Two new systems Weather forecast based on Artificial Intelligence (AI) just featured in the magazine Nature and both can greatly improve the quality of forecasts. In fact, for the time being they show comparable precision in their analysis to methods currently used by meteorologists. extraordinary reducing the time required to perform the forecast and the ability to predict certain phenomena that are difficult to predict are some of the advantages of these two methods.
Compared to the working hours required by the traditional system, now called the numerical method used by professionals, the new AI-based models are much faster and require only seconds. So much so that one of the new models developed by researchers from the Huawei company in Shenzhen (China), Get results 10,000 times faster and with a degree of accuracy similar to an integrated operational forecasting system The European Meteorological Prediction Center is the most powerful in the world today.
The new model, which can be used from any computer, was fed 39 years of meteorological data from the planet to predict temperature, wind, pressure and humidity conditions at different altitudes.
Better predict extreme events
The second of the new systems produced by artificial intelligence comes from Tsinghua University School of Software and University of California (USA), also located in China. Scientists at both universities developed a model. It predicts the precipitation that will occur in a very short time in a certain place., just six hours in advance. In fact, the model was christened as: Publish now.
In this case, the goal is to more accurately predict extreme weather events that could harm or endanger the population. hurricanes, hurricanes, or tsunamis. It is a system that combines physical rules with deep learning. It works from existing data and respects the underlying physical processes.
Testing in special cases to predict precipitation on land areas of only 2,000 square kilometers has yielded really encouraging results, because This new system improved prognosis in 70% of cases compared to the classical system.. In any case, the researchers agree that there is still room for improvement in this model.
The speed of the process has been improved so much, Key to reducing the economic budget currently needed for meteorological centers’ computer hardware. Researchers say it will be possible to get the same quality information faster for less money.
However, experts believe that, at least for now, AI should coexist with traditional predictions driven by flesh and blood professionals, which is hard to replace for now. Human intervention is still necessary, especially in the case of extreme events for which a full long data history is not available due to their exceptional nature.
Reference works:
https://www.nature.com/articles/s41586-023-06185-3
https://www.nature.com/articles/s41586-023-06184-4
…….
Contact address of the environment department: [email protected]