Google’s AI Model Can Predict Floods with a 7-Day Lead Time

In times of crisis, people rely on Google to get reliable and valuable information to help them defend themselves and their families. According to the latest updates, the tech company promises that the new technology could deliver earlier warnings of severe flooding to climate-sensitive areas around the world.

Google claims its artificial intelligence (AI) has accurately predicted river floods up to seven days in advance, providing residents in 80 countries with reliable forecasts. The findings were published in the scientific journal “Nature” and highlighted the promise of machine learning for one of the world’s most common natural disasters.

Floods are the most common natural disaster, and nearly 1.5 billion people, or about 19 percent of the world’s population, are directly exposed to significant hazards from major floods around the world, according to a Google blog post.

Google’s AI-based approach has greatly increased the accuracy of global flood forecasts, reducing the average forecast time from zero to five days. The paper detailed that the company trained machine learning models using historical events, river levels, elevation, topographic data and other data, producing local maps and running hundreds of thousands of simulations for each location. This thorough methodology allowed the models to reliably predict impending floods, even in areas with insufficient data.

The method has had a particularly large impact in underrepresented countries such as Africa and Asia, where flood forecasting has historically been difficult due to the lack of flow gauges in most rivers. Google’s AI overcame this challenge by leveraging available data and applying a machine learning model to pools where data was not available.

Google has made these flood forecasts available to a population of 460 million in 80 countries through multiple channels, including Google Search, Google Maps, Android notifications, and its own Flood Hub web app, launched in 2022.

Google plans to continue exploring the possibilities of machine learning in the development of better flood forecasting models. The company has been working with academic researchers to improve the AI-based technology, with the ultimate goal of creating a global end-to-end flood forecasting platform.