Exhibit 6-1: Actions & Tests for Ozone Nonattainment & Maintenance Areas
Exhibit 6-2: Actions & Tests for Carbon Monoxide Nonattainment & Maintenance Areas
Exhibit 6-3: Integration of the Travel & Emissions Modeling Processes for Regional Analysis
Exhibit 6-4: Conventional Regional Travel Models
Exhibit 6-5: Comparison of MOBILE Default & Area-specific VMT Mixes
In addition to the criteria listed in 40 CFR 93.109(b) - Table 1 of the transportation conformity rule, the conformity determination is also based on the criteria specific to the nonattainment areas, which are summarized in Exhibits 6-1 and 6-2 for ozone and CO, respectively. In order to demonstrate conformity, ozone and CO nonattainment and maintenance areas need to perform regional analysis as shown in Exhibits 6-1 and 6-2. Detailed discussion of these requirements is provided in the transportation conformity rule and in Chapter 5 of this document.
|Nonattainment Area||Actions/Tests||Sec. (§)||Applicable Time Period|
|All ozone nonattainment and maintenance areas||
Latest planning assumptions
Latest emissions model
TCMs in an approved SIP
|All ozone areas, if an adequate or approved SIP budget exists||Motor vehicle emissions budget test||93.118||After EPA has declared a SIP motor vehicle emissions budget to be adequate for transportation conformity purposes|
Marginal and below ozone, if no adequate or approved budget exists
If build=no build
8-Hr: Interim emissions tests (build-no-greater-than-no-build test OR no-greater-than-2002 test) and/or budget test using 1-hour budgets
No regional emissions analysis is required
|93.119||If no adequate or approved budget exists|
|Moderate and above, if no adequate or approved budget exists||8-Hr: Interim emissions tests (build-less-than-no-build test AND less-than-2002 test) and/or budget test using 1-hour budgets||93.119||Prior to adequate SIP budgets. If EPA declares motor vehicle emissions budget in submitted control strategy SIP inadequate for transportation conformity and no existing adequate or approved motor vehicle emissions budgets|
|Moderate and above with three years of clean data that have not submitted a maintenance plan and EPA has determined are not subject to reasonable further progress and attainment demonstration requirements||
Must satisfy one of the following:
|Until maintenance plan is submitted|
|Actions/Tests ||Sec. (§)||Applicable Time Period|
|All CO nonattainment and maintenance areas||
Latest planning assumptions
Latest emissions model
TCMs in approved SIPs
Hot Spot test
|All CO nonattainment and maintenance areas, if an adequate or approved SIP budget exists||
Motor vehicle emissions budget test
|93.118||After EPA finds the motor vehicle emissions budget in the submitted revised control strategy plan or maintenance plan adequate for transportation conformity purposes|
Moderate areas with design value of 12.7 ppm or below or not classified CO nonattainment areas (not required to submit an attainment demonstration and have not submitted a maintenance plan), if no adequate or approved budget exists
If Build=no build
Interim emissions test (either build-no-greater-than-no-build test or no-greater-than-1990 test)
No regional emissions analysis required
Until maintenance plan is submitted and EPA finds the motor vehicle emissions budget adequate for transportation conformity purposes
|Serious and moderate CO areas with design value greater than 12.7 ppm, if no adequate or approved budget exists||
Interim emissions tests (build/no-build test and less-than-1990 test)
|93.119||Prior to adequate SIP budgets. If EPA declares the motor vehicle emissions reduction budget in the revised submitted control strategy implementation plan or maintenance plan inadequate for transportation purposes and no previously established motor vehicle emissions budget exists in an approved SIP or previously submitted revised control strategy or maintenance plan|
In addition to the regional analysis criteria and requirements applied to all areas at all times (as discussed in Chapter 5), and in order to demonstrate conformity, moderate and below ozone and CO nonattainment areas are required to follow these specific criteria, as stated in the transportation conformity rule,
40 CFR §93.122(d)
In all areas not otherwise subject to paragraph (b) of this section, regional emissions analysis must use those procedures described in paragraph (b) of this section if the use of those procedures has been the previous practice of the MPO. Otherwise, areas not subject to paragraph (b) of this section may estimate regional emissions using any appropriate methods that account for VMT growth by, for example, extrapolating historical VMT or projecting future VMT by considering growth in population and historical growth trends for VMT per person. These methods must also consider future economic activity, transit alternatives, and transportation system policies.
Areas without network models may project vehicle miles traveled (VMT) using any appropriate methods that account for VMT growth by, for example, extrapolating historical VMT or projecting future VMT by considering growth in population and historical growth trends for VMT per person. These methods must also consider future economic activity, transit alternatives, and transportation system policies All VMT estimates must be "calibrated" to HPMS. Also, all methods related to VMT estimations must be decided through interagency consultation.
40 CFR §93.122(b)(3)
Highway Performance Monitoring Systems (HPMS) estimates of vehicle miles traveled (VMT) shall be considered the primary measure of VMT within the portion of the nonattainment or maintenance area and for the functional classes of roadways included in HPMS, for urban areas which are sampled on a separate urban area basis. For areas with network-based travel models, a factor (or factors) may be developed to reconcile and calibrate the network-based travel model estimates of VMT in the base year of its validation to the HPMS estimates for the same period. These factors may then be applied to the model estimates of future VMT. In this factoring process, consideration will be given to differences between HPMS and network-based travel models, such as differences in the facility coverage of the HPMS and the modeled network description. Locally developed count-based programs and other departures from these procedures are permitted subject to the interagency consultation procedures of §93.105(c)(1)(i).
The requirements for regional analysis in ozone and CO moderate and below nonattainment areas are listed below:
Areas without network models may project vehicle miles traveled (VMT) using reasonable methods that account for VMT growth by, for example, extrapolating historical VMT or projecting future VMT by considering growth in population and historical growth trends for VMT per person. These methods must also consider future economic activity, transit alternatives, and transportation system policies. All VMT estimates must be "calibrated" to HPMS. Also, all methods related to VMT estimations must be decided through interagency consultation.
In addition to the regional analysis criteria and requirements applied to all areas at all times (as discussed in Chapter 5), serious and above ozone and CO nonattainment areas with an urbanized population over 200,000 are required to follow the following specific criteria, as stated in the conformity rule, in order to demonstrate conformity.
40 CFR §93.122
The criteria and procedures for regional analyses for conformity determinations in serious and above ozone and CO nonattainment areas with an urbanized population over 200,000 must meet the following requirements:
Network models are required for regional emissions analysis conducted in serious and above ozone and CO nonattainment areas with an urbanized area population of 200,000 or more. All models were to have been in place by January 1, 1997. Furthermore, all other nonattainment and maintenance areas that have been using a network-based modeling analysis must continue to do so. In addition, whether or not an area is required to use a network model, all areas must use the consultation process to select regional models and assumptions, as required by 40 CFR§93.105(c). A two-year grace period is provided in the rule for those areas not previously subjected to this requirement:
40 CFR §93.122
The decision on setting the threshold for the network modeling requirements in urbanized areas with a population of 200,000 and more is based on several factors:
62 FR 43790, August 15, 1997
...EPA believes that network modeling requirements are most important for large urbanized areas…
Commenters agreed with EPA that network modeling is not always appropriate in rural or urban areas with smaller populations, and therefore, should not be required in these areas…
...§93.122(d) of the conformity rule requires areas that are already using network models to continue using them, even if they are not serious or above areas or have a population less than 200,000. EPA and DOT will consider the specific technical needs of smaller areas when developing future modeling guidance…
The 1997 conformity rule amendments streamlined the modeling criteria from eleven (58 FR 62230-31, Nov. 24, 1993) to six (40 CFR §93.122(b)(1)(i)-(vi)). EPA believes that the streamlined criteria and clarified rule language will assist areas in implementing the rule's network modeling provisions, and the retaining of these criteria establishes minimum acceptable practice.
EPA and DOT, as stated in the preamble to the 1997 conformity rule amendments (40 CFR, 62 FR 43791), committed to develop modeling guidance in the future to address some of the modeling requirements that are eliminated from the rule and to foster the exchange of information on current and future modeling improvements, through an open stakeholder process.
The six modeling criteria and their specific requirements are summarized below. (Refer to 40 CFR §93.122(b)(1)(i)-(vi) for complete discussion.)
EPA requires that models should be validated for all modes, including transit, bicycle, and pedestrians against "observed" counts. EPA has also qualified that validation against both peak- and off-peak counts is only required where it is possible. EPA is aware that not all areas collect peak- and off-peak counts. As a result, although EPA continues to believe that validation against peak- and off-peak counts is preferable, the rule only requires it where it is already possible given available data. The travel model validation base year must not be greater than 10 years before the date of the conformity determination.
EPA requires that model forecasts be analyzed for reasonableness and compared to historical trends and other factors, and that the results be documented. Historical trends in travel behavior may include factors such as changes in per capita vehicle trips and VMT, trip length, mode shares, and time-of-day travel.
EPA requires land use, population, employment, and other network-based model assumptions to be documented and based on the best available information. See EPA/USDOT joint guidance on Use of Latest Planning Assumptions in Conformity Determinations, January 18, 2001, (Appendix L) for more information.
The distribution of employment and residences for different transportation options must be reasonable. Appropriate consideration must be given to how major anticipated transportation system improvements might influence development and, in turn, how that might affect the forecasted distribution of population and employment used to estimate travel and emissions.
EPA intends that emissions be calculated on the basis of peak- and off-peak speeds separately and applied to peak- and off-peak final assigned volumes, regardless of whether these assigned volumes are based on peak- and off-peak modeling or are modeled on a 24-hour basis.
This network modeling requirement is based on provisions of the November 1993 conformity rule (58 FR 62230 and 62249, November 24, 1993), which required feedback of travel times resulting from traffic assignment to travel times used in trip distribution. Reasons for EPA to retain this requirement include: there is clear theoretical justification for feedback between traffic assignment and trip distribution, especially in congested areas, and full feedback is already widely available and used.
The rule requires that zone-to-zone travel impedances (which may include a combination of vehicle travel time, cost, travel times by other modes, etc.) used in trip distribution be in reasonable agreement with travel times that are estimated from final assigned traffic volumes to reflect the fact that speeds should be estimated by post-processing assigned volumes.
Network-based models must be reasonably sensitive to changes in the time(s), cost(s), and other factors affecting travel choices. The November 1993 conformity rule strongly encouraged a dependence of trip generation on the accessibility of destinations, but it was not specifically required. EPA continues to believe that such a trip generation requirement is not a widely available, minimum practice.
"Reasonable methods in accordance with good practice" must be used to estimate traffic speeds and delays in a manner that is sensitive to the estimated volume of travel on each roadway segment represented in the network-based travel model (40 CFR §93.122(b)(2)).
EPA believes that free-flow speeds on network links should be based on empirical observations, a requirement in the November 1993 conformity rule (58 FR 62230 and 62249, Nov. 24, 1993), and other network speed related issues are best handled in modeling guidance, where they can be fully discussed and can avoid misinterpretation  .
EPA and DOT emphasize that input network speed assumptions used in model application must be consistent with speed assumptions used in model development and calibration, and that these assumptions and calibration techniques should be documented.
EPA and DOT recognize that free-flow impedance inputs into traffic assignment may not reflect empirically observed free-flow speeds, because these input impedances may reflect considerations that affect travel behavior other than travel time, such as driver preferences for using specific classes of facilities. If free-flow impedance inputs used in traffic assignment deviate significantly from observed free-flow speeds, the documentation should include a discussion of the differences and rationale for adjustments made.
Since emissions estimates are extremely sensitive to vehicle speed, EPA and DOT recommend that speeds be estimated in a separate step after traffic assignment (also known as "post-processing"), using refined speed-volume relationships and final assigned traffic volumes. Post-processed speeds estimated in the validation year should be compared with speeds empirically observed during the peak- and off-peak periods. These comparisons may be made for typical facilities, for example, by facility class/area type category. Based on these comparisons, speed-volume relationships used for speed post-processing should be adjusted to obtain reasonable agreement with observed speeds. Regardless of the specific analytical technique, every effort must be made to ensure that speed estimates are credible and based on a reproducible and logical analytical procedure.
The Highway Performance Monitoring System (HPMS) is a national transportation database. It includes limited data on all public roads, more detailed data for a sample of the arterial and collector functional systems, and certain summary information for urbanized, small urban and rural areas. The HPMS provides data that reflects the extent, condition, performance, use, and operating characteristics of the Nation's highways. Please see the FHWA HPMS website for more information.
As described in the conformity rule  , for areas which are sampled on a separate urban area basis, HPMS estimates of vehicle miles traveled (VMT) shall be considered the primary measure of VMT within the portion of the nonattainment or maintenance area and for the functional classes of roadways included in HPMS. For areas with network-based travel models, a factor (or factors) may be developed to reconcile and calibrate the network-based travel model estimates of VMT in the base year of its validation to the HPMS estimates for the same period. These factors may then be applied to model estimates of future VMT. In this factoring process, consideration shall be given to differences between HPMS and network-based travel models, such as differences in the facility coverage of the HPMS and the modeled network description. Locally developed count-based programs and other departures from these procedures are permitted subject to the Interagency Consultation Procedures of 40 CFR §93.105(c)(1)(i), and also discussed in Chapter 2.
Communities that are not designated serious or above ozone or CO nonattainment areas may also use the HPMS procedure described above or other locally developed programs and procedures (e.g. count-based programs), subject to the interagency consultation process. A detailed discussion of the procedures to be followed in the application of the factoring recommendations is presented later in this Chapter.
Development of regional emissions estimates for plan or TIP conformity determinations require the integration of travel demand estimates and the emissions factors output by the most current emissions factor model. Exhibit 6-3 graphically displays an example of the integration process that is required to develop regional emissions estimates.
Exhibit 6-3 shows that estimates of VMT that are output by the regional travel demand model for the base year must be compared to VMT data for the same year obtained from HPMS. The base year estimates of VMT output by the model must be adjusted to match the HPMS VMT estimates on a roadway class-specific basis, resulting in HPMS-adjusted VMT estimates for each roadway class. EPA has developed guidance that provides additional details on the technical issues involved in this adjustment process  .
Estimates of congested speed for each roadway class, and output by the regional travel demand model, are input to the MOBILE emissions model to develop roadway class-specific emissions factors for the base year. These emissions factors are combined with the HPMS-adjusted VMT estimates to compute on-road emissions for each roadway class. The resulting emissions estimates are then summed to compute total regional emissions from on-road motor vehicles.
For each future year of interest (i.e., analysis year), the regional travel model must first be used to develop roadway-class specific VMT estimates. These estimates are then compared to the base year VMT estimates output by the travel model, to develop roadway-class specific growth rates. The base year HPMS-adjusted VMT estimates described above are then multiplied by these growth rates to compute future year HPMS-adjusted VMT estimates. Similar to the base year computation methodology, congested speeds output by the model are input to MOBILE to develop roadway class-specific emissions factors. The HPMS-adjusted VMT estimates for each roadway class are then multiplied by the emissions factors output by MOBILE, and the resulting emissions estimates summed across all roadway classes to compute total regional on-road emissions. This process is followed for each selected analysis year.
The above example provides one approach to computing HPMS-adjusted VMT estimates for each analysis year and scenario. There are, however, a number of other variations that could be pursued depending on the robustness of the available data and desired level of analysis. For example, additional temporal resolution could be introduced into the analysis by performing separate peak- and off-peak period computations (or even hourly computations if such data are available). Finer spatial resolution could also be pursued by performing the calculations on a link-specific basis, or by traffic analysis zone (TAZ) or other geographic limits.
It is also possible to limit the resources required for adjusting regional travel demand modeling results to available HPMS counts, by developing a single area wide adjustment factor rather than disaggregating the analysis by roadway class. However, this approach may lead to a loss in accuracy likely to result from developing and using a single growth rate to represent the change in travel across all roadways in an urban area.
FHWA issued a memorandum on December 21, 1995  to provide guidance on issues regarding model validation and reasonable expectations of current travel model capabilities. The FHWA guidance discussed that two different processes are used to verify that: 1) the model is doing what it is expected to do (calibration), and, 2) what the model is doing is correct (validation). Calibration and validation tests typically employ comparisons between modeled and measured estimates of vehicle volumes on specific links, ignoring other model estimates such as VMT, vehicle-hours traveled (VHT), congested speeds, travel times, and delay. However; in the development of emissions inventories, key travel parameters are trips, VMT, and speed.
To provide a reasonableness check on the network model forecasts, modelers typically define a set of screenlines. Model-estimated traffic volumes are then checked against actual counts of the traffic crossing the lines. It is desirable to establish at least two screenlines, which extend to the limits of the region (one approximately east-west and the other approximately north-south). Additional screenlines are often located along natural or constructed barriers (e.g. lakes, rivers, mountain ranges, freeways, canals) within the region. Generally, the more screenlines the better, with the exact number most often depending on the resources that the planning agencies have available for such traffic count activities.
Additional insights on this issue, as well as more sophisticated methods for evaluating the validity of regional travel models is contained in the NARC Modeling Manual  (Section 3.6, Model Development and Validation, pp. 3-86).
The conventional four-step travel demand forecasting model, as shown in Exhibit 6-4, has several limitations, including:
While these limitations are valid, it is important to remember the origin and purpose of existing travel demand models. By and large, they were originally developed to help determine the size of needed highways (i.e. given projected travel patterns, how much additional capacity would be needed and where). In contrast, these models are now being used to assess the impacts of alternative policies designed to reduce trips, alter trip lengths, increase vehicle occupancy and a variety of other transportation demand management strategies. These new uses call for accurate link-level speeds and volumes to address demands for increasingly accurate emissions impact estimates.
MOBILE is designed to generate emissions estimates based on a series of default assumptions (based on national estimates) that have been coded into the model. Its accuracy can, however, be significantly improved by input of area-specific assumptions when adequate data is available. An example of the process that can be used to develop such area-specific inputs is presented in Exhibit 6-5.
The quality of emissions modeling results is only as good as the underlying data. In cases where only limited local data are available, additional data collection may be needed to improve the accuracy of the modeling results. Before embarking on a program to improve the accuracy of model inputs, analysts should perform a limited sensitivity analysis on the model. The purpose of the analysis would be to determine where to best invest limited modeling resources. Sensitivity analyses of MOBILE6 conducted by FHWA and EPA can be found at www.fhwa.dot.gov/resourcecenter/teams/airquality/publications.cfm and www.epa.gov/otaq/m6.htm respectively.
Exhibit 6-5 shows that the underlying area-specific data (vehicle registration data and annual mileage accumulation rates) can be used to compute a local VMT mix. While the data described above are available in most urban areas, this analysis requires a certain level of resources to complete. Additional resources would be needed to perform similar analysis to develop other area-specific assumptions that can be input to the emissions model.
It thus comes down to a question of what level of resources is available and should be used to make modeling improvements. The opportunity exists to improve modeling results, but doing so may require significant resources. As a general rule, conformity-related emissions analysis should rely on assumptions as good as those used in SIP-related emissions analysis.
Default VMT Mix
The default VMT mix represents a typical urban mix, based on national data characterizing registration distribution, annual mileage accumulation rates by age for each vehicle type, etc. This default mix is used to compute a composite emissions factor in MOBILE from the vehicle class-specific (e.g. light-duty vehicles) emissions factors computed by the model.
Area-Specific VMT Mix
To compute an area-specific VMT Mix, vehicle registration data can be used to compute the fraction of the in-use fleet for each vehicle class contained in the model. Annual mileage accumulation rates can also be obtained in those areas where vehicle mileage is tracked (e.g., in an I/M Program area where vehicle mileage is recorded at the time of annual or biennial testing). Average mileage accumulation rates for each vehicle class can be computed from these data and combined with the vehicle class-specific fleet fractions obtained from the registration data to compute class-specific VMT fractions (i.e. VMT mix). The resulting VMT mix is then input to MOBILE.
How do modeling improvements relate to conformity requirements?
62 FR 43792, August 15, 1997
EPA recognizes the concerns about the implementation difficulties that may occur as a result of model improvements which may lead to problems associated with inconsistencies between the models used in conformity analysis and those used in SIP development. However, Clean Air Act §176(c)(1)(B)(iii) requires conformity determinations to "be based on the most recent estimates of emissions." EPA believes that areas must use the most current tools available at the time of the conformity determination, in accordance with the Clean Air Act. Using the best models and assumptions will also produce the best emissions estimates on which areas will base decisions regarding transportation and air quality. EPA also notes that areas already have the ability to use the consultation process to coordinate the introduction of transportation modeling improvements into their planning processes.
Regional travel forecasts are available for pm peak- and off-peak periods. At what level of temporal detail (e.g. hourly, daily, etc.) should regional emissions be calculated?
The transportation conformity rule (40 CFR §93.122[b][iv]) requires that all serious and above ozone and CO nonattainment areas with urbanized area populations of greater than 200,000 must use a capacity-sensitive assignment methodology that differentiates between peak- and off-peak link volumes and speeds, and uses speeds based on final assigned volumes. Thus, regional emissions must be computed on the basis of both peak and off-peak periods. Information on latest development of travel demand forecasting models is available on Travel Model Improvement Program website: http://www.fhwa.dot.gov/planning/tmip/
 Regional analysis is required. Network models are required for serious and above ozone and CO areas with urbanized populations greater than 200,000. Areas already using network models must continue to do so. For all others areas, best professional practice should be used.
 62 FR 43794, Aug. 15, 1997.
 The November 1993 requirement was read by some to require significant data collection efforts. In fact, EPA had simply intended that available empirical information be used instead of posted speed limits.
 40 CFR §93.122(b)(3)
 U.S. EPA, §187 VMT Forecasting and Tracking Guidance, Jan. 1993.
 FHWA Memorandum, Travel Modeling Guidance for Air Quality Analysis, Dec. 21, 1995. See Appendix G.
 Harvey, G and E. Deakin, A Manual of Regional Transportation Modeling Practice for Air Quality Analysis, prepared for NARC, July 1993