Geoscience Reference
In-Depth Information
module that employs the Analytic Hierarchy Process (AHP) to calculate a land
suitability index for each location. However, as emphasized by Church and Murray
(
2009
), special care should be given in handling data of different types (nominal,
ordinal, interval or ratio) and thorough analysis of the data transformation functions
and the methods for obtaining the weighting factors is required. The wealth of
spatial and attribute information available these days and the increasing capabilities
of modern GIS allows location modelers to utilize such systems in order to generate
data for a variety of location science models. Depending on the scope of the
location model in question, the modeler may employ GIS functions to aggregate
polygons, generate polygon centroids, derive service zones, calculate distances
between different objects and make this data directly available to location science
models. Moreover, as noted by Murray (
2010
), GIS functions may be used to
determine more complex spatial relationships such as adjacency, contiguity and/or
shape. For example, Murray and Kim (
2008
) developed a GIS-based procedure to
identify cliques of parcels or areal units, namely sets of parcels that are in conflict
with each other. These cliques were then used to generate constraints in an integer
programming formulation of the Anti Covering Location Problem which regards the
positioning of a maximally weighted set of facilities so that no two located facilities
are within a specified time or distance measure of each other.
19.4.2
Visualization of Results
Visual representation of large amounts of information is one of the most useful
aspects of GIS. Its role in the visualization of location model results has been
recognized by several researchers including Densham (
1994
) and ReVelle and Eiselt
(
2005
). Murray (
2010
) clearly states that the use of GIS for visualizing the results
of location models is far more complicated than a mere depiction of the sites
where facilities are located. By exploiting the graphical capabilities of modern
GIS, many more aspects of the solution may be represented. For instance, classical
GIS tools such as the construction of Voronoi diagrams or spider diagrams have
been utilized to represent additional aspects of location model results, namely the
coverage of each facility or the allocation of customers to facilities. Other methods
such as graduated circles or choropleth maps have also been used to represent
other attributes of the solution, e.g., the amount of demand left uncovered after the
facilities have been located or the level of cannibalization resulting from the location
of several competing facilities. These methods have been employed to good effect
in numerous applications including the ones reported by Vijay et al. (
2005
), Ghose
et al. (
2005
), Dobbins and Jenkins (
2011
), Suárez-Vega et al. (
2011
) and Pekin et al.
(
2013
).
The visualization of the results of location science models is vital in most
practical applications since it facilitates communication between the various stake-
holders involved. It allows analysts and clients to easily experiment with different
problem settings, directly compare alternative solutions simultaneously with respect
