuing need for SAR Earth Observation missions that provide systematic repeat observations to facilitate continuous monitoring of flood-affected regions.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2017
- Bibcode:
- 2017AGUFMNH23E2823D
- Keywords:
-
- 1910 Data assimilation;
- integration and fusion;
- INFORMATICS;
- 1922 Forecasting;
- INFORMATICS;
- 4313 Extreme events;
- NATURAL HAZARDS;
- 4331 Disaster relief;
- NATURAL HAZARDS