Explore IUCN data format

[timrobertson100]

The IUCN has guidelines on how to take Darwin Core and convert it into a format used in the IUCN mapping standards.

We should explore what GBIF can do to simplify this process, by offering a native download format (i.e. using their field names), providing a simple format that includes only the fields of interest, or enabling a more flexible "select your fields to download" format.

@andrewrodrigues is our IUCN liaison for this

[andrewrodrigues]

Attached are the list of required and optional standard attributes that should be submitted with an IUCN range map. The point attributes tab lists those attributes that are required and optional for point data and the respective equivalent Darwin Code term. Requested that download format would only comprise of those fields.
1539098999-IUCN_Standard_attributes_for_spatial_data_v1.6_2018.xlsx

[MattBlissett]

So, these are the fields, but there are a few that aren't actually Darwin Core.

genus — fine
specificEpithet — fine
occurrenceStatus — fine
establishmentMeans — fine
behavior — fine
dateLastModified  — fine as http://purl.org/dc/terms/modified
decimalLatitude — fine
decimalLongitude — fine
geodeticDatum — fine
institutionCode — fine
year — fine
associatedReferences — fine
basisOfRecord — fine
catalogNumber — fine
locationRemarks — fine
eventRemarks — fine
island — fine
infraspecificEpithet — fine
taxonRemarks — fine
informationWithheld  — fine
accessRights — fine
collectionID — exists, but we do not interpret this (we don't align it with identifiers we have in the registry)
recordNumber — fine
recordedBy — fine
day — fine
month — fine
countryCode — fine
locality — fine
minimumElevationInMeters — we have this uninterpreted, the result of our interpretation goes into the term http://rs.gbif.org/terms/1.0/elevation
maximumElevationInMeters — as for minimumElevationInMeters
verbatimElevation — fine
verbatimLatitude — fine
verbatimLongitude — fine, they might also want verbatimCoordinates
verbatimCoordinateSystem — fine
verbatimSRS — fine
coordinateUncertaintyInMeters — fine
georeferenceVerificationStatus — fine
georeferenceRemarks — fine
identificationCertainty — not a DWC term. http://rs.tdwg.org/dwc/terms/identificationQualifier ?
typeStatus — fine
subgenus — fine
obsYearQualify — not a DWC term. Overlap with a full ISO-8601 date range, which eventDate should hold (but we don't yet support) and verbatimEventDate may hold.
obsCompilerNotes — not a DWC term.
adminError — not a DWC term. GBIF issues field?
adminFixed — not a DWC term.
adminSourceFixed — not a DWC term.
adminChangeNotes — not a DWC term.

I assume they'll want text (tab-delimited) format, though we can now supply AVRO format too.

[MattBlissett]

@andrewrodrigues, I need

• [ ] Clarification of what the format should be -- a zipped tab-delimited file, similar to our "simple" format but with different fields?
• [ ] Confirmation of what (if anything) should be used for the missing non-DWC fields.

Given that, it will be quick (next day or so) before something is ready to look at on gbif-uat.org.

[timrobertson100]

I gather this will be brought up on an IUCN technical group meeting shortly. We'll aim to get a contact who understands the IUCN format well and implications of these decisions and complete the mapping. Once reviewed by the working group we anticipate that GBIF will deploy the IUCN download format for any query on GBIF.