Doctoral Dissertations

Jun Hu

Date of Award

8-1992

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Geography

Major Professor

Bruce A. Ralston

Committee Members

Thomas Bell, Henry Herzog, Cheng Liu, Theodore Schmudde

Abstract

The location of landfills belongs to a class of complex spatial problems called semi-structured location problems. Since landfills are necessary but undesirable public facilities, many conflicting factors have to be considered when choosing their locations. For example, landfills should be located near major population centers to reduce the cost of transporting solid waste. On the other hand, landfills should be also located away from population centers to reduce public opposition to the environmental risks and noxiousness associated with landfills. Some of these siting factors are, however, intangible and difficult to define. These include unarticulated preferences, political motivations, and the psychological impacts of being exposed to risks. Traditional location modeling approaches are found unsatisfactory to locate landfills because of these conflicting and intangible factors. This study adopts spatial decision support system (SDSS) as the approach to deal with landfill location problems. The basic idea is to include decision makers as part of a multi-pass solution process. Decision makers provide preference information to the SDSS during each iteration. The SDSS generates corresponding solutions based on decision makers' preference information, and then presents the solutions to decision makers to evaluate. The continuous interactions between decision makers and the SDSS are expected to reveal and incorporate heretofore intangible information into the solution process. A specific SDSS for locating landfills, the SDSSLF, has been developed in this study. The SDSSLF is a menu-driven computer software system based on personal computer. It consists of four basic components: decision makers, a multiobjective location model, GIS, and interfaces. At the heart of the SDSSLF is an interactive decision making framework called the reference point method (RPM). The RPM is used to integrate the basic components of the SDSSLF, and determine when and what kind of preference information decision makers should provide to the system. Decision makers supply their preference information as aspiration and reservation values. Based on these values, the RPM formulates a special scalarizing achievement function, optimizes this function, and generates a small set of best worst solutions for decision makers to evaluate. The RPM also provides decision makers the Utopian and nadir points derived from optimizing each objective individually. These points can be used by decision makers as the reference information for individual objectives and as the initial aspiration and reservation values. The SDSSLF has been tested on two types of landfill location problems. The first problem has only a small number of possible location configurations (less than seven). Decision makers are able to directly evaluate all possible location configurations. Some functions of the SDSSLF, such as its graphical display capabilities, can be used to help decision makers interpret each locational solution. The second landfill location problem has a large number of possible locational configurations. It is impossible for decision makers to evaluate all solutions at one time. The SDSSLF is used to generate and present only a few solutions each time for decision makers to evaluate. A final compromise solution is generated through the multi-pass interactions between decision makers and the SDSSLF. Data from the Sevier County, Tennessee landfill location problem (SCLLP) was used in the second testing problem. A decision maker familiar with the SCLLP was asked to evaluate the SDSSLF. He found the SDSSLF to be a useful system for landfill locations in practice. The most interesting feature of the SDSSLF is that decision makers can work with the system, express their preferences, and obtain corresponding location solutions. This allows decision makers to explore the geographic effect of different assumptions and scenarios, and is more likely to generate a politically acceptable final location decision than has heretofore been the case.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS