Developing County-Level Commodity-Flow Models Incorporating Land-Use Characteristics and Economic Factors for Utah

Abstract

Reliably forecasting freight demand has become essential in freight transportation management to systematically plan for future transportation facility needs. Commodity-based freight demand modeling techniques have been in the mainstream because these techniques overcome the weakness of applying a typical four-step demand forecasting modeling process to freight transportation and capture more accurately the fundamental economic mechanisms that drive freight movement. However, none of the existing models actively use commodity-flow data that have become available to public through the Internet in the past decade. The authors used such data and found good correlation between commodity-flow and aggregate land use type in a previous study. However, it was found that aggregate land-use type data were not readily available to public. Instead, data that describe characteristics of particular land-use such as employment data, business pattern data, and wage data were readily available. Hence, models that use land-use characteristics instead of aggregate land-use type data were developed in this study by using the commodity-flow data available from the 2002 Commodity Flow Survey conducted by the Bureau of Transportation Statistics and U.S. Census Bureau, and by using other data available to the public through the Internet. These models were used to estimate commodity flow in 2007, and their estimates were compared with commodity-flow values found in the 2007 Commodity Flow Survey. Results of this comparison showed that the models could produce decent estimates of commodity flow at the county level and be of practical use because the model inputs are available free through the Internet.