Abstract:Interaction exists between land use and traffic congestion. In order to quantify the non-stationary impact of land use on traffic congestion, this paper uses ride-hailing data from Chengdu, which are published by the GAIA open dataset. The study employs the GWR model to reveal this impact. The findings show that as tested by Sigma, AICc, and Adjusted R2, the GWR model is superior to the OLS model in analyzing the impact of the built environment on traffic congestion. Further, spatial heterogeneity exists in both the land use distribution and the traffic congestion distribution. For the selected time period, catering points in suburban areas have little impact on traffic congestion, whereas in the lively region, they have an alleviating effect on traffic congestion. The more concentrated the distribution of companies and residential areas are, the more serious the traffic congestion is, and these effects are more significant in the suburbs.

A Data-Driven Study on the Impact of Land Use on Traffic Congestion Based on GWR Model