top of page
Writer's picturemeowdini

AI Enhances Flood Warnings but Cannot Erase Disaster Risk

In September, parts of Europe were devastated by severe flooding, despite accurate weather predictions powered by AI. While advanced AI-driven systems, such as Google's GraphCast, predicted the intense rains, the true scale of destruction came as a surprise. This reflects a broader issue: while AI is revolutionizing weather forecasting, it alone cannot prevent disasters or mitigate their effects.


A flooded European city, showcasing the aftermath of extreme rainfall, worsened by climate change.
AI-based forecasts predicted the intense rains, but Europe was still devastated by floods, highlighting gaps in disaster prevention. Photo: Unsplash

Artificial intelligence has supercharged the field of meteorology by analyzing years of historical data and identifying weather patterns. AI systems can make faster and more accurate predictions than traditional numerical methods and at a lower cost. However, the September floods revealed gaps not in the forecasting itself but in how effectively the information is communicated and acted upon.

One of the challenges is that AI predictions are only as reliable as the data they receive. With climate change causing more frequent and intense extreme weather events, it’s difficult to account for all the variables at play. Andrew Charlton-Perez, a meteorology professor at the University of Reading, explains that while AI models can outperform physics-based models in some areas, they can also struggle when faced with unprecedented events.


In Europe, the accuracy of AI forecasts, such as those from the European Centre for Medium-Range Weather Forecasts (ECMWF), was evident ahead of the floods. The system predicted 300-400 millimeters of rain in affected regions, but even with accurate data, the impact of such rare, severe weather events is difficult to predict and manage.



Experts stress that communication is key to ensuring that AI predictions translate into meaningful action. Shruti Nath, a postdoctoral researcher in weather prediction at Oxford University, argues that forecasts need to be communicated clearly to emphasize the severity of potential impacts. This could prompt governments and businesses to allocate more resources toward preventative measures.

Governments also face challenges in using AI to its full potential. In the U.S., for example, businesses and authorities are better equipped with risk management tools, while European governments tend to rely on outdated data providers. AI models, like those developed by Norway’s 7Analytics, offer highly detailed predictions, but they require substantial computing power, energy, and investment—resources that many governments are not yet committing.


Moreover, AI models, while crucial in forecasting, have their own environmental impact. Large-scale computing systems consume significant energy and water, contributing to the emissions driving climate change. To mitigate the effects of extreme weather, experts like Friederike Otto from Imperial College London stress that cutting fossil fuel consumption and investing in sustainable infrastructure is essential.

AI may enhance flood warnings, but ultimately, the solution to climate-driven disasters lies in implementing sustainable policies and reducing the use of carbon-intensive resources


Source: Reuters

Comments


bottom of page