Spatial analysis and relocation choice modeling of motor freight transportation and warehousing establishments

This dissertation consists of three consecutive studies of the motor freight transportation and warehousing industry, using the New York Metropolitan area as the case study. First, this research systematically explored the spatial distribution pattern or so-called the landscape of the motor freight transportation and warehousing establishments in the New York metropolitan area. It was compared with the landscapes of the industry in other five large metropolitan areas, namely Chicago (Illinois), Houston (Texas), Los Angeles (California), Miami (Florida), and Seattle (Washington). The resulting kernel density maps and the landscape metrics revealed that the landscape of the motor freight transportation and warehousing industry in New York is similar to the one in Houston, showing a uni-centered clustering pattern with the largest dense patch centered at the downtown. Other areas exhibit different patterns due to the different geographical boundaries, transportation network structures, and spatial distributions of major freight activity generators.;Furthermore, a clustering analysis was conducted to identify any possible trends of business agglomeration and to find out why. The Nearest Neighbor Ratio test was used to verify the existence of spatial clusters, followed by two separate clustering analyses for all establishments in the entire study area and the ones in the Manhattan area, the central business district of the area. As found, the establishments tend to cluster, particularly in the Manhattan area, to take advantage of the economies of scale. In addition, the analysis of the accessibility to transportation networks reveals that these establishments tend to be located close to highways and railroads. Moreover, the accessibility to major intermodal freight terminals (such as airports, seaports, railroad and highway terminals) were also analyzed to explore the impact of these terminals in affecting the spatial distribution pattern of the industry.;In addition to the aggregate analysis at the spatial and temporal levels, discrete choice models were developed to identify factors that may explain the relocation decisions of individual establishments. The modeling results identified eight significant factors that affect the relocation decisions, among which the size measure of a location alternative, the accessibility of it to the nearest highway ramp, accessibility to intermodal freight terminals, and the average sales volume of the transportation and warehousing industry have positive effects on the likelihood for an establishment to choose that location alternative, while the distance from the previous location to the location alternative, the labor cost of the location, and the sales volume of the establishment seem to discourage an establishment to relocate.;The findings and implications can be used as valuable inputs to better plan and manage freight terminals and urban distribution centers. In addition, the proposed methods can be integrated into freight demand forecasting to improve the state of the art in freight planning and travel demand forecasting. In terms of methodology, this research is among a very few explorations to introduce the spatial analysis and relocation choice behavior of the freight transportation industry to the freight travel demand forecasting practices.