Skip to content

Well-Known Text Representation of Coordinate Reference Systems

Serialize coordinate reference systems and coordinate operations in a standardized nested text grammar that preserves their defining geodetic, coordinate-system, unit, axis, and operation semantics.

Version
v2 · 2026-09-06 · History
Domain-specific #
3107
Origin domain
geographic information science
Subdomain
spatial referencing standards
Aliases
WKT-CRS, CRS WKT, Well-known text for CRS, WKT2

Core Idea

Well-Known Text for Coordinate Reference Systems (WKT-CRS) is a standardized text grammar for representing the defining content of coordinate reference systems and coordinate operations. A WKT string nests typed elements—such as a geodetic reference frame, ellipsoid, coordinate system, axes, units, conversion method, parameters, identifiers, scope, and extent—so software and people can exchange an interpretable CRS definition without relying only on a short external code.[1]

The grammar is a representation of the ISO 19111 conceptual model, not the model itself. It serializes enough content for a receiver to understand and use common CRS and operation definitions, while deliberately omitting some provenance and database-management metadata. The standard therefore warns that a WKT string is not, by itself, the preferred authoritative store for definitions.[2]

The recognition invariant is declared WKT-CRS version + standard keyword grammar + balanced nested object structure + semantically valid CRS or coordinate-operation components + explicit axes and units + conformance checks beyond successful parsing.

Structural Signature

  • Top-level object: geodetic, projected, vertical, temporal, engineering, compound, or bound CRS, or a coordinate operation.
  • Reference frame/datum: terrestrial, vertical, dynamic, ensemble, or other applicable frame information.
  • Reference surface: ellipsoid and prime meridian where applicable.
  • Coordinate system: type, dimensionality, ordered axes, directions, and units.
  • Conversion/projection: method and named parameters for derived CRSs.
  • Operation context: source CRS, target CRS, method, parameters, accuracy, scope, and extent where supported.
  • Nested grammar: uppercase keywords, delimiters, quoted names, numerals, and ordered child productions.
  • Identifiers: optional authority and code references that complement—not replace—the content.
  • Version profile: WKT1 dialect or a specified WKT2/WKT-CRS revision.
  • Conformance: lexical, syntactic, structural, and semantic requirements.
  • Round-trip obligation: parse and re-emit without silently changing axis, unit, datum, or operation meaning.

What It Is Not

It is not Well-Known Text for geometry. Geometry WKT encodes shapes such as points and polygons; WKT-CRS encodes spatial reference definitions and coordinate operations. A valid POINT(...) string is not a CRS.

It is not an EPSG code, PROJ pipeline, JSON schema, or XML geography markup, although software may convert among those representations. It is not a coordinate transformation algorithm: an operation element names methods and parameters, while the mathematics comes from the referenced method definition. It is not “just a label”; axis order, unit scale, datum realization, and operation direction can change numerical meaning.

Scope of Application

WKT-CRS appears in geospatial files, databases, GIS APIs, map services, metadata, interchange formats, and coordinate-transformation software. It represents geographic/geodetic, geocentric, projected, vertical, temporal, engineering, compound, and bound reference systems, plus many conversions and transformations.[3]

WKT2 adds model fidelity and consistency absent from fragmented WKT1 practice, including stronger axis/unit semantics and richer dynamic, ensemble, temporal, scope, extent, and operation structures. Actual interoperability still depends on the exact revision and implementation profile; “supports WKT” is insufficient when one product emits an ESRI-flavored WKT1 dialect and another expects standards-conforming WKT2.

Clarity

Parsing proves only that text fits a grammar. Semantic validation asks whether required children exist, dimensions and axes agree, units suit quantity types, identifiers match content, and operation parameters conform to the named method. Operational validation asks whether downstream software preserves those commitments.

Names are human-readable and may be redundant; identifiers enable registry comparison but do not license ignoring conflicting embedded values. The safest comparison is not byte equality: equivalent strings can differ in whitespace, optional elements, names, order where permitted, or identifier use. Conversely, superficially similar strings can differ materially in axis order or datum realization.[4]

Manages Complexity

The nested grammar packages a multi-part geodetic object into one portable string. Typed keywords expose component roles, identifiers connect to registries, and a formal syntax supports generic parsers rather than product-specific field extraction. Version markers and conformance tests make compatibility assumptions auditable.

Serialization moves complexity into schema fidelity and conversion. Lossy WKT1↔WKT2 mapping, omitted metadata, unsupported dynamic CRSs, renamed parameters, implicit units, and axis normalization can change meaning. Robust systems retain source strings, parsed object models, warnings, and authority provenance.

Abstract Reasoning

  1. Identify whether the object is a CRS, operation, or geometry and choose the corresponding standard.
  2. Declare the WKT-CRS revision and any implementation profile.
  3. Build or obtain a CRS object conforming to the ISO 19111 conceptual model.
  4. Serialize every required frame, ellipsoid, coordinate-system, axis, unit, conversion, and parameter element.
  5. Add identifiers, scope, extent, accuracy, and remarks where applicable.
  6. Parse with an independent conforming implementation.
  7. Validate semantic constraints, not only delimiter balance.
  8. Compare the parsed object with an authoritative registry definition where one is claimed.
  9. Test coordinate behavior, version conversion, and parse–format–parse round trips.
  10. Report any loss, dialect substitution, or unsupported element.

Knowledge Transfer

The transferable principle is to serialize a rich conceptual object with typed, nested syntax and explicit conformance rules. The proposed immediate parent is Specification Language because WKT-CRS is a formal domain language that constrains representations and interoperable interpretation.

The analogy to general markup is incomplete: geodetic component semantics and coordinate-operation obligations determine whether two strings mean the same thing.

Examples

Geographic CRS. A GEODCRS object includes a reference frame/datum, ellipsoid, ellipsoidal coordinate system, latitude and longitude axes with order, angular units, and perhaps an EPSG identifier.

Projected CRS. A PROJCRS nests its base geographic CRS, a conversion method and parameters, Cartesian axes, and linear units.

Non-example. POINT (10 20) is geometry WKT; without a CRS association, it does not say whether values are longitude/latitude, easting/northing, degrees, metres, or something else.

Structural Tensions

  • Human readability versus exhaustive metadata.
  • Self-contained content versus registry identifiers.
  • Version evolution versus backward compatibility.
  • Flexible formatting versus semantic equivalence.
  • Strict conformance versus installed dialects.
  • Round-trip stability versus normalization and information loss.

Structural–Framed Character

Keywords, grammar productions, nesting, required roles, and conformance are structural. Version choice, registry policy, naming, optional metadata, software tolerance, and migration strategy are implementation-framed.

Structural Core vs. Domain Accent

The portable core is formal serialization of a typed conceptual object. Datums, reference frames, ellipsoids, axes, units, projections, operation methods, and geospatial standards are constitutive domain accent, so the abstraction is domain-specific.

Specification Language is the proposed immediate parent. Coordinate Reference System supplies the represented object; Parsing, Serialization, Identifier, and Unit support important roles. World Geographic Reference System is an instance/family, not coverage.

The prospective queue contains one strict edge to domain_specific:specification_language. No live DAG mutation is authorized.

Relationships to Other Abstractions

Local relationship map for Well-Known Text Representation of Coordinate Reference SystemsParents appear above the current abstraction, mutual partners to the right, and children below. Node labels state whether each abstraction is prime or domain-specific; colors identify relation types.Well-Known Text Repr…DOMAINDomain-specific abstraction: Specification language — is a kind ofSpecificationlanguageDOMAIN

Current abstraction Well-Known Text Representation of Coordinate Reference Systems Domain-specific

Parents (1) — more general patterns this builds on

  • Well-Known Text Representation of Coordinate Reference Systems is a kind of Specification language Domain-specific

    Specification Language is the proposed immediate parent.

Hierarchy paths (3) — routes to 3 parentless roots

Neighborhood in Abstraction Space

Well-Known Text Representation of Coordinate Reference Systems sits in a sparse region of the domain-specific corpus (94th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (1565 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With

  • Geometry WKT for points, lines, and polygons.
  • An EPSG identifier or registry record.
  • PROJ strings or executable transformation pipelines.
  • Generic text markup lacking CRS semantics.
  • WKT1 and WKT2 treated as losslessly interchangeable.
  • Syntactic parse success treated as semantic conformance.

References

[1] Open Geospatial Consortium, Geographic Information—Well-Known Text Representation of Coordinate Reference Systems, OGC 18-010r11, version 2.1.11, 2025, official standard. registry ↩a ↩b

[2] International Organization for Standardization, ISO 19162:2019 Geographic Information—Well-Known Text Representation of Coordinate Reference Systems, 2019, standard record. registry ↩

[3] International Organization for Standardization, ISO 19111:2019 Geographic Information—Referencing by Coordinates, 2019. registry ↩

[4] Open Geospatial Consortium, Geographic Information—Well-Known Text Representation of Coordinate Reference Systems, OGC 18-010r7, version 2.0.6, 2019, HTML edition. registry ↩