Urban and metropolitan models
Urban and metropolitan models is a general category incorporating travel demand models for an entire city or metropolitan region. These models may take many forms. Increasingly there is divergence in modeling approaches between models developed for and applied in small and mid-sized regions and those developed for and applied in the largest U.S. cities. One common feature is that these models are designed to forecast transportation demand across several modes (not simply auto demand) within a city or metropolitan region. Urban and Metropolitan models are primarily focused on ground transportation demand generated by the inhabitants within these regions, although these models do account for external travel and commercial goods movements (typically truck movements) at different levels of sophistication.
# Background and History
A brief history of urban and metropolitan modeling. In the 1950s, agencies were created, often as offshoots of state or county-level highway departments, to forecast travel demand at the regional level. This was viewed as particularly critical because the Interstate Highway System was being built and the significant impacts on regional travel patterns were becoming apparent. J. Douglas Carroll was one of the key figures in establishing metropolitan-level travel forecasts beginning in the early 1950s. These models were early versions of four-step models.
The scale of metropolitan models increased during the 1970s, and efforts were undertaken to develop land use models, most of which were deemed unsuccessful until the late 1980s. The choice models at the heart of the four-step model became more elaborate and took transit and, later, active modes (i.e. walking and cycling) into consideration. With the increase in computing power, the most complex models have moved from aggregate trip-based models into Activity-based models that incorporate micro-simulation techniques.
# Governmental Context
In the United States, Metropolitan Planning Organizations are responsible for the use of (and maintenance) of regional travel demand models. Many urban regions in Western Europe, Canada and increasingly Asia have one agency responsible for regional transportation planning and maintenance of regional travel demand models. Urban and Metropolitan models are designed to guide decision makers about the impacts of future transportation plans, particularly in the context of regional transportation planning.
In the U.S., the outputs of these models are also used to assess whether regions are making progress towards their air quality and greenhouse gas emissions targets. As urban and metropolitan model have been asked to inform more complex policy questions, they have evolved and increased in complexity. Urban and regional planning agencies operate within a regulatory framework that, in the U.S. at least, is largely governed by federal law. Due to the need to inform stakeholders on a wide range of policy questions, it is not uncommon for urban and metropolitan transportation models to be designed to provide information above and beyond the minimum requirements set out in federal law. Decision makers have continued to increase the number of policies that they expect to be informed by their metropolitan transportation models.
# Current Practice
In North America, the regional models for majority of small and mid-sized MPOs are 4-step trip-based models. Increasingly, large regions in the U.S. are migrating to activity-based models. In the most general terms, urban and metropolitan models must be populated with trip makers (either through an aggregate zone-based approach or through population microsimulation), the travel demand is generated, the trips (or tours) should be elaborated (involving destination choice and mode choice) and the network demand is assigned. There are a wide variety of options available to urban and regional planners. For example, in North America, destination choice precedes mode choice, whereas many European models reverses this sequence. In addition to the recent state of the practice report (opens new window), many MPOs have extensive documentation on-line, collectively forming an important resource for planners and transportation modelers.
# Modeling issues faced by small- and medium-sized MPOs
Small and mid-sized MPOs face major resource constraints, which include difficulties in staffing, time pressure on staff who have any transportation modeling experience, shortage of local data and a related reliance on pre-existing data sources (particularly from the U.S. Census) and standard model settings embedded in the major demand forecasting software packages.
In 2004, the state of the practice and important issues facing small and medium-sized MPOs was summarized in a peer exchange (opens new window).
# Modeling issues faced by large MPOs
Staff retention (and conversely staff turnover) are often problematic even at large MPOs. Documentation of models is often insufficient for long-term models that have undergone partial upgrades or simply been adjusted in an ad hoc fashion.
Region-specific data collection is usually undertaken periodically, though limited resources (and an increase in non-response rates) often prove to be a challenge. Large MPOs have traditionally sponsored the collection of travel diaries in order to calibrate their transportation demand models, though some MPOs are investigating alternative approaches to gather necessary data.
Large MPOs often face greater scrutiny than small or mid-sized MPOs, and may have additional statutory obligations, particularly if they serve a region classified by the EPA to be a nonattainment area (opens new window). Furthmore, large MPOs occasionally face legal challenges over their planning processes (Garrett and Wachs, 2006). This has generally led to policy makers asking models to answer a wider range of questions about transportation policies and potential interventions. Some of these questions cannot be adequately answered with 4-step models, which in turn has led some large MPOs to move into Activity-based modeling and microsimulation. Large MPOs may also face pressure to develop combined land use-transport models, as the two are closely intertwined and many observers question the utility of modeling transportation in the absence of land use feedbacks.
# Practical Considerations
The best practices in metropolitan tranportation planning was summarized in a 2012 report (opens new window) sponsored by the U.S. DOT and the Volpe Center.
It is fair to say that few MPOs have managed to implement all of the best practices in their travel demand models. This may be due to data limitations, staff and other resource constraints, or simply because certain aspects of "best practice" may have limited relevance for a particular MPO, which is particularly common for small- and medium-sized MPOs. Special Report 288 (opens new window) on the state of the practice in metropolitan modeling was published by TRB in 2007. One common finding was that previous forecasts of basic demographic information such as population and employment were often considerably off the mark, which in turn largely invalidated the transportation forecasts.
While this should not be taken as a how-to manual, this resource page will list key concerns for planners looking to develop urban or metropolitan models, such as whether the zone structure is suitable for the region, what types of information to include on a travel survey, what modes should be incorporated into the mode choice module and so forth.
# Sketch planning
Sketch planning models are used to provide quick-turn-around, order-of-magnitude estimates of a limited, focused range of policy variables and effects. Sketch plan models provide quicker turn-around in general by focusing modeling and analysis on a specific segment of the transportation or travel behavior process, normally without attempting to iterate or equilibrate estimates. Sketch planning tools in use focus on emissions, intelligent transportation systems, tolling or value pricing, urban design, health benefits of active modes, and land use/transportation interactions. Because some sketch planning models are simpler to run and may require less intensive or detailed input data, they are most often used by small and medium sized MPO’s, although they are used by MPO’s of all sizes. They may also be used to reduce the number of alternatives that need to be run through a more comprehensive model for more detailed analysis. However, sketch planning models and tools may not be rigorous enough to be used for environmental analyses and certainly not to conduct air quality and emissions modeling.
# Linkages to state-level models
Although they cover a smaller territory, metropolitan models are frequently more detailed than state-level models, in the sense that certain modes (walk, bike and even transit) may not be covered in a state-wide or state-level model, although commuter and passenger rail and even aviation modes might be included. State-level models often are restricted to modeling transportation flows between metropolitan regions within a state (i.e. long-distance travel), though others seek to capture flows within metropolitan regions (at a somewhat aggregated level). State-level models typically pay special attention to truck flows throughout the state.
# Advanced Topics in Urban and Metropolitan Modeling
Since the late 1990s, some of the larger metropolitan regions adopted advanced models. The most common type of advanced regional transportation model are a type of model called activity-based models. In the U.S., these models were microsimulation-based, starting from a synthetic population and generating individual trip records.
More recently, a new generation of land use models has been developed that operate at the regional scale. Several metropolitan regions have concerted efforts underway to develop a framework where land use models and transportation models are integrated and pass (or iterate) data back and forth.
M. Garrett and M. Wachs. (1996) Transportation Planning on Trial: The Clean Air Act and Travel Forecasting. Sage Publications,
National Academies Committee for Determination of the State of the Practice in Metropolitan Area Travel Forecasting. (2007) Metropolitan Travel Forecasting: Current Practice and Future Direction -- Special Report 288 (opens new window)
U.S Department of Transportation/Volpe Center. (2012) Best Planning Practices (opens new window): Metropolitan Transportation Plans.
M. Wachs. (1995) "The Political Context of Transportation Policy," in Susan Hanson, ed., The Geography Of Urban Transportation, 2nd edition. New York: Guilford Press, Chapter 11, pp. 269-286.