| Popis: |
This report describes the types of data used for disaster risk reduction (DRR) and provides two country case studies, for Fiji and Sudan, with an in-depth look at the DRR datasets and associated metadata used by each country. These datasets were assessed against 15 FAIR (Findable, Accessible, Interoperable, Reusable) data metrics to identify which elements of FAIR were met. The report also provides a broader context giving details on the national, regional, and global agencies providing or hosting DRR data as well as initiatives aiming to increase the FAIRness of DRR data. Both of our case study countries are using remote sensing data which were assessed as having the richest metadata and met most of the FAIR metrics used in the assessment. Strategies for exploiting this data are discussed as they have great potential to provide up to date information during an emergency and to fill gaps in DRR data. An essential task for any scientific discipline is the establishment of common standards and terminologies. Historically, standards have differed considerably with agencies creating standards and vocabularies based on their own use cases and priorities; consequently, there is currently no universal standard used by all DRR practitioners. We discuss the most widely used standard definitions and provide suggestions for harmonising standards. As both the United National Nations Office for Disaster Risk Reduction (UNDRR) and the World Meteorological Organisation (WMO) have been working toward improving the FAIRness and consistency of DRR data, we describe their efforts and outline their lessons learned and recommendations. Our next deliverable, which discusses metadata standards, controlled vocabularies, and ontologies, will add to this discussion. While the current report focuses entirely on the DRR research area, DRR research is interdisciplinary by nature, encompassing researchers from earth sciences, climate change and environmental sciences, social studies, cultural information, and others. A key recommendation from the UNDRR is that there should be interdisciplinary collaboration when setting standards and definitions; therefore, increasing FAIRness in DRR has the potential to increase FAIRness across many related disciplines. The study found that the data used by Fiji and Sudan for DRR is missing many key FAIR data elements. Hazard data tended to score highest for FAIRness, particularly hazard data originating from satellites. In contrast, vulnerability and exposure data were the least FAIR with little metadata and limited machine readability. However, there are some excellent regional and global initiatives aimed at increasing the level of FAIRness in DRR data. The UNDRR is currently reinventing its DRR database to provide a much more coherent and consistent view of the state of DRR both globally and nationally. We applaud this project and believe that significant effort should be made by the global and regional agencies to work together to provide standards, controlled vocabularies, data distribution platforms, resources and guidance for all people working to reduce the impact of disasters. |
