Demand modeling of successful park and ride planning: multivariate spatial regressive analysis
Park and ride facilities are designed to efficiently intercept traffic flow toward metropolitan business districts and help relieve traffic congestion in the central business areas. An attempt in this research was made to develop a successful park and ride demand model based on the distribution of park and ride usage, by applying geographic information system (GIS) and other spatial statistical packages. The Minneapolis-St. Paul (Twin Cities) Metropolitan Area was selected as a study area for this research because its park and ride system has grown to become one of the nation's largest systems in terms of the number of facilities and total capacity. Recently, the Twin Cities Metropolitan Area started to consider designating large-scale park and ride facilties in the region. There is a need to conduct a research for achieving successful park and ride planning. This research involves multivariate regression demand forecast, spatial cluster identification, and spatial autoregressive analyses. Factors considered in the model for assessing the success and failure of park and ride facilities include socioeconomic characteristics, transportation network features, user behavior, and spatial statistics. These factors are analyzed for predicting the park and ride usages, i.e., the number of parked lots in the park and ride sites during the last visit in FY 2004. In general, the contribution of this research for successful park and ride facilities demand forecast is to integrate spatial association with quantitative statistics by a series of GIS-based statistical techniques for future practice in the field of transportation facility planning.