Provenance is the account of where a record came from: who issued it, when, by what method, from which sources and through which transformations it passed before reaching the reader. Without it a record cannot be checked, weighed or corrected, only trusted or ignored. This page sets out how provenance is captured and expressed, why lineage chains break, and what the people who rely on registry records need to see before they do.
GCRI · Established · service level R0
Record provenance is the account of where a record came from: who issued it, when, by what method, from which sources and through which transformations it passed. Without it a record cannot be checked, weighed or corrected. The W3C PROV model gives a standard way to express it, and lineage often breaks at hand-offs between systems.
For record provenance, in GCRI's spatial-intelligence reading, the map opens on true-colour imagery, night lights and population density with true colour from NASA GIBS, framed on World; every other NASA layer, live feed and forecast waits in the layer list.
The interactive map needs JavaScript; the layer list and sources are below.
Layers 3
People and hazards, for this page
Keyboard, with the map focused
Hold Ctrl or ⌘ and scroll to zoom
Latest available
NASA GIBS / EOSDIS · ASTER GDEM (METI, NASA) · Blue Marble (NASA) · © OpenStreetMap contributors (ODbL) · Natural Earth
Loading the latest NASA imagery.
Drag to move. Ctrl or ⌘ and scroll, pinch or the buttons to zoom; hover for the values under the pointer; reset returns to the region.
For record provenance, in GCRI's spatial-intelligence reading, the map opens on true-colour imagery, night lights and population density with true colour from NASA GIBS, framed on World; every other NASA layer, live feed and forecast waits in the layer list.Satellite and model products as NASA publishes them, each with its own date; indicators, not assessments of any place; claim ceiling INTEGRATED.
Layers and dates
Issuer, date, method and source lineage. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.
Records pass between spreadsheets, databases, aggregators and exports. Each step can drop the issuer, the source reference or the method note. Once the chain is cut, later copies look authoritative while nothing connects them to what they were derived from.
Organisation names change, merge and are spelt many ways, and people share names. Without persistent identifiers for issuers, two bodies are recorded as one or one as two, and responsibility for an assertion can no longer be established.
The method behind a field changes, through a new coding rule, a new instrument or a revised definition, while the field name stays the same. A series then joins incompatible values, and no reader can see where the join is.
An unsourced claim is copied by an aggregator, quoted in a report and indexed by search engines. Each reuse adds apparent authority and removes context, until the claim is cited back to its origin as if it were independent confirmation.
Each term as this page uses it. Where a standard-setting body defines the term, the wording follows that body.
| Term | Definition |
|---|---|
| Data lineage | The traceable path of data from its origin through each transformation to its current form. |
| W3C PROV | A family of W3C specifications for representing provenance as entities, activities and agents. |
| Content hash | A fixed-length fingerprint computed from a record's content, which changes if any byte changes. |
Each indicator carries its unit, its cadence, the kind of source behind it and the evidence label a value would carry. No values are shown: a programme supplies them under its own authority.
| Indicator | Unit | Cadence | Source type | Label |
|---|---|---|---|---|
| Records with complete issuer, date and method | share of records | monthly | Schema validation | OBSERVED |
| Lineage depth to a primary source | number of derivation steps | per record | Provenance graph | OBSERVED |
| Unresolved source references | count | weekly | Resolver checks | OBSERVED |
| Transformations without a recorded agent | count | monthly | Provenance graph audit | OBSERVED |
| Records with content hashes or signatures | share of records | monthly | Integrity check | OBSERVED |
| Issuers with a persistent identifier | share of issuers | quarterly | Identifier reconciliation | OBSERVED |
| Provenance inferred rather than declared | share of provenance statements | quarterly | Curation review | ESTIMATED |
The topic as a computational graph: named sources feed the methods, an evidence review gates the outputs, and the decisions stay with the authorities that hold them.
Fig. 02Nodes carry their type, two metric rows and a state; connectors show what feeds what; dashed frames name the owners; the timeline is the cadence. Design of the pipeline, not a record of a run.
Named public datasets and institutional records, each with its licence class. Sources by agreement are used only under their terms and never republished.
| Source | What it gives | Licence class |
|---|---|---|
| W3C PROV-O and PROV-DM | The standard vocabulary of entities, activities and agents for expressing provenance graphs | W3C document licence, free to implement |
| DataCite Metadata Schema | Creator, contributor, date, relation and related-identifier fields for research outputs | CC0 for metadata |
| ROR (Research Organization Registry) | Persistent identifiers for research organisations, with name variants, relationships and succession | CC0, no restrictions |
| ORCID public data file | Persistent identifiers for researchers, with linked works, affiliations and funding | CC0 |
| GLEIF Legal Entity Identifier data | Identifiers, legal names and parent relationships for legal entities | CC0, no restrictions |
| PREMIS Data Dictionary | Preservation metadata for events, agents and rights on digital objects | Free to use, maintained by the Library of Congress |
| RO-Crate | A specification for packaging research data with its provenance in JSON-LD | Open specification, attribution |
Nexus Registry prepares evidence. The decisions below are taken by the authorities named, under their own rules, and never by the platform.
| Who decides | The question | Evidence this page serves |
|---|---|---|
| Data stewards | Whether a record's lineage is complete enough to publish or must be held for curation | Provenance completeness, unresolved references and transformations without a recorded agent |
| Journal and report editors | Whether a cited registry record can support a claim in a manuscript | Issuer identity, method, date and the depth of lineage to a primary source |
| Research integrity officers | Whether an assertion was altered, misattributed or fabricated | Time-stamped provenance graph, content hashes and the full edit history |
| National statistics offices | Whether administrative or registry records can enter official statistics | Source lineage, method stability over time and documented transformations |
Lineage is usually lost at boundaries: when a dataset is exported to a spreadsheet, when a figure is retyped into a report, when two sources are merged without recording which value came from where. Each step can be innocent, yet together they produce numbers whose origin no one can reconstruct. Archival science calls the related principle respect for provenance: records keep their link to the body that created them and to their original order.
Modern provenance models describe three kinds of thing: entities such as datasets and reports, activities such as a model run or a review, and agents such as people, institutions and software. The W3C PROV family of specifications expresses these relationships in a form machines can read, and the Registry's lineage design is intended to follow it, as described on provenance and attribution.
Each record names who issued it and under what authority, so a reader knows whose word it is.
Transformations, model runs and reviews are recorded with their method version and date.
Cryptographic hashes can show that a file has not changed since it was recorded, a technique from cybersecurity and cryptography.
Provenance shows where a record came from, not whether its content is right. Lineage is only as complete as the steps that were captured, and self-declared provenance can be false. Restricted sources can be named but not inspected.
The current service level is R0: information and enquiries. An enquiry receives a reference and names the next step; applications, payments, compute allocations and registrations open only when their team, terms and review capacity exist.