Spatial form and dynamics of urban growth

Abstract

The world's population continues to grow and much of the growth in the late 20th century has occurred as urban development in major metropolitan regions. Hence, there is a need to better understand how cites grow so that, among other issues, ‘quality of life’ in urban environments is maintained at a sustainable level. Cities are complex dynamic systems whose evolution reflects a myriad of social economic, political and physical interactions, and the study of cities necessarily encompasses a wide range of disciplines. This thesis focuses on the dynamic spatial characteristics of urban growth and change. That is, the city as seen in plan view from above and how that view changes over time as the city grows. In the first section, satellite imagery and digital air photos are used to explore and map two dimensional urban form and dynamics. The purpose of the work in this section twofold. Firstly, to use imagery to explore the spatial composition of urban form. The primary question was: what are the components of urban form and can the urban matrix be broken down into primary bio-physical components such as impervious surfaces, vegetation and soil. Secondly, by understanding the physical composition of the urban matrix, classification techniques were developed that allowed urban form to be mapped from remotely sensed imagery and the dynamics of urban growth explored using time series of classified images. The method provides a new and applicable approach to urban composition analysis. Landsat Thematic :rv1apper image data provided accurate (83% adjusted overall accuracy) classification of broad land cover types and their change over time for a rapidly growing region of Eastern Australia. Thus providing a base information set for understanding the relationship between the dynamics of growth and the resulting form. A stochastically constrained cellular model of urban growth is developed in the second section of the thesis. The design of the model places it within a suite of models currently popular in the literature whose conceptual design stems from new theories of complex systems, in particular, notions of self-organization. Models of this kind are local or micro models and provide a way in which the same phenomena of growth at different levels and scales can be synthesized into an ordered understanding. The purpose of the work in this section was firstly to explore the principles and validity of the underlying concepts in model design, focusing particularly on self-organization, and secondly, to develop an operational model that is constrained by key bio-physical and socio-economic factors that influence urban dynamics and form. The results of the work provided new insights into the self-organizing characteristics of urban growth associated with transport networks. Correlations of 60% to 70% for comparison between predicted and observed growth was achieved based on the data set developed in section one. In the third section of the thesis, attention is turned to the development of models that can be applied to address issues of urban sustainability. The purpose of the work in this section was to synthesize the work in the previous two sections to develop an operational model that can be applied in a urban planning context. In this section, the cellular model developed in section two is integrated with a spatial optimization model that can be applied to optimally allocate land to meet population projections, given some environmental targets or thresholds. The resulting model and software can be seen as a prototype modelling tool that incorporates a unique combination of remote sensing, cellular modelling, and optimization that could be incorporated in an spatial decision support system for urban growth. The modelling approach was applied to produce spatial realizations for population projections for a rapidly growing region of Eastern Australia. The thesis concludes with a synthesis of the three broad sections of work in the context of current research questions associated with monitoring, modelling, and planning the spatial dynamic characteristics of urban growth. Broad research questions associated with understanding urban spatial dynamics, the synthesis of scales, the identification of key socio-economic and bio-physical factors, and planning support tools are addressed with respect to findings from the work carried out in the thesis.