by Ruth Duerr, NSIDC
This is an exciting time for scientific data management, a field that up to just a few years ago desperately needed more attention, but which now shows signs of coming into its own as people finally realize that digital materials of all types are ephemeral and that unless concrete near-term steps are taken to prevent it, much of the scientific and cultural heritage of the last 50 or so years could simply be lost.
Many of the problems with properly managing scientific data are cultural and structural. While people tended to agree that scientific data need careful management, the world of science does not consistently embody those principles in its values and structures. In short, data management lacked a clear mandate.
Much of this sea change is happening within the agencies of the US Federal government itself. In 2010, NSF introduced a requirement that all proposals submitted be accompanied by a two-page data management plan, and now other agencies are following suite. In the Earth sciences, NASA now calls for a data management plan for all of its Earth science missions, projects, grants and cooperative agreements and has developed a Provenance Content standard; and the USGS now requires that proposals to the USGS National Climate Change and Wildlife Science Center or any of the Department of the Interiors Climate Science Centers require a Proposal Data Management Plan. But the most exciting news I’ve heard recently is the work that NOAA is undertaking to improve their data management practices not just by their grantees, but also within NOAA’s systems and culture itself.
I was invited to attend the third annual NOAA Environmental Data Management conference to lead a session discussing implementation of NOAA’s new data management planning directive, as well as to teach a short course on creating data citations. The Data Conservancy (DC) and the Federation of Earth Science Information Partners (ESIP), who have been collaborating on the development of data management training materials, sponsored my participation in the conference. There I learned about five other documents, procedural directives in NOAA’s vernacular, which cover much of NOAA’s data-related activities that are either approved or in the works over the near term. As Jeff de La Beaujardière’s, NOAA’s Data Management Architect, presentation described them, these documents are:
- “Data Management Planning Procedural Directive – Plan, in advance, how you will preserve, document and distribute your data
- Archive Procedural Directive – What to archive, how to submit to archive
- Data Documentation Procedural Directive – How to apply ISO 19115 metadata for discovery, use & understanding
- Data Sharing by NOAA Grantees Procedural Directive – State how you will share your data and share within two years
- Data Access & Discovery Procedural Directive (in preparation) – Provide on-line services so your data can be found and retrieved
- Data Citation Procedural Directive (in preparation) – Use unique identifiers to allow data to be referenced and tracked”
Moreover, NOAA is taking a very pragmatic approach to the development and implementation of these guidance documents. Initial documents are drafted and approved, then tested in the real world for a period of a year or so to see what, if any, changes are needed to facilitate broad implementation within the agency. NOAA also has been very proactive about providing training for their staff. For example, recently NOAA has embarked on a series of webinars where staff can learn more or less everything there is to know about how to implement the ISO 19115 series of metadata standards – what tools are available, how to use them, what are the best practices within NOAA for filling out each section of the standard, etc. All in all, quite an impressive series of steps and commitments, commitments which should serve the agency well over the upcoming years.