Classified Traffic Volume Reduction Modeling under Winter Hazards: Insights from High-Resolution WIM and Weather Data in Alberta, Canada

This study investigates the marginal and combined impacts of temperature and snowfall on classified traffic volumes using 5 years of weigh-in-motion (WIM) and weather data from two highway sites (RD3D and LV4D) along Alberta’s Calgary–Edmonton corridor. Traffic data were normalized using observed and expected daily volume factors, and dummy-variable regression models were developed to estimate weather-induced reductions across total traffic, passenger cars, and trucks. Marginal analyses showed that colder temperatures (CC1–CC7) reduced total traffic and passenger cars by up to 19%, while trucks declined less (6%–15%). Snowfall had a stronger effect: at RD3D, total traffic dropped by 4%–31% and passenger cars by up to 38%, while trucks decreased by 19%; at LV4D, maximum reductions were 27% for total traffic, 35% for passenger cars, and 11% for trucks. Combined effects revealed nonlinear thresholds where losses accelerated sharply. At RD3D, snowfall above 8 cm coupled with extreme cold (CC6–CC7) produced reductions of 49% for total traffic, 72% for passenger cars, and 17% for trucks. At LV4D, similar thresholds occurred between 7 and 8.4 cm, with maximum losses reaching 48%, 64%, and 19%, respectively. Passenger cars were consistently the most sensitive to adverse conditions, while trucks showed greater resilience but still declined under severe cold and heavy snow. These findings identify critical snowfall thresholds, 7–8.4 cm and 10.8–16.2 cm, that are operationally significant for winter road management. The developed models provide transportation agencies with a practical framework to forecast weather-related traffic reductions, optimize snow removal strategies, and enhance roadway safety and resilience during severe winter events.