Causes of sprawl: A portrait from space Burchfield M., Oversman H., Puga M., Turner M., 2005 I. Introduction nSprawl definition –Residential Sprawl –Commercial Sprawl nAdvantages and Disadvantages of Sprawl n Urban Sprawl n“Low-density development beyond the edge of service and employment, which separates where people live from where they shop, work, recreate and educate — thus requiring cars to move between zones.” (http://www.sierraclub.org/) n nUrban Sprawl - беспорядочно застроенная территория n n Residential Sprawl Single-family detached homes… …occupied by households that commute to work, and …built at low density …beyond walking distance of goods and services …more than critical response time from fire services Heavy reliance on private automobiles as the primary transportation mode Residential sprawl orthomap_MGISISA1FSMAP126722612771 Low density, auto-dependent development outside compact urban and village centers, along highways and in rural countryside. nNo centralized planning or control of land uses nSignificant fiscal disparities among localities nReliance on a “trickle-down” or filtering process to provide housing to low-income housholds n Urban Sprawl 7 Urban sprawl n“Leapfrog" development which occurs when developers choose to build on less expensive land farther away from the city, bypassing vacant land located closer to the city n n n Europe versus U.S. Cities: Sprawl FG13_21 European cities, including this hypothetical U.K. example, tend to restrict suburban development, thereby concentrating new development in and around existing concentrations. This leaves large rings of open space, so-called greenbelts. nSegregation of land use types into different zones n nStrip or ribbon development, which involves extensive commercial development in a linear pattern, which contributes to traffic congestion n Urban Sprawl Commercial sprawl Auto-oriented development… …built at a low floor area ratio …in strips along major routes or in isolated business parks …separated from other land uses. Commercial sprawl ADVANTAGES AND DISADVANTAGES OF SPRAWL •Social •Economic •Environmental •Health • •…. ADVANTEGES OF SPRAWL Advantages ·improvement of life quality in range of flat (housing estate with block of flats vs. own detached house) ·improvement of life quality in range of environment and landscape (housing estate with block of flats vs. own detached house) ·nearness of recreation space; ·improvement of life quality in terms of safety ADVANTEGES OF SPRAWL falowiec stara zabudowa a 060 domki w łapinie ·improvement of life quality in range of flat (housing estate with block of flats vs. own detached house) 4 ADVANTEGES OF SPRAWL ·improvement of life quality in range of environment and landscape (housing estate with block of flats vs. own detached house) ·nearness of recreation space; ·improvement of life quality in terms of safety DISADVANTEGES OF SPRAWL •Social •Economical •Environmental •Health • •… Some recent conclusions about sprawl… n“The vast majority of metropolitan areas experienced a significant decline in metropolitan density (between 1982-1997) and therefore can be described as sprawling.” (Fulton et al. 2001) n“Sprawl is ubiquitous and it is continuing to expand.” (Glaeser and Kahn 2003) n“Many extended urbanized areas are very sprawl-like on some dimensions, but not so sprawl-like on others.” (Galster et al. 2005) n“The extent of sprawl has remained roughly unchanged between 1976 and 1992.” (Burchfield et al. 2006) II. Measurements of Sprawl § Methodology •Analyzing Landscape Change with Satellite Remote Sensing and Geographic Information Systems • Data Source • Land Use and Land cover Digital Data (derived from high attitude aerial photography, 1972) • National Land Cover Data (derived from satellite imagery, 1990) • • Land Use Classification • Urban Land • Sprawl Index • • NW suburb of Washington DC1 Data Source: Satellite Imagery Data units nSquare cells of 30×30 meters situated on a regular grid (8.7 billions cells) nFor each cells predominant land cover was assigned nLand Cover Codes –Residential development; –Commercial and industrial development and transportation networks; –Water; –Bare rock and sand; –Forest; –Range and grassland; –Agricultural land; –Wetlands Urban Land in US Urban Land in US n1.9 % of the United States was developed in 1992 (Burchfield, 2005) –(2.5 % of US was classified as “urban” by Census Bureau in 1990) ntwo-thirds of this developed land is already in urban use in 1976 nDeveloped area grew by 48 percent over 16 years n n Urban Land in US: Large Differences across States State % of state land area urbanized by 1992 % of state non urban land urbanized by 1976- 1992 Arizona 0.79 0.35 DC 68.13 2.8 Massachusetts 17.34 5.7 Wayoming 0.21 0.09 Sprawl Index nCell-based measure of sprawl: n nSI = average % undeveloped land within 1 km2 of residential cells in metropolitan area Limitations of analysis Use remotely sensed data (NLCD) to detect urban land development between 1976-1992 Compare disparate datasets Measurement is limited to single spatial scale Analysis is aspatial: average measure of sprawl for metro area is regressed on metropolitan characteristics Ignore land use pattern outside metro areas 25 Untitled 240 undev. pix 1200 total pix = 20% Urban 0-50% Developed pixels are dark gray 560 m Sprawl Index 26 Untitled 960 undev. pix 1200 total pix = 80% Developed pixels are dark gray 560 m Sprawl Index Suburb 50 – 80% 27 Untitled 1160 undev. pix 1200 total pix = 97% Rural 90 – 100% Developed pixels are dark gray 560 m Sprawl Index Sprawl Index nSpatial scale choice (1 km2 - radius 560 m) –Only 0.3 % of residential development was more than 1 km from other residential development in 1992 Limitations of analysis Use remotely sensed data (NLCD) to detect urban land development between 1976-1992 Compare disparate datasets Measurement is limited to single spatial scale Analysis is aspatial: average measure of sprawl for metro area is regressed on metropolitan characteristics Ignore land use pattern outside metro areas Sprawl Across United states nSprawl Index (1992) =0.43 –Measure of sprawl shows that 43 percent of the square kilometer surrounding an average residential development is undeveloped nSprawl Index (1976) =0.42 –Average residential development was essentially unchanged between 1976 and 1992 Limitations of analysis Use remotely sensed data (NLCD) to detect urban land development between 1976-1992 Compare disparate datasets Measurement is limited to single spatial scale Analysis is aspatial: average measure of sprawl for metro area is regressed on metropolitan characteristics Ignore land use pattern outside metro areas Sprawl as scattered residential development nCell-based measure of sprawl = average %undeveloped land within 1km2 of residential cells in metro “The extent of (residential) sprawl has remained roughly unchanged between 1976 and 1992.” Limitations of analysis Use remotely sensed data (NLCD) to detect urban land development between 1976-1992 Compare disparate datasets Measurement is limited to single spatial scale Analysis is aspatial: average measure of sprawl for metro area is regressed on metropolitan characteristics Ignore land use pattern outside metro areas Sprawl as scattered residential development nFlow of new residential development between 1976 and n1992 was biased towards sprawling areas nCell-based measure of sprawl = average %undeveloped land within 1km2 of newly residential cells in metro Limitations of analysis Use remotely sensed data (NLCD) to detect urban land development between 1976-1992 Compare disparate datasets Measurement is limited to single spatial scale Analysis is aspatial: average measure of sprawl for metro area is regressed on metropolitan characteristics Ignore land use pattern outside metro areas Sprawl as scattered commercial development nFlow of new residential development between 1976 and n1992 was biased towards sprawling areas nCell-based measure of sprawl = average %undeveloped land within 1km2 of commercial cells in metro Limitations of analysis Use remotely sensed data (NLCD) to detect urban land development between 1976-1992 Compare disparate datasets Measurement is limited to single spatial scale Analysis is aspatial: average measure of sprawl for metro area is regressed on metropolitan characteristics Ignore land use pattern outside metro areas Sprawl Index for metropolitan areas Metropolitan area SI, 1992 SI, 1976 Atlanta 55.57 57.77 Boston 47.64 44.72 Miami 20.73 20.23 Los Angeles 35.41 32.95 New York 28.75 33.92 “Sprawl varied dramatically across metropolitan areas.” “Even at the metropolitan area level the extent of sprawl is very stable over time” Another Sprawl Measures nMedian Lot Size (S1) nMiles Driven per Person (S2) n% Employment over 3 miles from CBD (S3) Correlation matrix for various sprawl measures nS1 - Median Lot Size (S1) nS2 - Miles Driven per Person (S2) nS3 - % Employment over 3 miles from CBD – (USA Today’s sprawl index published in 2001) SI S1 S2 S3 SI 1.000 S1 0.521 1.000 S2 0.271 0.187 1.000 S3 -0.070 0.011 -0.073 1.000 Urban sprawl variables (other authors) nResidential density nNeighborhood mix of homes, jobs, and services; nAccessibility of the street network. nOverall mobility nPublic transport nRoad traffic nAccessibilities nStrength of activity centers and downtowns n... n Some critique (E. Irwin, N. Bockstael and H. J. Cho) nMeasurement of sprawl is highly dependent on data and spatial scale of analysis nAggregate pattern analysis masks important variations III. Causes of Sprawl nThe Monocentric City Model •proportion of jobs located in CBD and employment decentralization •transport costs •Open space & amenities nSpace is not Featureless Plain •Aquifers •Physical landscape nPolitical Geography •Jurisdictional fragmentation •Zoning Monocentric city nThe monocentric city model assumes that all employment in the city takes place at a single center, the central business district. nResidential development around that center is then shaped by the tradeoff between convenient commuting close to the center and affordable housing further away. nSubstitution in response to declining land and housing prices leads to larger dwellings with lower capital to land ratios (i.e., less tall, more spacious units and larger yards) as one moves away from the center. Monocentric city nCities specializing in sectors where employment tends to be more centralized will be more compact. n nLower transport costs within a city will result in more dispersed development. Monocentric city nStandard monocentric city model cannot explain leapfrog development. nAmenity value of public open space: –characteristics that make public open space more attractive will increase sprawl; –cities that have been growing faster will tend to experience less sprawl; nHousing is durable and redevelopment costly => construction lag –leapfrogging is greater the greater the uncertainty about future urban growth When Space Is Not a Featureless Plain nExample. Municipal water systems –wherever aquifers underly the urban fringe, household water can be obtained without the large increasing returns associated with public water systems and this facilitates scattered development. –We would expect rugged terrain to naturally encourage scattered development. – In contrast, high mountains in the urban fringe are likely to make development more compact. –Cities with a pleasant temperate climate experience more sprawl. – Political Geography nJurisdictional Fragmentation nZoning –To the extent that there are unincorporated areas on the urban fringe, developers can escape municipal regulation by building outside municipal boundaries, and this facilitates sprawl. –sprawl should be more prevalent where local tax payers pay a smaller share of local government expenses. Causes of Sprawl nCities will sprawl more if: n they specialize in sectors where employment is not typically located close to the city center, nthey were built around the car rather than around public transport, n they have experienced slow population growth, nthere is greater uncertainty regarding their future population growth, Causes of Sprawl nCities will sprawl more if: naquifers underlie a greater fraction of their urban fringe, n they are not surrounded by high mountains, n terrain in their urban fringe is rugged, n their climate is temperate, n they begin with substantial unincorporated areas on the urban fringe, n local tax payers pay a smaller share of local government expenses. The determinants of sprawl nSample: N=275 metropolitan areas nSI newly – Sprawl Index for newly developed cells 1976-1992 nSI 1992 – Sprawl Index, 1992 nX1 – Centralized sector employment, 1977 nX2 – Steetcar passenger per capita, 1992 nX3 – Mean % pop.growth 1920-70 nX4 – S.d. pop.growth nX5 – % of urban fringe overlaying aquifers nX6 – Elevation range in urban fringe nX7 – Terrain ruggedness index nX8 – Mean cooling degree-days nX9 – Mean heating degree-days nX10 – % of urban fringe incorporate 1980 nX11 – Intergov. Transfer, % of local revenues, 1967 n n The determinants of sprawl (1) nSI newly – Sprawl Index for newly developed cells 1976-1992 nX1 – Centralized sector employment, 1977 nX2 – Steetcar passenger per capita, 1992 nX3 – Mean % pop.growth 1920-70 nX4 – S.d. pop.growth nX5 – % of urban fringe overlaying aquifers nX6 – Elevation range in urban fringe nX7 – Terrain ruggedness index nX8 – Mean cooling degree-days nX9 – Mean heating degree-days nX10 – % of urban fringe incorporate 1980 nX11 – Intergov. Transfer, % of local revenues, 1967 n n The determinants of sprawl (4) nSI 92– Sprawl Index,1992 nX1 – Centralized sector employment, 1977 nX2 – Steetcar passenger per capita, 1992 nX3 – Mean % pop.growth 1920-70 nX4 – S.d. pop.growth nX5 – % of urban fringe overlaying aquifers nX6 – Elevation range in urban fringe nX7 – Terrain ruggedness index nX8 – Mean cooling degree-days nX9 – Mean heating degree-days nX10 – % of urban fringe incorporate 1980 nX11 – Intergov. Transfer, % of local revenues, 1967 n n