Abstract Manager

The current traffic growth forecasting approaches are based on a variety of approaches, including historical growth factors, travel survey-based approaches, and fuel consumption-based approaches. The historical growth factor approach is revealed to be the most popular method in different jurisdictions across North America. However, the main limitation of this approach is that growth factors are not sensitive to changing socioeconomic conditions over time, while it is well-known that demographics and economic indices are the key factors that contribute to the continued growth in traffic. To address these challenges, an innovative study was carried out to develop and implement a new approach to forecasting traffic growth rates in the Regional Municipality of Peel, Ontario. The regression techniques model was selected as the preferred approach, which can describe the relationship between traffic growth rate and its driving forces, such as historical traffic volume data, socioeconomic data, including population and employment estimates per Traffic Analysis Zone (TAZ), road classification, and the latest road network shapefile. As part of the modelling exercise, a vector of trip distributions from the Region’s EMME model was utilized to accurately capture the impact of independent variables within each TAZ. Using the regression technique, several link-level and TAZ-level models were created. The models’ goodness-of-fit was assessed using different criteria, such as the coefficient of variables, R², and the magnitude of error terms. The selected link-level and TAZ-level models revealed R² values, ranging from 76% to 91%, with high levels of confidence intervals and low margins or errors. In the final step, the traffic volumes and annual growth rates were estimated for the 5-, 10-, and 20-year horizon. After calibrated models were developed, a web-based GIS tool was developed to visualize the annual growth rates on the Regional roadway network.

This study was the first of its kind, aiming to create a robust model to predict the traffic growth rates within a large-scale municipality in North America. This unique characteristic of such modelling approach enables the Region to forecast traffic growth rates and background traffic growth rates for specific development projects, without the need to re-run the EMME model. The developed regression model can be recalibrated for other municipalities in Canada using the most up-to-date information such as traffic volumes as well as demographic and socioeconomic data. The GIS-based platform can provide a unique opportunity for the end-users to review the growth rates within their jurisdiction and make informed decisions.