The following question(s) were asked in the Collection Management Systems Webinar and will be answered here.
David Shorthouse: I noticed absence of the word “occurrence” (the intersection of an event and a material sample) or even occurrenceID (persistence of this intersection) from John’s slides. Would collections management systems be responsible for creating occurrences or could that responsibility shift to GBIF in the new data model? What might this mean for gbifID?
Response:
In the first webinar we talked about how dwc:Occurrences map to the Unified Model. As you say, Occurrences are a cross-walk across primary entities in the Unified Model and are not entities on their own. They are a view that includes, minimally, Location, Event, Entity, Identification, and Taxon. They might also relate to lots of other things. As such, in the new model occurrenceIDs do not have much meaning, there is no such independent concept that requires an identifier a priori. They are conclusions derived from primary data. With the new model, data publishers no longer need to (though they still can if they want to) artificially stuff data into Occurrences and provide an identifier for this abstract concept that doesn’t really exist in their data management systems (unless they did so to satisfy the requirement for sharing these data via Darwin Core archives).
I imagine that GBIF will continue to provide gbifids for Occurrences, since they are such a prominent and useful product and having a simple way to refer to them has value.