Citation

Kim D, Chanchlani HM. Transp. Res. Rec. 2018; 2672(35): 90-100.

Affiliation

Department of Urban and Regional Planning, California State Polytechnic University, Pomona, Pomona, CA Corresponding Author: Address correspondence to Dohyung Kim: dohyungkim@cpp.edu

Copyright

(Copyright © 2018, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

10.1177/0361198118793496

PMID

unavailable

Abstract

Although it is reasonable to hypothesize that the spatial pattern of pedestrian collisions changes by the time of day, there has been a lack of research that addresses this spatiotemporal variation because of a shortage of pedestrian volume data by time of day. The goal of this paper is to examine the contribution of the built environment to the spatiotemporal variations by employing a big data set of pedestrian count data that is based on cell phone signals on the streets. This paper constructs eight spatial error models that are based on a 3-h time period and a 300 m by 300 m (984 feet) grid cell. Of the twelve built environmental factors, the employment and restaurant factors show consistent correlation with pedestrian collisions. Local roads, office land uses, and bars do not present a correlation. Although the outputs of the models do not show distinct patterns of correlation between built environment and pedestrian collisions by time period, they indicate that the built environment contributes to pedestrian collisions more during nighttime than daytime. This distinction is primarily caused by transportation factors associated with pedestrian collisions when people journey back home during the night period. They include bus stops, rail transit stations, and arterial roads. School is a factor that positively correlates with pedestrian collisions during the school exit time, while population density negatively correlates with collisions during periods with lower pedestrian and automobile volumes. Employment density shows a negative correlation with pedestrian collisions, regardless of the time period.

Language: en