Spatial Clustering of Pedestrian Activity and the Built Environment Characteristics

Abstract

Promoting pedestrian activity has attracted increasing attention as an important strategy for the improvement of public health and urban revitalization. The impact on physical activity underpinned by built environment has been studied substantially; however, few studies had focused on the geographically varying relationships between pedestrian activity and the built environment characteristics. Built upon previous work, this study looks at the spatial patterns of pedestrian counts and the built environment contributors along two major streets in Buffalo, New York using global and local spatial autocorrelation tests and geographically weighted regression. Pedestrian generators, job density and land use mix are included as independent variables in order to study the impact on them due to the characteristics of built environment. Our findings suggest that (1) there are statistically significant clusters of street intersections with high pedestrian counts along the streets selected in our study; (2) there are some optimal sizes of clusters of pedestrian generators, which attract more pedestrians; (3) geographically weighted Poisson model helps to analyze the geographically varying relationships between the built environment and pedestrian activity with a more pronounced goodness of fit. This research contributes to the understanding of the spatial patterns of pedestrian activity and the geographically varying relationship between the built environment and pedestrian counts. Hopefully this research will help to guide and focus the minds of policy makers and urban planners alike to introduce street vitality through the modifications of the built environment, so as to improve the quality of life in their neighborhoods.