Cybergeo (Nov 2007)
GIS-based accessibility analysis for network optimal location model
Abstract
Location analysis has become ‘intrinsically linked with GIS-based analysis and modelling.’ Optimal location models are often used for decision making to locate facilities to maximise access to clients/demand within a defined continuous area of possibilities. But what happens if the potential location of demand is spatially confined and best location model is applied to choose between different stationary facility sites? This is a question concerning the mining industry in the UK, who wants to restore exhausted quarries (facilities) into energy-crop fields. A small number of quarries have to be selected within a larger panel of potential quarries sites to be restored. These selected quarries have vocation to supply a power station with energy-crop. We defined the quarries of best location to be converted into energy crops supplier sites by minimising the transport costs (required criteria) of crops to a power station. We derived transport cost from distance measurements and time of travel. Therefore, the objective function of the model minimized the demand distance/time of travel (expressing transport cost) between the demand node (Power station) and the nearest quarry(ies). Based on the shortest path distance criteria, the demand of crops required for the power station was allocated to the quarries until the demand was satisfied. Doing so, the cluster of quarries of best location (minimum distance/time) to be restored in energy-crop fields was identified in order to minimize the transport cost.Several ways to assess the time of travel between quarries and power site were explored and tested in the model of quarries selection. Much attention was given to measuring accessibility. First, based on metric distance the length of roads was used to find the least cost path. Multi-criteria distance-time measures were computed according to the length of the road, the travel speed according to the type of road, the gradient slope of the road derived from a Digital Elevation Model (DEM) values and the ‘slow down effect’ of urban areas. The results of the different routing processes were compared. Our findings underline the use of multi-criteria travel time calculations in order to improve accuracy of accessibility and location models. The method was tested on a sample area in the North of England and could be extended and applied to the entire UK with some accommodations. This is presented as part of the methodology and discussed in the conclusion.
Keywords
