Data Collections Development

Data Types

We accept engineering datasets, as well as social and behavioural sciences datasets, derived from research conducted in the context of natural hazards. In the area of engineering the primary focus is on data generated through simulation, hybrid simulation, experimental and field research methods regarding the impacts of wind, earthquake, and storm surge hazards. We also accept data reports, publications of Jupyter notebooks, code, scripts, lectures, and learning materials. In social and behavioural sciences (SBE), accepted datasets encompass the study of the human dimensions of hazards and disasters. As the field and the expertise of the community evolves we have expanded our focus to include datasets related to COVID-19, Fire Hazards, and Sustainable Material Management. 

Users that deposit data that does not correspond to the accepted types will be alerted when possible prior to publication so they can remove their data from DDR, and will not be allowed to publish it. If a dataset non-compliant with the Collections Development policy gets published with a corresponding DOI, we will contact the authors and work with them to remove the data and leave a tombstone explaining why the data is not available. Curators review both in process and published data on a monthly basis. In both cases we will work with the authors to find an adequate repository for their dataset. 

Data Size

Researchers in the natural hazards community generate very large datasets during large-scale experiments, simulations, and field research projects. At the moment the DDR does not pose limitations on the amount of data to be published, but we do recommend to be selective and publish data that is relevant to a project completeness and is adequately described for reuse. Our Data Curation Best Practices include recommendations to achieve a quality data publication. We are observing trends in relation to sizes and subsequent data reuse of these products, which will inform if and how we will implement data size publication limit policies.

Data Formats

We do not pose file format restrictions. The natural hazards research community utilizes diverse research methods to generate and record data in both open and proprietary formats, and there is continual update of equipment used in the field. We do encourage our users to convert to open formats when possible. The DDR follows the Library of Congress Recommended Format Statement and has guidance in place to convert proprietary formats to open formats for long term preservation; see our Accepted Data Formats Best Practices for more information. However, conversion can present challenges; Matlab, for example, allows saving complex data structures, yet not all of the files stored can be converted to a csv or a text file without losing some clarity and simplicity for handling and reusing the data. In addition, some proprietary formats such as jpeg, and excel have been considered standards for research and teaching for the last two decades. In attention to these reasons, we allow users to publish the data in both proprietary and open formats. Through our Fedora repository we keep file format identification information of all the datasets stored in DDR.