El-Khodary, Ihab Ahmad Fahmy2006-07-282006-07-2819971997http://hdl.handle.net/10012/72Canada and the U.S.A. are the world's largest trade partners and 60 percent of this trade is between Ontario and the U.S.A. The volume of this trade between the two countries is expected to increase four-fold by the year 2020 over the 1989 levels. This transborder movement is primarily served by the trucking industry. Therefore, the expected increase in commodity flow will induce an increase in truck traffic. Consequently, congestion at border crossings could result, as well as increase in energy consumption and air pollution levels. In order to forecast the commodity flows and hence address future congestion and pollution concerns, it is necessary to determine the factors and variables influencing these flows within the region under study. The research described in this thesis investigated the generation and distribution factors of shipments for the for-hire trucking industry, and the mode choice variables between rail and truck for the transborder flows. This was achieved through implementing the multi-stage model approach which focuses on transport submodels. Several data sources were employed in the analyses which included: (i) demographic and international trade data from Statistics Canada's publications, (ii) commodity flow data for the for-hire trucking industry compiled by Statistics Canada from the "Annual Motor Carrier of Freight Survey", (iii) flow data between Ontario and the American states by mode from the U.S. Bureau of Census, and (iv) mode specific data collected through a self designed survey questionnaire. The commodity flow data used was in terms of shipments which is basically a quantity of goods transported by a carrier, and therefore could vary from a few kilograms to several truckloads. The analyses were performed for the top five commodity groups exchanged between Ontario and the U.S. and these were chemical products (VI), pulp of wood, paper and paperboard (X), base metals (XV), machinery, mechanical appliances and electrical equipment (XVI), and vehicles, transport equipment and parts (XVII). The for-hire truck shipment flow structures indicated that Southern Ontario is a centre of shipping activities, with Toronto being the nucleus. The principal trucking spatial linkages between Ontario and the U.S.A. are with the crescent of states accessible through the Niagara frontier and the Windsor-Detroit gateways. Three factors have been identified to influence the shipping distribution patterns of commodities, which are: nature of products, demand for final products, and strong inter-industry linkages. Thus, normal transport models such as the gravity model were not capable of replicating the existing flow distribution behaviour. Therefore, such models were determined to be inappropriate for use in estimating future truck volumes, whereas a Fratar type expansion of existing commodity interaction patterns was more appropriate. Generation models for estimating the number of for-hire truck shipments produced and attracted in terms of several socioeconomic variables. The best forecasting variables included: - population for the consumption of commodity group XVI, - total labour force for the production of commodity group XVI and the consumption of commodity groups VI, X and XVII, - manufacturing labour force for the production of commodity groups VI, X, XV and XVII, and - construction labour force for the consumption of commodity group XV. Therefore, regression models in terms of socioeconomic variables could be calibrated and used to forecast zonal truck shipment productions and attractions. Finally, analysis of aggregate mode share data revealed that commodity type and volume influenced the mode choice, whereas there was no evident relationship between rail share and both distance and seasonal variation. Also, based on data acquired through a shipper survey, it was determined that auto parts manufacturers in Ontario consider just-in-time, transit reliability and freight cost as the most important factors that influence their mode choice decisions. Unfortunately, the unavailability of sufficient data did not permit capturing these variables within a mode choice model.application/pdf16786737 bytesapplication/pdfenCopyright: 1997, El-Khodary, Ihab Ahmad Fahmy. All rights reserved.Harvested from Collections CanadaDoctoral Thesis