By Steve Rizer
The Open Geospatial Consortium’s (OGC) membership has adopted a new version of the OGC City Geography Markup Language (CityGML) Encoding Standard. CityGML is a community-defined information model and XML-based encoding for the representation, storage, and exchange of virtual 3D city and landscape models. CityGML is designed to provide a standard model and mechanism for describing 3D objects regarding their geometry, topology, semantics, and appearance and defines five different levels of detail. CityGML is considered highly scalable, and datasets can include different urban entities supporting the general trend toward modeling individual buildings, whole sites, districts, cities, regions, and countries.
Compared to Version 1.0, CityGML Version 2.0 defines additional feature types and new feature properties, including new thematic modules for tunnels and bridges; the ability to model footprint and roof edge representations for buildings in order to allow users to derive 3D models from existing 2D building data; and generic attribute sets, allowing users a more powerful way to customize CityGML without the need for additional coding, according to OGC.
OGC stressed that CityGML provides more than 3D content for visualization by diverse applications. “It allows users to share virtual 3D city and landscape models for sophisticated analysis and display tasks in application domains such as environmental simulations, energy demand estimations, city life-cycle management, urban facility management, real estate appraisal, disaster management, pedestrian navigation, robotics, urban data mining, and location-based marketing. Because CityGML is based on the OGC Geography Markup Language Encoding Standard, it can be used with the whole family of OGC web services for data accessing, processing, and cataloging.
“CityGML has been implemented in many software solutions and is in use in many projects around the world,” OGC stated. “In National Spatial Data Infrastructure programs in the Netherlands, Germany, France, Malaysia, Abu Dhabi, and other countries, CityGML provides an important platform for the transition from 2D to 3D data. It also plays an important role in bridging urban information models with building information models to improve interoperability among information systems used in the design, construction, ownership, and operation of buildings and capital projects.”
OGC is an international consortium of more than 435 companies, government agencies, research organizations, and universities participating in a consensus process to develop publicly available geospatial standards. OGC Standards support interoperable solutions that “geo-enable” the web, wireless and location-based services, and mainstream IT.
OGC Official Provides Additional Information
In an email interview with Construction Project Controls and BIM Report (CPC/BIM), Lance McKee, a senior staff writer for OGC, provided the following additional details:
CPC/BIM: Which types of professionals are expected to use CityGML Version 2.0? Architects? Local government officials?
McKee: Basically the same as CityGML 1.0 -- local government officials and urban and/or environmental planners. But now, also users from new application fields like energy planning, city life-cycle management, and smart city initiatives. Architects ‘should’ use CityGML (especially to embed their designs into a neighborhood consisting of well-defined entities), but we expect that it will take a long time for the AEC industry to widely adopt workflows based on open standards.
CPC/BIM: What statistics, if any, can you provide regarding the number of people/organizations that have used at least one version of CityGML? Any similar estimates for the number of projects where CityGML was used?
McKee: Very difficult to answer…. A growing number of cities are using the OGC City Geography Markup Language (CityGML) Encoding Standard to build 3D city models on top of their 2D data, and these models will increasingly be used for energy analysis. CityGML defines an information model for the representation of 3D city and landscape objects that can be shared over different applications. Some cities, including -- Zurich, Switzerland; Paris; Marseille, France; and Rotterdam, the Netherlands -- manage their 3D city models exclusively in CityGML, and the Netherlands has made CityGML part of their national data model. In contrast to other 3D vector formats, CityGML provides five levels of detail, and it is based on a rich, general-purpose semantic model as well as reference system, geometry, and graphics content. The semantic model allows for employing virtual 3D city models for sophisticated analysis tasks. Estimators of heating energy consumption in buildings have been performed using the 3D city model of Berlin, and this has also been done in other cities. More and more projects are using CityGML on an international level. Simply use Google or Google Scholar and search for ‘CityGML.’
CPC/BIM: In what other ways is Version 2.0 considered an improvement over Version 1.0?
McKee: See the list of changes in the [following] specification document:
Additions in CityGML 2.0
CityGML 2.0 is a major revision of the previous version 1.0 of this International Standard (OGC Doc. No. 08- 007r1) and introduces substantial additions and new features to the thematic model of CityGML. The main endeavor of the revision process was to ensure backwards compatibility both on the level of the conceptual model and on the level of CityGML instance documents. However, some changes could not be implemented consistent with directives for minor revisions and backwards compatibility as enforced by OGC policy (cf. OGC Doc. No. 135r11). The major version number change to 2.0 is therefore a consequence of conforming to the OGC versioning policy without having to abandon any changes or additions which reflect requests from the CityGML community.
CityGML 2.0 is backwards compatible with version 1.0 in the following sense: each valid 1.0 instance is a valid 2.0 instance provided that the CityGML namespaces and schema locations in the document are changed to their actual 2.0 values. This step is required because the CityGML version number is encoded in these values, but no further actions have to be taken. Hence, there is a simple migration path from existing CityGML 1.0 instances to valid 2.0 instances.
The following clauses provide an overview of what is new in CityGML 2.0:
New thematic modules for the representation of bridges and tunnels. Bridges and tunnels are important objects in city and landscape models. They are an essential part of the transportation infrastructure and are often easily recognizable landmarks of a city. CityGML 1.0 has been lacking thematic modules dedicated to bridges and tunnels, and thus such objects had to be modeled and exchanged using a GenericCityObject as proxy (cf. chapter 10.12). CityGML 2.0 now introduces two new thematic modules for the explicit representation of bridges and tunnels, which complement the thematic model of CityGML: the Bridge module (cf. chapter 10.4) and the Tunnel module (cf. chapter 10.5). Bridges and tunnels can be represented in LOD 1 – 4 and the underlying data models have a coherent structure with the Building model. For example, bridges and tunnels can be decomposed into parts, thematic boundary surfaces with openings are available to semantically classify parts of the shell, and installations as well as interi- or built structures can be represented.
This coherent model structure facilitates the similar understanding of semantic entities and helps to reduce software implementation efforts. Both the Bridge and the Tunnel model introduce further concepts and model elements which are specific to bridges and tunnels, respectively.
Additions to existing thematic modules
CityGML Core module (cf. chapter 10.1) Two new optional attributes have been added to the abstract base class core: CityObject within the CityGML Core module: relativeToTerrain and relativeToWater. These attributes denote the feature’s location with respect to the terrain and water surface in a qualitative way and thus facilitate simple and efficient queries (e.g., for the number of subsurface buildings) without the need for an additional digital terrain model or a model of the water body.
Building module (cf. chapter 10.3)
LOD0 representation. Buildings can now be represented in LOD0 by footprint and/or roof edge polygons. This allows the easy integration of existing 2D data and of roof reconstructions from aerial and satellite imagery into a 3D city model. The representations are restricted to horizontal, three-dimensional surfaces.
Additional thematic boundary surfaces. In order to semantically classify parts of the outer building shell which are neither horizontal wall surfaces nor parts of the roof, two additional boundary surfaces are introduced: OuterFloorSurface and OuterCeilingSurface.
Additional relations to thematic boundary surfaces. In addition to AbstractBuilding and Room, the surface geometries of BuildingInstallation and AbstractBuilding and Room, the surface geometries of BuildingInstallation and IntBuildingInstallation features can now be semantically classified using thematic boundary surfaces. For example, this facilitates the semantic differentiation between roof and wall surfaces of dormers which are modeled as BuildingInstallation.
Additional use of implicit geometries Implicit geometries (cf. chapter 8.3) are now available for the representation of Opening, BuildingInstallation, and IntBuildingInstallation in addition to BuildingFurniture. A prototypical geometry for these city objects can thus be stored once and instantiated at different locations in the 3D city model.
Generics module (cf. chapter 10.12). Two generic attributes have been added to the Generics module: MeasureAttribute and GenericAttributeSet. A MeasureAttribute facilitates the representation of measured values together with a reference to the employed unit. A GenericAttributeSet is a named collection of arbitrary generic attributes. It provides an optional codeSpace attribute to denote the authority organization who defined the attribute set.
LandUse module (cf. chapter 10.10). The scope of the feature type LandUse has been broadened to comprise both areas of the Earth’s surface dedicated to a specific land use and areas of the Earth’s surface having a specific land cover with or without vegetation.
CPC/BIM: What is your organization doing to promote the availability of Version 2.0?
McKee: The OGC will continue to promote CityGML 2.0 at conferences and OGC Technical Committee meetings and also in blogs, articles, and columns. Vendors who implement CityGML usually promote the use of CityGML.
CPC/BIM: When may Version 3.0 be proposed, considered, and adopted?
McKee: The development of V3.0 has not been started yet. The OGC Technical Committee's CityGML Standards Working Group will collect experiences with the adoption of V2 and remaining issues from V1 (and maybe new ones with V2).
CPC/BIM: What potential improvements to Version 2.0 could Version 3.0 conceivably include?
McKee: Closer integration with BIM. Also, some missing concepts from The Industry Foundation Classes (IFC) could be integrated, like volumetric wall, slab objects.