The SLEUTH urban growth model as forecasting and decision-making tool

[摘要] Accelerating urban growth places increasing pressure not only on the efficiency of infrastructureand service provision, but also on the natural environment. City managers are delegated the task ofidentifying problem areas that arise from this phenomenon and planning the strategies with which toalleviate them. It is with this in mind that the research investigates the implementation of an urbangrowth model, SLEUTH, as a support tool in the planning and decision making process. Theseinvestigations are carried out on historical urban data for the region falling under the control of theCape Metropolitan Authority. The primary aim of the research was to simulate future urbanexpansion of Cape Town based on past growth patterns by making use of cellular automatamethodology in the SLEUTH modeling platform.The following objectives were explored, namely to: a) determine the impact of urbanization on thestudy area, b) identify strategies for managing urban growth from literature, c) apply cellularautomata as a modeling tool and decision-making aid, d) formulate an urban growth policy based onstrategies from literature, and e) justify SLEUTH as the desired modeling framework fromliterature. An extensive data base for the study area was acquired from the product of a jointinitiative between the private and public sector, called 'Urban Monitoring. The data base included:a) five historical urban extent images (1977, 1988, 1993, 1996 and 1998); b) an official urban bufferzone or 'urban edge', c) a Metropolitan Open Space System (MOSS) database, d) two roadnetworks, and d) a Digital Elevation Model (DEM). Each dataset was converted to raster format inArcEdit and finally .gif images were created of each data layer for compliance with SLEUTHrequirements. SLEUTH processed this historic data to calibrate the growth variables for best fit ofobserved historic growth. An urban growth forecast was run based on the calibration parameters.Findings suggest SLEUTH can be applied successfully and produce realistic projection of urbanexpansion. A comparison between modelled and real urban area revealed 76% model accuracy. Theresearch then attempts to mimic urban growth policy in the modeling environment, with mixedresults.

[发布机构] Stellenbosch University