05.08.2023 – 02:20
Huawei
Shenzhen, China (ots/PRNewswire)
July 2023 will likely go down in history as the hottest month on record and possibly the warmest in 120,000 years. With the increasing warming of the climate, the probability of extreme weather events increases. Conventional weather forecasts require enormous computing power to deliver useful results. A new AI-powered weather model is now available to the public that is fundamentally changing weather forecasting.
Pangu-Weather, a weather forecasting AI model developed by HUAWEI CLOUD, enables more precise weather forecasting with a 10,000X increase in calculation speed that enables global weather forecasting within a few seconds. This facilitates early predictions and preparatory measures in the run-up to extreme weather conditions. These results were published in the peer-reviewed scientific publication on July 5, 2023 Nature published.
Pangu-Weather is the first AI forecast model with a higher precision than traditional numerical forecasting methods and is made available to the public free of charge for the first time on the ECMWF (European Center for Medium-Range Weather Forecasts) website. This will provide global weather forecasting professionals, meteorologists, weather enthusiasts and the general public with a platform to view the Pangu-Weather Model’s 10-day global weather forecasts.
A “game changer” for conventional weather forecasts
Besides providing 10-day weather forecasts, the ECMWF also has one Message published, which compares forecasts from Pangu-Weather and the ECMWF IFS (a leading global NWP system) from April to July 2023.
According to the report, the advent of machine learning (ML) methods such as Pangu-Weather could be “a game changer for the gradual and rather slow evolution of traditional numerical weather forecasting (NWP) methods,” whose predictive ability, according to the World Meteorological Organization (WMO ) has improved by about one day per decade. This is due to the high computational cost of producing a forecast on standard NWP systems. ML models have the potential to revolutionize weather forecasting because they require significantly less computational effort and in are very competitive in terms of accuracy.
dr Tian Qi, Chief Scientist at HUAWEI CLOUD AI Field, IEEE Fellow and Academician of the International Eurasian Academy of Sciences, explains: “Weather forecasting is one of the most important application scenarios for scientific calculations, since meteorological forecasts are highly complex and it is difficult to integrate all aspects of the mathematical and physical knowledge to cover. Currently, Pangu-Weather complements the work of the forecasting system, and its main capability is to forecast the evolution of atmospheric conditions.”
Predicting extreme weather conditions with a proven high level of accuracy
The forecasting capabilities of the Pangu Weather model have been tested in extreme situations, such as E.g. Storm Eunice, which swept through north-west Europe in February 2022, and when the UK exceeded 40°C for the first time in summer 2022. These two examples show that data-based models are capable of predicting extreme weather situations and guidelines for medium-term forecasts.
The Pangu Weather forecast includes geopotential, specific humidity, wind speed and temperature – essential data to predict the evolution of weather systems, storm movement, air quality and weather patterns. Pangu-Weather was also used to predict the path of this year’s sixth Typhoon Khanun.
The ECMWF has long called for increased efforts from the global weather forecasting community to use AI models as additional components of their forecasting systems and to further explore the strengths and weaknesses of such models to support weather management.
dr Tian Qi reports, “Our goal is to provide a next-generation weather forecasting framework using AI technologies to optimize the existing forecasting systems.”
Photo – https://mma.prnewswire.com/media/2169705/1.jpg
View original content: https://www.prnewswire.com/news-releases/huaweis-ki-modell-pangu-weather-sagt-wetterereignisse-in-sekundenschnelle-vorher–jetzt-kostenlos-fur-die-offentlichkeit-zuganglich-301894040.html
Press contact:
Yifan Jim,
Tel.: +86-18620346367
Original content from: Huawei, transmitted by news aktuell