Working Group IV — Analysis — Position Paper: Spatial Data Analysis in 3D GIS

Abstract

One of major challenging tasks of 3D GIS is to support spatial analysis among different types of real 3D objects. The analysis functions in 3D require more complex algorithms than 2D functions, and have a considerable influence on the computational complexity. In order to maintain a good performance, not only are the algorithms implemented efficiently, but also the 3D spatial objects are represented by a suitable 3D data model. However, it is a difficult task to select an appropriate data structure designed for the characteristics of the applications, for example, objects of interest, resolution, required spatial analysis, etc. (Zlatanova et al. 2004). A model designed for 3D spatial analysis may not exhibit good performance on 3D visualization and navigation. In other words, different data models might be suitable for the execution of specific tasks but not others. In order to maximize efficiency and effectiveness in the provision of operations, Oosterom et al. (2002) proposed multiple topological models maintained in one database by describing the objects, rules and constraints of each model in a metadata table. Metric and position operations such as area or volume computations are realised on the geometric model, while spatial relationship operations such as ‘meet’ and ‘overlap’ are performed on the topological model. However, it is necessary to find out whether the developed 3D data models are designed for 3D spatial analysis.