Using GIS-based measurements and MLR for understanding the effect of street network characteristics on walking

The walkability of the street networks is critical to improving the walking trips and health levels of urban residents. However, few studies have correlated the quantitative evaluation of street network characteristics with pedestrian spatial perception, and integrated it into a composite pedestrian environment to determine their effects on walking. Using the old city of Fuzhou as a case study, this study aims to deepen the understanding of the effect of street network characteristics on walking by using GIS-based measurements and multiple linear regression (MLR), meanwhile, a multi-source dataset was used to quantify the characteristics of pedestrian environments. The key findings suggest that street network characteristics had a potential driving effect on walking; walking volume on the weekday was negatively associated with Straightness (a measure of street network centrality) and positively associated with the density of bus stops; walking volume on the weekend was positively associated with population density and Reach (a measure of street network accessibility); the density of street length and street intersection, land use mix, configuration of park green space were statistically insignificant factors. These can help provide a basis for more targeted walkability enhancement in urban and street planning and promote the utilization of GIS and other technical methods.