Data Sources

This project treats public trust as part of the product. Every public record needs a source type, and user-submitted reports never appear before review.

The first release is deliberately conservative: it favors fewer records with visible attribution over a large unverified crawl. Production data should keep species-level sources, occurrence-level sources, image licensing, and import logs in sync.

Curated species profiles

Published species profiles must include a source type and a stable reference URL. v1 starts with a small high-risk list so each profile can be checked before expansion.

GBIF-linked taxonomy and records

GBIF taxon keys, species pages, and occurrence records are used when configured. Import runs store status and failures so stale or incomplete syncs can be reviewed.

Authority references

GISD, GRIIS, DAISIE, and regional authority pages are kept as references for invasive status, regional context, and source attribution.

Occurrence-level attribution

Every public point should keep source type, source URL, and source record ID when available. Hidden or pending points are excluded from the public map and nearby search.

Reviewed user reports

Community submissions are stored as pending records first. They only appear publicly after an administrator approves them and creates a reviewed occurrence point.

Image licensing

Production species images should use public R2 URLs and include credit, license, and original image source URL before launch. Demo seed images should be replaced with species-specific licensed media.

CARTO / OpenStreetMap basemap

The interactive map uses MapLibre GL JS with CARTO raster tiles and OpenStreetMap data. Map attribution remains visible inside the map UI.

Publication and correction policy

  • Public pages only read published species and public occurrence points.
  • Rejected and pending reports remain hidden but are retained for moderation history.
  • Records can be corrected or hidden when a source changes, a report is disputed, or licensing is incomplete.
  • Questions, takedown requests, and data corrections can be sent to the support email in the footer.