This guide explains how to enter occurrence records so they validate cleanly and export to a GBIF-ready Darwin Core Archive. You don't need to know Darwin Core by heart — hover any field label in the entry form to see a tooltip.
The three field tiers
The Add form groups fields by how important they are:
- Required — the minimum GBIF needs. Smart defaults are pre-filled; usually only the Scientific Name needs your attention.
- Recommended — optional but strongly improve data quality and discoverability (locality, coordinates, recorder, media).
- All Darwin Core fields — the full set of 229 terms, grouped by theme. Expand "Show all Darwin Core fields" to reach them.
Required fields
- Scientific Name — lowest-rank name, with authorship if known (e.g.
Pinus sylvestris L.). It is checked live against the GBIF backbone taxonomy. - Event Date — see Dates below.
- Basis of Record — how the record was obtained (e.g.
HumanObservation,PreservedSpecimen). - Occurrence Status —
presentorabsent. - Occurrence ID — a globally unique identifier; let the app generate it (see Identifiers).
Dates & times
Darwin Core eventDate must be ISO 8601, and the app stores and exports it that way:
YYYY,YYYY-MM, orYYYY-MM-DD- A range with a slash:
2024-06-15/2024-06-20
Dates shown elsewhere in the interface (record history, dataset metadata) are displayed in the European DD/MM/YYYY style — this is display only and does not change how dates are stored or exported.
Coordinates & uncertainty
- Latitude -90 to 90, Longitude -180 to 180, in decimal degrees (WGS84).
- Set the Geodetic Datum (default
WGS84) whenever you give coordinates. - Coordinate Uncertainty (m) must be greater than 0. Leave it blank if unknown — but do not enter
0(a common slip when coordinates are picked from a map), which is rejected as invalid.
Identifiers & UUIDs
Event, occurrence and measurement IDs are generated for you as deterministic UUID v5 values, derived from the record's content (project, date, location, protocol, etc.). The same inputs always produce the same ID, so identifiers stay consistent across systems and re-imports. When the inputs needed for a deterministic ID are incomplete, a random UUID v4 is used instead. Use the ↻ button next to an ID field to (re)generate one.
If your project configures an ID prefix, generated identifiers are namespaced like PREFIX-<uuid> — still a valid Darwin Core identifier, just easier to attribute to your project.
Adding images & other media
To attach an image, audio or video: host the file somewhere public (an institutional server, a repository, cloud storage with a public link, …), then paste that link into Media URL (mediaIdentifier). Fill the related media fields (type, format, title, creator, licence) where you can. A valid http(s) URL is shown as a thumbnail in the records table and rendered in the record view.
Controlled vocabularies
Many fields (Basis of Record, Sex, Life Stage, Event Type, Establishment Means, …) accept only specific Darwin Core values. These are offered as dropdowns; if you import data, a value outside the vocabulary is flagged during validation.
Import & export
- Import CSV, multi-sheet Excel (
.xlsx) or a Darwin Core Archive. Download a CSV or Excel template (with the correct DwC headers and dropdowns) from the toolbar. - Export a GBIF-compliant Darwin Core Archive (
.zip) per dataset, containingoccurrence.txt,multimedia.txt,meta.xmlandeml.xml.