详情

原文传递 Spatial Analysis of Intersection Bicycle and Pedestrian Counts

题名：	Spatial Analysis of Intersection Bicycle and Pedestrian Counts
作者：	Sullivan, J.; Lu, G.; Troy, A.
关键词：	Comprehensive Data Handling System##Spatial analysis##Intersection control##Pedestrian counts##Bicycles##Traffic volumes##Traffic circulation##Geographically weighted regression (GWR)##Geospatial analyses##
摘要：	Encouraging travelers to walk and bike in lieu of motorized modes of travel benefits both the traveler and the community at large. Maximizing these system benefits is critically important for the state and municipalities, especially when funding for transportation is scarce. In order to make better funding decisions for nonmotorized transportation infrastructure, it is first necessary to understand comprehensively the current walking and biking behavior of a region’s inhabitants. This study investigates the linkage between non-motorized traffic volumes and the built environment by focusing on a larger set of road intersection-based counts of the PM-peak hours. The dearth of effective methods to address the spatial dependencies present in these comprehensive data sets motivated this geospatial study to determine (a) whether spatial dependency exists for non-motorized traffic volumes, and (b) whether a significant spatial relationship could be identified between non-motorized traffic volumes and specific built-environment characteristics once the spatial dependency was accounted for. Addressing this nonrandom factor in spatial based counts is an essential step to attaining a robust understanding of bicycle and pedestrian travel throughout a region. Some of the technical information covered in this report was also compiled in a conference paper (Lu et. al., 2012). For a better prediction of motorized and non-motorized travel on multimodal facilities, spatial dependency must be considered because traffic volume at one monitoring station is related to the volume at neighboring stations due to the routing and the continuity in the network due to area-wide traffic circulation and common origins and destinations. A few studies have acknowledged this spatial dependency. Of them, Eom et al. (2006) researched annual average daily traffic (AADT) using spatial Kriging estimation. The spatial model outperforms that of the ordinary least-square (OLS) model. Zhao and Park (2004) analyzed AADT in gridlike networks utilizing geographically weighted regression (GWR) that compensates for spatial dependency by estimating local model parameters. They found GWR models were more accurate than OLS models and useful for studying the effects of the variables at different locations. A smaller group of studies have conducted geospatial analyses of walking and bicycling with appropriate recognition of spatial dependency. Zahran et al. (2008).
总页数：	27
报告类型：	科技报告