Ideation and Evaluation in Environmental Planning

186 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1983.Typically, environmental plans involve substantial quantities of valuable resources and affect a large number of people. The final plan is selected by the decision-maker(s), but the objectives behind the selection are sometimes poorly defined. Since judgements can be affected by an individual's reaction to his surroundings, it is argued that the role of the analyst is to improve the information available to the decision-maker(s). Therefore, ideation methods, i.e., methods of generating alternative solutions by manipulating the mathematical model of the planning problem, and evaluation methods, i.e., techniques of processing the generated information, become desirable tools for the analyst. In this thesis, various ideation methods are developed and unified using vector space theory, and cluster analysis is used to evaluate alternative solutions.The geometric conceptualization based on vector space theory leads to an analogy between dissimilar solutions and orthogonal (perpendicular) vectors. The notion of orthogonality can be used in two ways to generate alternatives: to find solution vectors that are orthogonal to existing solutions, and to search for solutions in directions orthogonal to the directions previously explored. A regional wastewater system planning example, a school district planning example and a land use planning example are used to illustrate the ideation methods.The Gibbs inner product between two decision vectors is analogous to the dot product between two vectors in a three dimensional geometric space, and provides a simple means of comparing two decision vectors. The Gibbs inner product measure is shown to possess applicability, interpretability, and computational feasibility for ideation. On the other hand, metric measures and the product moment correlation coefficient are shown to be inappropriate and computationally inefficient for ideation.The results from the ideation process have to be communicated effectively to the decision-maker(s). The differences in the decision space among various land use plans are reduced to a presentable form using hierarchical clustering methods. Assuming that two solutions with similar decision space representations achieve similar levels of different objectives, the clustering process may lead to useful classifications of the alternatives without forcing the analyst to make value judgements.U of I OnlyRestricted to the U of I community idenfinitely during batch ingest of legacy ETD