A new system called Aardvark Weather and based solely on artificial intelligence (AI), promises to transform current practices for predicting the weather.
It is expected that this new system can offer precise forecasts in a matter of minutes with a fraction of the computing resources of traditional systems.
Features of the system that can predict the weather in minutes
This system presents superior efficiency as it uses only 10% of the input data from traditional models, surpassing the U.S. GFS system in multiple variables.
It is a unique model, replacing the entire traditional weather prediction chain with machine learning, using data from satellites, weather stations, and sensors.
It runs on desktop computers, democratizing access to accurate weather forecasts, especially in developing countries and regions with data scarcity.
Benefits and applications
The Aardvark system replaces the entire weather prediction chain with a single machine learning model, which takes observations from satellites, weather stations, and other sensors to develop global and local forecasts.
It can quickly adapt to create tailored forecasts for specific sectors or locations, whether predicting temperatures for African agriculture or wind speed for a renewable energy company in Europe.
It is expected to improve decision-making for policymakers, emergency planners, and businesses.
Collaborations and development
Aardvark has been developed by an interdisciplinary team at the University of Cambridge, the Alan Turing Institute, Microsoft Research, and the European Centre for Medium-Range Weather Forecasts.
This innovative approach could transform global meteorology and expand the reach of advanced technologies in communities with fewer resources.
The future of climate prediction
Despite current limitations, the future of weather prediction with AI looks promising. Researchers believe that AI could someday completely outperform traditional physical models.
Scientists are focusing on improving the accuracy of existing physical models, which could then be used to train AIs, allowing them to replicate results more quickly and efficiently.
While this approach is still in its early stages, many are optimistic about the potential of AI to transform meteorology, reducing the need for supercomputers and making predictions more accessible and affordable.
Do you already know our YouTube channel? Subscribe!
