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 3 - What Methodologies are Applicable for your Situation?

Given the variety of methods available for GHG analysis, it is important to recognize that not all methods are applicable in all circumstances. There are a number of key factors that affect the appropriate selection of a methodology. As shown in the figure below, key questions that the analyst should consider are:

Figure 6. Key Factors to Consider in Selecting a GHG Estimation Method
The diagram poses three questions to the analyst: 'What is the goal of the analysis?' 'What data, tools, and resources do we have available?' and 'What variables do we want to analyze?' Under the first question, the analyst should consider the type of analysis and time frame (inventory - year(s), forecast - year(s), strategy analysis - what specific strategies to consider); geographic scope (state, metropolitan area, local); emissions and sources to be included (CO2 only or also CH4 & N2O; all transportation, on-road vehicles, or subset; operational, lifecycle, construction & maintenance); and analysis precision required (regulatory/compliance, approximate/sketch plan). Under the second question, the analyst should consider data availability (motor fuel sales; HPMS, travel surveys, odometer surveys; roadway speed limits, traffic speed surveys; fleet mix data); modeling capabilities (non-network-based approaches, network-based 3-step travel model, network 4-step or activity-based model); and resources available (staff time, budget). Under the third question, the analyst should consider fuels and vehicle technologies (fleet mix changes, vehicle technology changes, alternative fuel penetration); travel demand (land use changes, transportation demand management, freight flow changes); operations and speed factors (traffic congestion levels, vehicle idling, speeds and eco-driving); and cross-cutting factors to analyze or consider (fuel prices, population and employment levels, and other demographic and economic factors).

These factors are discussed below.

3.1. Analysis Purpose

The first question to consider is, "What is the goal of the analysis?" Answering this question will typically include several components, addressing:

A State DOT or MPO might also want to consider whether it is interested in measuring the transportation activity of residents living within its jurisdictional boundaries or in all travel that occurs within its boundaries - including "through traffic" that does not originate or stop in the area. This can be a particular issue for areas that would like to allocate emissions to particular jurisdictions within their planning area (e.g., allocating transportation emissions to each county within a state), since allocating emissions between origins, destinations, and pass through areas can become complicated.

Reconciling Fuel-based and VMT-based Estimates

At the state-level, GHG estimates may be developed by different state departments using different data. For example, an environmental department may develop a fuel-based inventory and projections as part of a multi-sector inventory, whereas a State DOT may develop estimates and projections using VMT-data. These methodologies may produce different results, and some areas have struggled to reconcile fuel-based inventories and VMT-based inventories/forecasts of GHG emissions. It is important to keep these concerns in mind early on so that agreement is made on a common approach, or procedures are developed to reconcile estimates developed using different methodologies.

Type of Analysis

Recognize uncertainties associated with forecasts
Forecasting by nature involves more uncertainties than developing an inventory where "real world" data (e.g., on fuel consumption or vehicle travel, speeds, and other characteristics) are collected as inputs to calculating emissions. The further away the time horizon for analysis, the more uncertainty is typically introduced into the analysis, particularly as it relates to future fuel prices and vehicle technologies. GHG emissions are often analyzed using a long time horizon (e.g., 2050), and consequently, it may be useful to examine multiple scenarios for some of the key factors, such as vehicle technologies, which will affect emissions levels.

One of the most important factors in deciding on a methodology is to determine what type of analysis is to be conducted. As noted in Section 2, there are three common types of analyses, which often are conducted in combination.

Consider relationships among strategies

Different transportation strategies may have effects that are synergistic or antagonistic with each other, so it is important to consider a combination of factors when analyzing transportation strategies. However, if reduced fuel costs result in increased desire to drive then there may be a negative impact of VMT reduction strategies.

If a State DOT or an MPO wants to consider VMT reduction strategies or changes in congestion levels and speeds as part of forecasting GHG emissions, then understanding the level of accuracy and sensitivity of any model and assumptions they are using is essential. More sophisticated models will do a better job of accounting for factors like land use, urban form, and transit, bicycle, and pedestrian investments than simple travel forecasting models. If the model is not sensitive to certain factors, then separate analyses of strategies may be required.

Geographic Scope

The geographic scope tends to influence the type of method that is selected for GHG modeling, based on data availability. At the state level, fuel-based methods are often used given the availability of state-level fuel consumption data; at the MPO level, fuel sales data may not be available.

Most MPOs have access to data and travel models that allow them to use a VMT-based method for estimating and forecasting GHG emissions. For those MPOs with conventional or advanced network-based models, VMT forecasting is often routine and built into existing regional transportation planning or air quality conformity processes. 1 However, GHG analyses can entail a much longer timeframe - e.g., out to forty or more years from the present year - which is inherently less certain than predicting VMT for the shorter timeframes required for conformity and transportation planning. Network-based travel models are typically not available at the state level. Absent a model, HPMS data are available for base year estimates and forecasts can be made using extrapolation based on past trends and population growth, and can be applied at both the MPO and state levels.

What GHG emissions should be "counted"?

In developing GHG inventories and forecasts, an important policy question to consider is "what emissions should be 'counted' or assigned to a state or region?" There are a couple of options:

  • Emissions associated with fuel sold within the boundaries. This is the approach commonly used within fuel-based inventories; however, it raises some questions about whether a state or region should be responsible for fuel that is sold in an area but is used outside of it.
  • Emissions from all travel on the transportation system within the boundaries. This is the approach that is applied for transportation conformity purposes - emissions from all travel on the roadway network, whether due to local traffic or interstate through trips, are analyzed and reported.
  • Emissions generated by households and/or businesses within the boundaries. A State DOT or MPO might instead be interested in focusing on GHG emissions due to the transportation activity of residents living or working within their jurisdictional boundaries (possibly including business and freight travel, as related to household goods consumption), excluding the effect of "through trips. " This can be useful for areas that would like to allocate emissions to particular jurisdictions within their planning area (e.g., allocating transportation emissions to each county within a state). In this case, it is necessary to understand trip origins and destinations, and assign a portion of emissions to each, while removing the effect of trips that entirely pass through an area.

There are some advantages and challenges to each approach, so it will be important to agree upon what is the most desired way of counting emissions.

Emissions and Sources Included

Those conducting a GHG analysis need to determine whether their analysis will incorporate all GHGs or only CO2. Not all methodologies presented in the Handbook consider all GHGs - for instance, fuel-based methods are limited to only considering CO2 emissions, while VMT-based methods can account for all major transportation emissions sources. If an agency would like to or is required to consider all GHGs, this may determine which methodology is selected or whether additional analysis is required.

In addition, it is important to consider what sources of emissions are included: all transportation sources, all on-road vehicles, or only certain types of motor vehicles, such as light-duty vehicles (e.g., automobiles and light-duty trucks, such as pick-up trucks, sport utility vehicles, and minivans). Finally, while this Handbook is primarily concerned with emissions from fuel combustion during vehicle operation (pump-to-wheel emissions), it is important to acknowledge that all transportation fuels and modes have some sort of upstream emissions associated - notably, the energy used to produce, refine, and transport fuel. However, there is not currently an accepted or widespread methodology for quantifying lifecycle emissions from the transportation sector. Two options applicable to particular strategies -- the Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) model for alternative fuels and the American Public Transportation Association's (APTA) for electric emissions from transit -- are presented in Section 8.1.

Analysis Precision/Accuracy Required

An agency that requires a precise and accurate analysis (e.g., an area subject to GHG reduction targets or under regulatory requirements to analyze GHG emissions and strategies) will select a more sophisticated methodology (e.g., a methodology that requires a robust and detailed analysis or forecast of VMT, estimates vehicle fleet mix, and uses an emissions model such as MOVES or EMFAC. ) As a case in point, California's Senate Bill (SB) 375 requires the use of model-based VMT forecasts in combination with EMFAC to estimate GHG emissions. The VMT estimates in this example must be stratified by speed bin and include special trip type accounting to isolate internal versus external trips. An agency that wants to have a general sense of GHG emissions in their region outside of any regulatory context may select a less precise methodology. For example, a less precise approach could be to use simple spreadsheets to generate GHG emissions (e.g., use vehicle, household, and land use data to generate VMT and apply emissions factors). Another option could be to use VMT estimates with MOVES, relying on default information and other flexibilities within the model to streamline the analysis, as described in EPA guidance. 2

3.2. Availability of Data, Tools, and Resources

The availability of data, existing tools that are being used, and the resources available for analysis, including technical expertise with models, are also important considerations in selecting a method to use and deciding how to apply that method. For instance, a simple spreadsheet analysis may be selected by smaller agencies with limited staff, resources, and modeling experience, whereas a sophisticated model may be utilized by agencies that have experience in modeling and the resources to run the model.

Available Data or Ability to Collect Data

As noted earlier, fuel data are most often readily available at the state level, but often not available at the MPO level. As a result, using a fuel-based methodology is much more common at a state level than at a metropolitan area level. Moreover, MPOs are more likely to select a VMT-based methodology, particularly because many large MPOs already have experience using VMT data to generate emissions levels through conformity experience. Those MPOs that have experience with conformity will likely already have VMT as well as fleet data available - although predictions of future VMT and fleet characteristics - especially over long timeframes - is still necessary. States and smaller MPOs that do not have a travel demand model to generate VMT or information on factors such as vehicle speeds may opt to choose a method that relies on vehicle, household, or land use data, or HPMS data.

Understand Data Quality

In addition to knowing what data are available, it is important to consider the quality of data and what the data represent. For instance, in most areas, data on the activity of heavy-duty trucks is limited, and it is often not possible to distinguish between light commercial and passenger trucks based on data available from vehicle registration data, even though this distinction can have a major impact on vehicle activity and emissions. Since estimates of VMT from the Highway Performance Monitoring System (HPMS) are derived from traffic counters on sample segments, these estimates are subject to sampling error. Moreover, standard methods are used to factor traffic counts for a number of days to represent an average annual daily traffic count for a roadway segment, and these adjustments have some level of error associated with them.

Existing Travel and Emissions Modeling Capabilities

Another key factor in selecting a method for analyzing GHG emissions is what modeling capabilities either already exist within the agency or could be easily obtained and applied. Nearly all large and mid-size MPOs (those with population greater than 200,000) use a network-based travel model. However, most State DOTs do not have a statewide travel model, and some smaller MPOs (those with population less than 200,000) may not. In a 2004 survey, 15 percent of small MPOs reported no modeling capabilities at all. 3 According to a recent Government Accountability Office (GAO) survey, about half of the MPOs do their own travel modeling, while the rest rely on consultants or their State DOT. 4

In general, larger MPOs are more likely to develop and operate models in-house, and smaller MPOs, if they use a model, are more likely to require outside technical assistance. Some of the most sophisticated MPOs are using or developing more advanced activity-based travel models. The capabilities of existing travel models will determine the ability to conduct GHG scenario analyses and whether off-model approaches are needed to analyze strategies.

In addition, experience with transportation conformity and emissions analysis can be a key factor in selecting a method to use. Areas conducting conformity analysis typically will be able to apply the same methods they use for criteria pollutant analysis for analyzing GHG emissions, and rely upon most of the same data sources and models; however, some additional analyses may be conducted specific to GHGs. For instance, conformity methods are commonly based on a typical weekday, and additional analyses could be conducted to account for weekend travel. Areas not subject to conformity likely will have limited experience collecting data inputs used in emissions models, and may rely on simpler approaches.

MOVES Model Preferred

It should be noted that beginning in March 2013, MOVES will be required for all new regional conformity analyses. Many states and regions are currently in the process of transitioning to MOVES from previously used models. For this reason, and other advantages discussed below, MOVES is the preferred model for estimating GHG emissions, but it is not mandatory for GHG estimation. Additional explanation of the preferred role of MOVES in GHG emissions estimates is provided in Section 5.3.

Time and Budget Resources

The time and budget available for GHG analysis will influence the type of analysis selected. Developing a GHG inventory and forecast can be time intensive and may require a significant level of effort depending on the method selected and the experience of the agency. Some MPOs only have a few staff on hand and limited budgets for GHG analysis, whereas other MPOs and State DOTs have a larger staff base available to perform a GHG analysis. As noted above, experience with conformity is a factor that may influence the time and budget resources necessary for GHG analysis. MPOs that have experience with conformity may already have the modeling capabilities and familiarity with emissions modeling tools such as MOVES for GHG analysis, which will reduce the time and budget necessary for GHG analysis. However, MPOs that do not have conformity experience may have to spend more time learning, developing, or using tools for GHG analysis, or may need to reach outside the agency for assistance.

3.3. Variables and Strategies of Interest

It is important to consider whether the methodology enables analysis of all the variables or factors of interest when considering any given transportation strategy.

Fuel and Vehicle Technologies: Strategies that influence the future fleet mix, vehicle technologies, and alternative fuels are important because they have a significant impact on emissions. Although these factors are largely dependent upon national and state policies outside of the transportation planning process (such as federal Corporate Average Fuel Efficiency standards), states and MPOs have some opportunities to support and advance strategies, such as incentives for purchasing fuel efficient vehicles. Moreover, assumptions made about future vehicle and fuel technologies will affect the GHG reduction effectiveness of VMT reduction strategies.

Travel Demand: For strategies aimed at reducing VMT, it is important to understand how the VMT forecasts are developed, since most travel forecasts, whether developed through a network-based travel demand model or non-network based approaches, lack the capability to evaluate many travel strategies. Specifically, many travel models are unable to address strategies such as neighborhood-scale land use and urban design, and employer site-based TDM strategies.

Vehicle Operations and Speeds: Transportation operations and congestion relief strategies, and eco-driving programs, affect emissions through changes in vehicle operations, speeds, and/or congestion-related delays. In order to assess the effects of these changes on GHG emissions, the approach needs to account for changes in vehicle speeds or idling, and the emissions factors need to account for these factors as well (e.g., not just applying a simple emissions factor to all VMT without regard to speeds).

Cross-cutting Factors: Exogenous factors related to the economy, fuel prices, and demographic and societal factors can impact transportation emissions, particularly when developing long-range forecasts. For instance, while states can influence fuel prices through taxation, fuel prices are largely driven by global market forces. They are subject to high levels of uncertainty and have effects on both vehicle travel and vehicle purchase decisions. 5 Therefore, when strategies are analyzed, it would be appropriate to conduct sensitivity analyses for these variables (e.g., using high and low values for future fuel prices).

3.4. Identifying an Appropriate Methodology

In order to identify appropriate methods for your circumstances, it may be helpful to review several sections of the Handbook. Since most sections discuss a range of different levels of sophistication for each type of methodology, there are often multiple options to consider in terms of data inputs and applications of the approach. Each section begins with a summary table highlighting applicability at different geographic levels and key attributes. The table below provides a high-level summary, which may be helpful to provide an initial starting point for selecting a methodology. Each methodology begins with a summary box that highlights key considerations.

Table 2. Summary Table to Assist Users in Identifying Appropriate Handbook Sections

Type of Method

Methodology (Handbook section)

Purpose

Geographic Scope

Sources Included

Travel Model Required

Data, Tools, and Resources Required

Strategies Considered

Fuel-based Methods

Fuel-based inventory (Section 4.1)

Inventory

State (typically)

May include all transport sources

No

Fuel sales data, fleet mix (optional)

N/A

Fuel-based forecasts (Section 4.2)

Forecast

State (typically)

May include all transport sources

No

Fuel projections

May account for economic and vehicle technology factors, including effects of fuel prices and regulations.
(Not designed to address individual transportation investments or strategies. )

VMT-based Methods: Estimating VMT

Relying on Vehicle, Household, or Land Use Data (Section 5.1) to estimate VMT

Inventory, forecast, or strategy analysis

State, MPO, local

On-road vehicles

No

Vehicle data: odometer data, vehicle stock data
Household travel data: results from household travel survey
Land use data: land use areas, trip generation rates, demographic and socioeconomic data

May account for land use, demographic changes, and vehicle ownership changes.
(Not designed to address individual transportation investments or strategies. )

Relying on HPMS data and/or a network-based travel model (Section 5.2) to estimate VMT

Inventory, forecast, or strategy analysis

State, MPO, local

On-road vehicles

Applicable with or without a model

HPMS: HPMS VMT data, VMT by vehicle type and within vehicle type groupings
Network Model: network model output

Travel model forecasts may account for changes in transportation investments, land use, and pricing.
(Not typically able to address some types of TDM measures, operational strategies, and eco-driving. )

VMT-based Methods: Estimating Emissions

Developing Emissions Factors & Emissions Inventories (Section 5.3)

Inventory or forecast

State, MPO, local

On-road vehicles

Applicable with or without a model

Simple Factors: VMT
Look-up Tables: VMT, fleet characteristic s, speed bins (optional)
MOVES or EMFAC model: VMT by vehicle type, vehicle population and age distribution, VMT by speed bin

MOVES and EMFAC can account for effects of changes in vehicle travel and congestion and speeds.
MOVES or EMFAC can be used with VMT estimates from any source; can account for any transportation strategy that is incorporated in the VMT estimate although it may not always be possible to distinguish the distinct impacts of individual strategies.

Alternative GHG Estimation Approaches

Commodity Flow Based Methods to Estimate Freight Truck Emissions (Section 6.1)

Inventory or forecast

State, some regions, possible at county-level

Freight trucks

No

Commodity flow estimates, truck survey data (optional)
May require development of OD truck trip table

Largely designed for basic inventories or forecasts, accounting for changes in goods movement.
(Not designed to address strategies affecting time or location of truck travel (e.g., peak hour restrictions).

Energy and Emissions Reduction Policy Analysis Tool (EERPAT) (Section 6.2)

Scenario/strategy analysis

State

On-road vehicles

No

Demographic, land use, and strategy-related data required as inputs.

Land use, transportation demand, vehicle technology, fuels, and price changes.

Specific Transp. Strategy Analysis Methods [note: many of these tools estimate travel or fuel consumption effects, and must be combined with methods to estimate GHGs]

Transportation demand management strategies (Section 7.1)

Scenario/strategy analysis

State, regional, county-level

On-road vehicles

Applicable with or without a model

Sample tools and approaches include: COMMUTER, TRIMMS, and BAAQMD tool

VMT-reduction and travel time shift strategies, such as transit improvements, ridesharing, bicycle and pedestrian improvements, transit and parking pricing, and employer trip reduction.

Land use strategies (Section 7.2)

Scenario/strategy analysis

Regional or county-level (typically)

On-road vehicles

Applicable with or without a model

Sample tools and approaches include: INDEX, PLACE3S, URBEMIS, Sustainable Communities Model, MetroQuest, and CommunityViz

Land use changes, including land use mixing, increased density, and pedestrian accessibility.
Impact of "4 Ds" (density, diversity, design, and destinations).

Transportation system management and eco-driving strategies (Section 7.3)

Scenario/strategy analysis

State, regional, county-level

On-road vehicles

Applicable with or without a model

Sketch planning tools, deterministic tools, and traffic simulation tools. Sample tools include the ITS Deployment Analysis System (IDAS) and Screening for ITS (SCRITS).

Traffic surveillance, work zone management, electronic toll collection, traffic incident management, road weather management, emergency management, and traveler information services.

Freight strategies (Section 7.4)

Scenario/strategy analysis

State and regional

Freight trucks

Applicable with or without a model

Sketch analyses, US EPA SmartWay Transport Partnership tools

Idle reduction programs and policies, mode shift strategies, and strategies affecting pricing and time or location of truck travel.

Additional Considera-tions in GHG Analysis

Lifecycle Emissions Analysis Methods (Section 8.1)

Inventory or forecast

State, regional, possible at the county-level

On-road vehicles and related up-stream sources (e.g., fuel processing) and electric utilities

Applicable with or without a model

Alternative fuels using GREET model: fuel mix, inventory results
Electric Transit Emissions: ridership, passenger load

Fuel and vehicle technology strategies.

Planning Level Analysis of Construction and Maintenance Emissions (Section 8.2)

Inventory or forecast

State, regional, possible at the county level

Infrastructure construction and maintenance emissions

Applicable with or without a model

Type and length of activity (e.g., lane miles constructed)

Alternative construction materials and techniques.


1 It should be noted that most MPO models are based on a typical weekday and do not include weekend travel. Typically, a factor decided by the MPO is used to convert weekday VMT to annual VMT, including weekends, based on available travel data. It is good practice to account for weekend travel in GHG analysis, since emissions are affected by factors that differ between weekday and weekend conditions.

2 Refer to EPA's latest guidance on using MOVES to estimate GHG emissions, found at: http://www.epa.gov/otaq/stateresources/ghgtravel.htm.

3 Transportation Research Board, "TRB Special Report 288: Metropolitan Travel Forecasting - Current Practice and Future Direction. " October 2007.

4 U.S. Government Accountability Office, "Metropolitan Planning Organizations: Options Exist to Enhance Transportation Planning Capacity and Federal Oversight," Report GAO-09-868, September 2009.

5 According to data from the U.S. Energy Information Administration, Annual Energy Outlook 2011: light-duty vehicle CO2 emissions in 2030 would range from 15 percent below to 13 percent above its baseline ("reference case") forecast under a high oil price and low oil price scenario, respectively.

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