Global trade patterns are continuously changing as economies and trade policies develop and interact. Recent developments and forecasts suggest changes in such patterns are likely to continue. Global trade patterns ultimately manifest themselves in freight flows on global and domestic transportation systems, but the translation of economic flows into transportation patterns is not straightforward. Moreover, while countries may benefit from global trade, the transportation impacts are felt locally, as passenger and freight movements compete for domestic infrastructure capacity.

This thesis introduces a joint transportation and trade modelling framework to analyze the effects of changing global trade patterns on domestic freight operations. A unique multi-scale modelling framework confronts the notion that it is convenient but unrealistic to draw a geographic boundary around the economy and freight transportation system. Innovative harmonization, transformation, estimation, and optimization-based methods are developed to jointly model trade at the global, national, and regional levels. Multinomial logit (MNL) models are used to include the influence of transportation disutility in global trade choice behaviours, extending the multi-scale model into the domain of random-utility-based multi-regional input-output (RUBMRIO) models, which are first comprehensively introduced and reviewed.

To demonstrate the feasibility of the proposed approach, a Canada-centric model was developed that includes forty-eight countries, Canada's ten provinces and three territories, and Ontario's eleven economic regions. Validation results show that it is possible to link spatial scales with a reasonable degree of accuracy. And while Canada's industrial outputs are the sum of its provincial and regional outputs, individual provinces and regions are not just microcosms of the larger country's behaviour. Example applications of food and paper product demand shocks from the United States, as well as doubling and halving of global transportation costs, demonstrate that the economic impacts from scenarios related to global trade and their implications for freight demand and traffic patterns differ for each province of Canada and region of Ontario.