Chinese scientists developed an artificial intelligence (AI)-based model with 3D neural networks for accurate and global weather forecasting, according to a recent research paper in the journal Nature.
The most accurate current forecasting system is the numerical weather prediction (NWP) method although it is computationally expensive. Daily weather forecasts, extreme disaster warnings, and climate change predictions are all realized by the NWP method, which relies on high performance computing and complex physical models.
Tian Qi Chief AI scientist with Chinese cloud service provider Huawei Cloud says that “ conventional NWP method requires four to five hours of computation on a supercomputer cluster with 3,000 servers to forecast global weather for the next 10 days”.
AI-based methods have recently shown some potential in speeding up weather forecasting by orders of magnitude.A large meteorological model research and development team from Huawei Cloud designed a 3D neural network adapted to the Earth’s coordinate system to process is complex.
The large-scale Pangu-Weather meteorological model, trained on nearly 40 years of global data, acquired 100 million level parameters in two months. Pangu-Weather takes only 1.4 seconds to complete a 24-hour global weather forecast, including potential humidity, wind speed, temperature and sea level pressure, among other values. Its prediction speed is 10,000 times faster than traditional numerical methods.
During Super Typhoon Mawar this May, Pangu-Weather performed admirably by predicting the spin path five days in advance. Bi Kaifeng, first author of the research paper, admits the shortcomings of AI-based weather forecasting, saying it is still highly dependent on reanalysis data and needs to improve its ability to predict extreme weather.
Read Now:Priyanka Chopra asked to share ‘three unknown facts’ about herself in an interview
