dwcready

Is coordinateUncertaintyInMeters required for GBIF?

Darwin Core · dwc:coordinateUncertaintyInMeters

What you are seeing — your archive ingested cleanly and you are wondering whether the blank uncertainty column matters.

The short answer

No, it is not required. Your archive will ingest, your records will be served, and GBIF's validator will not mention it. GBIF publishes an issue code for an uncertainty value that is invalid — COORDINATE_UNCERTAINTY_METERS_INVALID — but there is no code at all for one that is simply absent.

That absence is why it goes missing so consistently. We audited 5,064 occurrence records that were already published to GBIF and being served, and 1,972 of them carried no coordinateUncertaintyInMeters at all. Nothing in the pipeline had any reason to say so.

It matters to the person downstream, not to the ingest. A record at a named lake with 5 km of uncertainty and a GPS fix with 10 m of uncertainty are the same row once the field is blank, and anyone modelling a distribution has to either discard your records or silently assume a precision you never claimed. Filling it in is the difference between data that can be filtered and data that has to be trusted.

There is a related trap worth knowing about. If you record a coordinate to five decimal places and an uncertainty of 5,000 m, those two fields contradict each other — the precision implies about a metre. We report that as FS_PRECISION_EXCEEDS_UNCERTAINTY, and it fired on 1,230 of the same already-published records.

What to put in it

  1. A GPS fix under open sky: 3–10 m, or whatever your unit reported.
  2. A point read off a 1:50,000 map: on the order of 100 m.
  3. A named locality with no coordinate — a lake, a park, a township — the radius of the named place, which is often kilometres. This is the honest number, and a large value is not an embarrassment.
  4. Never leave it blank to mean 'exact'. Blank means unknown, and every downstream user has to treat it that way.
Find out how many of your records are missing it

Drop the file you already have. We will tell you how many rows carry no uncertainty, how many state a precision their uncertainty contradicts, and which columns we could not interpret at all. Free until 1 September, no account, and your records are discarded after the check — we keep the column mapping, never the rows.

Check the file you have →

Why we can say that

5,064records already published to GBIF
0findings in GBIF's vocabulary
4,268findings from ours
31.7%came back clean

We audited a file GBIF had already accepted and served. Their validator answers will this load; these had loaded. Figures read from ops/data.json, not typed into this page. We do not issue a rating of any kind — every finding names the rule it applied and where that rule comes from.

Related