Sustainability Enhancement of Inclement Weather Dummy Factor Model: A Study of Big Traffic Data in Canada

Abstract

Transportation agencies in cold regions are tasked with developing traffic models and verifying their accuracy. Few prior studies have examined the spatial transferability of model coefficients across different types of highway segments. This study utilizes traffic data from six weigh-in-motion sites in Alberta, Canada, to develop and test winter traffic models for three vehicle types. The research empirically assesses the spatial transferability of a winter model developed for one site by applying its coefficients to other sites with varying road functionalities within the highway network. The findings indicate that winter traffic models can be transferred to other road segments with differing levels of accuracy. This demonstrates that an appropriate model structure can be selected for each vehicle class based on observed road functionality. Consequently, transportation agencies can use and maintain these developed models without the need for additional traffic monitoring sites.