Now AI can predict weather, no supercomputer is needed

Researchers launched an artificial intelligence (AI) based weather forecasting system called Aardvark Weather which is beneficial in many different ways compared to traditional methods. This completely AI-powered system has been said to be capable of replacing the current system that relies on customized supercomputers and a team of experts, i.e. no heavy supercomputer is required to predict weather anymore.

The speed and accuracy of forecasting will improve and hyperlocal (local) forecasting will become possible. This AI system has been created through a partnership of researchers across various academic institutions and technology companies.

This research paper has been published in the journal Nature, which explains how an integrated AI model can replace supercomputers and expert monitoring. It is worth noting that last year Google introduced its GenCast model, which improves the accuracy and speed of weather forecasting by replacing numerical solvers using AI, however, the scope of Aardvark Weather is much bigger than that, as it aims to replace the entire weather forecasting system.

This AI system has been developed jointly by researchers from the University of Cambridge, Microsoft Research AI for Science, Google DeepMind, the University of Toronto, The Alan Turing Institute, and the European Center for Medium-Range Weather Forecasts (ECMWF).

Aardvark Weather collects data from weather satellites, weather balloons, stations, and other sensors and processes and analyzes this data to produce both global and local forecasts. While traditional supercomputers and meteorological experts take a few hours to do this task, this AI system can do it in just a few minutes.