It is commonly accepted in the transportation planning community that local data is the only legitimate data that can be used to forecast travel. Therefore, large, highly aggregated data sets are often discredited and deemed an inappropriate data source without extensive consideration. Additionally, since freight flow databases are generally large and highly aggregated, freight transportation traditionally has not been explicitly included in the process; rather, freight is implicitly included through the application of a factor related to passenger travel. This paper presents the results obtained modeling freight transportation in an urban area using a highly aggregated, publicly available, freight flow database, known to have limitations. The paper discusses approaches to maximize the use of the aggregated freight flow data at various scales, disaggregation factors included in the process, and methods to overcome known limitations as part of the methodology to format the data for entry into a traditional travel model. The paper applies statistical validation techniques proving the freight volumes obtained from the aggregated data within a traditional transportation model do, in fact, provide reasonable matches to the existing counts, demonstrating that rejection of the data is not warranted. The paper concludes that a highly aggregated freight data set can be used in transportation planning activities, achieving acceptable levels of accuracy. Use of the highly aggregated data set, considering the result of a validated and acceptable model, is a preferable outcome to the options of ignoring freight in the modeling process or accepting freight is simply a portion of passenger travel.
Developing Validated Freight Transportation Models Utilizing Highly Aggregated Data – Harris.doc