Scientists at the University of Cambridge have created a new AI system called Aardvark Weather, which can predict weather much faster and with far less computing power than current methods. Unlike traditional systems that use multiple steps and supercomputers, Aardvark runs on a desktop and delivers accurate forecasts in minutes.
Supported by The Alan Turing Institute, Microsoft Research, and The European Centre for Medium-Range Weather Forecasts, Aardvark is a breakthrough in forecasting. It learns directly from real-world data—like satellite images and weather stations—allowing it to predict global and local weather efficiently.
Professor Richard Turner, who led the research, said:
“Aardvark is thousands of times faster than previous weather forecasting systems. It can provide more flexible, cost-effective, and accurate predictions.”
Traditional weather forecasting requires powerful computers and large teams. Some tech companies like Google, Microsoft, and Huawei have improved parts of the process with AI, but Aardvark replaces the entire system with a single AI model.
Despite using only 10% of the data that current models require, Aardvark has already outperformed the United States national GFS forecasting system on several key factors. It can also predict extreme weather events like hurricanes, wildfires, and tornadoes, making it useful for both developed and developing countries.
Dr. Scott Hosking from The Alan Turing Institute emphasized its impact:
“Aardvark makes weather prediction accessible to everyone. By shifting forecasts from supercomputers to desktop computers, we can help countries with limited resources.”
The next step for Aardvark is to improve its accuracy further and expand its use globally, including in developing countries where advanced weather prediction is often unavailable. Experts believe AI-powered forecasting could transform how industries—such as agriculture and renewable energy—prepare for changing weather conditions.
Source: preventionweb
