Screenshot of weather forecast on CCTV

As new technologies become increasingly popular in China, the nation is hoping to see more AI talents in weather forecast, which is a less a popular area compared to self-driving technology, but solves more down-to-earth problems.

The essence of weather forecasting is calculating millions of possibilities based on known meteorological conditions, said Zong Zhiping, deputy director of the National Meteorological Center of China Meteorological Administration.

“It’s a predictive science and it’s impossible to be 100% accurate,” Zong said. “To make weather forecasts more precise is a process of continuously narrowing the gap between calculations and reality. This is a difficult problem in our work,” he added.

AI provides a new solution. “Weather forecasting involves massive amounts of different types of space-time data, which is appropriate for AI,” said Chen Min, deputy director of CMA’s Institute of Urban Meteorology in Beijing. Chen believes that algorithms and supercomputers can provide quickly and automatically provide realistic results.

However, there is a talent shortage for AI talents in weather forecasting, as most AI jobseekers tend to look at fancier scenarios such as self-driving technology.

AI is already helping researchers to make far more precise forecasts of ordinary and regular weather conditions in China. In Beijing, AI helps with temperature forecasting, while Guangdong province uses AI to predict short-term immediate precipitation.

However, experts said that in some catastrophic, extreme, or transitional weather conditions, meteorology knowledge and accumulated forecasting experience from researchers is still needed to conduct the most accurate forecast.