ABSTRACT
Coronavirus Disease 2019 (COVID-19) has become one of the most serious global health crises in decades and tremendously influence the human mobility. Many residents changed their travel behavior during and after the pandemic, especially for a certain percentage of public transport users who chose to drive their owned vehicles. Thus, urban roadway congestion has been getting worse, and the spatiotemporal congestion patterns has changed significantly. Understanding spatiotemporal heterogeneity of urban roadway congestion during and post the pandemic is essential for mobility management. In this study, an analytical framework was proposed to investigate the spatiotemporal heterogeneity of urban roadway congestion in Shanghai, China. First, the matrix of average speed in each traffic analysis zones (TAZs) was calculated to extract spatiotemporal heterogeneity variation features. Second, the heterogenous component of each TAZ was extracted from the overall traffic characteristics using robust principal component analysis (RPCA). Third, clustering analysis was employed to explain the spatiotemporal distribution of heterogeneous traffic characteristics. Finally, fluctuation features of these characteristics were analyzed by iterated cumulative sums of squares (ICSS). The case study results suggested that the urban road traffic state evolution was complicated and varied significantly in different zones and periods during the long-term pandemic. Compared with suburban areas, traffic conditions in city central areas are more susceptible to the pandemic and other events. In some areas, the heterogeneous component shows opposite characteristics on working days and holidays with others. The key time nodes of state change for different areas have commonness and individuality. The proposed analytical framework and empirical results contribute to the policy decision-making of urban road transportation system during and post the COVID-19 pandemic.