A team of international researchers, including experts from WorldPop, has released a groundbreaking dataset documenting over one million official early warning records across mainland China.
The Chinese Multi-Hazard Early Warning Dataset provides a standardized archive of over one million warning records issued between 1 January 2022 and 31 December 2025. Led by WorldPop visiting researcher Dr Haiyan Liu and Principal Research Fellow Dr Shengjie Lai, the study addresses a critical “institutional gap” between the scientific detection of a hazard and the dissemination of a public alert.
The dataset tracks high-resolution, hourly records across four administrative tiers, at national, provincial, city, and county levels. It employs a unified classification method covering hydrometeorological, geophysical, oceanographic, and environmental hazards, each graded by four levels of severity.
Accurate, contemporary data is vital for effective disaster risk reduction. By making warning timing and administrative hierarchy jointly observable at scale, this database provides essential empirical resources to improve risk communication and societal resilience during extreme events.
This work was supported by the National Natural Science Foundation of China, and an open grant from The Laboratory of Resources and Environmental Information System (LREIS), Chinese Academy of Sciences.
Image: Spatial distribution of warning counts by hazard type. a-d, Provincial level counts of environmental (a), geophysical (b), hydrometeorological (c), and oceanographic (d) warnings. e-h, Corresponding counts for the city level dataset. i-l, Corresponding counts for the county level dataset.
