JAKARTA - The sandstorm in Gansu, northwest China, is not just a yellow sky and dirty streets. For residents, it can mean a chaotic journey, interrupted breathing, and increased health risks.
Citing a Xinhua report, Friday, June 5, China is now using artificial intelligence or AI to speed up and sharpen forecasts of dust storms and air pollution.
The system is called the AI-driven Global Aerosol-Meteorology Forecasting System or AI-GAMFS. This model was developed by Chinese scientists to read the movement of dust and pollution particles in the atmosphere faster.
With the system, Gansu residents can receive warnings several days before a storm comes. They can prepare masks, postpone travel, or reduce outdoor activities.
Spring has been a vulnerable time for Gansu. Dust storms can come suddenly, covering the city with a yellow haze, disrupting transportation, and endangering health.
According to Gui Ke, a senior researcher at the Chinese Academy of Meteorological Sciences, the old forecasting system usually calculates weather and aerosol elements separately.
Aerosols are very small particles that float in the air. Their shape can be dust, smoke, or fine particles such as PM2.5 which are harmful to breathing.
"Aerosol forecasting is much more complex and requires more computing than ordinary weather forecasting," Gui told Xinhua.
He explained that the system must read many aerosol sources, chemical changes in the air, and their interaction with the weather system on various scales.
AI makes the process more integrated. Dust and pollution particles are read together with weather factors such as temperature, air pressure, and wind speed. In this way, atmospheric changes can be simulated more accurately.
Another advantage is speed. Traditional numerical prediction models use a cluster of supercomputers to calculate complex physics equations. The process can take hours to make global forecasts several times a day.
AI-GAMFS runs on a GPU, a graphics processing unit commonly used for heavy computing. The system can generate global forecasts in just 36 seconds, more than 100 times faster than traditional methods.
The technology is starting to be used in the field. According to Duan Haixia, a senior expert at the Lanzhou Arid Meteorology Institute under the China Meteorological Administration, the system has helped predict more than 10 major dust events in northern China since late last year.
This model is also able to make environmental weather forecasts for the next three to five days.
The benefits do not stop at storm forecasts. This system can also support more specific health warnings, for example, asking allergy sufferers to wear N95 masks or helping hospitals prepare for a surge in respiratory diseases.
Xinhua's report said that AI-GAMFS has been used at the National Meteorological Center and more than 10 provincial meteorological departments, including Gansu and Shaanxi.
The system also entered the MAZU public early warning platform of the Chinese meteorological authority. This platform is used to deliver disaster warnings to the public.
The model is made as an open source according to international standards so that developing countries can use it as a solution for low-cost and high-precision aerosol forecasting.
The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)
