(Georgia Institute of Technology, 2022-05-25)
Brandel-Tanis, Freyja Alice
In March 2017, an overpass on I–85 in Atlanta caught fire and collapsed, disrupting traffic for 43 days while the Georgia Department of Transportation (GDOT) rebuilt the bridge. During this time, transit ridership increased as commuters reacted to the changes in travel time, thanks in part to concerted efforts to expand service by the Metropolitan Atlanta Rapid Transit Authority (MARTA). Ridership declined after GDOT restored service but remained higher than pre-disaster levels, requiring further research to understand how long the effect lasted.
Multiple linear regression models are used to investigate the relationship between 2019 ridership and origins and destinations affected by the bridge collapse. Travel time matrices from the Atlanta Regional Commission’s (ARC) Activity Based Model (ABM) are used to identify Traffic Analysis Zones (TAZs) with notable service impacts and choose comparable regions. The weighted trip counts from the ARC’s 2010 and 2019 onboard transit surveys map transit trips to origin and destination TAZs. When controlling for MARTA’s service quantity, residential and employment population, and the percent of households without access to a vehicle (choice riders), the models found a significant relationship between the region impacted by the bridge collapse and an increase in MARTA rail trips and MARTA trips by patrons who could have used a vehicle. A significant increase in choice rail ridership from the impacted TAZs, those most likely to have switched during the network disruption of 2017, suggests that the bridge collapse’s impact on MARTA riders lasted until at least the fall of 2019, over two years after the inciting network disruption.