Three-dimensional geospatial compact abstract cell complexes data structure for urban air pollution modelling

Abstract

Understanding urban air pollution is important for sustainable urban development. Vizualization of urban air pollution modelling has been provided through the advancement of two-dimensional (2D) to three-dimensional (3D) Geographical Information System (GIS). However, current 3D GIS is still in its developing phase and focuses more on the geometry of the 3D city models. Presently, there have only been a few attempts in studying the topological relationships between 3D city objects as key elements in geospatial science. These topological relationships are vital for 3D traversal and 3D nearest neighbours information between 3D city objects in GIS. To address the lack of topological relationships in 3D GIS, Compact Abstract Cell Complexes (CACC) as a new topological data structure was developed in this research. CACC was developed by analyzing the existing data structure and identifying the limitations. Firstly, it was designed to store all the cycles that exist in different dimensions with the aim of articulating the connectivity in 3D city models. Secondly, it stored the nearest neighbours information by implementing the space-filling curve method called the Hilbert Curve. Functionally, CACC has the full traversal ability inside single and multiple connected components. Performance experiments proved that CACC requires 60% to 80% less disk storage as compared to other comparable data structures. The results have demonstrated that it is the most compact and requires minimal disk storage as well as being 90% faster in search queries of large 3D datasets. Upon validation, the implementation of CACC in urban air pollution modeling demonstrated the expediency of having a 3D topological data structure for 3D city models application. The research has developed and proved that CACC can identify topological relationships between 3D objects. Furthermore, its application can be extended to other 3D applications such as Building Information Model (BIM), Computer Aided Design (CAD), Computer Aided Engineering (CAE), web standards design, geospatial database design and disaster management and planning.