How to Add Accurate Source and Collection Metadata

Metadata should let another person decide whether a Geobble resource is appropriate before they reuse it. Record what the evidence supports and keep shared Collection context separate from Source-specific facts.

Start with evidence, not the filename

Gather the authoritative provider page, licence, technical documentation, version or dates, citation guidance and any processing notes, then inspect the actual Source as well: its geographic coverage, data model, important fields and representative records.

Do not infer a licence, date or creator merely because it appears plausible from the filename or organisation name.

Add Collection-level context

A Collection describes the family of Sources. Depending on the resource, Geobble supports Collection metadata such as:

title, description and longer explanatory text;
subject tags and country codes;
geographic scope and collection type;
update cadence and version label;
original creator and source URL;
licence name and licence URL;
review status and review dates.

Use these fields for information that genuinely applies to the family, but if individual Sources have different dates, licences or processing histories, do not flatten those differences into one Collection statement.

Remember that every Source has a required primary Collection and its public visibility derives from that Collection. Review Collection visibility with the member Sources in mind.

Add Source-specific metadata

A Source can carry more specific catalogue and provenance information, including:

title, description and explanatory text;
source type and geographic/admin scope;
temporal start/end and reference date;
original creator and original URL;
licence name and licence URL;
retrieval timestamp;
processing notes;
limitations;
recommended citation text;
review status and review dates.

Describe what one row, feature or raster represents, especially when the dataset name is broader than the actual contents. If the Geobble Source is a processed copy, distinguish the original producer from the account that filtered, corrected, enriched or otherwise prepared it.

Separate original evidence from Geobble processing

Suppose an agency publishes a facility register and your team filters closed facilities, corrects coordinates and adds service categories. The agency remains the original creator of the upstream dataset, while your Source metadata should also explain the transformations that produced the prepared copy.

Do not write the metadata in a way that implies the original provider authored your corrections.

The same distinction applies to derived Sources created from joins, routing, aggregation, enrichment or extraction. Record the upstream context and the transformation that changed the meaning or structure of the data.

Review metadata as a hand-off

After saving, try to answer these questions without relying on the Project conversation:

Can a reader identify what the resource represents?
Can they tell which geography and period it covers?
Can they identify the original producer?
Is the licence attached to the exact dataset rather than only to the organisation?
Can they tell whether the Source is original, copied, enriched or derived?
Is the most important limitation visible?
Can they cite the Source or return to the upstream resource?

Update the metadata when the Source is refreshed, replaced or materially transformed, because reusing the same title does not make a new boundary release or revised field definition equivalent to the old one.