Earth observation-driven analysis of flood extent in mangrove areas under tropical storms

Abstract

Mangrove forests are coastal ecosystems that provide a wide range of ecosystem services. Among these, their role in coastal protection is especially relevant, as mangroves have the potential to reduce storm-driven flooding by slowing down surge, dissipating short-wave energy, and retaining water. With tropical storms intensifying due to climate change and mangrove ecosystems being increasingly under threat, understanding the capacity of mangroves to reduce flood extent under tropical storms has gained greater urgency. Yet, the complex interactions between ecological, geomorphological, and hydrodynamic variables that shape the capacity of mangroves to reduce flood extent make this difficult to assess. Hydrodynamic flood models face challenges in incorporating such complexity and are often constrained in their validation by the scarcity of observational flood data in tropical regions where mangroves are most prevalent. Remote sensing presents a promising yet underutilized opportunity to improve understanding of mangrove flood attenuation and support the validation of hydrodynamic models.

This research presents a novel approach consisting of (i) a conceptual guideline that structures the key variables influencing mangrove-induced flood attenuation and (ii) a remote sensing tool that maps flood extent using Sentinel-1 SAR C-band data in Google Earth Engine and visualizes relevant variables. Together, they form an approach that not only maps flood extent in mangrove areas following tropical storms, but also integrates key ecological, geomorphological, and hydrodynamic variables to contextualize the factors influencing flood attenuation by mangroves.

The approach was applied to a case study of Hurricane Irma (2017) in the mangrove region of the Everglades National Park, where the remote sensing tool detected a flood extent of 168,811 hectares, showing spatial resemblance to flood maps reported in literature. The flood detection model and its settings were refined through dry tests and a threshold sensitivity analysis. Applying the conceptual guideline to the case study area, guided by the outputs of the remote sensing tool, provided insights into elements affecting flood attenuation by mangroves in the region, including mangrove extent, storm track, precipitation, elevation, mangrove biomass and zonation, and channel patterns. Furthermore, comparing the flood extent derived from the remote sensing tool with that of a bathtub inundation model enabled the identification of a high-potential area where mangrove impact on flood extent appears most likely, providing a targeted basis for future research.

These results highlight the value of remote sensing as an accessible and globally applicable tool for generating flood extent maps and identifying areas with high potential for flood reduction by mangroves following tropical storms. The developed remote sensing tool represents a first step toward demonstrating how remote sensing can strengthen the validation of hydrodynamic flood models. Since such models can isolate the influence of mangroves on flooding, integrating remote sensing-derived flood extents as validation can enhance their accuracy and support future research into the flood attenuation capacity of mangroves. Ultimately, combined with the insights provided by the conceptual guideline, this can contribute to the wider goal of improving the understanding of the role of mangroves in coastal flood protection during tropical storms in the face of climate change.