Search is set by default to the publications collection, so initial results are drawn from UNEP’s official publications. To search across all repository collections, select the full repository from the dropdown next to the search field.

Searches can be made by keyword, title, author, subject or other metadata. For greater precision, enclose an exact phrase in quotation marks – for example, searching for "climate change adaptation" will return results matching that phrase. Results can then be narrowed using filters, or facets, including collection, resource type, subject, author, date, language and other available metadata.

Each record is catalogued using a harmonized metadata model based on Dublin Core, with structured fields for title, authorship, date, language, document type, geographic coverage, Sustainable Development Goals, identifiers, relationships and subject terms.

Subject indexing uses controlled multilingual vocabularies, with GEMET as the primary taxonomy and AGROVOC and UNBIS extending coverage and interoperability. This standardized metadata supports accurate search and filtering, citation, cross-platform interoperability, machine discovery and long-term management of UNEP knowledge resources.

Multilingual access

Different language versions of the same publication are maintained within a single repository record, supporting consistent metadata, linking and discovery across languages.

Artificial intelligence is also used to support multilingual access and translation. In line with United Nations guidance on the responsible use of artificial intelligence, content translated with the assistance of AI is subject to human quality control before publication. UNEP translators review and validate AI-assisted translations before publication.

Scenario Note

Date

Authors

Journal Title

Journal ISSN

Volume Title

Export

Abstract

Citation

URI

Collections

Endorsement

Review

Supplemented By

Referenced By