Key points are not available for this paper at this time.
The Emergency Events Database (EM-DAT) compiles global disaster data resulting from technological and natural hazards. EM-DAT details the human and economic impacts from 1900 to present, with systematic recording since 1988. Serving the humanitarian, disaster risk reduction, and academic sectors, EM-DAT's transition to open access and climate change concerns have expanded its reach and visibility. The dataset, freely available for non-commercial purposes, offers structured tabular data categorized by hazard type and standardized by individual disaster events at the country level. Data collection and validation involve daily monitoring and manual processing of textual documents from predefined sources, searching for additional sources, and conducting periodic thematic updates. With over 26,000 unique disaster entries, approximately two-thirds of EM-DAT records relate to natural hazards, predominantly floods. Although initially rooted in Health Sciences, EM-DAT's historical use cases increasingly reflect applications in Earth, Environmental, and Social Sciences, paralleling heightened global concern around climate impacts and disaster risk understanding. The evolution of EM-DAT's data content mirrors societal and technological advancements in disaster reporting. Despite these improvements, known inconsistencies and biases in data quality have been reported. Aside from providing a detailed description of EM-DAT, this article acknowledges existing limitations, discusses their implications for research and decision-making, and identifies future opportunities to enhance disaster loss databases. • Global disaster database with 26,000+ entries tracking human and economic impacts from 1900. • First scientific documentation since 1992 of EM-DAT's methods and data sources. • EM-DAT Archive implements FAIR principles with enhanced validation and online documentation. • Database limitations reflect global disaster reporting gaps.
Delforge et al. (Wed,) studied this question.