Effects of Covid-19 pandemic restrictions on zonal transit demand: Evidence from a low-density city

Abstract

Studies investigating the effects of COVID-19 pandemic on transit ridership of low-density cities are scarce. There exist three unanswered questions in the context of low-density cities: (i) how much patronage losses can be attributed to pandemic restrictions? (ii) which land use zones are more vulnerable to demand declines? and (iii) what factors contribute to zonal vulnerability? The present study intends to answer these questions by investigating zonal level boardings of bus transit system in the city of Winnipeg, Canada. For this purpose, two multivariate adaptive regression splines (MARS) models were developed: (i) a time-series MARS model based on historical transit demand patterns at zonal level, and (ii) a regressive MARS model to predict demand decline as a function of land use, socio-demographic, and zonal-level variables. The magnitude of the demand decline was found to be highest in April 2020, with a total loss of 1.74 million boardings attributable to the COVID-19 pandemic. Among 840 zones, transit usage in commercial (50 %–60 % reduction), education (80 % reduction), and recreational (60 %–80 % reduction) zones are most affected by pandemic restrictions. The findings are valuable for transit officials of low-density cities to effectively plan response strategies for long-term operational disruptions due to pandemic situations.