Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration

Office of Planning, Environment, & Realty (HEP)
PlanningEnvironmentReal Estate

HEP Events Guidance Publications Awards Contacts

Handbook for Estimating Transportation Greenhouse Gases for Integration into the Planning Process

Chapter 2 - Overview: Estimating GHG Emissions in the Planning Process

PDF files can be viewed with the Acrobat® Reader®

2.1. Introduction to Transportation GHG Emissions

The transportation sector is one of the largest sources of GHG emissions in the U.S. , comprising 27 percent of U.S. GHG emissions in 2010. 1 In some states, transportation emissions comprise a significantly larger share of GHG emissions and in other states a smaller share. For example, in Washington State, transportation accounted for 45 percent of the state's total GHG emissions in 2008, not because transportation GHG per capita is higher than other states but because GHG emissions from the electricity sector are relatively low due to Washington's heavy reliance on hydroelectric power. 2 The graph below shows that nationally, transportation has historically been the second largest contributor of GHGs behind the electric power industry.

Figure 1. U.S. GHG Emissions Allocated to Economic Sectors, 1990 to 2010

A graph showing U.S. GHG emissions allocated to the electric power industry, transportation, industry, agriculture, commercial, and residential sectors between the years 1990 and 2010. Emissions from the electric power industry have increased from about 1,900 Tg of CO2 equivalent to 2,250 Tg. Emissions from the transportation sector have increased from about 1,500 Tg to 1,750 Tg. Emissions from industry have declined from about 1,500 Tg to about 1,400 Tg. Emissions from the agriculture, commercial, and residential sectors have stayed basically flat at around 400 Tg.

Source: U.S. EPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2010, Figure 2-12. Note: Does not include U.S. territories

The focus of the Handbook is on on-road GHG emissions since they comprise the majority of transportation emissions and since the statewide and metropolitan transportation planning largely focuses on surface transportation. Nationally, on-road sources account for about 84 percent of transportation GHG emissions, as shown in Figure 2. However, it is important to note that off-road emissions sources (e.g., ports and airports) may be important contributors to transportation GHG emissions in some states and metropolitan areas.

Figure 2. U.S. GHG Emissions from Transportation Sources, 2010 (CO2 equivalent)

A pie chart showing the percentage of transportation GHG emissions that come from passenger cars (43%), light-duty trucks (19%), medium- and heavy-duty trucks (22%), aircrafts (8%), ships and boats (2%), and all other transportation sources (6%).

Source: Developed using data from U.S. EPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2010.

Carbon dioxide (CO2) is the primary GHG associated with the combustion of transportation fuels, accounting for over 95 percent of transportation GHG emissions based on global warming potential. CO2 is emitted in direct proportion to fuel consumption, with different emissions levels associated with different fuel types.

Other notable GHGs include methane (CH4) and nitrous oxide (N2O), which together account for two percent of transportation GHG emissions, and hydrofluorocarbons (HFCs), which comprise approximately three percent of transportation GHG emissions. N2O and CH4 are not directly related to fuel consumption, but instead are dependent on engine operating conditions (i.e., vehicle speeds) and emissions control technologies. In addition, HFCs are emitted from vehicle air conditioners and refrigeration used in some freight shipments; these emissions do not come from the tailpipe, and depend on factors such as the age of the vehicle and how often air conditioners are used. Given the relatively small percentage of these gases in comparison to CO2, these emissions are often not calculated in simple analyses. However, their potential global warming impact per unit of gas is many times that of CO2 and therefore regions may want to calculate N2O and CH4 directly along with CO2, particularly if they already have experience modeling emissions.

CO2 emissions from transportation can be calculated based on the amount of fuel - gasoline, diesel, and other fuels - used by motor vehicles and other transportation sources. This simple concept becomes more complex though, when trying to capture the variety of factors that affect fuel consumption as generally depicted in the flow chart below.

Figure 3. Key Factors Influencing On-Road CO2 Emissions

A diagram showing different factors and arrows connecting these factors. The arrows show that 'vehicle fleet' factors - composition (light-duty, heavy-duty vehicles), characteristics (e.g., age), and fuel type (e.g., gasoline, diesel, electric) - affect 'fuel economy' and 'carbon content of fuel.' 'Transportation system investments and strategies' (multimodal options, system quality, and system management and operations) affect 'vehicle miles traveled (VMT)' and 'vehicle operations' (speeds, stops/starts, idling). 'External factors' (land use and urban form, fuel prices, and economic growth) in turn affect 'vehicle fleet' and 'VMT.' A double arrow connects 'transportation system investments and strategies' and 'external factors affecting travel demand.' Arrows show that 'VMT' affects 'vehicle operations,' which in turn affects 'fuel economy.' Together, 'VMT' and 'fuel economy' affect 'fuel consumption.' 'Fuel consumption' and 'carbon content of fuel' together affect 'CO2 emissions.'

The amount of fuel consumed by vehicles depends on a wide range of factors, including the amount of vehicle travel and the fuel economy of those vehicles, which in turn depends on how they are operated and the vehicle characteristics. The type of fuel burned (e.g., gasoline, diesel, compressed natural gas, biofuel) also affects the amount of CO2 that is emitted, based on the carbon content of the fuel. Extrapolating many of these variables into the future is challenging and requires numerous assumptions that can have a significant effect on forecasts. For GHG forecasts, vehicle and fuel assumptions have significant potential to affect results, so particular attention must be paid to these assumptions, by both analysts and the users of GHG forecasts.

Estimations of GHG emissions can rely on similar methods to those used for analyses of criteria air pollutant emissions. These methods include analyzing VMT with emissions factors, or using an emissions model, such as EPA's MOVES model, which is EPA's preferred tool for developing on-road GHG emission inventories at the state and local levels. 3

2.2. Why Estimate GHG Emissions in Transportation Planning?

Transportation planning is a cooperative process to make decisions about transportation investments and strategies for operating, managing, maintaining, and financing the transportation system in such a way as to advance the area's long-term goals. Transportation planning involves many steps, from developing a vision and goals through developing a long range plan (LRP), sometimes referred to as a long range transportation plan (LRTP). The overall transportation decisionmaking process involves additional steps, including programming investments, project development, systems operations, and on-going monitoring of system performance (as shown in Figure 4).

Figure 4. The Transportation Decisionmaking Process

A diagram demonstrating the steps involved in the planning process. The center of the diagram shows arrows connecting steps: 'Regional vision and goals' leads to 'alternate improvement strategies (operations, capital),' which connects to 'evaluation & prioritization of strategies,' which leads to 'development of transportation plan (LRP),' followed by 'development of transportation improvement programs (S/TIP),' which connects to 'project development,' which is followed by 'systems operations (implementation),' and finally leads to 'monitor system performance (data).' Critical factors and inputs are shown on the left and right of the diagram. These inputs shown are: economic development, public involvement, data, fiscal constraint, safety, non-discrimination, air quality, and environmental issues. Feedback loops encircle the entire process, connecting 'monitor system performance (data)' back to 'regional vision and goals' and vice versa.
Source: FHWA/FTA, "The Transportation Planning Process: Key Issues - A Briefing Book for Transportation Decisionmakers, Officials, and Staff. " FHWA-HEP-07-039, September 2007. Available at:

Although Federal regulations do not require analyzing GHG emissions as part of statewide or metropolitan transportation planning, many actions to address GHG emissions can be initiated at the state and regional level. Therefore, there are several reasons for State DOTs and MPOs to consider estimating GHG emissions as part of the planning process.

Incorporating GHG Emissions in the Planning Process

GHG consideration can be incorporated into the transportation planning process at numerous points, as shown in Figure 5, and described below.

Figure 5. Potential Points for Addressing GHG Emissions in the Planning Process
A diagram showing 6 steps in the planning process as boxes connected by linear arrows in the following order: stakeholder identification and initial outreach; establish vision, goals, and objectives; define performance measures, data availability, and needs; develop a baseline; develop alternative plan scenarios; and develop plan and TIP/STIP. Feedback occurs between the third box and the second, and between the sixth box and the fifth. Feedback also occurs between the sixth step (develop plan and TIP/STIP) and the first (stakeholder identification and initial outreach).

Source: Adapted from: FHWA, "Integrating Climate Change into the Transportation Planning Process," July 2008.

The public should be actively involved throughout the process, including in selecting a preferred "vision" for the community and defining preferred outcomes of that vision. If that vision includes reducing GHG emissions from transportation, the public and other stakeholders have important roles in helping to achieve them. Involving the public up front and gaining their acceptance of the benefits of programs to mitigate climate change can provide continued support for policies and programs to address GHGs.

Examples of Areas Integrating GHGs in Planning

A number of states and metropolitan areas are including climate change considerations in their transportation plans and programming documents, and have integrated GHG emissions analysis in the planning process. Just a few examples are highlighted below.

State of Maryland. "The 2009 Maryland Transportation Plan notes that Maryland is beginning to address climate change through its participation in the Statewide Climate Action Plan required by the Governor and Maryland legislature under state law. The Plan sets statewide goals for reducing GHG emissions 25% below 2006 levels by 2020. Reducing transportation GHG emissions will likely require a range of transportation and land use policy options, including increasing the use of cleaner fuels, transitioning State vehicle fleets to high efficiency clean emission vehicles, providing robust and expanded transit service, promoting land use options that reduce vehicle trip lengths and the need for single occupant vehicle use. MDOT and its Modal Administrations are implementing key policies aimed at reducing GHG emissions described in the Climate Action Plan. The Maryland Transit Administration (MTA) prepared the MARC Growth and Investment Plan to expand and improve commuter rail service in Maryland, and is working to increase transit ridership across the State. The State Highway Administration is working to ensure safe walking and bicycling conditions whenever highway facilities are being improved, and promotes ridesharing through its provision of park-and-ride facilities. In addition, they reduce vehicular emissions by implementing travel flow improvements. MDOT programs aimed at reducing GHG emissions include travel demand management (TDM) strategies to reduce traffic congestion, VMT growth, and single occupancy vehicle travel, and transit-oriented development around high quality transit service. The MTA is also promoting work trip reduction alternatives such as rideshare matching and employer incentives for initiating or expanding TDM options for their employees through Commuter Choice Maryland. In addition to reducing GHG emissions, these policies also help relieve transportation system congestion, improve quality of life and access to jobs, and stimulate community reinvestment.S" 7

Philadelphia Region. The Delaware Valley Regional Planning Commission (DVRPC), the MPO for the Philadelphia region, has included a key plan principle to "Build an Energy Efficient Economy" in in its long range land use and transportation plan, called "Connections: The Regional Plan for a Sustainable Future. " The plan sets a goal to "Reduce GHG Emissions" by 50 percent from 2005 levels by 2035 (across all sectors) while building an energy efficient economy. The document notes that DVRPC's regional GHG emissions inventory estimates that in 2005 the region produced just over 90 million metric tons of CO2, roughly 1.5 percent of the U.S. total. It states, "Recent studies identify the need to reduce global GHG emissions by 80% by 2050 to keep climate change within an acceptable range. A 50% reduction by 2035 would put us on track to achieve this goal. " It also notes that

"More than simply an environmental imperative, the act of reducing greenhouse gas emissions is also an economic opportunity... Building an energy-efficient economy will: Create a steady supply of sustainable jobs in emerging, high-growth industries; Provide new green collar jobs for those currently underemployed; Reduce airborne pollutants to acceptable levels; and Save residents on household energy and transportation costs; Save local governments in reduced energy expenditures. "8
For more information about integrating GHG considerations into transportation planning, see:

San Francisco Bay Area. Plan Bay Area is an integrated long-range transportation and land-use/housing plan for the San Francisco Bay Area. It includes the Bay Area's Regional Transportation Plan, which the Metropolitan Transportation Commission (MTC) updates every four years, and the Association of Bay Area Governments' (ABAG's) demographic and economic forecast, which is updated every two years. Taken together, the land use patterns and transportation investments aim to reduce GHG emissions for cars and light-duty trucks in the nine-county region. Due for adoption in spring 2013, Plan Bay Area covers the time period through 2040. MTC used the EMFAC (California Emissions Factors) model in conjunction with its regional travel demand model to generate GHG estimates. MTC has been motivated to estimate regional GHG emissions for several reasons. The public values this information, and California state law requires that MPOs demonstrate per capita GHG reductions in their regional transportation plans. MTC has been calculating GHG emissions for nearly a decade. As Plan Bay Area is being developed, performance targets have been selected against which to measure and evaluate various land use scenarios and transportation investments and policies. After consulting with experts, stakeholders and the public, ABAG and the MTC adopted 10 targets. The first is a GHG reduction target required by Senate Bill 375, "The California Sustainable Communities and Climate Protection Act of 2008. " MTC must demonstrate that its long range plan will reduce per-capita CO2 emissions from cars and light-duty trucks 7% by 2020 and 15% by 2035, compared to 2005 levels. 9

2.3. Types of GHG Analyses in the Context of Statewide/Metropolitan Transportation Planning

As states and regions attempt to understand GHG emissions levels associated with on-road sources, transportation planners may be called upon to provide information on current and past levels of emissions and their sources as well as information on what future emissions are likely to be under multiple scenarios. The understanding of GHG emissions levels can help states and MPOs achieve performance goals and targets by addressing emission reductions strategies through a comprehensive process in a consistent, coordinated manner. Strategic implementation of investments in the multimodal transportation system can reduce GHG emissions while achieving a balanced, environmentally responsible transportation network.

This information can be important in the transportation planning process, since one of the factors to be considered in both statewide and metropolitan transportation planning is: "protect and enhance the environment, promote energy conservation, improve the quality of life, and promote consistency between transportation improvements and State and local planned growth and economic development patterns. "10 Moreover, states and MPOs must consider including appropriate environmental mitigation activities in their long range transportation plans. Many areas have developed goals related to environmental quality within their transportation plans, or have developed goals for GHG reduction as part of separate climate action plans that can be referenced in the long range transportation plan.

GHG analysis falls into three broad categories based on the analysis purpose and timeframe of interest:

These types of analyses are closely related, and some states or MPOs engaged in GHG analysis will perform all of them in order to assess past and current emissions levels, to identify anticipated trends in emissions in the future, to explore the impacts of plans and programs, and to examine the potential impacts of alternative strategies.

Note on Terminology: What is meant by an emissions "inventory"?

As used in this Handbook, an emissions "inventory" refers to the level of emissions in a current or past year, while a "forecast" refers to the level of emissions in a future year, either under a "business as usual" scenario or under any number of alternatives.

It is important to note that air quality planners often use the term emissions "inventory" to refer to any analysis quantifying a total amount of emissions for any calendar year, in contrast to an emissions "rate" or "factor" (such as grams per mile). Under this definition, an inventory could be for 1990, 2010, or 2020, and could include multiple scenarios (e.g., a 2030 "business as usual" inventory and a 2030 alternative scenario inventory). Transportation planners, on the other hand, may prefer to think of "inventories" as present or past, and "forecasts" as future predictions, given the differences in data sources that may be used and levels of uncertainty with each. Regardless of how the terms are typically used, it is important for individuals and agencies to have a clear understanding of terms when discussing their analyses.

Inventory Development

An inventory of GHG emissions from transportation (or on-road specific) sources provides information on the magnitude of emissions and their sources. An inventory is usually performed for a recent year, depending on data availability. Inventories may also be calculated for a more distant-past baseline year specified in legislation or executive mandates or policies, and as such can be critically important in developing strategies and measuring progress over time to meet those mandates. The level of detail and accuracy that an inventory provides is determined by the methodology used and by the accuracy of key input data (e.g., vehicle fleet characteristics in the inventory year, VMT, speeds, operating conditions, etc. ). General steps for developing a GHG inventory are noted below. 11

General Steps for Developing a GHG Inventory

  • Set boundaries - Define the geographic boundaries of analysis.
  • Define scope - Decide which emissions source categories (e.g., on-road sources only, or all transportation sources) and subcategories (e.g., light-duty vehicles, heavy-duty vehicles, buses), as well as which specific GHGs (CO2 only, or also N2O and CH4) should be included. Also determine how to account for GHGs from travel that starts or ends outside the geographic boundaries.
  • Choose analysis method - Depending on the data available and purpose of the inventory, choose a top-down (fuel-based), bottom-up (VMT-based), or hybrid approach.
  • Set a baseline year - Select a baseline year to provide a benchmark to compare progress going forward, considering whether data for that year are available, the chosen year is representative, and the baseline is coordinated with baseline years used in other inventories. (Note: In some cases, legislation or executive direction will specify the baseline year.)
  • Collect input data and conduct analysis - Gather necessary data and use appropriate tools for the analysis. If necessary data are incomplete or have limitations, as will often be the case, make appropriate assumptions.
  • Document results and how they were derived - Having complete documentation of methodologies used is critical when comparing inventories or forecasts conducted in future years to the current estimate. Document all assumptions, caveats and limitations.

For more information, EPA's State and Local Climate and Energy Program provides technical assistance, analytical tools, and outreach support; available at

Forecasts / Analyses of Alternative Scenarios

As states and MPOs try to understand the potential impact of their decisions on the transportation network and the natural and human environment, they may want to identify anticipated trends in GHG emissions levels or assess potential effects of different long range transportation plan scenarios on GHG emissions. To do this, they will need to estimate future emissions. These estimations can take the form of forecasts analyzing future alternative scenarios. In the case of a forecast, the organization will typically analyze emissions under a business-as-usual scenario. This can reflect anticipated changes in fuel economy, fleet composition, travel patterns, and other variables likely to impact emissions.

In order to identify ways to reduce emissions, the agency may choose to analyze alternative scenarios that estimate the anticipated impact of various policy choices or implementation strategies. For example, an MPO creating its metropolitan transportation plan might forecast emissions for a twenty-year planning horizon under current trends and also forecast what would happen under different implementation strategies being explored as part of the metropolitan transportation plan. In addition, the MPO could provide alternative analyses using different assumptions about future vehicle technology and fuels.

General Steps for Developing a GHG Forecast

  • Determine forecast year(s) - Select one or more milestone years in the future. The selection of milestone year(s) may be influenced by: (a) legislative or executive branch GHG targets and laws and (b) the need to synchronize with planning timeframes of the state or area. Also, consider whether to analyze GHGs on a cumulative basis, rather than for a specific forecast year, since climate change impacts are based on cumulative GHGs, over decades.
  • Choose analysis method - Depending on the data available and purpose of the forecast, select a method that matches the appropriate level of detail and accuracy for the analysis purpose.
  • Collect input data and conduct analysis - Gather necessary data and use appropriate tools for the analysis.
  • Conduct additional strategy analyses - Depending on the sophistication of the analysis method and existing modeling tools (e.g., the level of sophistication of the travel demand model), conduct additional "off-model" analyses to adjust the forecast, if needed.
  • Document the results - Clearly document the results, including assumptions and any limitations or caveats. Identify key areas of sensitivity affecting results.

GHG Strategy Analysis

Analyzing the effects of GHG strategies may be part of the overall forecasting process, particularly if different scenarios for the future are being explored. It is important to note, however, that standard travel forecasting approaches are not well geared toward analyzing certain types of strategies, such as strategies that reduce non-recurring delay, reduce heavy-duty vehicle idling (e.g., truck stop electrification), introduce a low carbon fuel standard, and many others. Consequently, it may be important to conduct specific analyses of GHG reduction strategies or packages of strategies as part of the planning process. Moreover, as agencies move toward more performance-based planning and programming approaches, they may wish to analyze the GHG effects of different projects and programs in order to help prioritize investments for funding. A State DOT or MPO also may wish to analyze the GHG emissions benefits of existing projects or strategies, based on collected data on impacts, in order to help understand their effectiveness and to help inform future decisionmaking.

General Steps for Conducting GHG Strategy Analysis

  • Identify the specific strategies to be analyzed - The analysis may be for one or more strategies and policies.
  • Select appropriate analysis tools - These tools typically will focus on estimating the changes in vehicle technology and fuels, VMT, vehicle operating conditions, or vehicle fleet in response to the strategies. One tool may be able to analyze a set of strategies simultaneously, or separate analyses using multiple tools may be needed.
  • Consider the effects of combinations of strategies - It is important to consider the effects of different combinations of strategies, since some strategies may be synergistic while others are antagonistic.
  • Collect or develop input data and conduct analysis - Gather necessary data and use appropriate tools for the analysis. Convert changes in travel characteristics (such as changes in travel demand, speeds, congestion) or other factors (such as changes in vehicle technologies) to changes in emissions.
  • Document the results and how they were derived - Document the methodology used, assumptions about strategies and baseline conditions, and caveats and limitations.

2.4. Overview of Primary Methods

Most efforts to estimate transportation GHG emissions fall into the following primary categories of methods:

Each type of method serves certain needs better than others and has strengths and weaknesses in application, due to data requirements, outputs produced, and sensitivity to different factors. Fuel-based inventories and forecasts are typically best for state-level analysis due to the availability of state fuel sales data, while VMT-based methodologies may be used at multiple levels. Both State DOTs and MPOs generally have methods to estimate VMT, and most MPOs have travel demand models to forecast future VMT under different scenarios for use in the planning process. Each type of method can typically be applied at different levels of sophistication, based on the amount and quality of data available and the purpose and needs of the analysis.

Table 1 highlights the primary types of methods and approaches that can be used, along with key strengths and limitations. A summary of each general type of method follows.

Table 1. Methods and their General Strengths and Limitations

Methods for Estimating and Forecasting GHG Emissions

Type of Method




Fuel-based Methods

Simple spreadsheet inventory or forecasts: Collect fuel data and multiply by emissions factor (based on carbon content of fuel) or use EPA's State Inventory Tool; for projections use State Inventory Projection Tool or simple growth factor

  • Simple
  • Data generally accessible
  • Can be used for all modes (to the extent that state or MPO-specific data are available)
  • Only produces estimates by fuel type, not vehicle type
  • Fuel sales may not match well with actual travel activity, particularly in smaller geographic areas
  • Projections for future years are not as precise if based on simple growth factors
  • Method only addresses CO2, not other GHGs (e.g., nitrous oxide, methane)

More refined inventory or forecasts: Disaggregate by vehicle type or geography; account for multiple factors in forecasts

  • Relatively simple
  • Provides more detailed breakdown
  • Fuel sales may not match well with actual travel activity, particularly in smaller geographic areas
  • Many assumptions need to be made to develop projections
  • Method only addresses CO2, not other GHGs (e.g., nitrous oxide, methane)

VMT-based Methods

Simple approach: Develop estimates/forecasts of VMT relying on vehicle, household, economic activity, pricing, and land use data and apply simple emissions factors*

  • Relatively simple
  • VMT estimates are generally available
  • Well-geared toward areas without network travel models or experience with emissions analysis
  • Does not account for impacts of congestion, speeds, or eco-driving behavior of motorists
  • May not account for significant variation in vehicle fuel efficiency and fuel types across the passenger and freight fleets

More sophisticated: Rely on HPMS data and/or a network-based travel model to develop estimates/forecasts of VMT broken out by major facility type and/or speed bin and apply emissions factors based on look-up tables*

  • Relatively simple but provides more robust analysis (accounting for impacts of speed changes)
  • Well-geared toward areas with or without network travel models
  • Does not account for full range of factors that may be addressed in emissions models (although emissions factors typically will be developed using an emissions model, but applied using simplifying assumptions)
  • Requires extra effort to attribute VMT shares to different vehicle/fuel types, to reflect variations in GHG/mile for both passenger and freight vehicles

Emissions Modeling: Develop estimates/forecasts of VMT and use MOVES (or EMFAC in California)

Simple approach: Rely on model's defaults for inputs other than VMT

Most sophisticated: Customize inputs for the specific area being modeled*

  • Most robust ability to address all of the factors that influence GHG emissions
  • MOVES is EPA's preferred tool for developing on-road GHG emission inventories at the state and local level
  • Local data will need to be assembled, unless the emissions model has already been used for SIP or conformity analysis, or relying on default data (note: default data reduces precision of analysis)

Alternative GHG Estimation Approaches

Commodity Flow-based Methods

  • Enhances an existing inventory that does not include freight
  • Difficult to forecast since emissions are often driven by factors that are external to a state or region

FHWA's Energy and Emissions Reduction Policy Analysis Tool (EERPAT)

  • Provides policy sensitivity for different GHG mitigation measures
  • Can evaluate future changes in land use and is sensitive to external changes in the price of fuel
  • Can incorporate changes in technology
  • Can be used to assess the overlapping effects of bundles of GHG mitigation strategies
  • Is relatively well-suited to statewide transportation GHG analysis
  • There are a large number of model inputs and some may be difficult to obtain
  • Model applies to statewide analysis only
  • Model's VMT estimates are not as accurate as a network-based model

Specific Transportation Strategy Analysis Methods

These approaches include off-model tools that capture the effect of GHG mitigation strategies that cannot be analyzed through travel demand models (e.g., Commuter Model) - EPA has developed an approach called the Travel Efficiency Assessment Method (TEAM), which encompasses use of these types of methods

  • Relatively easy to use
  • Some tools and approaches (e.g., COMMUTER Model, TRIMMS) can analyze the impacts of TDM and TCM strategies in one package
  • TEAM builds directly on outputs from existing travel demand models and uses existing modeling tools
  • These approaches generally require application of emissions factors. Lack of familiarity with MOVES by some users could be a limitation.
  • Some methods involve simple calculations or rely on relationships drawn from national literature, which may not be accurate in all locations

Additional Considerations in GHG Analysis

Lifecycle analysis methods: using Argonne National Laboratory's Greenhouse Gases, Regulated Emissions, and Energy use in Transportation (GREET) model or other approaches

  • Provides a fuller understanding of the net impact of strategies (accounts for emissions associated with tailpipe and upstream emissions)
  • GREET allows the user to compare the lifecycle emissions attributable to conventional and alternative fuels
  • For GREET, units of grams of CO2 equivalent per mile (gCO2e/mi) limit the user to a pre-determined fuel economy
  • The GREET tool does not capture the emissions from so-called indirect land use change attributed to sources such as corn and soybeans

Construction and maintenance emissions methods - spreadsheet tools can be used to estimate GHG emissions associated with materials and equipment used in construction and maintenance of roadways

  • Allows transportation agencies to consider the GHG impacts of roadway planning and construction, rather than just considering tailpipe emissions from vehicles using roadways
  • Existing tools work best to analyze individual projects using detailed engineering data. Tools are not well equipped for system level analyses of construction and maintenance in long range planning

* Within these general VMT methods, there are a number of different approaches available to estimate VMT as well as to develop inputs for calculating emissions factors; each of these approaches has its own strengths and limitations, depending on data availability and modeling capability. Refer to Chapter 5 for more details.

Fuel-based Methods

Fuel-based methods typically rely on estimates of fuel sales, and directly convert fuel use estimates into CO2 emissions estimates based on the carbon content of each fuel. The basic equation for estimating CO2 emissions is:

Fuel Consumed x Emissions Factors = CO2 Emissions

Note: The emissions factor will depend on the fuel type (e.g., motor gasoline, diesel). Fuel consumption is often expressed in gallons and emissions factors in CO2/gallon.

Fuel-based Methodologies Reviewed in this Handbook. Fuel-based methods can be used to develop:

  • inventories (Section 4.1) and
  • forecasts (Section 4.2).

For both fuel-based inventories and forecasts, a basic approach produces estimates of CO2 by fuel type. More refined approaches involve additional steps to allocate emissions to vehicle types or geographic areas, or incorporate additional factors into forecasts.

Fuel-based methods are most applicable where fuel data are available and fuel purchased in the geographic area is used by vehicles operating within the same area. This tends to be the case at a larger scale such as at the state level for developing GHG inventories based on historical data on fuel consumption.

A challenge with fuel-based methods includes the potential that fuel sales may not match directly with travel activity (they may differ due to interstate trucking, through traffic, and other factors). This often raises a policy question about whether GHG should be measured based on where fuel is sold, where travel occurs, or based on the generators of trips (e.g., households and businesses). Moreover, forecasting fuel consumption is often a challenge as the cost of oil fluctuates, and fuel-based methods often are not sensitive to economic development, demographic, and land use allocation policies, transportation investments and strategies, and non-highway transportation user costs (e.g., transit fares and parking costs).

VMT-based Methods

VMT-based methods focus on quantifying the amount of vehicle travel and then connecting this information to an estimate of emissions using emission factors or an emissions model. The basic equation for estimating emissions is:

VMT-based Methodologies Reviewed in this Handbook. All VMT-based methods involve estimating vehicle travel and applying emissions factors or an emissions model. This Handbook identifies several different sources of VMT estimates, as well as different approaches to developing emissions estimates. The Handbook is divided into the following sections:

  • Relying on vehicle, household, economic, pricing, and land use data to estimate VMT (Section 5.1)
  • Relying on HPMS data and/or a network-based travel model to estimate VMT (Section 5.2)
  • Developing emissions factors and emissions inventories (Section 5.3) - This section discusses various ways to estimate emissions, ranging from use of simple look-up tables of emissions factors to use of EPA's MOVES model.

Handbook users are advised to look both at sections discussing VMT estimation and emissions estimation.

VMT x Emissions Factors = GHG Emissions

Note: The emissions factor (typically presented in grams per mile) will depend on vehicle type, classes within vehicle types, technology/fuel type, speeds, and operating conditions. Different emissions factors are available for CO2, N2O, and CH4.

There are many different techniques and levels of sophistication that can be used both for estimating VMT and for developing emissions factors.

VMT can be estimated based on a range of data sources such as data from the HPMS, vehicle odometer readings, household travel survey data, or land use-based vehicle trip generation estimates. VMT forecasts can be developed using network-based travel forecasting models, in which a range of factors such as transportation investments, land use, and modal options affect estimates of future VMT. In non-network based methods, simplified approaches may be used to forecast future travel demand, such as applying simple growth factors. Non-network methods are sometimes referred to as 'sketch' planning models although these methods may include relatively robust models of travel behavior to reflect a wide range of variables.

Emissions factors, meanwhile, also can be estimated at various levels of sophistication, ranging from simply applying an average GHG emissions factor to VMT (based on average vehicle fuel economy), to more sophisticated methods accounting for other variables (e.g., vehicle speeds and fleet mix), to use of emissions models like EPA's MOVES model and the Emission FACtors model (EMFAC), developed by the California Air Resources Board and used by agencies within the State of California. MOVES and EMFAC account for a wide range of factors (including vehicle age, road type, drive cycles, and other factors). Policy analysts may wish to match the level of sophistication of the emissions analysis to be somewhat commensurate with the level of sophistication in the VMT analysis.

Alternative GHG Estimation Approaches Reviewed in this Handbook.
  • Commodity Flow Based Methods to Estimate Freight Truck Emissions (Section 6.1)
  • Energy and Emissions Reduction Policy Analysis Tool (EERPAT) (Section 6.2)

Alternative GHG Estimation Approaches

Although fuel-based and VMT-based methods are most common for developing GHG inventories and forecasts, other emissions estimation techniques may be applied. Specifically, emissions from freight trucks, as well as other freight modes, can be difficult to forecast since they are often affected by economic factors not accounted for in travel demand forecasting models, which tend to focus on household travel. As a result, one approach profiled in this Handbook is to develop GHG emissions estimates based on commodity flow data.

In addition, most of the traditional travel and fuel forecasting methods do not easily enable the user to account for a full range of transportation GHG reduction strategies, including policies such as pricing, incentives for energy efficient vehicles, and land use policies. The Energy and Emissions Reduction Policy Analysis Tool (EERPAT) was designed specifically to help agencies conduct a screening-level analysis of a wide range of emissions reduction strategies.

Specific Transportation Strategy Analysis Methods

Specific Transportation Strategy Analysis Methods Reviewed in this Handbook.

  • Transportation demand management strategies (Section 7.1)
  • Land use strategies (Section 7.2)
  • Transportation system management and eco-driving strategies (Section 7.3)
  • Freight strategies (Section 7.4)

Most travel forecasts, whether through a network-based travel demand model or non-network based approaches, lack the capability to evaluate many GHG reduction strategies such as changes in small-scale land use density, land use mix, pedestrian environment, and transit accessibility; parking price changes, pay-as-you-drive auto insurance; and most other pricing strategies; employer trip reduction programs and other transportation demand management (TDM) strategies; and eco-driving programs and transportation system management and operations strategies, such as improved incident management and traffic signal coordination. A range of tools and approaches can be used to analyze the effects of GHG reduction strategies that cannot be directly accounted for in standard travel forecasting methods. These "off-model" analyses often use simple spreadsheet calculations. Similar approaches are widely used in the transportation conformity process to calculate the emissions benefits of strategies that cannot otherwise be accounted for in the travel forecasting process.

Additional Considerations in GHG Analysis: Lifecycle Analysis and GHG Emissions from Transportation Construction and Maintenance

Additional Considerations Reviewed in this Handbook.

  • Lifecycle Emissions Analysis Methods (Section 8.1)
  • Construction and Maintenance Emissions Analysis (Section 8.2)

While most of the methodologies discussed in this Handbook focus on GHGs emitted directly from motor vehicles, analysts should be aware that transportation activities also generate other emissions. The field of LCA (also known as lifecycle analysis) is concerned with understanding the full environmental impacts associated with all the stages of a project or product life. Within the transportation planning context, lifecycle GHG analysis includes not only direct emissions from motor vehicles but also emissions associated with upstream activities, such as fuel production and distribution. These can be important issues, particularly when examining strategies related to alternative fuels or electricity use in transportation, since powering motor vehicles or rail with electricity will generate some GHG emissions at the powerplant source of electricity.

In addition, some State DOTs and MPOs are interested in considering the emissions associated with transportation infrastructure construction and maintenance activities. These activities produce emissions, and there are strategies available to reduce these emissions (for instance, reduced roadside mowing, use of low emissions construction equipment, use of warm-mix asphalt).

2.5. A Note on Travel Demand Models

A number of GHG estimation techniques rely (especially for future estimates) on the availability of a suitable network-based travel demand model. Such models are built to evaluate long term regional changes in travel activity due to socio-economic changes (economic development, population shifts) and modifications of the transportation system. They are frequently used by MPOs and State Transportation Agencies in cost/benefit analysis for major highway or transit investments (e.g. through the FTA New Starts program), for alternative analysis and prioritization in long range transportation plans, and for air quality conformity analysis.

Evaluating the suitability of an existing travel model will require due diligence to determine the resolution of the travel demand model, its ability to generate consistent aggregate estimates of future travel, and the compatibility of its modeling and scenario assumptions with the policy goals of the GHG analysis. Using an existing travel model is not guaranteed to support a "better" GHG analysis if the model's assumptions and limitations do not coincide with the needs of the GHG analysis. For example, evaluating multiple scenarios for a rapid analysis of many possible policy initiatives may be prohibitively expensive and time consuming if attempted with a complex travel model, and may not yield substantially better results for the purposes of initial screening than a well-developed strategic model such as FHWA's EERPAT.

Analysts who are considering using travel demand model outputs to support greenhouse gas analysis should make an effort to understand the available travel model - what its inputs are (and what they are not), what the model does with them, and how the model outputs are intended be interpreted. The remainder of this section reviews the structure of travel demand models, how they are applied, and what kind of information they generate. This overview is necessarily brief; additional resources that provide more in-depth discussions of the travel demand modeling process are discussed in Chapter 5 (page 63).

For decades, the standard technology for travel demand modeling has been what is called a trip-based model, or a "four step" (or sometimes "three step" ) model, and most regions that do travel demand modeling still operate such a model for regional planning purposes. The four "steps" include trip generation (how many people are traveling for different purposes), trip distribution (where and how far those trips go), mode split (which classifies the trips by mode of travel, including various transit modes or carpooling) and trip assignment (identifying the highway and transit facilities on which those trips occur). In a 3-step model, the mode split step is eliminated or simplified. Trips are typically analyzed by purpose (for example, home-to-work, or home-to-shopping, or even heavy trucks), reflecting the observation that people typically travel different distances and at different times of day for different purposes.

Trip-based models originally focused on reproducing summary vehicle counts (and transit ridership), often in the form of average annualized daily traffic or average weekday traffic (from which VMT estimates are derived). Trip-based models depend heavily on statistical analysis of regional traffic patterns. Because these models treat trips as independent statistical events, they rely heavily on implicit assumptions about the correlations between trips in different zones and for different purposes. Such assumptions are often insensitive to various types of demographic shifts (e.g. changes to travel patterns shrinking household size, or due to reduced vehicle ownership), and to certain changes in the transportation system (e.g. peak-period tolling which may induce shifts to different time periods as well as changes to vehicle occupancy).

Over the years, various enhancements to the basic four-step model have been applied, such as adjusting trip assignment estimates due to congestion (so trips will accrue on alternate routes when the most direct route is congested), and altering the distribution of trips to reflect the undesirability of destinations that are relatively inaccessible by any mode or that require traversal of highly congested facilities. In addition, such models are often split into time periods to permit more refined analysis of "within day" daily travel patterns (peak versus off-peak travel).

In addition, trip-based models (as well as the more complex tour- and activity-based models) have been used with dynamic traffic assignment (DTA) approaches to examine the dynamics of rush-hour congestion in more detail. But DTA outputs are not typically presented in the form of traffic or VMT summaries at the aggregate level required for GHG analysis and may require post-processing to generate consistent traffic estimates for GHG analysis.

Over the last two decades, new techniques for improving travel model sensitivity to linked trips and household dynamics have emerged. Commonly referred to as "tour-based" or "activity-based" models, the distinctive feature of such models is their focus on analyzing the joint behavior of individual travelers (tour- and activity-based models are thus collectively referred to here as "traveler based" models). Traveler-based models explore more detailed demographic impacts of system changes, by comprehensively linking trips to the characteristics of travelers and their local environment so that better account can be made of the influence of household interactions, variations in travel choice among individuals within a various demographic groups, and linked constraints on travel.

Such models often include considerably finer detail about the travel network and the local environment that individual travelers experience. They may also support detailed analysis of bicycle or pedestrian modes (and shifts to those modes due to system enhancements, which may be relevant for GHG analysis). Traveler-based models are often also implemented as "simulation models" , computing outcomes by mapping out likely sequences of correlated trips taken by synthesized individuals and households over a certain period of time. The simulation approach can be very useful for evaluating operational performance and correlated effects of related transportation system enhancements, but simulation increases the model runtime considerably compared to trip-based models. More importantly, such models may have statistical limitations with respect to estimating cumulative statistics such as average daily or annual VMT. For example: since simulations represent a snapshot of activity in a period of time and do not intrinsically develop estimates of cumulative activity - one day is not necessarily the same as another due to statistical variations in the inputs, and thus adding up a year's worth of days based on a one-day snapshot may magnify small statistical errors. Consequently, it may be desirable to perform multiple model runs with different "random seeds" (starting points) in order to develop an accurate estimate of total annual travel. Analysts contemplating the use of a simulation-based model should carefully consider the implications of the model's statistical assumptions and the intended uses of the model outputs.

The question of statistical assumptions (and more broadly, the statistical uncertainty of any travel model) is by no means unique to simulation models. In important respects, traveler-based models using a simulation framework are expected to perform better than trip-based models, for example by being less susceptible to errors in core assumptions about trip linkages and the receptiveness of different elements of the population to improved performance on various system elements. But there is still little accepted science about how to estimate the uncertainty of forecasts. Though traveler-based models reduce the effect of hidden assumptions about trip linkages and response to system changes, they introduce a potential additional statistical burden of requiring multiple model runs in order to evaluate uncertainties in constructing detailed synthetic populations and in computing aggregate measures of the travel activity of such populations.

1 U.S. EPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2010, April 2012; EPA-430-R-12-001,

2 Washington Department of Ecology, Washington State Greenhouse Gas Emissions Inventory 1990-2008, December 2010,

3 U.S. EPA, "Using MOVES for Estimating State and Local Inventories of On-Road Greenhouse Gas Emissions and Energy Consumption: Public Draft," January 2012.

4 23 USC 150(b)(6).

5 Additional information on the PEL approach is available through the FHWA website, "Planning and Environment Linkages: Program Overview,"

6 U.S. EPA, State and Local Climate and Energy Program,

7 Maryland Department of Transportation, Maryland Transportation Plan 2009,

8 DVRPC, Connections: The Regional Plan for a Sustainable Future (The Long-Range Plan for the Greater Philadelphia Region),

9 ABAG, the Bay Area Air Quality Management District (BAAQMD), the Bay Conservation and Development Commission (BCDC) and MTC, One Bay Area,

10 23 USC §134 (metropolitan) and 23 USC §135 (statewide).

11 Adapted from U.S. EPA, State and Local Climate and Energy Program, "Developing a Greenhouse Gas Inventory. " Available at:

Updated: 03/27/2014
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000