[ Contents ]
Having now produced a cost surface, it is possible to generate accessibility maps to any target locations within this region. Targets can be vector points, but some software programs will produce a better result if they are rasterised first. In this first example, the accessibility to schools is examined (figures 8 and 9)
The exact algorithm used to generate the accessibility map can be written by the user, but most GIS packages contain a basic least cost distance tool. This will work outwards from each specified target, calculating the least cumulative cost moving from the targets for any cell in the raster and using those least costs to generate a new raster, the accessibility map. This raster can contain large, floating point numbers and it is necessary to reclassify it for use. In this case study, using ArcGIS 9, the raster resolution was 1000 and so the values were divided by 60,000 to account for the resolution and give time in hours rather than minutes. This output was further reclassified in to a simple five level ordinal scale for clarity.
This basic school accessibility map shows that straight line distances bear little or no relationship to how accessible somewhere like a school really is. It also shows the overriding influence of the road network on accessibility (Figure 10).
Exactly the same process can be used to generate accessibility maps to any chosen target or groups of targets. The next two examples shows travels times to regional hospitals and to information resources at Village knowledge Centres (VKC) (figures 11 and 12)
In all three cases, the distribution of the targets are different and the accessibility maps also differ. However, the gross underlying pattern is similar and is largely determined by the road network, which is to be expected. This also indicates that the model is fairly robust and not overly sensitive to changes of inputs
