Understanding Analytical Tools
This section investigates the various analysis tools and methods which offer the means to quantify the suggested performance measures.We draw both on team expertise for this inventory and such key resources as the Atlanta Benefits and Burdens Study, NCHRP Project 8-36(11), and NCHRP Report 532.
The objective will be to identify tools and approaches that will be particularly relevant in analyzing the selected case studies, to compare those approaches with local capabilities, and to identify enhancements that will allow the local tools to be used most effectively in pursuing the necessary analysis.A particular need will be to be able to discern impacts on particular population subgroups using GIS tools.A technical memorandum will document the various analytic tools considered, their strengths and weaknesses, data requirements, and recommendations for local application.
The purpose of this task has been to investigate the various analytic tools and methods that will offer the means to quantify the Measures of Impact previously identified under Task 2 (BREJT memorandum to FHWA, June 23, 2006). Concurrently, the BREJT project is in the process of identifying three case study topics that will be addressed in detail in Phase II. These case studies will furnish a range of pertinent issues that will be investigated by the BREJT team.At this time, the three case studies have only been generically identified.Subsequent meetings with the community-based case study teams will establish in greater specificity the exact nature of the respective issues, and hence provide better guidance on the particular measures of impact which will be quantified, and hence, the types of analytic approaches that will be required.
Our goal in this report is to identify those tools and approaches that will be particularly relevant in analyzing the case study issues as framed, but also in the real-world context of having those issues evolve as their study progresses.It is envisioned that a variety of tools and approaches may be called upon as familiarity with the case study issue sharpens, which may be expected to lead to either different questions being raised or more detail desired on existing questions. It will further be our purpose to illustrate in the Toolkit how such typical EJT issues evolve and how the capabilities of the tools used must increase in capability correspondingly.
This report describes the process and criteria, which have been used to identify the set of analytic tools we expect to work with in the study.We present a framework that illustrates how these different techniques will be associated with particular types of levels of analyses.We then describe the tools that are currently available to us in the Baltimore region, discuss their primary uses and limitations, and address special data requirements or application procedures that may be necessary to use them.Finally, we examine the potential analysis needs of the candidate case studies, and indicate what measures of impact and what analytic tool/approach would likely be used for that assessment.
The following factors have been taken into consideration when seeking to identify the analytic tools that will be evaluated and used in this study:
- They should constitute a range of capabilities, which are appropriate to the level, and needs of the analysis.Simple methods should be available for analyses that do not require a high degree of detail and provide quick response; at the same time, more complex methods should be available to address issues of greater substance or more complex impact measures. Moreover, a desired capability of the package of tools is the ability to increase focus and detail on a particular issue as more insight is gained or as definition of the issue is refined.
- They should attempt to make maximum effective use of existing methods, databases, and organizational expertise.While key analysis questions or impact measures will not be decided solely by the current capabilities of tools and data, leverage in creative and effective use of commonly available resources will be an objective highlighted in this study.
- The tools should be capable of dealing with distributional effects: A primary consideration in environmental justice analyses is whether the incidence of a benefit or an impact falls disproportionately on one population group versus another, or versus the population as a whole.
- They should allow forecasting or prediction of impacts or effects in relation to a transportation system change, problem solution alternatives, or alternative long-term scenarios.
- They should offer the capability to visualize conditions or impacts in order to facilitate understanding and meaningful dialogue toward resolving the problem.
As with the Task 2 review of Measures of Impact, a targeted review of literature was conducted to provide insight and support for the types of analytic approaches that will be considered in addressing the issues framed by this study. In this regard, the three studies identified in the Task 2 memorandum once again proved to be the most relevant source documents for our review. This is primarily because of their focus on the identification of analytic tools and procedures in specific application to EJ issues.While other EJ literature may refer to analytic procedures used in conjunction with specific EJ issues (e.g., SCAG’s development of accessibility or tax burden measures), and provide insight into the practical experience of application, these core studies provide a structured synthesis of the techniques in relation to how and where they would be used, and in relation to the full range of EJ issues that are or should be addressed.
These core studies were previously described in the Task 2 memo. They are:
- Transportation Benefits and Burdens in the Atlanta Region, Cambridge Systematics, et alia for USDOT (May 2002).
- NCHRP Project 8-36(11): Technical Methods to Support Analyses of Environmental Justice Issues, Cambridge Systematics (April 2002).
- NCHRP Report 532: Effective Methods for Environmental Justice Assessment, Forkenbrock and Sheeley (2004).
Because of its position in the sequence of studies, NCHRP Report 532 is the most complete and comprehensive guide to the set of tools and their applications, although there are important details in the other studies that warrant separate attention.For example, NCHRP 8-36(11) provides considerable detail on the use of GIS tools and complementary applications, including population synthesis and household micro simulation approaches. The other major difference between Report 532 and NCHRP 8-36(11) is that the latter was designed intentionally to look at the capabilities of existing tools and procedures, while the former also has the objective of identifying tools, which may not be in current or common practice, but could be drawn upon for particular applications.
Perhaps the most compelling aspect of NCHRP Report 532 is its structure for packaging and recommending tools and procedures in conjunction with the setting, the question, and the level of detail needed for the analysis.This is similar to the way in which we have envisioned the EJ Toolkit from the beginning, wherein different issues–based either on geographic scale, the type of decision being supported, or the stage of the investigation–will have different analysis needs. Initial investigations, when the problem is being diagnosed and the underlying factors identified, are best served by quick and easy screening tools. However, when the investigation moves into more specificity, a shift to other measures of impact or sharper levels of geographic separation or a need for more accuracy in an estimated impact cause a shift in the complexity, data and expertise of the analytic tools which are necessary.
A simplified overview of the type of guidance in analytic tool selection provided by NCHRP Report 532.For each of 11 impact areas (transportation has been split into accessibility and choice subcategories), Report 532 suggests approaches for particular types of questions and different levels of detail. For each impact category they describe:
- The method or methods;
- The conditions under which it should be used;
- The type of analysis it should be used for, ranging from “screening” to “highly detailed”;
- The planning context in which it is most appropriate, i.e., project level, corridor level, system level, or community level;
- The data requirements, ranging from low to high; and,
- The type of expertise that is required to apply the method, including familiarity with particular software tools or models.
The report not only provides a convenient tabular summary, such as has been presented in reduced fashion here, but written step-by-step instructions on each approach. Because this resource covers more than 300 pages, it can only be described here at a very summary level, and is hereby incorporated by reference for use by the EJT Toolkit.
Generalizing from guidance provided by Report 532, the set of tools and procedures associated with the three degrees of application detail may be broadly summarized as follows:
- Assembly and review of existing data or forecasts;
- Published reports and tables;
- Creation or analysis of maps;
- Visual (field) inspections;
- Simple surveys, interviews, or focus groups.
- Standard four-step regional transportation planning model;
- Use of GIS to create maps for locating projects or impacts in relation to population subgroups at a TAZ or census tract level;
- Corridor traffic flow simulation models and analyses;
- Transportation emissions forecasting models;
- Physical measurement of noise, pollution, runoff impacts;
- Visual preference surveys;
- Formal surveys or operational data collection.
- Enhanced travel forecasting models (including activity-based methods);
- Population synthesis and household micro simulation approaches using detailed GIS;
- Pollution surface models to gauge air pollution exposure;
- Regression or other advanced statistical analysis methods to isolate and quantify contributing factors; and,
- Integrated transportation/land use models (such as DRAM/EMPAL, UrbanSim, or PECAS).
Importance of Geographic Information System (GIS) Tools
Perhaps transcending all of the other analysis tools and procedures, GIS technology is critical to effective review and evaluation of environmental justice issues. The principal reason for this is that the ability to spatially identify the location of population subgroups is paramount to dealing addressing distributive effects of a policy, plan or service. The ability to juxtapose layers of demographic information onto layers which detail the transportation system (or other prominent physical or socioeconomic feature) makes it possible to directly link transportation or travel-related information with the areas and people who are affected by it.
NCHRP 8-36(11) does a very good job in detailing the capabilities of GIS for performing EJ analyses, as well as providing insight into a substantial number of application strategies. GIS is still only coming into its own as a legitimate planning tool. For well over a decade, planners and demographers have used GIS as a tool for storing, manipulating and displaying data. However, its integration into mainline transportation planning is still evolving. Because virtually all conventional transportation (four-step) planning models operate at TAZ level of geographic detail, the ability offered by GIS to picture conditions down to the level of an individual parcel is beyond the practical use range of most current travel models.
For example, GIS can be used to compile information on the characteristics of an area circumscribing a ¼ or ½-mile radius around, say, a transit station, through geo-referencing with the respective layers upon which that buffer is imposed. To assemble the characteristics it either accumulates data at a smaller level of detail by adding up parcel (or point) information, or by interpolating from a larger unit of geography, such as a census tract or TAZ.However, if the travel model is designed to function at a TAZ level, it cannot extract much value from this more focused input. In most EJ studies, protected populations are identified based on the average condition of the TAZ, i.e., where the percentage of persons or households in that TAZ exceeds a specified threshold for race, income, auto ownership, or some other defining variable. While this is a good start toward accounting for distributional effects, it tends to overlook the often-significant differences that may occur within a TAZ based on where particular households actually live.
This unrealized potential is not to diminish the many valuable uses for GIS in EJ-related analyses, and there are methods in use that are already taking advantage of the greater spatial resolution of GIS in transportation planning. Drawing upon NCHRP 8-36(11), listed below are various GIS applications that will be considered in pursuit of the BREJT case studies and development of the EJT Toolkit.
Mapping and Visualization
GIS is most commonly used for querying spatial databases to find locations that fit criteria, mapping demographics, displaying trends or historical data, displaying assets like transportation infrastructure, visualizing areas and points of capital investment. Maps are always an important first step in analysis, revealing patterns that may not be obvious from numerical data. They also facilitate in the formation of hypotheses for statistical testing. Simple choropleth, or graduated color, maps of racial or economic characteristics by block group or even TAZ can make it easy to discern spatial patterns in the location of key population subgroups–an essential starting point for EJ analyses. This also is a convenient way to show how the location of these groups may have shifted over time, or where they are located in relation to transportation facilities or centers of opportunity. Slightly more complex maps can be developed that combine different variables into a single display, such as travel flows by origin/destination (using desire lines), volume (variable line width), and choice of mode (via pie chart).
A simple but effective mapping tool is to define a “buffer”, or select subarea, adjacent to some item of interest–e.g., a household, a transportation facility–and use GIS to accumulate information from the respective layers falling within that geography to create a profile for that space. The typical buffer consists of an area within a particular radius from the point of interest, often ¼ or ½ mile for transportation purposes, such as defining the walk-shed for transit. Combined with census information, this can be a way to determine the location of populations with EJ characteristics in proximity to an existing or proposed transportation facility or service. Unfortunately, buffer analysis is not useful for travel demand analyses, since the conventional travel demand models cannot deal with geographic breakdowns smaller than a TAZ. However, as will be discussed below, travel analyses based on individual survey households or using population synthesis methods can take advantage of this higher degree of resolution.
An interesting application of GIS is in creating surface maps, which depict conditions not only two dimensions but in a vertical dimension as well, using orthographic projection techniques. These surface maps might resemble topographical relief maps, showing differences in elevation corresponding to hills and valleys, but in this case they represent the magnitude of some demographic or transportation variable in relation to an x,y location on a map. This provides immediate visual recognition of the differences in the variable by geographic location. By co-locating more than one variable, say subtracting level of transit access from a surface of percent minority population, it is possible to begin to see if patterns of service have a systematic relationship with the demographic variable. In this case, if the peaks for minority population are increased, it would suggest that transit access is poorer in minority areas, and signal the need for a closer inspection. Mathematical tests can be applied to these paired relationships also to determine whether there is a statistical pattern inherent in the display, and the magnitude of that relationship.
GIS can be used to create measures of dispersion or concentration of a characteristic in an area, using indices that relate the characteristics of a given point with that of those surrounding it. These spatial statistics can be calculated to describe the location, centrality or dispersion of a spatially distributed variable. NCHRP 8-36(11) gives the example of a “population-weighted centroid” measure to describe how the distribution of the black population in Atlanta changed between 1980 and 1990. Also described is a “nearest neighbor” statistic that describes how clustered or spread out a population is, and whether that pattern is random or reflects a discernable trend. An index of “dissimilarity”, which measures the degree to which two variables are distributed differently over space, was used to gauge the degree of racial segregation in the Atlanta area. The degree to which the spatial distribution of minorities and non-minorities is similar or different can help assess the degree to which transportation needs are being addressed by the existing or proposed system. There are many such measures that can be used to explore the connection between spatial location, particular population groups, and transportation service levels.
Household Micro simulation Modeling
As earlier discussed, the majority of current travel demand models are based on the concept of traffic analysis zones, or TAZs. While TAZs vary in size, with densely-populated areas represented by much smaller TAZs, the general characteristic is that the TAZ is an aggregate representation of the characteristics of the population (and employment) located in that spatial area.Techniques are applied by transportation planners to stratify key conditions in a TAZ, such as household size, income, or auto ownership, but this additional detail is only for the purpose of improving the estimates of trip generation or auto ownership for the TAZ. In the end, the TAZ becomes the unit of analysis in transportation modeling. Among the shortcomings this approach imparts are that (1) the socioeconomic differences among the households in the TAZ are important factors differentiating travel needs and demands; (2) even within the same TAZ, availability of transportation alternatives can be quite different; and (3) among these key characteristics are the attributes of the protected populations (race, income, vehicle ownership, elderly, disabled, etc.).
Micro-simulation modeling approaches attempt to overcome these limitations of aggregation by basing analysis and forecasts on a sample of households or individuals that represent a larger population group. Statistical weighting methods are then used to “enumerate” the effects determined through the sample to the overall population. The advantage of this approach, in addition to greater accuracy in modeling (through use of more complex logit or activity-based models) is that the travel benefits (or impacts) associated with a transportation change can be tracked across any population characteristic that is included in the sample used for the model. Historically, this has been done for income level, since income is a key travel prediction variable, however, the characteristics of the sample can quite easily be broadened to include characteristics of race and ethnicity as well as other variables of interest.
In applying household sample micro simulation, a choice is to use either a sample of households obtained from a recent regional household travel survey, or to create a larger synthetic sample using information from the Census’ Public Use Micro data Sample (PUMS).Use of the household survey is relatively straightforward, assuming that the sample is of sufficient size, coverage and composition that it supports the types of analysis desired (Baltimore has a 2001 survey with 3,500 households represented, and extensive GIS information has already been compiled for this sample). A “synthetic” sample would be composed of a hypothetical set of households with characteristics that as a whole match those of a larger population group. Often the decision to use the synthetic sample approach is to create a larger sample, of sufficient size to support a statistically reliable analysis of the issue in question, and also to provide a revised trip table for use in running travel assignments. A population synthesizer routine is used to create the synthetic population of households from the source census files.
VMT Generation Analysis
Central to the understanding of the concentration of transportation related pollution are determination of the factors that drive VMT. VMT can be induced by a host of land use characteristics and economic and recreational opportunities across space. How each element illicit VMT demand and how VMT translates into generated pollution need to be determined. This has a number of practical uses: (1) determination of the measured link between trip generation and pollution helps better understand the environmental impacts associated with a given transportation system; (2) determination of the factors that affect VMT demand helps predict future travel demands on the basis of projected change in identified causal factors; and (3) as a matter of policy, the environmental and health impacts of VMT driving forces and their relationship to pollution can be well understood in a measured way, and hence helps identify key causal factors to focus policy actions. Econometric models and regression based statistical analysis can help achieve both the determinants of VMT demand in a transportation zone and how VMT is related to pollution generation.
Hedonic Valuation Analysis
From a broader perspective, the distribution impact of transportation generated pollution on different segments of communities is important consideration. To determine the actual economic value of the distributional impact, however, requires advanced valuation analysis. The key policy questions to consider here are (1) what is the economic value of the negative impact of transportation related pollution? (2) what is the economic value of the health index deterioration in communities due to transportation generated pollution? And (3) what is the economic value of policies that help improve pollution and health index by a given percentage across communities? All these are critical policy questions that need to be addressed in overall transportation design and implementation. A hedonic valuation approach helps design econometric models that will allow estimating the economic value of pollution concentration impacts and health index deterioration on the basis of their impact on property values. Property values are determined by physical attribute of the property and also by the environmental and health quality of communities. A hedonic valuation approach helps segment out the value of environmental and health quality from property values.
Considerations for Evaluation and Selecting Analytical Tools
The need to compile all available useful information and existing studies and forecasts as part of an initial analyses is a critical first to understanding Environmental Justice in Transportation. The following list is based on experiential BREJT project.
- Existing schematic maps portraying the minority and low income communities, as developed for a 2005 review of a regional bus service restructuring plan;
- Accessibility analyses and graphics developed by BMC for use in its latest Regional Transportation Plan “Transportation 2030”;
- Information on transit routes and schedules, ridership, and boardings, along with plans for future transit service modifications or new services (ongoing rail transit studies);
- Information on volumes and congestion of existing streets and highways, along with plans for enhancements in LRP and most recent TIP;
- Aerial photos of the Baltimore region;
- Funding breakdowns for transportation projects and programs by location;
- Crash or accident and fatality statistics;
- Myriad reports and studies prepared by BMC, Maryland state agencies, or local transportation or planning agencies.
Regional Transportation Planning Model
The Baltimore Metropolitan Council (BMC) uses a conventional four-step TAZ based model for analyzing transportation plans, policies and projects in the Baltimore region. This model encompasses Baltimore City and five surrounding counties: Anne Arundel, Baltimore, Carroll, Harford, and Howard.Based in the TP+/Viper software environment, the model operates on a system of 1,440 zones and projects trip making in relation to six trip purposes: Home-based Work, Home-based School, Home-based Shop, Home-Based Other, Work-based Other, and Non-home Based. The model is used for a wide variety of applications, including long-range forecasting, evaluation of projects for the regional Transportation Improvement Program, air quality conformity, and numerous tasks typical for an MPO.During 2004-2005, BMC initiated a state-of-the-practice review of its model by a TMIP expert panel. Part of the reason for this review is that BMC was engaged in analysis of a proposed new rail transit line (Red Line), and wished to ensure that the model would meet the stringent requirements necessary to submit an application for federal New Starts funding. Subsequent to receiving the recommendations of the TMIP panel, BMC has been undertaking important enhancements to the model, particularly in the areas of trip generation, distribution and mode choice, taking advantage of new data from its 2001 regional household travel survey. The model has been used for past EJ analyses in conjunction with the regional long range plan and in a special study of the impacts of the Maryland Transit Administration’s Greater Baltimore Bus Initiative, a radical restructuring of the regional bus system. In this latter case, the model was used to analyze changes in travel time and accessibility due to the restructuring; linked to spatial information from GIS as to the location of minority and low-income populations, it was possible to ascertain whether the burdens from the service adjustments (mainly cuts) were equitably distributed.
Corridor Simulation Models
Certain investigations may find it important to examine the impacts of a transportation system change on traffic flow in a corridor.Corridor simulation models have the purpose of simulating the micro-movements of vehicle traffic in relation to flow volumes, facility characteristics, signalization, or even exogenous events like accidents.Such simulations not only provide a means for estimating the impacts of system changes or events on traffic congestion and speeds, but also are a vital input to estimating air pollution impacts since certain emissions are sensitive to speed/acceleration parameters or have pronounced localized effects. One option to fill this analysis need is the CORSIM model which is maintained and operated by BMC. Its uses are primarily for evaluating congestion conditions and congestion mitigation strategies, but it may find application for accessibility analysis, safety, or air quality/health impacts. The BREJT team may also have access to a fairly new and highly visual traffic simulation program in the form of Trans Modeler, a new feature of the TransCAD software package that will also be used by the team.
Transportation emissions in the BMC region are estimated by BMC using the latest version of EPA’s emissions factor model, MOBILE6.2.Air quality in the Baltimore region exceeds the national standards for 8-hour ozone and for fine particulate matter (PM 2.5), and the region is in a “maintenance” phase with regard to the carbon monoxide (CO) standard. As a result, the region must demonstrate conformity of its regional Transportation Improvement Program (TIP) and its Regional Transportation Plan (RTP) with the standards for 8-hour ozone, PM 2.5, and CO for the years 2010, 2020 and 2030. The MOBILE model takes inputs from the regional transportation model conveying traffic volumes and speeds on regional transportation facilities and computes resultant emissions in relation to information on the age and mix of the regional vehicle fleet, with adjustments for fuel additives, vehicle technology, inspection and maintenance programs, and climatic conditions. The combination regional travel model and MOBILE emissions model may be used to analyze any of a wide range of mitigation strategies, by first determining the effect on travel (changes in VMT or speed) and then reapplying the emissions factor relationships.Certain strategies, however, do not lend themselves to analysis in this fashion, for reasons of compatibility with the travel or emissions models, and hence must be analyzed with alternative–frequently simpler–“off-model” types of procedures. It is important to note that the described emissions procedures do not result in estimates of actual air quality concentrations. Physical measures of emissions concentrations are determined through monitoring sites, and approximation of air quality in relation to transportation activity is done through atmospheric diffusion models, typically by the state departments of the environment.
Geographic Information Systems
The broad-reaching importance of GIS in environmental justice analysis has been already discussed.Fortunately, the options for the BREJT study appear to be good ones.BMC maintains an extensive GIS system based in ArcView/ArcGIS, with data layers containing copious information from census, individual jurisdictions, state agencies, and even private vendors.Aerial photographs cover most if not all of the region in high resolution reflecting conditions post-2000. All transportation features and facilities exist in point and line based layers, and transportation activity (volumes, speeds, congestion, and modal use) can be displayed in relation to the facilities and services, as well as desired demographic or physical geographic features. A second capability is available through the TransCAD software program, which is discussed below.
FHWA’s STEAM (Surface Transportation Efficiency Analysis) Model
This tool was developed by the FHWA as a sketch-planning tool to assist planners in developing the types of economic efficiency and other evaluative information needed for comparing the impacts of various transportation system investment options.This capability was intended to fill a void in planning practice created by the ISTEA Act’s requirements for comparing project options on more than the standard travel and cost measures. STEAM enables calculation of such key impacts as emissions, accidents, noise, energy use, and congestion delay. These impacts are calculated using inputs gathered from either the four-step regional transportation planning model or a variety of other sources. The program, which is spreadsheet-based, allows for considerable flexibility in relation to level of detail desired and quality of available data, such that the user can tailor the application to its particular needs.
TransCAD Transportation GIS Software
TransCAD is a fairly unique software package that combines GIS capabilities with transportation planning and analysis functions.While TransCAD can perform transportation analysis at the level of traffic analysis zones, its ability to manipulate data at the much finer levels of points, lines and polygons means that it can be readily applied to much finer levels of spatial resolution.Because of the aforementioned aggregation problems with TAZs, this offers an important capability for environmental justice applications. In particular, TransCAD employs built-in features that allow transition to household micro simulation modeling, which provides much more realistic estimates of impacts by socioeconomic groups and individual trip movements (which is useful in vehicle micro simulation). Combined with excellent graphical and visualization capabilities, TransCAD offers an exceptional media both for analyzing complex transportation, equity and environmental issues and for clearly communicating those results to a very diverse audience.TransCAD was earlier applied in Baltimore [DATE/CITATION] in a joint study involving BMC, Environmental Defense, and Caliper Corp., the developer of TransCAD, at which time considerable experience was gained in its application requirements and capabilities.
BMC has acquired the TransCAD software, as has Morgan State University, and we expect to employ this software in a wide range of applications in the BREJT case studies.Indeed, as there may be uncertainty to the on-demand availability of BMC’s standard tools or staff, the BREJT team intends to establish TransCAD as its primary analysis platform for the case studies and the Toolkit. This will not only provide us great flexibility in the types of analysis we may wish to undertake, but will permit us to assess the value of applying household micro simulation approaches in comparison to similar analyses done through the standard TAZ based regional model.
PECAS Integrated Transportation-Land Use Model
Regional transportation planning agencies, like BMC, are taking increasing interest in a body of planning tools that attempt to account for the fact that transportation investment and land use planning decisions are highly interrelated.Clearly, a major capital investment in a transportation facility such as a highway or rail transit line, has a major impact on the economic attractiveness of the area served by virtue of improved accessibility (reduced time and cost to travel to the area). Integrated transportation and land use models attempt to use this value-added feature of transportation to project potential impacts on land use trends and economic development in relation to transportation investment and other key underlying variables. BMC has invested in one of the more advanced such models, PECAS, developed by Douglas Hunt and John Abraham of the University of Calgary, and since 2005 has been engaged in its implementation and testing in the Baltimore region. While the BREJT project may not develop a major interest in long-range scenario planning, PECAS is a powerful tool and its availability through BMC is relatively unique. It is conceivable that questions may arise in the case studies concerning the longer-term impact of various transportation investment policies on the location of jobs or housing for minority and low income residents, as well as the impact on redevelopment and economic revitalization in the older, urban portions of the region where EJ populations are in greatest concentration.
MEASURE Pollution Surface Model
A major missing link in environmental justice studies has been the ability to quantify the relationship between transportation activity, vehicle emissions, and the impact on human health. Developing transportation-related indicators to measure public health impacts is actually a requirement under Title VI. While transportation may impact health in many ways, for example vehicle/pedestrian conflicts, noise and exhaust odors, perhaps the most pernicious is transportation’s contribution to air pollution. Poor air quality has a detrimental effect on persons with asthma or other pulmonary health problems, children and the elderly. A large and growing body of empirical research is able to demonstrate an epidemiological link between the proximity of exposure to air pollution concentrations and higher incidence rates of such health abnormalities as asthma, emphysema, and cancer. Another body of evidence shows that minority and low-income populations are most likely to live and work closer to these sources of air pollution, and hence face greater health risks.
NCHRP Report 532 provides a comprehensive review of the role of air quality in environmental justice, and lays out the issues, challenges, and potential analytic approaches for dealing with this impact.It points out that transportation-related air pollution’s effect on communities can occur in two primary ways:
- Through increased ground-level concentrations of pollutants like carbon monoxide (CO) or particulate matter (PM) caused by motor vehicle traffic and congestion; and,
- Through atmospheric concentrations of ozone and particulate-causing pollutants like VOCs, NOx, SOx, and also CO.
From an environmental justice perspective, the conventional transportation and emissions model based methods are not particularly good for dealing with either type of effect. Ground-level effects are analyzed using hot spot or micro-scale techniques that relate vehicle activity levels at roadway intersections with readings at pollution measurement receptor sites. While the models are fairly accurate at creating this linkage between activity and receptor reading, they are not able to project what concentrations are or will be in non-receptor areas, e.g., along sidewalks or inside neighborhoods. And for the measurement of atmospheric pollutants, the standard regional air quality models used for transportation conformity cannot distinguish whether concentrations are greater in some areas than others. Hence, the ability of standard tools and data to tie transportation activity to pollution concentrations in particular geographic areas–i.e., those with EJ populations–is much challenged. This obviously compromises the second element in the analysis chain, linking transportation activity with health, even though the link between pollution concentrations and health has been well demonstrated.
An interesting solution to this may be the application of the “pollution surface” concept, introduced as the fourth method in Chapter 3 of Report 532.In this approach, conventional transportation and emissions models are used to estimate emissions based on roadway geometry, traffic volumes and vehicle fleet emissions characteristics. Transportation activity is then linked to pollution concentrations as measured through receptors, spatially distributed in such a manner as to record emissions concentrations by time period over a broad sample of receptor sites. Statistical methods are then used to predict pollution levels across this defined “surface” by fitting regression models to observations at monitoring sites with known values for predictor variables such as land use, population, and vehicle miles traveled. Creation of this pollution surface makes it possible to estimate pollution concentrations and duration in particular geographic areas, which makes this a very encouraging approach for evaluating exposure for EJ populations.
Report 532 introduces a prototype model known as MEASURE (Mobile Emission Assessment System for Urban and Regional Evaluation), developed by the Georgia Institute of Technology with support from the EPA and FHWA, as a working example of a pollution surface approach.The MEASURE model operates in a GIS framework, which allows it to not only produce more accurate estimates of emissions than conventional MOBILE6 approaches, but also provide better spatial and temporal resolution of the emissions and be sensitive to how transportation system changes can affect emissions rates. Because pollutionmonitoring networks are typically sparse, Report 532 advises that a larger monitoring network and a larger number of samples over time will yield a more accurate model.