A new weather forecasting system called Aardvark Weather, based solely on Artificial Intelligence (AI), can provide accurate forecasts dozens of times faster and using thousands of times less computing power than current systems.
The system is presented in a study published in the journal Nature and constitutes a model for a completely new approach to weather forecasting, with the potential to transform current practices.
Aardvark was developed by researchers at the University of Cambridge, with support from the Alan Turing Institute (both UK), Microsoft Research and the European Centre for Medium-Range Weather Forecasting.
Currently, weather forecasts are generated using a complex set of steps, each of which takes several hours to execute on custom-built supercomputers, explains the Alan Turing Institute.
Research from Huawei, Google and Microsoft has already shown that one component of this chain, the numerical solver (which calculates how weather evolves over time), can be replaced by AI, resulting in faster and more accurate predictions.
The European Centre for Medium-Range Weather Forecasts is applying this combination of AI and traditional approaches.
The Aardvark system, however, replaces the entire weather forecasting chain with a single machine learning model, which uses observations from satellites, weather stations and other sensors to produce global and local forecasts.
With this AI-powered approach, forecasts can be produced in minutes using a desktop computer, the institute adds. Using just 10 percent of the input data of existing systems, Aardvark outperforms the U.S. national forecast system GFS on many variables and is also competitive with forecasts from the U.S. Weather Service, which uses input data from dozens of weather models and analysis from expert human meteorologists.
Aardvark learns directly from the data, so it can be quickly adapted to produce customized forecasts for specific industries or locations, whether it’s predicting temperatures for African agriculture or wind speeds for a renewable energy company in Europe.
“Unlocking the potential of AI will transform decision-making for everyone from policymakers and emergency planners to industries that rely on accurate climate predictions,” said Scott Hosking, a researcher at the Alan Turing Institute. “Aardvark’s breakthrough isn’t just about speed, it’s also about access,” he added.
According to the researcher, transferring weather forecasting from supercomputers to desktop computers could "democratize forecasting, making these powerful technologies available to developing countries and regions of the world with data scarcity."
No comments:
Post a Comment