ISO 19115 Metadata Explained

This article opens the new Data Standards & Specifications sub-hub with the foundational metadata standard underlying every spatial-data infrastructure (SDI).

What ISO 19115 is

ISO 19115 is the international standard for geographic-information metadata — the descriptive information explaining what a geographic dataset is, who created it, when, where, how, and under what conditions.

The standard provides a conceptual model and schema for metadata; specific encodings (XML, JSON, RDF) implement the model concretely. The goal: dataset discoverability, interpretation, and appropriate use across organizations and over time.

Structure of the standard

The ISO 19115 family includes multiple parts:

ISO 19115-1:2014

Part 1: Fundamentals — the core conceptual model and abstract schema. Defines the metadata classes and their relationships. Most "ISO 19115" references mean this part.

ISO 19115-2

Part 2: Extensions for acquisition and processing — additional metadata classes specific to imagery and gridded data (satellite imagery, DEMs, remote sensing). Used alongside Part 1 for raster-heavy catalogs.

ISO 19115-3:2016

XML schema implementation — replaced the older ISO 19139 (2007) as the standard XML encoding. Streamlined namespaces, better support for modern features.

Related standards

• ISO 19119: services metadata.
• ISO 19135: registry of codes and items.
• ISO 19139 (deprecated): older XML implementation of ISO 19115 superseded by 19115-3.

What a metadata record contains

ISO 19115 metadata is comprehensive. The major sections:

Identification

• Title: human-readable dataset name.
• Abstract: 1-3 paragraph description.
• Purpose: why the dataset was created.
• Dataset identifier: globally unique identifier (UUID, DOI, etc.).
• Citation: how to cite the dataset in publications.

Topic categories

A controlled vocabulary of 19 topic categories in ISO 19115:

farming, biota, boundaries, climatologyMeteorologyAtmosphere,
economy, elevation, environment, geoscientificInformation,
health, imageryBaseMapsEarthCover, intelligenceMilitary,
inlandWaters, location, oceans, planningCadastre,
society, structure, transportation, utilitiesCommunication

Records select one or more applicable categories. The controlled vocabulary enables consistent catalog browsing.

Spatial extent

• Bounding box: geographic min/max in lat/lon.
• Polygon: precise boundary geometry.
• Named region: gazetteer reference (country, state, etc.).
• Vertical extent: depth/elevation range.

Temporal extent

• Date range: validity period.
• Frequency of update: continuous, daily, weekly, monthly, annually, asNeeded, irregular, notPlanned, unknown.
• Reference dates: creation, publication, revision.

Spatial reference system

CRS identification, typically by EPSG code:

EPSG:4326 (WGS 84 lat/lon)
EPSG:3857 (Web Mercator)
EPSG:25832 (ETRS89 / UTM zone 32N)

Resolution and scale

• Equivalent scale: e.g., 1:25,000 for topographic.
• Spatial resolution: pixel size for raster (e.g., 30 m for SRTM).
• Distance: linear ground equivalent.

Lineage

The processing history:

• Sources: input datasets used.
• Process steps: transformations applied.
• Software: tools used.
• Operator: who performed the processing.

Data quality

Five quality dimensions (ISO 19157):

• Positional accuracy: how close coordinates are to true positions.
• Attribute accuracy: correctness of attribute values.
• Completeness: features omitted or extra.
• Logical consistency: internal rules (e.g., topology, schema).
• Temporal accuracy: timestamp correctness.

Each dimension can be quantified or qualified.

Distribution

• Format: file type and version (Shapefile, GeoTIFF, GeoPackage, etc.).
• Online resource: URL for download, web service, or API.
• Transfer options: HTTP, FTP, OGC services, physical media.

Contact information

• Point of contact: organization or individual responsible.
• Originator: who created the data.
• Custodian: who manages the data now.
• Distributor: who provides access.

Use constraints

• License: CC-BY, ODbL, proprietary, etc.
• Access constraints: restricted, classified, fees.
• Security constraints: confidentiality classification.

Maintenance

• Frequency: how often updated.
• Next scheduled update: when revised version expected.
• Custodian: who maintains it.

Major profiles

Different communities have adopted profiles of ISO 19115 — specific subsets, extensions, or required elements:

INSPIRE (EU Directive 2007/2/EC)

INSPIRE (Infrastructure for Spatial Information in the European Community) is the EU's mandatory geospatial-data infrastructure framework. INSPIRE Metadata Regulation (2008) requires:

• ~30 mandatory ISO 19115 elements (subset of the full standard).
• Specific controlled vocabularies for INSPIRE- defined themes.
• Validation against the INSPIRE schema.
• Publication to the INSPIRE Geoportal.

EU member states must publish metadata for their public-sector geospatial data through INSPIRE- compliant catalogs.

NAP (North American Profile)

NAP (North American Profile of ISO 19115) is the joint US/Canada profile. Adopted by:

• US FGDC (Federal Geographic Data Committee).
• Natural Resources Canada.
• Many state and provincial agencies.

NAP includes US/Canada-specific extensions (citation styles, FGDC controlled vocabularies).

MEDIN

MEDIN (Marine Environmental Data and Information Network) is a UK profile for marine geospatial data. Required by UK marine agencies.

Other national profiles

• CGDI: Canadian Geospatial Data Infrastructure profile.
• ANZLIC: Australia/New Zealand profile.
• DGIWG: Defense Geospatial Information Working Group profile (NATO).

Each profile adapts ISO 19115 to specific national or sectoral needs.

Encodings

ISO 19115 is a conceptual standard; specific encodings implement it concretely.

ISO 19139 XML (legacy)

The original XML implementation (2007). Heavyweight schema with namespaces:

<?xml version="1.0" encoding="UTF-8"?>
<gmd:MD_Metadata
  xmlns:gmd="http://www.isotc211.org/2005/gmd"
  xmlns:gco="http://www.isotc211.org/2005/gco">
  <gmd:fileIdentifier>
    <gco:CharacterString>my-dataset-uuid</gco:CharacterString>
  </gmd:fileIdentifier>
  <gmd:language>
    <gmd:LanguageCode codeList="..." codeListValue="eng"/>
  </gmd:language>
  ...
</gmd:MD_Metadata>

The XML is verbose; a complete record can be hundreds of lines.

ISO 19115-3 XML (current)

The modernized XML implementation (2016). New namespaces (mdb, mri, gex, etc.), better structure, but still heavyweight XML.

DCAT-AP / GeoDCAT-AP (JSON-LD / RDF)

DCAT (Data Catalog Vocabulary) is a W3C standard for general data catalogs. GeoDCAT-AP extends DCAT for geospatial data, providing a JSON-LD / RDF encoding that maps to ISO 19115:

{
  "@type": "Dataset",
  "title": "Dataset name",
  "description": "Dataset description",
  "spatial": {
    "@type": "Location",
    "bbox": [...]
  },
  "temporal": "2024-01-01/2024-12-31",
  "publisher": {
    "@type": "Organization",
    "name": "..."
  },
  "license": "https://creativecommons.org/licenses/by/4.0/"
}

DCAT-AP is increasingly adopted in Europe alongside or instead of ISO 19139 XML. Easier to work with than XML; broader audience than geospatial-specialist tools.

schema.org Dataset

A web-search-engine-oriented encoding using schema.org Dataset type. Enables Google Dataset Search indexing.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Dataset",
  "name": "Dataset name",
  "description": "Description",
  "spatialCoverage": {...},
  "temporalCoverage": "2024",
  "creator": {...}
}
</script>

Less rich than ISO 19115 but enables search- engine discoverability.

Catalog systems

GeoNetwork

GeoNetwork is the dominant open-source ISO 19115 metadata catalog. Features:

• ISO 19115 / 19139 / 19115-3 support.
• Multiple profile support (INSPIRE, NAP, MEDIN, others).
• CSW (Catalog Service for Web) endpoint.
• OAI-PMH harvesting.
• Web GUI for metadata creation/editing.
• User permissions and workflows.

GeoNetwork is OSGeo project and an INSPIRE- compliant solution used by many EU member states.

INSPIRE Geoportal

The EU's aggregated catalog of geospatial metadata from all member states. Single search interface across thousands of national catalogs.

data.gov

The US federal open-data portal includes geospatial metadata, often via NAP profile.

CKAN

CKAN is a popular general-purpose open-data platform (not specifically geospatial). With GeoNetwork integration or DCAT-AP, CKAN can host ISO 19115 metadata.

Other catalogs

• pyCSW: Python implementation of CSW.
• GeoNode: integrated GIS publishing platform using GeoNetwork for metadata.
• PyCSW: pure-Python OGC CSW server.

Discovery via CSW

CSW (Catalog Service for Web) is the OGC standard for discovering metadata records in spatial-data catalogs. A CSW endpoint exposes HTTP-based queries:

GET /csw?service=CSW&request=GetCapabilities
GET /csw?service=CSW&request=GetRecords&typeNames=csw:Record&...
GET /csw?service=CSW&request=GetRecordById&id=...

Clients (GeoNetwork harvesters, browser apps, Python scripts via owslib) query CSW endpoints to discover and retrieve metadata.

OAI-PMH is an older alternative for metadata harvesting (originally for academic publications, adapted for geospatial use).

Implementation tools

Editing

• GeoNetwork editor: web-based form for creating/editing ISO 19115 records.
• QGIS metadata editor: built-in QGIS support for metadata authoring (via QGIS Layer Properties → Metadata).
• pygeometa: Python library for programmatic metadata generation.
• ISO19115Editor: standalone desktop editor.

Validation

• INSPIRE Validator: official EU validator for INSPIRE-conformant metadata.
• GeoNetwork built-in validation: validates against the ISO 19139 / 19115-3 schemas.
• xmllint: command-line XML schema validation.

Conversion

• pygeometa: convert between metadata formats (ISO 19139, ISO 19115-3, DCAT-AP, others).
• GeoDCAT-AP transformation: XSLT transformations.

The metadata problem

Creating good ISO 19115 metadata is labor- intensive:

• Dozens of fields to fill out per dataset.
• Required controlled vocabularies to look up.
• Quality validation and review.
• Maintenance as data changes.

Reality: many published metadata records are minimum-effort placeholders — title, abstract, bounding box, license, distribution URL, and not much else.

Tooling helps: automated metadata generation from data files (GDAL can extract some metadata), templates for common dataset types, organizational metadata standards reducing per-dataset work.

Modern alternatives

STAC

STAC (SpatioTemporal Asset Catalog, 2017+) is a simpler JSON-based catalog format for cloud- native geospatial data. Less rich than ISO 19115 but vastly simpler to implement.

Key differences:

• JSON-based (not XML).
• Item-oriented (each STAC item = single asset
  • metadata).
• Asset-oriented: explicit links to data files.
• Cloud-native: designed for COG, range requests.
• Extensible: STAC extensions for specific asset types.

STAC catalogs from Microsoft Planetary Computer, AWS Earth on AWS, Sentinel Hub demonstrate the appeal of STAC for satellite imagery and EO data.

ISO 19115 and STAC are complementary — ISO 19115 for institutional / regulatory needs, STAC for cloud-native workflows.

DCAT

DCAT (W3C Data Catalog Vocabulary) is broader than geospatial; GeoDCAT-AP extends DCAT for geospatial. JSON-LD / RDF encoding.

INSPIRE endorsed DCAT-AP in 2023 as an alternative to ISO 19139 XML — modernization of EU metadata practice.

schema.org Dataset

Search-engine-oriented metadata embedded in HTML pages via JSON-LD. Enables Google Dataset Search indexing.

Simpler than ISO 19115; broader audience (general web users finding datasets via Google).

Common misconceptions

“ISO 19115 is mandatory.” Required in specific regulatory contexts (INSPIRE for EU public sector; NAP for some US federal geospatial data). Optional for most private- sector and academic use. Many projects use simpler alternatives.

“XML is the only encoding.” Two XML encodings (19139 legacy and 19115-3 modern) plus DCAT-AP JSON-LD and schema.org JSON-LD. Multiple options for different audiences.

“ISO 19115 metadata is searchable like text.” Structured, not just text. The controlled vocabularies, typed fields, and spatial/temporal extents enable structured queries (CSW). Full-text search complements but doesn't replace structured access.

“Once created, metadata doesn't change.” Should be maintained. Dataset updates trigger metadata revisions (new dates, lineage entries, quality info). Many catalogs have stale metadata because maintenance is unfunded.

“ISO 19115 is dead.” Still active. INSPIRE requires it; major government geospatial agencies use it; GeoNetwork continues development. STAC and DCAT are alternatives but not replacements for many use cases.

“CSW is the only way to query.” OAI-PMH for harvesting; OGC API - Records (modern REST replacement for CSW) emerging; some catalogs offer custom REST APIs. Multiple protocols exist.

“Metadata costs more than data.” Often the case for small datasets. Comprehensive ISO 19115 metadata can take days to prepare; the underlying dataset may be quickly generated. Reduce the cost via templates, automation, and minimum-required profiles.

“Topic categories are restrictive.” Multiple categories can be assigned. A landslide-risk dataset might combine geoscientificInformation, environment, planningCadastre, society. The 19 categories cover broad domains.

“DCAT-AP replaces ISO 19115.” Complementary, not replacement. DCAT-AP profiles INSPIRE's ISO 19115 requirements into JSON-LD; the underlying metadata model is still ISO 19115-aligned. The encoding is the choice; the model is shared.

“GeoNetwork is the only choice.” Most common open-source choice but alternatives exist: pyCSW (Python), GeoNode (GeoNetwork-based but broader), commercial catalogs (ESRI Geoportal Server). Many organizations write custom catalog interfaces on top of ISO 19115-compliant metadata stores.

“Spatial extent must be a polygon.” Bounding box (rectangle) is the most common form; polygon and named region also allowed. For simple discovery, bounding box suffices.

“Metadata is for archivists.” For data users. Good metadata answers: is this the dataset I need? Can I use it for my purpose? Who should I contact? What restrictions apply? The audience is anyone considering using the data.

“Schema validation guarantees quality.” Validates structure, not content. A schema-valid metadata record can still have wrong information. Human review of metadata content remains essential.

“ISO 19115 is for governments only.” Used widely beyond government: academic research repositories, environmental NGOs, corporate spatial-data programs, international agencies. The standard is sector-agnostic.

“Modern STAC eliminates ISO 19115 use cases.” STAC excels for satellite imagery / EO data; ISO 19115 better for vector data, complex datasets, regulated environments. Each has its sweet spot; many organizations use both.

“You can't do ISO 19115 in JSON.” DCAT-AP JSON-LD and ISO 19115-3 also have JSON encodings. The XML legacy is real but JSON support has grown substantially in 2020+ implementations.