Interface of Early Warning Terminal for Djibouti Port.
In Djibouti, on the Horn of Africa, dockworker Mohamed Ali used to grapple with uncertainty day after day at the Doraleh Multi-Purpose Port — blistering heat, sudden gales and short bursts of torrential rain could strike without warning. Extreme weather events have caused shutdowns, equipment failures, and heatstroke cases. If such disasters could be accurately forecasted, preventive actions could be taken ahead of time.
In 2025, tailored to Djibouti's needs in disaster prevention and mitigation, MAZU, China's solution to the United Nations' Early Warnings for All campaign, released by the China Meteorological Administration (CMA), was put into use in the country. Since then, Chinese AI-based meteorological technology has taken root in the Horn of Africa, bridging the capability gaps in Djibouti's disaster prevention capabilities and building an intelligent safety shield for the country.
MAZU-Urban is formally put into operation in Djibouti.
AI Collaboration Addressing Djibouti's Pain Points in Disaster Prevention
Djibouti commands a strategic waterway linking the Red Sea to the Indian Ocean, and the Doraleh Multi-Purpose Port needs weather warnings for everything from cargo handling to the scheduling of ocean-going vessels. MAZU, China's early warning solution for all, launched in July 2025, has drawn widespread attention. It takes AI as its core technology, deeply integrates advanced algorithms with multi-source data, and is equipped with cloud-based early warning capabilities, open-source meteorological models and more. Djibouti's meteorological authorities wasted no time in proposing cooperation with China, hoping to introduce a city-level AI warning system adapted to local conditions. The Shanghai Meteorological Service (SMS) took on the project on the basis of its Urban Multi-Hazard Early Warning Center demonstration project, thus embarking on a path of cooperation on AI-powered disaster prevention and mitigation between China and Djibouti.
During the preparatory phase, the Chinese technical team conducted several rounds of online research and on-site surveys, identifying the core disaster-prevention needs of Djibouti's cities, ports and airports: closing the gap in fine-grained risk assessment for urban heat and strong winds, customizing a dedicated meteorological warning module for ports to solve the problem of delayed warnings for sudden hazards, and helping the Djiboutian side compile its datasets to lay a solid data foundation for localized AI training.
On December 18, 2025, SMS and Djibouti's National Meteorological Agency (ANM) officially signed the project handover certificate, marking the launch of the MAZU-Urban AI Agent for Urban Multi-Hazard Early Warning in Djibouti.
With the Power of AI, Dockworkers No Longer Fear Extreme Weather
Since the AI agent went live, local dockworkers no longer have to rely on luck when it comes to weather conditions, as their operating model has undergone a fundamental transformation. Drawing on remote-sensing data from China's Fengyun meteorological satellites, internationally exchanged observation data and the Fengqing AI forecasting model, and referring to China's grading standards for urban meteorological hazards, the AI agent set localized thresholds for high heat and strong winds tailored to Djibouti, and built a dedicated "port meteorological service system" and an "urban early warning smart dashboard." Extreme weather can now be flagged with graded warnings up to 24 hours in advance, with alerts reaching port management and frontline crews through both text messages and the dashboard.
Members of the Chinese technical team talk with an operations manager at the Doraleh Multi-Purpose Port.
"This is the kind of technological support we've long been hoping for," said a technician at Djibouti's ANM. "The AI agent is not just user-friendly; it has also markedly sped up our response to warnings. It has given us forecasting references we never had before, especially in tracking extreme weather incidents such as heat waves and heavy rain."
At present, in response to the needs raised by the Djiboutian side, the Chinese team quickly launched a dedicated upgrade of the system and developed a special alert dashboard exclusively for Djibouti. This module is now in the implementation and development stage.
From a Single Site to a "Meteorological Bridge" Benefiting the Whole Country
Relying on the standards of the MAZU system, Djibouti has proactively incorporated Chinese technological specifications as a reference for its future meteorological monitoring projects, advancing cooperation from the use of equipment to the joint development of standards.
Building on what has already been achieved, China and Djibouti have mapped out a three-phase cooperation plan. In the short term, the port service system and the customized features of the warning dashboard will be put into operation. In the medium term, a project to deploy smart meteorological monitoring and warning terminals will be implemented across Djibouti to continuously expand local sources of observation data. In the long term, AI-based urban disaster prevention capabilities will be extended to Djibouti's inland towns, helping the country build a nationwide, tiered early warning network.
The China-Djibouti cooperation model has continuously delivered its demonstration effect. Since its launch, MAZU has been put into operation in countries including Pakistan, Ethiopia and the Solomon Islands, turning China's experience in disaster prevention and mitigation into a global public good.
As a signature achievement of China's response to the United Nations' Early Warnings for All initiative and its efforts to advance the Sustainable Development Goals, MAZU has built a new kind of AI-powered "meteorological bridge" for global disaster prevention and mitigation through practical cooperation. MAZU will continue to bring inclusive meteorological services to the Horn of Africa and to more countries and regions around the world. (Gao Shaohua)
A single purchase
