3D Topographic Data Modelling: Why RigidityIs Preferable to PragmatismFriso PenningaDelft University of Technology, OTB, section GIS Technology,Jaffalaan 9, 2628 BX Delft, The [email protected]. In this paper two concepts for modelling 3D topography areintroduced. The first concept is a very pragmatic approach of 3D mod-elling, trying to model as much as possible in (less complicated) 2.5Dand use 3D modelling only in exceptional cases. The idea is to use a con-strained TIN in 2.5D and place 3D TENs on top or below this surface.As both data structures use the same simplexes (nodes, edges, triangles)this integration should be very well possible. At a conceptual level thisapproach seems suitable, but at design level serious problems occur. Toovercome these a rigid approach is developed, modelling all features ina 3D TEN, including the air above and earth beneath these topographicfeatures. This model is stored and maintained within a spatial database.Despite its more advanced concept, it is shown that this approach offershuge advantages compared to the initial pragmatic approach.1 IntroductionThis paper describes two modelling concepts developed within the research on 3DTopography as carried out within the GIS Technology group at Delft Universityof Technology. Initially the aim was a pragmatic approach, in which applicabilitywas one of the keywords. However, during the research we came to the conclusionthat our pragmatic approach seemed suitable at conceptual level, but that itcauses some serious modelling problems at design level. Based on the identifiedstrengths of this initial model a rigid approach is developed, which turned outto actually simplify most of our problems at design level, although it is moreadvanced at conceptual level.This paper starts with an introduction on the backgrounds of the 3D Topog-raphy research in Paragraph 2, including a short overview of relevant availabledata sets. Next the concepts of the initial pragmatic modelling approach areintroduced in Paragraph 3. Its drawbacks are addressed and some preliminaryconclusions on the initial modelling approach described. Based on these con-clusions the concepts of the new rigid approach are presented in Paragraph 4,followed by some remarks on the implementation of this approach. The paperends with final conclusions and future research in Paragraph 5.A.G. Cohn and D.M. Mark (Eds.): COSIT 2005, LNCS 3693, pp. 409–425, 2005.c Springer-Verlag Berlin Heidelberg 2005410 F. Penninga2 Backgrounds of the 3D Topography Research2.1 The Need for the Third DimensionMost current topographic products are limited to representing the real worldin only two dimensions. As the real world exists of three dimensional objects,which are becoming more and more complex due to increasing multiple land use,accurate topographic models have to cope with the third dimension. The overallgoal of this research is to extend current topographic modelling into the thirddimension. Applications of 3D modelling are not limited to the terrain surfaceand objects built directly on top or beneath it, as geological features and airtraffic or telecommunication corridors can be modelled too.Most initiatives on developing 3D GIS focus on supporting visualisation, of-ten in Virtual Reality-like environments. One of the objectives of this 3D mod-elling research is to enable 3D analysis as well, as this traditional GIS-strengthlacks until now in most 3D GIS approaches. Another important assumptionwithin this research follows from the required wide variety of applications oftopographic data. As topography is ranked high in the spatial data infrastruc-ture hierarchy, one cannot optimize the data model for one specific purpose.One has to be able to serve the complete range of user applications, regardlesswhether these applications require for instance optimal visualisation capabilitiesor optimal analytical capabilities.In 3D modelling one needs a 3D primitive (a volume) beside points, linesand faces to represent 3D objects accurately. Earlier research proposed amongstothers using simplexes (point, line, triangle, tetrahedron) (Carlson 1987), points,lines, surfaces and bodies (3D Formal Data Structure (FDS)) (Molenaar 1990a,Molenaar 1990b, Molenaar 1992), combining Constructive Solid Geometry (CSG)and a B-rep. (de Cambray 1993) and integrating a 2.5D Triangulated IrregularNetwork (TIN) with 3D FDS (Pilouk 1996). In applications polyhedrons are of-ten used as 3D primitive (Zlatanova 2000, Stoter 2004). These publications on3D modelling concepts are often limited to a conceptual description of the useof a 3D primitive, without addressing any of the actual problems concerning theuse of these 3D primitives as in analysis. As a result true implementations (be-sides some very small experiments) are rare, thus not proving actual usefulnessof the concepts.2.2 Current Available Data SetsThe initial approach to extend topographic modelling into 3D is both supplyand demand driven. The required data sets -both topography and height data-become available in a growing number of countries. 2D Topographic data setsare available and are being converted into object-oriented models, as this offershuge advantages in digital processes such as GIS analysis. Due to the grow-ing popularity of airborne laser scanning high resolution height data becomesavailable. Combining both types of data can lead to full 3D data. Within thepresented research Dutch data sets will be used in implementation tests. The3D Topographic Data Modelling: Why Rigidity Is Preferable to Pragmatism 411height information is available in a Digital Elevation Model (DEM) called theAHN (Actual Height model the Netherlands). As large parts of the Netherlandsare situated below sea level accurate large scale height data is of great impor-tance, thus leading to the development of the AHN, a height model with oneheight point for on average every 16m2. The most relevant topographic data set(1:10.000) is currently being converted by the Dutch Topographic Survey intoan object-oriented structure called TOP10NL.At the demand side developments as increasing multiple land use and risingawareness of the importance of sustainable urban development (both causedby urban space scarcity), increase the need for real 3D topographic data sets.At dataset level the major shortcomings of the current 2D products lie in theabsence of height information for buildings and other constructions, which isessential