China to extend AI weather warning system MAZU to 30 countries

July 24, 2026 (MLN): China is to extend its AI-powered MAZU meteorological early warning system to 30 countries over the next five years, part of a wider push to widen access to artificial intelligence tools for disaster preparedness.

The system, developed by the China Meteorological Administration (CMA), integrates cloud services, Fengyun satellite data, AI-based forecasting and impact-based warnings into a single platform designed to help vulnerable communities prepare for extreme weather before it strikes.

The commitment came at the 2026 World Artificial Intelligence Conference (WAIC), held in Shanghai, where Beijing positioned AI as a tool for climate resilience rather than only frontier computing, APP reported.

The upgraded, cloud-based version of MAZU now carries more than 200 products across 30 categories, with AI woven through the platform, according to the CMA.

Its scope has widened from conventional weather forecasting to impact-based forecasting, helping agencies gauge not just what weather is coming but how it could affect transport, agriculture, energy, aviation and other sectors.

The platform currently supports users in more than 40 countries, with tailored versions running in Pakistan, Ethiopia, the Solomon Islands, Jordan, Sri Lanka, Mongolia and Djibouti.

Deployment varies by country need. In Pakistan, Chinese and Pakistani meteorologists have jointly built specialized tools addressing glacial lake outburst floods, monsoon rainfall and storms, feeding in local radar and satellite data.

Jordan's version supports Arabic-language warning dissemination, Djibouti's covers port and urban management, Mongolia uses the system for sandstorm prediction and urban disaster response, while Ethiopia has extended it to agriculture, river basin management, aviation and marine services.

Supporting infrastructure includes permanent overseas cloud nodes in Singapore and Cairo run by the National Meteorological Information Center, while Fengyun satellite data now reach users in over 130 countries and regions.

AI has also been built into specialized tools such as a sand and dust forecasting model that runs considerably faster than conventional numerical systems, the CMA said.

CMA chief engineer Pan Jinjun said artificial intelligence had become a key driver of improved forecasting and early warning for hazardous weather, noting that AI models have shown strong results in typhoon forecasting even as limited extreme-weather data and reliability concerns remain hurdles before such systems can be fully trusted operationally.

Capacity-building has run alongside the technology push: nearly 1,000 meteorological professionals from more than 100 developing countries have taken part in China's early warning training programs in recent years, and a MAZU Scholarship launched in June is aimed at professionals from Belt and Road partner countries.

A newly introduced menu-based service platform lets agencies pick observation equipment, satellite products and AI forecasting tools to match their needs, with more than 100 international cooperation projects already on offer.

Copyright Mettis Link News