AI meets rising waters: How Google’s Flood Hub is saving lives with smarter forecasts
This case study belongs to a compendium of good practices and success stories in disaster risk reduction shared during the 2025 Global Platform for Disaster Risk Reduction (GP2025). These stories reflect the real-world progress being made by governments, communities, and organizations around the world to reduce risk and build resilience.
Google’s Flood Hub is reshaping how the world responds to one of its most devastating natural hazards. Developed by Google Research in 2022 and co-created with governments, NGOs, and UN partners, the platform uses artificial intelligence to forecast floods up to seven days in advance. Unlike traditional models that depend heavily on localized data—which many vulnerable regions lack—Flood Hub learns from global patterns of rainfall and river behavior, making it effective even in data-scarce settings.
In Bihar, India, a rigorous evaluation of an early warning system powered by Google’s flood forecasts found that communities receiving alerts experienced 30 percent lower medical costs after flooding. The reduction is attributed to earlier evacuation, improved preparedness, and proactive measures that prevented injuries and illnesses—showing the tangible benefits of timely, trusted information.
This is how Google's Flood Hub uses AI forecasting to power local-level flood warnings
The tool is free, open-access, and scalable. National agencies can integrate it into their early warning systems and adapt it to local languages, contexts, and needs. By combining advanced AI with community-level engagement, Flood Hub empowers local leadership, strengthens risk communication, and ultimately saves lives.
Source: As shared at the GP2025 Media Hub interview