Data and interoperability standards
Data Interoperability Standards for Risk: FAIR, OGC, ISO 19115
Data interoperability standards let hazard, exposure, vulnerability and loss data move between systems without manual rework. For risk information the core set is FAIR, OGC APIs, ISO 19115, W3C DCAT and PROV-O, schema.org, STAC and the Common Alerting Protocol.
Published Topic: Data and interoperability standards
Data interoperability standards are the shared formats, vocabularies, metadata schemas and interfaces that let risk data move between systems and institutions without manual rework. For hazard, exposure, vulnerability and loss information, the core set combines the FAIR principles [1], Open Geospatial Consortium (OGC) APIs [6], ISO 19115 metadata [4], W3C DCAT and PROV-O [2][7], schema.org [3], STAC [8] and the Common Alerting Protocol [12].
Key takeaways
- FAIR (Findable, Accessible, Interoperable, Reusable) sets goals; standards such as ISO 19115, DCAT and OGC APIs implement them [1][2][4][6].
- Geospatial metadata uses ISO 19115-1:2014, now under revision; web catalogues use DCAT 3, a W3C Recommendation since August 2024 [2][4][5].
- Provenance (PROV-O) and identifiers are what make risk data citable and auditable [7].
- In 2025, 119 countries reported multi-hazard early warning systems; warnings travel further when they use CAP [11][12].
- The Sendai Framework Monitor collects 38 indicators from up to 176 countries, a large interoperability exercise in its own right [9][10].
What is data interoperability and why does it matter for risk?
Interoperability means that data from one system can be found, read, understood and used by another with no special effort. It operates at several levels: technical (protocols and formats), syntactic (schemas), semantic (shared meaning of terms and codes) and organisational (licences, roles and agreements). Risk information needs all four, because a single assessment combines meteorological forecasts, census and building data, infrastructure networks and historical losses, each held by a different institution.
The FAIR principles, published in Scientific Data in 2016, framed the goal for research data and stressed machine-actionability: that software agents, not only humans, can find and use data [1]. The interoperability principle I1 asks that metadata and data "use a formal, accessible, shared, and broadly applicable language for knowledge representation" [1]. Risk agencies adopt FAIR because analysis now runs through automated pipelines and models rather than spreadsheets passed by email.
What are the FAIR principles in practice?
| FAIR goal | What it requires (summary) | Typical implementing standard |
|---|---|---|
| Findable | Persistent identifiers and rich, indexed metadata [1] | DOI or other persistent identifiers; DCAT catalogues [2]; schema.org Dataset [3] |
| Accessible | Retrieval by identifier using an open, standard protocol; metadata available even if data are restricted [1] | HTTP; OGC API Features [6]; STAC API [8] |
| Interoperable | Formal, shared vocabularies and qualified references [1] | ISO 19115-1 [4]; controlled vocabularies; RDF |
| Reusable | Clear licence, detailed provenance, community standards [1] | Licence fields; PROV-O [7] |
FAIR does not mean open. Sensitive exposure data, such as critical infrastructure locations or personal data, can be FAIR with restricted access, provided the metadata are discoverable and the access conditions are explicit [1].
Which standards make risk data interoperable?
Geospatial metadata: ISO 19115. ISO 19115-1:2014 defines the schema for describing geographic information: identification, extent, quality, spatial and temporal reference, lineage and distribution [4]. It was amended in 2018 and 2020, and a second edition is in development as ISO/CD 19115-1 [4][5].
Geospatial access: OGC APIs. OGC API Features Part 1: Core, version 1.0 of October 2019, defines web building blocks to "create, modify and query features on the Web" and was taken forward in ISO as ISO 19168-1 [6]. The OGC API family replaces older service interfaces with resource-oriented web APIs that developers can use without specialist geospatial software.
Earth observation catalogues: STAC. The SpatioTemporal Asset Catalog (STAC) specification provides a common language to describe geospatial information, organised into Item, Catalog, Collection and API parts [8]. It is widely used for satellite imagery and derived hazard layers.
Web catalogues: W3C DCAT. DCAT is an RDF vocabulary designed to make data catalogues on the Web interoperable; version 3 became a W3C Recommendation on 22 August 2024 [2]. National open data portals commonly expose DCAT.
Search discovery: schema.org. The schema.org Dataset type, "a body of structured information describing some topic(s) of interest", gives properties such as distribution, spatialCoverage, temporalCoverage, licence and variableMeasured that search engines read [3].
Provenance: W3C PROV-O. PROV-O, a W3C Recommendation of 30 April 2013, expresses provenance through entities, activities and agents [7]. For risk products, provenance answers which data, model version and analyst produced a figure.
Warnings: CAP. The Common Alerting Protocol 1.2, an OASIS Standard of July 2010, provides an "open, non-proprietary digital message format for all types of alerts and notifications" [12]; ITU-T carries it as Recommendation X.1303 [13].
How does interoperability support early warning?
Early warning is a chain. The UN Early Warnings for All initiative, launched in 2022 with the goal of protecting everyone by early warning systems by the end of 2027, organises it in four pillars: disaster risk knowledge and management; detection, observation, monitoring, analysis and forecasting; warning dissemination and communication; and preparedness and response [14]. Each hand-off between pillars is a data exchange.
The 2025 Global Status of Multi-Hazard Early Warning Systems report found that 119 countries, 60% of all countries, report having a multi-hazard early warning system, and that only 43% of small island developing States do [11]. Gaps are rarely only about sensors. Common failure points include:
- Hazard data that cannot be joined with exposure data, because of different coordinate systems, administrative boundaries or identifiers.
- Forecasts without impact information, because vulnerability and exposure datasets are not discoverable or licensed for operational use.
- Warnings re-typed for each channel, instead of a single CAP message distributed to broadcast, mobile and web [12].
- No feedback loop, because loss and impact data are not recorded in a structure that can be compared with the warning issued.
Interoperability standards address each point: shared metadata and APIs for the first two, CAP for the third, and loss databases aligned with Sendai indicators for the fourth.
What do the Sendai Framework monitoring data show?
The Sendai Framework for Disaster Risk Reduction 2015-2030 is monitored through the Sendai Framework Monitor, in which countries report on seven global targets through 38 indicators [9]. By October 2025, 176 countries had reported on the global targets; 87 reported across all seven targets and 84 reported data disaggregated by sex, age and disability [10]. Reported data also show 128 countries with multi-hazard early warning systems, against 55 in 2015 [10].
The figure differs from the 119 in the WMO and UNDRR status report [11] because the two use different reporting bases; cite each with its source. The disaggregation gap, fewer than half of reporting countries, is itself an interoperability problem: loss records often lack the common fields that would allow consistent breakdown.
Why is semantic interoperability the hardest layer?
Formats and APIs are the easy part: a file can be parsed and a service can be called. The harder question is whether two datasets mean the same thing. A "damaged building" in one loss database may be "affected" in another; "population exposed" may count residents in one model and daytime occupants in another. Shared formats do not resolve these differences.
Semantic interoperability requires controlled vocabularies with published definitions, code lists that are versioned, and crosswalks that record where terms match only partly. Metadata standards provide the slots for this information [4], and provenance records show which definition a producer used [7]. Without them, aggregated indicators can look consistent while combining incompatible measures.
How do the main standards compare?
| ستاندارد | Layer | Best for | Limitation |
|---|---|---|---|
| ISO 19115-1 [4] | Metadata | Detailed geospatial description, quality, lineage | Heavy for simple web discovery |
| DCAT 3 [2] | کاتالۆگ | Cross-portal dataset discovery | Limited spatial detail on its own |
| schema.org Dataset [3] | Web discovery | Search engine visibility | Not a full metadata record |
| STAC [8] | Catalogue and API | Earth observation assets | Focused on spatio-temporal assets |
| OGC API Features [6] | دەستگەیشتن | Feature data over web APIs | Raster and coverage data need other OGC APIs |
| PROV-O [7] | ڕەسەنایەتی | Lineage and accountability | Requires disciplined capture |
| CAP 1.2 [12] | Messaging | Alerts across channels | Does not carry full impact data |
How to make a risk dataset interoperable: step by step
- Assign a persistent identifier to the dataset and each version [1].
- Write a full metadata record in ISO 19115-1 for geospatial data [4], including extent, resolution, quality and lineage.
- Publish a catalogue entry in DCAT [2] and add schema.org Dataset markup to the landing page [3].
- Record provenance with PROV-O terms: source data, processing activity, model version and responsible agent [7].
- Expose standard access, such as OGC API Features or STAC, rather than bespoke downloads [6][8].
- State the licence and access conditions explicitly, including for restricted data [1].
- Align codes and boundaries with those used in national reporting, so data can feed Sendai indicators [9].
- Test with a second system before declaring the dataset interoperable.
Common mistakes
- Equating open with interoperable. A CSV on a website is open but not machine-interpretable without documented schema and meaning.
- Metadata written once and never updated. Versioned data need versioned metadata.
- Losing provenance in processing. Derived risk layers that drop their sources cannot be audited.
- Bespoke identifiers for places. Non-standard boundaries break joins with census and loss data.
- Ignoring licences. Unclear terms stop operational reuse in warning systems.
- Tables locked in PDFs. Statistics published only as document tables cannot feed models or dashboards without re-keying.
- No owner for the standard profile. When nobody maintains the agreed schema and code lists, each new partner adds its own variant and interoperability erodes.
Interoperability is therefore as much an organisational commitment as a technical one. Agreements on who maintains vocabularies, how versions are announced and how partners are notified of changes matter as much as the choice of format.
How Nexus Standards works on this
Nexus Standards publishes open standards for risk information on nexustandards.com, where their text, status and versions govern; every standard is at present a proposal under legal review, and adoption is decided by each jurisdiction [16]. Its NS-3b dataset and product metadata profile profiles STAC, DCAT and PROV-O rather than inventing a new schema [15], and the UNI catalogue and resolver gives each artefact an identifier that can be cited and resolved.
The programme's data and evidence and early warning pages describe the work in progress. NS-14 addresses conformance and crosswalks to external standards. A validator report under these standards is evidence for review, not a conformance claim, and Nexus Standards does not certify datasets or systems. Drafting is open to any person without membership or fee [16].
Further reading
- Wilkinson et al. (2016) on FAIR [1].
- W3C DCAT 3 [2] and PROV-O [7]; OGC API Features [6]; STAC [8].
- UNDRR Sendai Framework Monitor data [10] and the 2025 global status report on early warning [11].
Written by the Nexus Standards knowledge team. Last reviewed 2026-10-30. Corrections are welcome through the knowledge base report link on this page.
زاراوە سەرەکییەکان
- Interoperability
- The ability of data and systems to be exchanged and used together without special effort, because they share formats, vocabularies and access protocols. [1]
- FAIR principles
- Guiding principles, published in 2016, that make data and metadata Findable, Accessible, Interoperable and Reusable by both people and machines. [1]
- Metadata
- Structured information that describes a dataset, such as its content, extent, quality, lineage, licence and access; standardised for geographic data by ISO 19115-1. [4]
- DCAT
- A W3C RDF vocabulary designed to make data catalogues published on the Web interoperable; version 3 became a W3C Recommendation on 22 August 2024. [2]
- ڕەسەنایەتی
- Information about the entities, activities and agents involved in producing data, expressed in a standard form by the W3C PROV Ontology (PROV-O). [7]
- STAC
- A specification that gives a common language to describe geospatial information, mainly Earth observation assets, so it can be indexed and discovered. [8]
- Common Alerting Protocol (CAP)
- An open, non-proprietary digital message format for all types of alerts and notifications, standardised by OASIS (version 1.2, 2010) and by ITU-T as X.1303. [12]
Frequently asked questions
What are data interoperability standards?
They are agreed formats, vocabularies, metadata schemas and interfaces that let different systems exchange and use data without manual conversion. For risk data, typical examples are OGC APIs for geospatial access, ISO 19115 for metadata, W3C DCAT for catalogues and CAP for warnings.
What are the FAIR principles?
FAIR stands for Findable, Accessible, Interoperable and Reusable. Published by Wilkinson and colleagues in Scientific Data in 2016, the principles stress machine-actionability, so that software, not only people, can find and use data.
What is the difference between ISO 19115 and DCAT?
ISO 19115-1 is a detailed metadata standard for geographic information, including spatial extent, quality and lineage. DCAT is a lighter W3C vocabulary for describing datasets in web catalogues. Many portals keep ISO 19115 records and publish DCAT views for discovery.
Why does interoperability matter for early warning?
Early warning chains combine observations, forecasts, exposure data and alert dissemination run by different institutions. Without shared formats and identifiers, data arrive late or are re-keyed by hand. CAP, for example, lets one warning be distributed across many channels.
What data does the Sendai Framework Monitor collect?
Countries report against 38 indicators for the seven global targets of the Sendai Framework, covering disaster mortality, affected people, economic losses, infrastructure damage, strategies, cooperation and early warning. By October 2025, 176 countries had reported through it.
Is schema.org enough to make a dataset interoperable?
No. Schema.org Dataset markup improves web discovery, but it does not describe quality, lineage or spatial reference in the depth that risk analysis needs. It works best as a discovery layer on top of a richer metadata record.
سەرچاوەکان
- Scientific Data (Nature Portfolio). (2016). The FAIR Guiding Principles for scientific data management and stewardship (Wilkinson et al., Scientific Data 3:160018, doi:10.1038/sdata.2016.18).
- World Wide Web Consortium (W3C). (2024). Data Catalog Vocabulary (DCAT) — Version 3.
- Schema.org. (2026). Dataset (schema.org type).
- International Organization for Standardization. (2014). ISO 19115-1:2014 Geographic information — Metadata — Part 1: Fundamentals.
- International Organization for Standardization. (2026). ISO/CD 19115-1 Geographic information — Metadata — Part 1: Fundamentals (under development).
- Open Geospatial Consortium. (2019). OGC API — Features — Part 1: Core (17-069r3).
- World Wide Web Consortium (W3C). (2013). PROV-O: The PROV Ontology.
- STAC community. (2026). STAC: SpatioTemporal Asset Catalogs.
- United Nations Office for Disaster Risk Reduction (UNDRR). (2026). Monitoring the Sendai Framework.
- United Nations Office for Disaster Risk Reduction (UNDRR). (2026). Sendai Framework Monitor data and analytics.
- World Meteorological Organization and UNDRR. (2025). Global Status of Multi-Hazard Early Warning Systems 2025.
- OASIS. (2010). Common Alerting Protocol Version 1.2, OASIS Standard.
- International Telecommunication Union (ITU-T). (2012). X.1303: Common alerting protocol (CAP 1.2) — work item.
- United Nations. (2026). Early Warnings for All.
- Nexus Standards. (2026). NS-3b: Dataset and product metadata profile.
- Nexus Standards. (2026). Governance of the Nexus Standards.