Operating environmental impacts result from the operation of transportation vehicles. Operating impacts may include:
Emissions by pollutant type;
Air quality, including exposure to unhealthful pollutants, which is related to emissions;
Exposure to noise levels; and
Key factors determining environmental impacts include:
Transportation system performance parameters, such as VMT, speeds, and acceleration by vehicle type;
The geographic location of vehicle activity (important for air quality and noise impacts); and
Vehicle technology, including fuel efficiency, fuel type, emission controls, etc.
An integrated transportation-land use model is used to incorporate feedback from transportation investments into land development, and subsequently into transportation performance and transportation-related emissions.
A simplified air quality model is developed that simulates the dispersion of emissions from mobile and point sources. Pollutants are then overlaid on population data to obtain metrics of the air quality impacts of alternative regional development scenarios.
Transportation-related emissions data are decomposed within a GIS raster (grid cell) environment to pinpoint emissions spatially. A dispersion model is then applied to track pollutants and to measure their coincidence with population. Exposure to emissions is measured by socioeconomic group to develop indicators of equity.
GIS is used in conjunction with emissions and noise models to plot emissions and noise contours around transportation facilities. These are then overlaid on minority and low-income population data to develop measures of environmental justice.
Key emissions include criteria pollutants and precursors (VOCs, CO, NOx, PM, and SOx); toxics; and greenhouse gases, especially CO2. Emissions are an intermediate impact. Impacts of ultimate concern include the level of population exposure to unhealthful levels of criteria pollutants, as well as contributions to acidic precipitation or greenhouse gas effects.
Heuristic methods involve the multiplication of average emission rates by VMT, by vehicle type. This approach can be taken to estimate criteria pollutants or CO2 emissions from motor vehicles when a detailed analysis is not required. Emission rates can be obtained from the output of emission factor models applied for other purposes.
Emission factor models are used to develop emission rates (grams of pollutant per mile of travel or start) for a variety of pollutants. These are estimated for an average vehicle in the vehicle fleet, based on a variety of data on the vehicle fleet characteristics, temperatures, speeds, and other factors that influence emissions. Commonly used emission factor models the U.S. EPA's MOBILE and PART5 models and the California Air Resources Board's MVEI model. MOBILE and MVEI can estimate emissions of hydrocarbons, NOx, CO, and CO2. PART5 bears similarities to MOBILE and estimates PM emissions. A new version of MOBILE, MOBILE6, will be released in early 2001 to replace the existing MOBILE5 model. Outputs of emission factor models are then multiplied by VMT by speed, vehicle type, temperature, and/or other parameters to estimate overall vehicle fleet emissions. Emission factor models are widely applied in the U.S. for regulatory purposes.
If transportation inputs can be developed spatially (e.g., by traffic analysis zone), emission factors can be applied to estimate total emissions on a spatial basis. Emissions by location can then be used as inputs to regional airshed models that simulate the reaction of primary pollutants to form secondary pollutants (such as ozone), as well as the movement of these pollutants.
Considerable research has been undertaken recently to improve the state of practice in emissions modeling. In addition to continuing advances by EPA and CARB to the MOBILE and MVEI models, notable developments include:
The MEASURE model developed at the Georgia Institute of Technology. This model interfaces with travel demand model output in a GIS environment to determine emissions on a spatial basis. A further refinement is that socioeconomic data are used to estimate differences in the vehicle fleet by geographic area. This is significant because a neighborhood with older vehicles may have higher levels of emissions locally, which may in turn affect patterns of regional pollutants such as ozone. The MEASURE model has been applied in the Atlanta region.
The Comprehensive Modal Emissions Model developed for NCHRP Project 25-11 at the University of California - Riverside. These models refine the emissions estimation process by associating emissions with vehicle operating mode characteristics, such as speed/ acceleration profiles, and/or with physical characteristics such as power-to-weight ratio and transmission type. Modal and physical models have the potential for use in conjunction with traffic simulation models to develop more refined emissions estimates, and also to better estimate the effects of actions such as traffic flow improvement on emissions.
The Clean Air Act, which was last amended in 1990, requires EPA to set National Ambient Air Quality Standards for pollutants considered harmful to public health and the environment. The Clean Air Act established two types of national air quality standards. Primary standards set limits to protect public health, including the health of "sensitive" populations such as asthmatics, children, and the elderly. Secondary standards set limits to protect public welfare, including protection against decreased visibility, damage to animals, crops, vegetation, and buildings. EPA has established ambient standards for six "criteria" pollutants including carbon monoxide (CO), nitrogen dioxide (NO2), ozone (O3), lead (Pb), particulates less than 10 micrometers (PM10) and less than 2.5 micrometers (PM2.5), and sulfur dioxide (SO2).
Assessing the impacts of transportation activity on air quality requires translating emissions by location into criteria pollutants by location. Air quality models have been developed at the micro-scale, to examine air quality impacts adjacent to roadway segments or intersections, as well as at the regional scale, to examine the regionwide formation and transport of pollutants. Some emissions from vehicle exhaust, such as CO, are also criteria pollutants. Others, such as volatile organic compounds (VOC) and oxides of nitrogen (NOx), react to form criteria pollutants such as ozone.
For criteria pollutants from exhaust, it may be sufficient to model only the dispersion of pollutants from the source of emissions. For secondary pollutants such as ozone, it is also necessary to model reactions that take place in the atmosphere that lead to formation of the pollutant. In either case, modeling of air quality requires, at a minimum:
Knowledge of the spatial and temporal distribution of emissions;
Assumptions about meteorological conditions, such as wind speed and direction, temperature, humidity, and sunlight; and
Assumptions about the locations of "receptors," or the points at which air quality is being modeled.
A final step, which is usually not performed in air quality modeling, is to overlay the distribution of air pollutant concentrations with the distribution of human activity, to estimate the actual exposure of people to pollution.
Micro-scale transport/diffusion models, such as CALINE or CAL3QHC, are commonly used to assess the air quality impacts of actions such as intersection improvements or developments that increase traffic at particular intersections. These models require data on the traffic patterns on the roadway, wind speed and direction for typical meteorological conditions, and the geometry of the roadways and receptors. The models produce concentrations of pollutants by distance from the roadway. Micro-scale models are typically applied for CO, but may also be applied for NO2 and PM.
The Waterloo, Iowa case study illustrates how emission contours can be developed from microscale models, then overlaid with population data to estimate the number of people exposed to unacceptable concentrations of pollutants. The study also illustrates how census data can be used to estimate low-income and minority populations at the census block group level. The fine level of spatial detail allows the development of more precise measures describing the potential environmental justice impacts of alternative transportation projects.
Regional Gaussian dispersion models simulate the transport of pollutants through a region, usually across a two-dimensional grid. As with micro-scale models, wind speed and direction for typical meteorological conditions are used in conjunction with emissions by spatial location (generally by grid cell) to estimate concentrations in each grid cell. These models can be overlaid with population data to estimate exposure to pollutants such as NO2 and PM. Guassian dispersion models are much simpler to apply than regional airshed models, which also model reactions and may be three-dimensional. However, they cannot be used to determine exposure to pollutants, especially ozone, that are formed mainly through chemical reactions.
The Envision Utah and SPARTACUS case studies both provide examples of the application of Gaussian dispersion models at the regional level:
The Envision Utah project included the development of a model known as QMOD for northern Utah. The first component of QMOD is a model that simulates wind speed and direction over the area on an hourly basis. The second component is the capability to spatially disaggregate emissions from county-wide totals to an overlay of a regularly spaced (four-kilometer) grid. The model was used to develop metrics of population exposure to CO, PM, and ozone precursors.
In the SPARTACUS project in Europe, transportation-related emissions data are decomposed within a GIS 100-meter grid cell environment to pinpoint PM, NO2, and CO emissions spatially. A dispersion model is then applied to track pollutants and to measure their coincidence with population. Exposure to emissions is also measured by socioeconomic group to develop indicators of equity.
Regional airshed models also operate by taking inputs of typical meteorological conditions and emissions by grid cell, and simulating the movement of pollutants among grid cells. In addition, the models consider reactions between compounds that depend upon meteorological conditions. They are typically applied in the U.S. for regulatory modeling purposes, to simulate conditions under which ozone exceedances are likely to occur. The Urban Airshed Model is a commonly used model in the U.S.
Regional airshed models are data-intensive and time-consuming to develop and calibrate. Also, the relationships being modeled are quite complex and so current models have some limitations. As a result, regional airshed models are not normally used in conjunction with transportation models to model the impacts of alternative regional transportation investments or policies. With advances in computing power, scientific knowledge, and GIS data management techniques, the direct use of these models in transportation planning could become more common in the future. The MODELS-3 project being undertaken by the EPA is one effort to advance the state of practice in this area.
Noise can be defined as unwanted or detrimental sound. Traffic noise is a function of the volume, speed, and composition of traffic. The level of noise at any given point also depends on the distance from the source and physical objects that may absorb or reflect sound waves. Sound is measured in decibels (dB), which have a logarithmic scale so that an increase of 10 dB sounds twice as loud. When measuring noise, an adjustment factor is typically applied to weight high-pitched and low-pitched sounds to approximate human hearing of these sounds. The resulting measure is known as "A-weighted decibels" (dBA).
Noise from traffic can be disruptive to people both outdoors and indoors by causing sleep disturbance, communication interference, and general annoyance. Tolerance for noise can vary greatly from person to person. Communities often set acceptable noise thresholds based on the time of day and adjacent land uses. Common noise standards include L10, the noise level in dBA exceeded 10 percent of the time during specified hours; L50, the noise level exceeded 50 percent of the time; and Leq, a scale that converts a varying noise level to an equivalent constant noise level.
From a regional planning perspective, the exposure of population to traffic noise can be influenced by the location of major roadways; automobile and truck traffic volumes; the design of neighborhoods and buildings; and attenuation measures such as noise wall construction.
FHWA's Traffic Noise Prediction Model and STAMINA software were originally developed in 1977. A version of STAMINA, updated by the Minnesota DOT and known as MINNOISE, was applied in the Waterloo, Iowa case study. This model was used to compute Leq, L10, and L50 noise levels based on traffic volume, mix, and speed. Noise contours were plotted using a GIS and overlaid on low-income and minority population data to estimate exposure by population group.
The Sacramento Council of Governments (SACOG) conducted a regional noise assessment for the Environmental Impact Review of its 1999 Metropolitan Transportation Plan (MTP). SACOG employed the FHWA Highway Traffic Noise Prediction Model to conduct a cumulative noise analysis of the regional highway system. Inputs to the model included average daily traffic volume (ADT), the day/night traffic distribution, medium and heavy truck percentages, and vehicle speed. Network segments with similar traffic characteristics were aggregated, and noise levels 150 feet from the centerline were predicted. Tables and maps were then used to show the number of roadway segments in which traffic noise levels for the Proposed Plan Option were predicted to increase and decrease significantly, relative to the year 2022 No Project Plan Option.
FHWA developed an updated Traffic Noise Model (TNM) in 1998 to replace the Traffic Noise Prediction Model. In addition to the model software, FHWA has developed lookup tables from the TNM model for use in sketch-planning and screening applications. The lookup tables, which require only basic traffic and roadway geometry data, were developed by running the TNM under a variety of inputs. The TNM lookup tables are available at no charge from FHWA.
In the SPARTACUS project in Europe, transportation data are decomposed within a GIS 100-meter grid cell environment to pinpoint emissions on a spatial basis. A noise propagation model is then used to estimate noise levels by grid cell. These values are overlaid with population data to estimate exposure to noise by socioeconomic group. Noise propagation is estimated as a function of the type and density of development as well as the distance from a roadway.
The Surface Transportation Efficiency Model, STEAM, is a model developed by FHWA to estimate user benefits and costs of transportation projects, based on trip tables and networks from four-step travel demand models. STEAM includes default values to monetize noise impacts, based on VMT by vehicle type.
Energy consumption from transportation is primarily a function of vehicle-miles of travel and fuel efficiency by vehicle type. Other parameters, such as speed and acceleration, also influence energy consumption. However, their effect on regional transportation energy use is not usually considered because of the extra data requirements.
In addition to the transportation system, land use planning and urban design affect energy use. The density, mix, and arrangement of land uses in a community heavily influence the amount and mode of travel and, therefore, transportation energy use. These same urban characteristics also affect the amount of energy needed to heat and cool buildings and to build and operate community infrastructure. The relationship of energy consumption to land development patterns can be estimated by relating development by type and density to energy use. The more detail is available on the specific design characteristics of development (e.g., size of buildings, solar orientation, relative siting, topography), the more accurate the measurement of energy consumption.
Energy consumption can be important because of the negative environmental impacts associated with many types of energy use, as well as the economic and national security implications of dependence on foreign energy. In the case of transportation vehicles, energy consumption is almost entirely from petroleum-based fuels, of which a substantial proportion are from foreign oil reserves. Consumption of petroleum fuels is also directly related to emissions of carbon dioxide, a greenhouse gas.
This is a simple approach which multiplies vehicle-miles of travel (VMT) by fuel consumption rates to obtain total energy use. VMT for different transportation and/or land use scenarios may be obtained from the regional travel model or from other analysis methods. VMT should be identified for major classes of vehicles, including light-duty vehicles, trucks, and transit vehicles, to reflect the full energy impacts of each scenario.
A handful of custom models have been developed to assist state and local governments in identifying the energy impacts of different transportation and land use scenarios. These include:
The Tool for Evaluating Neighborhood Sustainability, developed for the Canadian Mortgage and Housing Corporation to evaluate greenhouse gas emissions for urban travel in different types of neighborhoods in Toronto. The tool is based on a multivariate regression model that predicts travel and greenhouse gas emissions on the basis of neighborhood attributes, sociodemographic data, and locational characteristics (e.g., distance from the CBD).
PLACE3S (PLAnning for Community Energy, Economic and Environmental Sustainability), a land use and urban design method created to help communities understand how their growth and development decisions affect energy consumption and other impacts. The PLACE3S methodology has been used by the San Diego Association of Governments (SANDAG) to assess the energy efficiency of the region's growth management strategy alternatives. The study used an integrated GIS land-use and transportation model to permit easy alteration of land use designations and to test how travel demand changes as land use changes. PLACE3S was also used for a similar exercise as part of regional transportation planning by the Lane County Council of Governments (Eugene-Springfield area) in Oregon.
INDEX, a GIS-based model that computes a variety of indicators of development patterns, including energy use. Like PLACE3S, INDEX is capable of estimating energy consumption from development patterns as well as from transportation infrastructure and travel.
The EPA Office of Transportation and Air Quality provides information and software in support of emission factor models including MOBILE5, MOBILE6, and PART5.
The EPA Support Center for Regulatory Air Models provides information and software for the Urban Airshed Model, CAL3QHC, and other air pollutant dispersion models.
National Cooperative Research Program Project 8-33, Quantifying air quality and other benefits and costs of Transportation Control Measures, reviews the state of practice and discusses needed advances in modeling emissions and air quality from transportation sources.
Guidance on estimating noise impacts of highway projects can be obtained from the Federal Highway Administration Office of Environment and Planning.