CityGML: An Open Standard for Interoperable 3D City Models

The rapid expansion of 3D visualization in web mapping platforms has drawn significant attention to digital representations of cities. With major technology companies introducing immersive urban views, the public appetite for interactive 3D environments is unmistakable. Yet most widely available 3D city models focus primarily on visual realism—delivering geometric representations that look compelling but offer limited analytical value.

These graphic-only models generally lack embedded semantic meaning and topological structure. As a result, they function well for visualization but fall short when applied to thematic queries, simulation, spatial analysis, or data mining. Another persistent obstacle is interoperability. Although open standards for web-based 3D geometry and geospatial exchange exist—such as X3D and Geography Markup Language 3 (GML3)—they remain relatively young and have not yet achieved universal adoption. While both rely on XML foundations that technically allow integration, harmonized implementation has lagged behind.

Where semantic and topological modeling are attempted, the absence of widely accepted application schemas creates further fragmentation. Without shared definitions of core entities, attributes, and relationships, organizations struggle to exchange and maintain consistent 3D city datasets across platforms.

Beyond Visualization: The Need for Standards

The interest demonstrated by leading technology firms underscores the demand for 3D city visualization. However, visual representation alone is only one layer of potential value. Urban planning, disaster response, infrastructure management, homeland security, and renewable energy planning all require robust, interoperable 3D data models capable of supporting analysis—not merely display.

Many of these applications could operate within a unified framework if a comprehensive standard existed. Particularly in high-impact domains such as emergency management and sustainable infrastructure development, immediate integration of 3D models created for other purposes would dramatically improve efficiency and coordination. Achieving this vision requires an open, harmonized 3D city model standard.

CityGML was developed to meet this need.

What CityGML Provides

CityGML is a standardized information model designed to represent 3D urban environments in a structured, interoperable manner. It defines object classes and relationships for key topographic elements within cities and regions, addressing geometric, topological, semantic, and visual characteristics.

The concept of “city” in CityGML extends beyond buildings to include terrain surfaces, vegetation, water bodies, transportation infrastructure, and urban furnishings. The model incorporates classification hierarchies, object aggregations, spatial relationships, and detailed appearance attributes. By embedding thematic information within the 3D structure, CityGML enables far more than visual exchange—it supports advanced analytical workflows.

CityGML is implemented as an application schema of GML3, the international standard for spatial data exchange developed within the Open Geospatial Consortium (OGC) and ISO TC211. While GML3 and OGC’s Web Feature Service (WFS) provide the syntactic foundation for sharing complex spatial datasets, CityGML defines the semantic profile specific to 3D urban modeling.

Integration with Other Open Standards

CityGML has been designed to complement rather than replace other open standards. Rendering of CityGML datasets can be achieved through established computer graphics formats such as VRML, GeoVRML, X3D, or Universal 3D (U3D). Its development has involved coordination with European spatial research organizations, including EuroSDR, to ensure compatibility and broad acceptance.

Within the OGC framework, CityGML discussions take place in the CAD/GIS/3D Interoperability Working Group. The specification has been positioned for submission to the OGC Technical Committee as a candidate OpenGIS standard, with the intent of free and open use upon adoption.

The initiative originated in 2002 through the Special Interest Group 3D (SIG 3D) under Germany’s Geodata Infrastructure North-Rhine Westphalia (GDI NRW). SIG 3D includes municipalities, companies, and research institutions dedicated to developing interoperable 3D models. A related proposal, the Web 3D Service (W3DS), has also emerged from this collaboration.

Supporters of CityGML include municipalities such as Berlin, Hamburg, Cologne, Düsseldorf, Recklinghausen, and Leverkusen; national mapping authorities like Ordnance Survey Great Britain; technology firms; and universities including Bonn, Potsdam, Dortmund, Stuttgart University of Applied Sciences, the Helmholtz Research Centre Karlsruhe, and the Fraunhofer Institute for Graphics Research.

Core Features of CityGML

CityGML integrates graphical representation with semantic and taxonomic information. It distinguishes five Levels of Detail (LoD), enabling increasing geometric and thematic complexity as needed. Objects may be represented at multiple LoDs within the same dataset, supporting flexible use across applications.

Future development may address dynamic objects, temporal aspects, historical states, and possibly Constructive Solid Geometry modeling principles. These extensions would further enhance CityGML’s capacity to represent evolving urban systems.

Tools and Resources

The CityGML project provides UML diagrams, XML schema definitions, example datasets, and viewer applications. Open-source tools such as Aristoteles—developed by the University of Bonn—allow visualization of GML3 datasets, including CityGML. Another application, LandXPlorer, enables interactive viewing and exchange of CityGML-based models, with free use available under defined conditions.

CityGML in OGC OWS-4

The fourth OGC Web Services testbed (OWS-4) serves as a proving ground for CityGML interoperability. The test scenario involves planning a temporary field hospital at an airport site in an emergency context. Participants integrate CAD data, GIS terrain information, orthophotos, transportation access routes, and aeronautical datasets through standards-compliant services.

A CAD Model Server provides building and infrastructure data. A Web Feature Service delivers terrain and site data encoded in GML, along with aeronautical information for helicopter operations. Through coordinated workflows, the testbed evaluates how CityGML supports cross-domain integration and operational decision-making.

Conclusion

As computational power, storage capacity, and network bandwidth continue to expand, the concept of a shared virtual urban environment becomes increasingly feasible. Industries ranging from architecture and engineering to public safety and urban governance stand to benefit from interoperable 3D city models.

CityGML represents a structured effort to harmonize geometric modeling, semantic structure, and topological consistency within an open standards ecosystem. Now entering rigorous testing phases, it invites engagement from stakeholders in geospatial data exchange, CAD/GIS integration, building information modeling, and workflow optimization.

If widely adopted, CityGML could provide the foundation for truly interoperable 3D city models—transforming digital urban representation from isolated visualizations into shared analytical infrastructure.