IN013-01
Evaluation, Accreditation, Certification... Oh, My!

Wednesday, 9 December 2020: 04:00
Virtual
Danie Kinkade, Woods Hole Oceanographic Institution, Woods Hole, MA, United States
Abstract:
More than a century ago, science acknowledged the importance and value data hold beyond their original, intended use (Cajal, 1897). Today, we continue to recognize the value of data; endeavoring to make them reusable in the pursuit of new knowledge, or for validating research results. But the nature of data has changed since the 1800’s, and we struggle under the size, speed and variety of data being produced today. We now look to digital data repositories to help ensure precious observations are stewarded long into the future; optimizing their management, discovery, access, and interoperability at scale. However, today’s multitude of geoscience data repositories range widely in purview, capabilities, and quality of operations.

In parallel, a broader data-related community has evolved; each stakeholder possessing its own role, perspective, and data needs (e.g., the funder wishing to maximize investment through data reuse; the peer reviewer requiring restricted access to assess validity of results). These needs are driving the proliferation of disparate efforts aimed at identifying quality repositories for their purposes. From operational models and standards, to criteria and certifications; even aspirational principles are being leveraged to evaluate repository infrastructure and practices. There is a growing expectation for repositories to demonstrate adherence to, or compliance with many of them. But, this comes at a cost to repository budgets and timelines. How can this be sustained and what are the implications of failure?

This presentation will highlight the operations and practices of an oceanographic repository as they relate to meeting emerging expectations for trustworthiness, and the broader implications of these growing assessments on a geoscience research community that relies on several long-lived repositories for their data publication workflows.