|
Research Project Description |
|
Beyond Uncertainty: Urban Models in Transportation
and Air Quality Planning
Project Number: 2407Principal Investigator:
Principal Investigator: Dr. Caroline Rodier, Mineta Research Associate,
Post-Doctoral Researcher, UC Berkeley
Institution: Telephone Number: Email Address: mti@mti.sjsu.edu Project Objective: This will be a three-part study that expands on and
synthesizes the finding of the Sacramento case study by addressing key
planning and policy questions surrounding uncertainty in travel and land use
models. Part One would employ a series of expert
interviews with federal, state, and local government officials and other
stakeholders involved in transportation, land use, and air quality planning
to identify (1) the critical information needed from the literature of
modeling uncertainty and (2) key barriers and potential incentives to improve
models and modeling practices. Part Two would mine and synthesize the results of the
literature on uncertainty in land use and travel models (i.e., the
Sacramento case study and other identified studies) in a manner that is most
relevant to officials and other stakeholders (based on part one). Results, for example, could include the
range of total error, relative importance of sources of errors, the
contribution of model attributes (and/or improvements), and the implications
of these results for policy and regulatory analysis. Part Three would illustrate an innovative modeling approach to policy
analysis that could enable stakeholders to think creatively about their policy
values, goals, and strategies without being distracted by technical debates
on the adequacy of the available tools. This approach would involve establishing
a policy benchmark (e.g., 10% transit mode share) that is significant within
the range of known model error and then working backwards to specify the
model inputs (e.g., transit expansion) to achieve the benchmark. The study
would use the SACMET and/or the MEPLAN model, confidence intervals from past
model validation studies, and community input for the policy benchmark.
Abstract: The transportation-related air quality
problems that travel and emissions models address are critical. Approximately
133 million Americans live in metropolitan areas with air pollution levels
above National Ambient Air Quality Standard (EPA, 2001). The 1990 Clean Air Act Amendments and the
resulting conformity regulations rely on travel and emissions models to be accurate
enough to demonstrate that regional transportation plans, which have 20-year
time horizons, conform to the emissions budgets set out in the approved state
implementation plans. However, it is
widely acknowledged that the forecasts produced by travel and emissions
models are significantly inaccurate, and thus these regulations assume an
implausible level of accuracy. Because of the limitations of travel
demand models and recent changes to regulatory requirements, state and
regional governments across the U.S. are beginning to implement more advanced
land use models and travel demand models (e.g., Sacramento, CA;
Springfield-Eugene and Portland, OR, and Salt Lake City, UT). TEA-21
urges transportation planning to consider the effects of transportation
policy decisions on land development. Conformity regulations require a
logical correspondence between future land use projections and transportation
plans in serious or worse non-attainment regions. Increasingly, citizen
groups in regions considering new beltway freeways and/or significant transit
investment want to understand the land development effects of proposed
projects. Land use models, however, are subject to many of the same sources of inaccuracy as travel demand models. Given the complexity and data requirements of these models, it may not be unreasonable to expect that their uncertainty may be equal to or greater than that of travel demand models. Moreover, it is possible that theoretical improvements with respect to the representation of the land use and transportation interaction in the simulation methods could be swamped by the errors of a more complex model set. Until recently, however, very few studies have been
conducted to quantify errors and their sources in travel and land use models,
the policy implications of these errors, and/or the respective advantages and
disadvantages of the different model capabilities. Indeed, uncertainty in models has traditionally been ignored
not only in the transportation profession, but in policy analysis in general
(Stopher and Meyberg, 1975; Hartgen, 1995; Morgan and Henrion, 1990). The Mineta Transportation Institute has
contributed funding to a multi-year case study of uncertainty in a
state-of-the-practice travel demand model (SACMET) and an advanced integrated
land use and transportation model (MEPLAN) in the Sacramento region. Both are official models of the region’s
MPO. Sensitivity analyses, validation
tests, and scenario comparisons have been systematically applied to both
models to assess total model error, identify the magnitude of key sources of
error, and compare the benefits of specific modeling capabilities for policy
analysis. This work has produced a
number of publications and reports (Rodier, 2002 & 2004; Rodier and
Johnston, 2002; Rodier et al., 2001, 2002a & 2002b; Hunt et al.
2001). In addition, a number of case
studies on uncertainty in travel and land use models have been conducted, for
example, in cities in Oregon (Condor and Lawton, 2002; Pradhan and Kockelman,
2002; and Waddell, 2000) and Vermont (Marshall and Grady, 2001). Description and Project dates:
|
|
