Aditi Sengupta, Nicholas Ward, James Stegen

There is consensus throughout the Earth system science research community that “open data” is of critical importance. However, discoverability and accessibility are often overlooked, raising the question of how useful archived, but not easily discoverable data are. As part of evaluating databases suitable for our own research data archival, we conducted a data discovery exercise (aggregators and repositories) with search parameters to evaluate (i) feasibility of discovering data, and (ii) number of relevant results found (defined by exact matches to our search). We found that search parameters need more options (and perhaps community driven development of thematic keyword search options), repositories affiliated with funding agencies/large scale research datasets were more likely to reveal relevant results, broad aggregators with poor metadata requirements yield the most irrelevant results, and current practices may drive smaller datasets to disappear thereby promoting a non-inclusive open data world that is not truly open for all. There are encouraging signs, however, whereby commitment to open data practices is leading to datasets becoming public--with due credits--prior to analysis and associated publication. Ideally, making data meet FAIR principles means more than depositing data as a journal or funding requirement: community buy-in and consensus is needed across the spectrum of data generators, hosts, and users to agree on how to best achieve the ideal of data being findable, accessible, interoperable, and reusable.