Repository Profiles for Atmospheric and Climate Sciences: Capabilities and Trends in Data Services

Abstract:

As digital research data proliferate and expectations for open access escalate, the landscape of data repositories is becoming more complex. For example, DataBib currently identifies 980 data repositories across the disciplines, with 117 categorized under Geosciences. In atmospheric and climate sciences, there are great expectations for the integration and reuse of data for advancing science. To realize this potential, resources are needed that explicate the range of repository options available for locating and depositing open data, their conditions of access and use, and the services and tools they provide.

This study profiled 38 open digital repositories in the atmospheric and climate sciences, analyzing each on 55 criteria through content analysis of their websites. The results provide a systematic way to assess and compare capabilities, services, and institutional characteristics and identify trends across repositories. Selected results from the more detailed outcomes to be presented:

• Most repositories offer guidance on data format(s) for submission and dissemination.
• 42% offer authorization-free access.
• More than half use some type of data identification system such as DOIs.
• Nearly half offer some data processing, with a similar number providing software or tools.
• 78.9% request that users cite or acknowledge datasets used and the data center.
• Only 21.1% recommend specific metadata standards, such as ISO 19115 or Dublin Core, with more than half utilizing a customized metadata scheme.
• Information was rarely provided on repository certification and accreditation and uneven for transfer of rights and data security.
• Few provided policy information on preservation, migration, reappraisal, disposal, or long-term sustainability.

As repository use increases, it will be important for institutions to make their procedures and policies explicit, to build trust with user communities and improve efficiencies in data sharing. Resources such as repository profiles will be essential for scientists to weigh options and understand trends in data services across the evolving network of repositories.