With rapid industrialization since 20th century, urban population dramatically increased, urbanized areas were horizontally expanded, and the number of cities also increased. Lowly dense and fragmented horizontal urban expansion in has been seen in a negative light so that this kind of urban change has been continuously paid attention as a target to be managed.
Urban sprawl can be defined as the urban development with low density, dispersion, and without systematical land use plan. In order to manage sprawl phenomenon effectively and to prepare alternatives, it is require to quantitatively measure the degree of urban sprawl. The measurement approaches of urban sprawl can be classified into three categories by analyzing existing researches; density-based, spatial structure-based, and comprehensive approaches. The density-based approach measures the spread degree of low density residential areas, and spatial structure-based approach measures the degree how urbanized areas are segmented and are structurally complicated. Finally, the comprehensive approach measures sprawl by considering every features related to it including density and spatial structure characteristics of urbanized areas.
This study regards low density and fragmentation as the key characteristics of urban sprawl. These two key characteristics have been grasped by density-based and spatial structure-based approaches respectively. As a result, density-based and spatial structure-based approaches only provide an understanding of one side of urban sprawl. It is possible to understand urban sprawl fully with comprehensive approach. In this approach, however, various characteristics of urban sprawl are measured separately and analyzers or decision makers have to evaluate synthetically individual measures. Therefore, the purpose of this dissertation is to develop an alternative method grasping two key characteristics of urban sprawl with a single value and intuitively interpreting the meaning of measure. For this, this study examines the major features and their limitation of density-based and spatial structure based approaches.
The third chapter of the dissertation evaluated the main characteristics of density-based approach. The sprawl index as a single indicator using only population density has been applied in a number of studies, because it is easy to understand and measure. However, the sprawl index might not be able to reflect the key characteristics of urban sprawl like leapfrog development and might be sensitive to modifiable areal unit problems (MAUP) and the threshold dividing high and low population densities. In this chapter, the characteristics of the sprawl index suggested by Lopez and Hynes were evaluated in various aspects such as change in spatial arrangement, scale and zoning effects and change in the density threshold. The main findings of this chapter are following. First, the sprawl index could reflect the change in population density but could not detect the change in the population distribution pattern. Second, the value of sprawl index was sensitively depended on the changes in the spatial scale and the zoning system. Third, the sprawl index also increased as the threshold dividing the low and high population density increases.
The fourth chapter examined what spatial characteristics of urbanized areas can detect with spatial structure-based approach. this study selected a set of landscape indices which might capture major spatial structure such as area, distribution, and shape of urbanized areas and applied them to Seoul Metropolitan Area. The selected landscape indices are class area, shannon’s diversity index, numbers of patches, aggregation index, and shape index. With these indices, it is possible to detect the change in the area and number of urbanized patches, fragmentation, and shaped-complexity of them. However, there is no single landscape index grasping all spatial structure of urbanized areas. In other words, in order to judge whether a metropolitan area is sprawled from spatial structure perspective, analysers or decision makers have to combine the results of individual landscape indices. In this process, some results are contradictory to each other. In addition, landscape indices could not detect the change of population density in urbanized areas between two periods urbanized area.
The fifth chapter developed an alternative method measuring urban sprawl. Two core components of defining urban sprawl are low-density and fragmented residential development. Sprawl measures based the density could not grasp the degree of fragmentation and sprawl measures based on spatial structure could not catch the variation of density. Therefore, the purpose of this chapter is to develop an alternative measure simultaneously reflecting density variation and the degree of fragmentation. To achieve this, Lopez and Hynes’s Sprawl Index which is based on population density and two spatial structure based indexes, that is joint count statistics and division index are combined with weights. These three indices have the same interpretation and range of measurement. We named this new alternative measure as a density-spatial structure sprawl index. The index is easily calculated using density, adjacency, and area of urbanized region and interpretation of the index is straightforward because indices with the same range and meaning of measurement are combined. Computational results show that the developed index can remedy the shortcomings of density-based and spatial structure-based sprawl approaches.
This dissertation might make a great contribution to in-depth understanding of urban sprawl from quantitative perspective and be the catalyst for future urban sprawl research. In addition, the suggested approach which can measure urban sprawl comprehensively and intuitively interpret results might practically provide useful information for the management of a metropolitan area struggling with rapid urban sprawl.