The China Meteorological Administration (CMA) has launched a new AI-driven multi-hazard early warning system, MAZU-Urban, which it says is a ‘joint action’ for the UN’s Early Warnings for All initiative. It is intended to create an “early warning service network” and share “practical experience and technological achievements” with global partners.
The MAZU framework aims to support global progress toward the UN goal of universal early-warning coverage by sharing China’s operational experience, tools and digital technologies with international partners. As part of the launch, CMA has donated the MAZU-Urban system to meteorological representatives from Djibouti and Mongolia.
Developed by the Shanghai Meteorological Service together with the National Meteorological Centre, China Unicom Shanghai, the Shanghai Academy for Science and Intelligence, the Shanghai Artificial Intelligence Laboratory and others, MAZU-Urban integrates advanced algorithms, multisource data, cloud-based early warning system products and meteorological open-source models. It also incorporates an Africa-region numerical model developed in Shanghai.
The platform combines knowledge-enhanced AI, multimodal data fusion and deep reasoning techniques to produce region-specific forecast and warning products. It supports the full warning chain, including risk knowledge, monitoring and early warning, warning dissemination and emergency response.
In response to the practical needs of regions such as Macao, China and Mongolia, the R&D team developed role-specific and hazard-specific workflow prompts for disaster prevention and mitigation guidelines. These guide large language models to engage in in-depth reasoning and dynamically generate preventive strategies tailored to local conditions.
MAZU-Urban uses a “three-terminal integrated” architecture:
- An all-in-one machine for professional decision-makers;
- A tablet interface tailored to sectors such as ports and shipping;
- A mobile terminal that reaches the public, delivering intelligent Q&A services, disaster alerts and evacuation guidance.
The system has been in trial operation since January 2025 in 35 countries and territories across Asia, Africa and Oceania.
At the 2025 SDG Cities Global Conference, the Shanghai Meteorological Service outlined a new international co-construction plan for MAZU-Urban. Within China, the platform is also being expanded into climate-related health applications. In Shanghai’s Fengxian District, meteorological and health agencies are jointly developing risk forecasts for cardiovascular and cerebrovascular diseases under a new Health + Meteorology project.
A district-level climate-related health risk monitoring and assessment platform has been established to display real-time disease risk levels, high-risk population trends and localized hazard information, providing a reference for government and industry regulatory departments to assess and respond to future health risks.
In related news, CMA launches first AI-powered meteorological service