| Abstract |
Extreme weather conditions can adversely impact transport networks and driver behaviour, leading to variations in traffic volumes and travel times and increased accident rates. Emergency services that need to navigate to an accident site in the shortest possible time require real-time location-based weather and traffic information to coordinate their response. We therefore require historical and high-resolution temporal real-time data to identify districts and roads that are prone to different types of incidents during inclement weather and to better support emergency services in their decision-making. However, real-time assessment of the current transport network requires a dense sensor network that can provide high-resolution data using internet-enabled technology. In this research, we demonstrate how we obtain historical time-series and real-time data from sensors operated by the Tyne and Wear Urban Traffic and Management Control Centre and the Urban Observatory based at Newcastle upon Tyne, UK. In the study, we assess the impact of rainfall on traffic volume and travel time, and the cascading impacts during a storm event in Newcastle during early October 2021. We also estimate the economic cost of the storm, with regards to transport disruption, as the cost of travel, using the “value of time” based on Department for Transport guidelines (2021). Using spatial-temporal analysis, we chose three locations to demonstrate how traffic parameters varied at different times throughout the storm. We identified increases in travel times of up to 600% and decreases in traffic volume of up to 100% when compared to historical data. Further, we assessed cascading impacts at important traffic locations and their broader implications for city areas. We estimated that the storm's economic impact on one sensor location increased by up to 370% of the reference value. By analysing historical and real-time data, we detected and explained patterns in the data that would have remained uncovered if they had been examined individually. The combination of different data sources, such as traffic and weather, helps explain temporal fluctuations at locations where incidents were recorded near traffic detectors. We anticipate our study to be a starting point for stakeholders involved in incident response to identify bottleneck locations in the network to help prepare for similar future events. © 2023 |
