EM-DAT is FAIR

The Centre for Research on the Epidemiology of Disasters (CRED), maintaining the Emergency Events Database (EM-DAT), has released a new data product: the EM-DAT Emergency Events Database Archive available from the University of Louvain (UCLouvain) Dataverse under the CC-BY-NC-ND license. This initiative aligns with the FAIR data principles, representing a notable improvement for the EM-DAT disaster risk data and its application in scientific research.

FAIR?

In 2016, a group of researchers published the FAIR Guiding Principles for scientific data management and stewardship, reframing the modern standards of high-quality scientific data, metadata, and managing infrastructure. FAIR data principles represent a fundamental shift in how the scientific community approaches data management and sharing. FAIR stands for Findable, Accessible, Interoperable, and Reusable – principles that have become increasingly crucial in modern scientific research.

• Findable means data should be easily discoverable through clear metadata and persistent identifiers.

• Accessible implies that data can be retrieved through standardized protocols.

• Interoperable ensures data can be integrated with other datasets and systems.

• Reusable means data is well-documented and can be utilized for future research.

The FAIR principles were developed to recognize the value of scientific data as a resource that goes beyond its original purpose. In today’s data-driven research environment, where international collaboration is common, these principles ensure that scientific discoveries can be verified, built upon, and expanded by the wider research community. By promoting transparency, they help reduce duplicated efforts and accelerate scientific progress by allowing researchers worldwide to effectively locate and utilize existing datasets. Additionally, the FAIR principles are designed to be machine-friendly, facilitating automated data sharing and processing, which enhances the overall use and value of data.

Why FAIR Matters for the EM-DAT Disaster Data?

For over 35 years, EM-DAT has been a unique global reference for disaster impact data, documenting human and economic losses resulting from both natural and technological hazards worldwide. The living EM-DAT data, currently accessible from the EM-DAT download portal, is updated weekly and does not offer access to the historical versions of the data. In the early days of EM-DAT, accessing the information required traveling to Brussels to visit the archive room or requesting specific files to be sent via fax. However, in the spirit of progress, EM-DAT has always evolved, adapting its standards, monitoring methods, data models, and content to better meet the needs of the EM-DAT user community.

We understand that reconstructing the history of the EM-DAT database is a challenging detective task. Since its inception, as described by Guha-Sapir and Misson in 1992, the EM-DAT database has undergone changes. The lack of a clearly archived dataset and comprehensive documentation, which could have served as reference points over time, complicates the analysis of the EM-DAT-based research conducted thus far. Questions arise regarding which version of EM-DAT was in use, as well as the protocols and standards that were applicable at that time. Additionally, we must consider whether the quality and content of the data have evolved since then.

To address these types of questions, CRED initiated an extensive documentation effort for the project, resulting in the launch of the EM-DAT documentation portal in 2023. Now, the EM-DAT archive that has just been released also includes a copy of the comprehensive documentation linked to the archived version of the data. From now on, any research item using the EM-DAT archive will be clearly associated with a specific dataset and its corresponding documentation, enabling the reproducibility and validation of EM-DAT-related research items.

This archive will further serve as a significant milestone, enabling future reanalysis of the data and allowing for transparent tracking of data quality and content improvements. Understanding and managing disaster risk is an area that can significantly benefit from ongoing improvements in both data quality and quantity. It is, therefore, imperative to manage disaster data in a manner that distinguishes changes caused by enhancements in data from those resulting from shifts in disaster patterns. This transparency is essential for supporting informed disaster science and effective risk management in the future.

👉 Access the EM-DAT Archive here