University of CambridgeThe research team developed the AI weather forecastingsystems Aardvark Weather is expected to revolutionize weather forecasting, with computational speeds tens of times faster than existing methods, while requiring one-thousandth of the computational resources of current systems. The system is also supported by the Alan Turing Institute, Microsoft Research and the European Center for Medium-Range Weather Forecasts.
As foreign media outlet TechSpot reported today, the system works through theA single machine learning modelReplacing the traditional weather forecasting process, theProcesses data from multiple sources, including satellites and weather stations, on a single common desktop computerIn a short period of timein a few minutesGenerate global and local weather forecasts.
" Aardvark has revolutionized weather forecasting, making forecasts faster, cheaper, more widely applicable and more accurate." Richard Turner, Professor of Engineering at the University of Cambridge, said, " Aardvark's computational speedThousands of times faster than all previous methods. "
Although Aardvark relies on only a small portion of the existing system's data, it already outperforms the U.S. National GFS forecasting system on several key metrics and rivals the accuracy of the National Weather Service's forecasts - which often require multiple models and expert analysis.
Anna Allen, the paper's first author, from the University of Cambridge's Department of Computer Science and Technology, says this is just a preliminary result from Aardvark. This end-to-end learning approach could be widely applied toHurricanes, wildfires, tornadoesand other extreme weather forecasts, but also for more comprehensive Earth system forecasts such as air quality, ocean changes, and sea ice cover.
1AI has learned that Aardvark's greatest strengths areFlexibility and simplicity. Because it learns directly from data, it can be quickly adapted to fit the needs of a particular industry or region, theFor example, helping African agriculture predict temperatures, or wind forecasts for European renewable energy companies. In contrast, traditional weather forecasting systems often take years to adjust.
This technology is expected toChanging weather forecasting in developing countries. Due to the lack of expertise and computing resources in these areas, Aardvark was able to incorporate weather forecastingMoving from supercomputers to desktopsThe Government of the Republic of Korea has been working to make advanced forecasting technology available to more underdeveloped regions.
Aardvark will also play an important role in expanding weather forecasting capabilities in the future. Turner mentioned that the model is expected in the future to extend the weather forecastingAccurate forecast range increased to 8 daysThis is 3 days longer than the existing model.
