Understanding the spatio-temporally heterogeneous effects of built environment on urban travel emissions

Abstract

Transportation has become one of the fastest-growing fields for greenhouse gas emissions. It is important to promote the coordinated development of cities and transportation. To deeply understand the emission distribution for urban travel, this study first applies a map matching algorithm to correct the vehicle Global Positioning System (GPS) trajectories on the road network and calculates the travel emissions by the COmputer Programme to calculate Emissions from Road Transport (COPERT) model. Most studies on the impact of built environment on travel emissions only consider the spatially heterogeneity of variables. Although the traditional GTWR can consider the spatio-temporally heterogeneous effects, the spatio-temporal bandwidth is selected at the same scale in the operation process, which limits the relationship analysis among diverse variables. Therefore, this study adopts a unilateral geographically and temporally weighted regression model (UGTWR) and its multiscale extended model (MUGTWR) to estimate the spatio-temporally heterogeneous effects of the built environment on urban travel emissions. In addition, considering the scale effect and zoning effect of the spatial geographical unit selection on the results, we conducted experiments separately on three spatial units: sub-district scale, Traffic Analysis Zones (TAZs) and grid scale to compare and obtain the most appropriate partitioning schemes. The results show that, the fitting effects of UGTWR and MUGTWR are better than that of GTWR, indicating that considering flexible bandwidth selection can effectively improve the accuracy of simulation. Moreover, the temporal and spatial bandwidth values of each independent variable calculated by MUGTWR reflect the heterogeneous characteristics in space and time of different built environment factors, which can provide scientific suggestions for formulating urban planning. The research results can provide planning strategies for optimizing the allocation of local transportation resources and guiding low-carbon travel behaviors.