Modelling chronic winter hazards as a function of precipitation and temperature

In previous research, winter weather conditions like precipitation and cold were recognized as factors affecting the change in traffic volume, and a simulation has been conducted to model the effect of weather factor on variations in traffic flow. A winter weather hazard model considering natural chronic hazards in the winter season has been proposed. To achieve this purpose, this research formulated a dummy variable winter weather traffic model while considering explanatory variables such as expected daily volume, snowfall, and temperature. This model was derived using six-year traffic data that were collected on a weigh-in-motion site on Highway 44, in Alberta, Canada. Precipitation and cold data collected from a weather station were linked with traffic data to model traffic variations caused by weather factors. The performance of the model, on the ground of temporal transferability and model specification, was tested using data from different years. The temporal transferability test confirmed that the model can be successfully applied regardless of the year. It was also revealed that each vehicle class prefers a different model specification in estimating correct traffic volume.