Exploring the Influence of Rainfall on a Stochastic Evolution of Traffic Conditions

Given the two major traffic regimes (i.e., free-flow and congested), dynamic evolution of traffic regimes comprises of four main transition processes; including traffic breakdown (free-flow to congested regime), remaining in the free-flow state, recovery process (congested to free-flow regime), and remaining in the congested state. Understanding these stochastic transition phenomena and their temporal dependence is crucial in developing effective countermeasures to the congestion problem. This study presents an empirical analysis that evaluates the influence of rainy weather on the dynamic evolution of traffic conditions using a Time-varying Markov Chain (TMC) model. The 2015 traffic and precipitation data for the I-295 freeway in Jacksonville, Florida were used in the analysis. Using the Gaussian mixture model, 68 mph and 66 mph were the identified thresholds for free-flow regimes during morning and evening peak periods, respectively. The results from the TMC model suggested that precipitation and traffic flow significantly influence the stochastic dynamic evolution of traffic conditions at a 95% Bayesian credible interval. The breakdown probability increases by 8.49 and 4.7 in the presence of rain compared to the absence of rain for morning and evening peak periods, respectively. For the process of staying in the congested regime, the probability was higher by 2.6 and 1.08 on precipitation than when there is no precipitation during the morning peak and evening peak period, respectively. These findings are anticipated to enhance the understanding of the transition process of different traffic conditions over time, which in turn will facilitate in developing effective congestion countermeasures.

• Supplemental Notes:
  • This paper was sponsored by TRB committee ABJ80 Standing Committee on Statistical Methods.
• Authors:
  • Kidando, Emmanuel
  • Moses, Ren
  • Kitali, Angela E
  • Kwigizile, Valerian
  • Lyimo, Sia Macmillan
  • Chimba, Deo
  • Sando, Thobias
• Conference:
  • Transportation Research Board 97th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2018-1-7 to 2018-1-11
• Date: 2018

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 17p

Subject/Index Terms

• TRT Terms: Markov chains; Periods of the day; Rainfall; Stochastic processes; Traffic congestion; Traffic flow
• Geographic Terms: Jacksonville (Florida)
• Subject Areas: Highways; Operations and Traffic Management; Planning and Forecasting;

Filing Info

• Accession Number: 01659720
• Record Type: Publication
• Report/Paper Numbers: 18-06187
• Files: TRIS, TRB, ATRI
• Created Date: Feb 9 2018 4:52PM