472 / 2019-02-28 10:47:37

A Grid-Based Method to Understand the Influence of Spatial Factors on Pedestrian Fatalities: A Case Study of Curitiba, Brazil

Road Safety,ArcGIS Spatial Analysis,South America

全体主题 > Spatial data collection, Analysis, and Visualization

Pedestrians represented 34% of all urban displacements in Brazil in 2016 (ANTP, 2016). However, figures demonstrate that the number of pedestrian fatalities is still considerable in most of the country’s largest cities. The main goal of this study is to spatially analyse the pedestrian exposure to traffic accidents in Curitiba, Brazil, by understanding the relationship between the accidents’ density – obtained through the Kernel Density Estimator (KDE) – and the selected spatial factors of the built environment (population den- sity, per capita income, road density and inverted road hierarchy). The pro- posed methodology encompassed the creation of a grid shapefile associating all variables, followed by a linear regression analysis. The results demon- strate that, despite low calculated R2 values, the expected direction of the relationship between variables was verified – a directly proportional rela- tionship with population density, road density and road hierarchy, and an inversely proportional relationship with per capita income – thus reinforcing that the methodology enabled a reasonable spatial association.