Journal of Transportation Engineering · 2001 · 27 citations · 13 references
PlanningTransportation PlanningUrban PlanningSpatial ModelingTransport PlanningTransportation GeographyCensus Data AnalysisTransport ModelingTransport ModellingTransportation Land Use PlanningTransportation-land Use IntegrationUrban TransportBlock GroupsIntegrated Transport-land Use ModelsLand Use DataLand Use InventoryRoadway Network PlanningTraffic AnalysisTraffic Demand ModelingTraffic Modeling
Statewide transportation planning models cover larger areas than urban models, making land‑use inventories and household surveys prohibitively expensive, so researchers use census geography such as tracts or block groups and decennial census data as alternatives. This study investigates which census geography level and centroid placement best support statewide planning, and whether the resulting differences justify the additional effort of using finer TAZs and network detail. Using Idaho’s statewide traffic demand model, the authors evaluate how different TAZ structures and roadway network details influence the model’s ability to replicate annual average daily traffic counts. The results offer a quantitative guide for planners to select an appropriate zoning system for statewide traffic demand models.
Because a statewide transportation planning model covers a much larger area than an urban transportation planning model, land-use inventories and a household survey, which are often used to collect socioeconomic and traffic data for a UTP model, are prohibitively expensive for a STP model. The literature suggests alternatives of using the geography of census such as census tracts or block groups as Traffic Analysis Zones (TAZs) and using decennial Census for the socioeconomic data needed for statewide traffic demand modeling. The use of these alternatives raises new research questions for transportation planners. Which level of the census geography is better for statewide planning? Where should the centroids of TAZs be located given that the TAZs are heterogeneous in land-use activities? How much difference will the change of TAZ size and centroid location make in the results? Would the difference justify the extra effort in using more detailed TAZs and network? The primary purpose of this study is to determine how different TAZ structures and different roadway network details affect the ability of a STP model to replicate annual average daily traffic counts on the network. Idaho statewide traffic demand model is adopted as a case study. Results of this study can provide transportation planners a quantitative guide in selecting a zoning system for their statewide traffic demand models.
13
The Effects of Data Aggregation in Statistical Analysis
William A. V. Clark, Karen L. Avery · Geographical Analysis · 1976 · 233 citations
Computing, Biostatistics, Ecology +15