be located in areas where analysis factors satisfy preset criteria. Still, one may ask, what should be done with the areas where only some of the criteria are satisfied? One key aspect of suitability analysis is the preference level of each variable. An area could be determined as suitable even though it satisfies only some of the criteria if the unsatisfied portions have lower preference levels. In this case, a trade-off can be made between less-preferable and more-preferable variables to make the case that a particular place may be suitable for an activity. Nonetheless, who makes this determination and how preference levels should be set are important questions related to the allocation of decision-making power. Whose judgment should prevail: that of the professional plan- ner or engineer, the community represented by politicians, or some combination of the expert and those most affected by the proposed new system? Sandercock argues that over time, this modern-expert model has evolved into several models, which corresponds with a rise in public participation and a decline of absolute expert power in the planning processes (1). Experts no longer focus solely on tech- nical and system performance criteria; they also aim to become allies of their communities and maximize community preferences. One possible approach to sort out, analyze, and optimize multiple- stakeholder decision making is to use a multicriteria decision-making tool that would allow for interaction between experts and the public. Experts should be able to set the suitability variables while allowing the public to directly provide input to set the preference level. The city–county of Honolulu is known not only for its swaying palms and white sandy beaches; it also is an excellent place to do research because of its isolated location. The entire state of Hawaii is located in the middle of the Pacific Ocean, and its four counties are sep- arated by water. It has an extremely centralized system of government, which also provides an ideal environment for evaluating various problems (2). Data for this study are from different sources with different spa- tial unit of analyses, which makes them difficult to compare. One possible solution to adequately describe and analyze conditions is to create a grid-based mapping system with the use of geographic information systems that would allow maps with different boundary sizes to be converted to a uniform spatial unit of analysis. Database information from the original maps could be manipulated and stored in each cell of this vector grid map system, which then would con- tain a rich source of information that could help researchers locate alignments for the transit system. The decision-making analysis procedure and grid-based mapping system are two elements used to conduct the suitability analysis. Nonetheless, those elements are independent systems that must be cou- pled with another system to create a coherent result. This study couples analytic hierarchy process (AHP) with geographic information systems Analytic Hierarchy Process and Geographic Information Systems to Identify Optimal Transit Alignments I Made Brunner, Karl Kim, and Eric Yamashita 59 Honolulu, Hawaii, like other urbanized centers, is struggling to solve problems of transportation supply and demand. One proposed solution is construction of a fixed-rail transit system along the east–west corridor, to connect the western part of the county to the downtown work zones. Setting up a transit alignment that meets technical, social, economical, and environmental considerations is a challenge. This study coupled the technologies of analytic hierarchy process (AHP) and geographic infor- mation systems (GIS) to assist in the decision-making process of deter- mining optimal transit alignment between the competing Salt Lake and Airport alignments in Honolulu. The goal of the AHP structure was to determine the location for the transit alignment: technical, social, eco- nomical, and environmental considerations are the criteria, and suitabil- ity levels are the alternatives. A uniform grid structure was developed: thematic maps related to the independent variables were converted into a 0.1-square-mile grid map. The grid data and survey results were entered into the AHP structure to produce an index of suitability that could be plotted in the GIS environment to indicate optimal alignments for the rail transit system on the basis of public preferences and technical criteria. Like other urban areas, the City and County of Honolulu, Hawaii, struggles to solve transportation problems. With a population of 876,156 persons and an average density of approximately 1,500 per- sons per square mile in 2000, Honolulu is one of the most densely populated cities in the United States. In some areas, such as in Waikiki and along the H-1 freeway, the density is more than 10,000 persons per square mile. Traffic congestion has become common in Honolulu. The H-1 freeway connecting both the leeward side of Oahu and Hawaii Kai to downtown is congested during the morning and afternoon peak travel times. The City and County of Honolulu has proposed building a mass transit system that will have the ability to move people in the con- gested east–west corridor between Kapolei, Ala Moana Shopping Center, and the University of Hawaii at Manoa and possibly to the Hawaii Kai area. The corridor will become the alignment or route for the transit system. The location rationale is based on several fac- tors, including population density, location of employment, land use, locations of commercial and other public activities, connectiv- ity to different transportation modes, walking and cycling transport conditions, and suitability for park-and-ride facilities. The rail should Department of Urban and Regional Planning, University of Hawaii at Manoa, 2424 Maile Way, No. 107, Honolulu, HI 96822. Corresponding author: I M. Brunner, ibrunner@hawaii.edu. Transportation Research Record: Journal of the Transportation Research Board, No. 2215, Transportation Research Board of the National Academies, Washington, D.C., 2011, pp. 59–66. DOI: 10.3141/2215-06