Preferred streets: assessing the impact of the street environment on cycling behaviors using the geographically weighted regression

Abstract

Cycling transport systems are an important way to reduce the city’s carbon emissions. Street renovation and renewal policies aim to encourage cycling transport by improving the cycling environment. However, most existing research studies the relationship between the street environment and cycling behavior from a global perspective, ignoring geospatial heterogeneity. Also, methods evaluating the cycling environment based on the frequency of cycling ignore the difference between spontaneous and necessary trips, hiding the problems that exist in streets with a high frequency of cycling. Therefore, the preferred streets index was proposed to evaluate the street cycling environment based on the difference between the cycling trajectory and the shortest path. Geographically weighted regression was used to explore the local effects of street environments on cycling behavior. The experimental results on Xiamen Island show that the type of street and the density of bicycle parking spots have a positive impact on cycling, while the effect of the availability of streetlights, availability of traffic lights, and POI density on cycling was determined by the geographic context of the street. These results provide concrete guidance for improving the cycling environment and enrich the evaluation methods for the cycling environment.