When you think of apples the first thing that enters your mind is probably not TMODEL2. However, that is exactly what we used to model the flows of agriculture produced in our Washington State Counties.
Last year, Bucher, Willis & Ratliff was contracted to develop a travel model for the Quad County Regional Transportation Planning Organization (RTPO). RTPOs have been created in Washington State to carry out the transportation requirements of a state-wide growth management act. One requirement is the development of a coordinated comprehensive transportation plan.
Kittitas, Grant, Lincoln and Adams Counties comprise the Quad County RTPO. The Quad County area has a population of nearly 115,000 spread over 9,200 square miles. The area is primarily of a rural character. Some of the finest agricultural products produced come from this part of the country, including asparagus, potatoes, wheat, barley, and, of course, apples.
The transportation system provides diverse opportunities for the movement of goods and people. The highway system includes Interstate 90 which travels through all four counties linking many communities in the region with Spokane and Seattle. Other major routes include U.S. Highways 2, 97, and 395, along with a number of state highways and local routes. A rail system is also in place to ship freight. The Columbia River provides a third major option for freight movement.
Nearly all travel forecasting methods are based on the concept that travel demand and the transportation facility supply interact in a transportation network providing a travel flow pattern. Travel demand models, like TMODEL2, use mathematical formulations that relate estimates of travel to variables which characterize the system. The travel model produces demand estimates which are typically distributed and assigned to a network representing streets, highways, and other modes.
TMODEL2 was used to produce traffic volumes on the Quad County street and highway network. The more traditional urban modeling variables such as households and employment were inventoried and coded to the zone system. It became clear that because of lower population and employment levels, transportation priorities would not be identified by examining volume- capacity relationships on links or nodes. Rather, the RTPO needed to know what major routes needed to be maintained or upgraded to accommodate the large number of truck movements. A second larger issue involved the impacts on the road system through larger modal cost shifts which could encourage movement by barge or by train.
Freight movements were simulated by calculating the origins of agricultural truck trips by estimating acreage and calculating the number of truck trips needed to ship the goods from farm to storage. From there, the movement from storage to processing was calculated by examining the amount of goods shipped by truck, by barge, and by train to destinations in Spokane, along the Columbia River, Seattle, Moses Lake, Wenatchee, and elsewhere. This process was modeled for dry crops and irrigated crops separately, due to their different characteristics. Recreational travel was also modeled because recreation and tourism activities in the area generate a large number of trips which either originate or are destined within the region.
Trip distribution required a strong information base of the location and amounts of goods being shipped. The non-truck trips were assigned to the transportation network which included water as well as highway and rail modes. The results were compared to state and county traffic counts and truck counts. The model was calibrated to not only produce accurate total traffic volumes, but to also reflect accurate truck trip movements.
The future year assignments included estimation of increases in agricultural production, changes in distribution and in mode. For example, an increase in barge shipment from the Columbia River ports of Pasco and Richland was examined as were increases in rail shipment from storage facilities. Other impacts included examining traffic flows resulting from additional economic development in a number of communities in the region. The application of TMODEL2 for non- urban and for freight modeling was a success. Future year highway improvement locations were identified. The model also provides a policy tool to test multimodal impact. Similar to urban travel modeling, the keys to a successful model are good socio-economic data (or agricultural data), a well defined zone and highways system, and knowledge of the travel distribution patterns. So, while TMODEL2 may not have a strong tie to the apple industry, we found that in this instance it was a valuable tool.