An important linkage covered in this course was that of combining the visual mapping superiority of GIS models with the computational powers of the more sophisticated market-based, spatial competition models. Using GIS models as preprocessors and postprocessors to models like MEPLAN has several advantages. First, it allows for better presentation of outputs, more disaggregate analysis (parcels rather than zones or large grid cells), and better assessment of environmental impacts such as the development of sensitive areas. As a preprocessor, the GIS models could be used (and have been used in the Sacramento area) to input general plan designations, zoning, built land, and land availability for each MEPLAN zone.
The paper for this week was on MPOs using PECAS and GIS models for policy analysis and what important policy outputs would be represented and which ones would not. Meanwhile, the students started working on the Final Paper, which was to perform a more complex policy comparison using MEPLAN with the now-fixed GUI.
In , Clay lectured on part of his dissertation research proposal, which is to estimate a probabilistic or logit model of large development projects on the urban edge. MEPLAN, PECAS, and UrbanSim all simulate development of floorspace in small increments in each zone or cell. This type of modeling misses those lumpy events where powerful developers get rezonings of agricultural lands for large projects on the urban edge. These large projects then affect subsequent development patterns. Johnston lectured on pathways for MPO development of travel models, separate land use models, and eventually urban models, as applied to small, medium, and large MPOs in the U.S. This concept is based on Hunt et al. (2001).