A high-resolution global flood hazard model.

Area of Science:

• Environmental Science
• Geoscience
• Hydrology

Background:

• Flood hazard research predominantly originates from developed nations, leaving data-scarce regions underserved.
• Growing populations and economies in developing nations increase the demand for accurate flood risk data.
• Existing flood models struggle with global applicability due to data scarcity and methodological limitations.

Purpose of the Study:

• To develop a globally applicable flood hazard model addressing key challenges in data-scarce regions.
• To create high-resolution (∼90 m) flood hazard maps for terrestrial areas between 56°S and 60°N.
• To validate the model's performance against existing high-resolution flood data.

Main Methods:

• A novel framework methodology integrating cross-disciplinary advances to overcome six key modeling challenges.
• Development of a global flood hazard model with an automatically parameterized subgrid channel network.
• Validation using high-resolution government flood hazard datasets from the UK and Canada.

Main Results:

• The global model accurately captures 66–75% of at-risk areas identified in benchmark datasets with minimal false positives.
• Aggregated to ∼1 km resolution, the model achieves a mean absolute error of ∼5% in flooded fraction.
• Explicit inclusion of channel networks significantly improves model performance compared to 2-D only or independently developed models.

Conclusions:

• The developed global flood hazard model offers a robust solution for flood risk assessment in data-scarce regions.
• The model's performance is validated, demonstrating its capability to inform flood management strategies worldwide.
• Future improvements in global terrain datasets are expected to further enhance the model's predictive accuracy, especially in urban areas.