REVISION OF TRAVEL DEMAND MODELS TO ENABLE ANALYSIS OF A TYPICAL LAND USE PATTERN

Among the problems associated with the traditional four-step transportation demand modeling process is the inability of the models to evaluate innovative changes in the factors affecting travel behavior. These factors may include travel demand management methods, atypical demographic changes, and alternative urban design and land use options. For the most part, existing models are capable of using only variables easily measured and typically associated with travel behavior such as travel time, cost, and socioeconomic measures. Since the models are estimated from travel survey data, they are designed to analyze travel demand within the range of behavior reported in the survey. In Portland, Oregon, alternative land use patterns for future suburban growth, travel demand management measures, and public transportation improvements are being considered as a way to slow growth in automobile travel. The land use options are different from typical suburban development patterns, and the existing travel demand models for Portland, developed by the Metropolitan Service District (Metro), were not designed to analyze the travel demand effects of such developments. As part of the project Making the Land Use, Transportation, and Air Quality Connection (LUTRAQ) for 1000 Friends of Oregon, an environmental advocacy organization, Cambridge Systematics and Metro revised the travel models to include variables representing characteristics of that are different from typical suburban land use. The variables tested include measures of residential and employment density and heterogeneity and of the quality of the pedestrian environment. Models revised included mode choice (including walk/bike, auto, and modes) and auto ownership. The original models had been calibrated to replicate the survey data across a range of different variables including location, trip length, and various household characteristics. The additional variables used in the revised models improved the ability to replicate the survey data, not only across ranges of these variables but also across ranges of employment density and pedestrian environment quality in the Portland area. The revised models were used to analyze the effects of the Alternative in suburban Washington County, consisting of alternative land use patterns such as transit oriented developments (TODs), improved service, and travel demand management actions designed to discourage commuting by single occupant automobiles. The LUTRAQ alternative was compared to other scenarios, including a no action alternative. The results from the model runs demonstrated that the LUTRAQ alternative was successful in reducing automobile travel and increasing walking, bicycling, and use compared to the other alternatives.