Deutsch English
← back to archive
AI & Technology · 481 words · MN-260604-0066

AI Transforms Sandstorm Forecasting in Gansu, China

with Speed and Precision

As spring arrives in northwest China's Gansu Province, residents are no longer caught off guard by sudden dust storms that once enveloped cities in a yellow haze, disrupted travel, and posed serious health risks. Thanks to a technological breakthrough, citizens now receive precise, real-time warnings days before a storm hits, allowing them to take simple precautions like wearing face masks or rescheduling travel plans to greatly mitigate the impact on daily life and public health.

This shift is driven by the AI-driven Global Aerosol-Meteorology Forecasting System (AI-GAMFS). Developed by Chinese scientists, the artificial intelligence model is dramatically improving both the accuracy and speed of dust and air pollution forecasting. According to Gui Ke, an associate researcher at the Chinese Academy of Meteorological Sciences (CAMS), traditional forecasting models calculate meteorological elements separately from aerosols—microscopic solid particles or liquid droplets suspended in the atmosphere, such as dust, PM2.5, and smoke. "Forecasting aerosols is far more complex and computationally expensive than traditional weather forecasting," Gui explained, noting that the system must simultaneously parse multiple aerosol sources, complex chemical transformations, and their multi-scale interactions with weather systems.

By applying AI, the new model dynamically links suspended aerosol particles with meteorological factors like temperature, wind speed, and pressure as a unified whole. This holistic approach enables the system to simulate the evolution of the atmosphere with much greater precision, dramatically enhancing forecast accuracy. In addition to precision, the AI model delivers unparalleled speed. While traditional numerical forecasting relies on massive supercomputer clusters to solve complex physical equations—often taking hours to run a global forecast just a few times a day—the AI-driven system operates on graphics processing units and can generate a global forecast in just 36 seconds, making it over 100 times faster than traditional methods.

The technology is already transitioning from the lab to real-world application. Duan Haixia, a chief expert at the Lanzhou Institute of Arid Meteorology of the China Meteorological Administration (CMA), stated that the institute has accurately predicted over 10 major dust events across northern China since late last year, leveraging the model's ability to deliver high-precision environmental weather forecasts for the next three to five days. Beyond tracking storms, the system powers personalized public health alerts, such as warning allergy sufferers to wear N95 masks or helping hospitals prepare for spikes in respiratory illnesses.

Currently, AI-GAMFS has been deployed at the National Meteorological Center and over 10 provincial meteorological departments, including those in Gansu and Shaanxi. It has also been integrated into the CMA's "MAZU" public early warning cloud platform, which is designed to deliver disaster alerts to the public. Fully open-sourced in line with international standards, the model now provides developing countries worldwide with a low-cost, high-precision aerosol forecasting solution.