Characterizing and Measuring Urban Sprawl
MEASURING URBAN SPRAWL AND COMPACTNESS: CASE STUDY ORLANDO, USA
SIM S.(1), MESEV V.(
Characterizing and Measuring Urban Sprawl
The literature characterizing sprawl is extensive including ―the scattering of urban settlements‖ (Harvey and Clark, 1971). Altshuler and Gomez-Ibanez (1993) describes sprawl as ―continuous low density residential development on the metropolitan fringe, ribbon low density development along major suburban highways, and development that leapfrogs past undeveloped land to leave a patchwork of developed and undeveloped tracts.‖ The term sprawl refers to a range from low-density urban development all the way to discontinuous and dispersed or even decentralized forms of urban expansion. At this point, it is important to review the most widely debated concerns and efforts to characterize sprawl as an urban form.
First, many definitions describe sprawl as a development pattern that consumes more open space and produces more impervious area causing a less environmentally sustainable form compared to a compact development. A growing trend of smart growth suburb depicts new urbanism ideas for handling sprawl such as mixed-use zoning, pedestrian-friendly streets, transit, and town centers. The concern here is how to decide whether a tract of development is actually sprawl compared to compact development. Secondly, sprawl represents a stage in the development process. The EPA (2001) report notes, ―at a metropolitan scale, sprawl may be said to occur when the rate at which land is converted to non-agricultural uses exceeds the rate of population growth.‖ Therefore, sprawl should be considered in a space-time context, and not as a simple increase of urban lands (EPA, 2001). This suggests that sprawl should be measured on a development pattern continuum over time. Thirdly, many characteristics associated with sprawl relate to density either using population or housing data. Some studies (Barnes, 2001; Galster et al, 2001 and Harvey et al, 1971) suggested the expanded dimensions of sprawl such as continuity, concentration, clustering, centrality, nuclearity, mixed use and proximity. Finally, urban sprawl causes the loss of informal open space and wildlife habitats. Some of the fastest rates of loss have been occurring at the interface of urban core and rural areas at the metropolitan region.
Developing a Measurable Definition of Compactness and Sprawl
Despite a lack of rigorously defined idea of sprawl, the term is often defined by four land use characteristics: low density; scattered development (i.e. decentralized pattern); commercial strip development; and leapfrog development (Tsai et al, 2009). The last three characteristics are related to the spatial arrangement of the urban area. These phenomena often occur at a metropolitan area. In contrast to sprawl, compactness can be defined as more energy efficient and less polluting. Advocates claim that a compact development is more sustainable for economic, environmental and social dimensions. Ewing (1997) defined compactness as some concentration of employment and housing as well as a mixture of land uses. Galster et al (2001) defined compactness as the certain degree of clustered developments and measured it by the amount of land developed in each square mile. Tsai (2009) described compactness as often involving the concentration of development as apposed to sprawling development. There are similarity and differences of sprawl and compactness in terms of measurement of urban development at a metropolitan level. Hanson et al (2001) employ metrics to describe land development along six geographical dimensions: density; continuity; concentration; centrality; nuclearity; and diversity. Tsai (2009) used a set of four dimensions of metropolitan form - metropolitan size, density, degree of equal distribution and degree of clustering and utilized Moran, Geary, Gini, and Entropy indicators to measure those four dimensions for both population and employment distribution. With these issues in mind, six dimensions of sprawl indicators were selected.
Size: the urban land area which has often been used as a simple index of sprawl. The idea of using the urban land size for sprawl causes more land consumption than compact development. Because sprawl is characterized by an increase in the built-up area along the urban and rural fringe, this attribute gives considerable information for understanding the behavior of such sprawl. Larger urban size values indicate a greater degree of sprawl.
Density: this can be measured by the population per unit of developed land. There are numerous density-based measurements that use population and employment data. Hasse (2004) suggested that density of new urbanization should be used as a measure of land consumption for new urban growth, Schneider et al (2009) used density of built-up land by measuring ratio of amount of urban land to all land by percentage. This index provided information on whether new land development is low or high density. Higher ratio of low-density development to total development indicates sprawl whereas lower rations indicate more compact or smarter growth patterns. In our study, the density indicator provides a measure of land consumption for new urban growth per capita. Population increment was calculated by comparing the difference between years from the following, 1974, 1985, 1993 and 2003. The amount of new urban growth in 1985, 1993 and 2003 was extracted from FWC land use and land cover maps and was then generated by intersecting urban areas in the earlier year. The new urban density indicator was calculated by normalizing the amount of new growth by the concurrent increase in population. Larger per capita new growth density values indicate a greater degree of sprawl.
Continuity: The index of continuity is a measure of the size or the distance of newly developed areas from the main developed part of a metropolitan area (Tsai, 2005). Typically a more sprawled landscape experiences decreasing continuity of built-up land. We adopted and used the definition and the model from the Urban Landscape Analysis Tool (ULAT) developed by Parent et al (2009). New development is any built-up land added between two time periods (T1 and T2). The classification of new development in T2 is based on location relative to the urban footprint in T1 (Figure 2). The urban footprint in T1 is defined as impervious surfaces and any open space likely to impact the city. The classification of new development is divided into infilling, edge-expansion or spontaneous growth. Infilling development is defined as “newly developed pixels that are in the urbanized area of the previous time period.” Extension development is defined as “newly developed pixels that are in the fringe area of the previous time period.” And leapfrog development is defined as “newly developed pixels that are outside of the rural area of the previous time period.” Any built-up areas in T2 where there is no contiguity with the previous developed areas (T1) are considered to be leapfrog development areas. Smaller infilling percentage values indicate a greater degree of sprawl.
Scattering: this is a measure for characterizing how developed land parcels are isolated from each other. Concentration is the extent to which development is confined to a relatively small portion of a jurisdictions’ total land area. A wide variety of techniques are employed in measuring concentration including, Shannon’s entropy (Yeh and Li, 2001) and the global Moran‘s I coefficient (Tsai, 2004). Some techniques are based on distance measures. Examples include the Euclidean nearest neighbor (ENN) distance, proximity (PLADJ) and Clumpy. These are landscape fragmentation measures available by open source GIS software, such as FRAGSTATS. For this study, we used mean Euclidean nearest-neighbor
distance (ENN_MN) to measure the degree of scattering. Euclidean nearest neighbor distance (ENN) has been used extensively to quantify patch isolation and is defined as the shortest straight-line distance between urban patches (McGarigal et al, 2002). Larger ENN values indicate a greater degree of sprawl.
Shape/Fractal dimension: shape complexity is another important attribute of sprawl. Since we are interested in compactness, shape indicators can be used to describe compact shapes with low values. Fragstats (McGarigal and Marks, 1995) provides diverse measures based on perimeter-area relationships. The simplest shape index is a straightforward perimeter-area ratio (PARA). This shape index (SHAPE) measures the complexity of patch shape compared to a standard shape (square or almost square) of the same size, and therefore alleviates the size dependency problem of PARA. Another other basic type of shape index based on perimeter-area relationships is the fractal dimension index. Perimeter-area fractal dimension (PAFRAC) is appealing because it reflects shape complexity across a range of spatial scales (patch sizes). Larger PAFRAC values indicate a greater degree of sprawl. For example, small and large patches alike have simple geometric shapes, then PAFRAC will be relatively low, indicating that patch perimeter increases relatively slowly as patch area increases.