Citation

Qu X, Xiao X, Zhu X, Wu H, Gong J, Li D. Int. J. Appl. Earth Obs. Geoinf. 2023; 123: e103472.

Copyright

(Copyright © 2023, Elsevier Publishing)

DOI

10.1016/j.jag.2023.103472

PMID

unavailable

Abstract

Road traffic safety issues pose a serious threat to urban health, and how the built environment affects traffic crashes has been hotly discussed. However, there are limitations in the selection of built environment factors according to the classic 5Ds system in most previous studies. Few studies have considered the influence exerted by visibility factors in intersection level and road diversity at the community level on traffic crashes, and the multi-level structure of crash data may be ignored. Therefore, the proposed visibility analysis method considering visible triangles and visual sensitivity is used to quantify the visual environment features of the intersections. Based on the proposed six-dimensional factors (6Ds) of the built environment (considering the intersection visibility degree and road diversity) and the hierarchical linear model (HLM), this paper proposes the 6Ds-HLM framework to analyze how road traffic crashes are influenced by the built environment factors. The results show that (1) the full model of the proposed 6Ds-HLM which contained the variables of intersection level and community level considering road diversity and the cross-level interactive effect performed best. (2) Interactive effects between intersection visibility and community built environment are founded, and significant 6Ds variables are identified using HLM: a) There was a positive cross-effect between intervisibility and road diversity. The viewshed area of the middle visual sensitive region also presented significant positive cross effects with traffic congestion speed, entertainment venue density, and road diversity. b) There was a significant negative cross-effect between the viewshed area of the middle visual sensitive region and school density.

Language: en

Keywords

Built environment factors; Hierarchical linear model; Interactive effect; Intersection visibility analysis; Traffic crash