Public Health Risk

Transportation Equity Cooperative Research Program

An Approach for Identifying Environmental Justice Communities at Risk to Mobile-Source Related Air Pollution By Timothy J. Buckley, Ph.D. The Ohio State University College of Public Health Division of Environmental Health Sciences A333C Starling Loving Hall 320 W. 10th Avenue Columbus, Ohio 43210-1240 And Timothy C. Matisziw, Ph.D. The Ohio State University Center for Urban and Regional Analysis 0126 Derby Hall 154 N. Oval Mall Columbus, Ohio 43210-1361 Contributor Glenn Robinson, MA. MM., Project PI Baltimore Region Environmental Justice in Transportation Project School of Engineering and Institute for Urban Research Morgan State University 1700 East Cold Spring Lane Baltimore, MD 21251 February 22, 2008

1

Background and Rationale. Traffic-related air pollution has been implicated as a serious public health threat by a growing and increasingly convincing body of epidemiologic literature, which has linked traffic pollutant exposure with non-specific mortality (Friedman et al. 2001), cancer (Pearson et al. 2000; Knox 2005), and a variety of cardiovascular (Bigert et al. 2003) and respiratory effects (Friedman et al. 2001; Brunekreef et al. 1997; Wjst et al. 1993; Weiland et al. 1994). In addition, risks from this exposure are disproportionately borne by racial minority and socio-economically disadvantaged subpopulations (Green et al. 2004; Apelberg et al. 2005; Gunier et al. 2003). While the adverse health consequences, epidemiology, and social disparities are already compelling, it is clear that further elucidation is necessary of the magnitude, chemical composition, and variability of human exposure, and source-to-effect mechanisms. Community exposure to a complex array of traffic-related pollutants is determined by vehicle volume, as well as varied emissions characteristics of vehicles, such as differing tailpipe emissions, heat soak, tire and brake wear, and road dust re-suspension. This underlying variability in emissions drives highly dynamic concentrations of traffic-related pollutants, which are further modified by meteorology, source proximity, and human time-activity patterns. Automobiles and trucks are a major source of air pollution including such toxins and irritants as carbon monoxide (CO), nitrogen oxides (NOx), volatile organic compounds (VOCs), particulate matter, and particle-bound polycyclic aromatic hydrocarbons (PAHs). In the urban environment, high-density traffic is brought in close proximity to densely populated communities. This is particularly true in some older East Coast cities like Baltimore where row-house neighborhoods are within a couple of meters of heavily traveled urban corridors. Environmental justice is a term used to describe the movement concerned with inequities in the distribution of adverse environmental and health consequences of industrial activities and environmental policies (U.S.EPA 2004a). The movement grew from early observations that a seemingly unequal burden of pollution fell on disenfranchised and disadvantaged communities, often characterized by lower incomes and high proportions of minorities (Brown 1995). With the issuance of Presidential Executive Order 12898 in 1994, achieving “environmental justice” was integrated into the missions of all federal agencies (Clinton 1994). The U.S. Environmental Protection Agency (EPA) defines environmental justice to mean that “no group of people, including a racial, ethnic, or a socioeconomic group” should be disproportionately affected by “industrial, municipal, and commercial operations or the execution of federal, state, local, and tribal programs and policies” (U.S.EPA 2004a). There is ample evidence that minority and low-income communities bear a disproportionate burden of exposure to many environmental contaminants (Brown 1995; Institute of Medicine 1999), including air pollution (Samet et al. 2001; Schweitzer and Valenzuela 2004). Because nationwide ambient monitoring data are available for the criteria air pollutants (carbon monoxide, lead, nitrogen dioxide, ozone, particulate matter, and sulfur dioxide), we have some means for assessing exposure and risk in disadvantaged and minority communities. However, considerably less is known about

2

the distribution of exposure to and risk from the wide range of hazardous air pollutants (HAPs or “air toxics”) identified by Congress in the Clean Air Act Amendments (1990), because nationwide ambient monitoring is not possible due to the sheer number of pollutants and their diverse chemical properties (Caldwell et al. 1998; Morello-Frosch et al. 2000; Woodruff et al. 1998). A recent analysis of modeled national estimates suggests that ambient concentrations of HAPs exceed benchmark risk levels for cancer and non-cancer endpoints in many areas of the country (Caldwell et al. 1998; Woodruff et al. 1998; Woodruff et al. 2000). Furthermore, several recent studies have documented a disproportionate burden of air toxics exposure and/or risk falling on minority and low-income populations. These studies have included varying sources of exposure, including high traffic density (Green et al. 2004; Gunier et al. 2003), location of Toxic Release Inventory (TRI) and other treatment, storage, and disposal facilities (Morello-Frosch et al. 2002; Pastor et al. 2001; Perlin et al. 2001), and modeled estimates from EPA’s CEP (Lopez 2002; Morello-Frosch et al. 2002). Given the compelling evidence of a health threat that is exacerbated by environmental injustice, we have developed a strategy for identifying communities at risk using available public data. The identification of such communities is a necessary first step to empower communities, design epidemiological studies to further elucidate the threat, and implement intervention studies to address the threat. Assessing Transportation-Related Health Risk in Baltimore, Maryland As discussed earlier, the impact of transportation on public health is a growing concern in many metropolitan areas world-wide. Here we focus on Baltimore, Maryland a major metropolitan area in the United States to illustrate how existing information can be used in identifying communities at risk. Baltimore is typical of many old large east coast cities with a housing stock that is dominated by row homes built in close proximity to busy urban arterial roadways. In this section we describe existing relevant data and their analysis for the purpose of identifying communities at risk from transportationrelated air pollution. Using these existing data, we have developed a risk index to identify communities at risk due both to socio-economic status and proximity to traffic emissions. Data Central to identifying communities at risk is the acquisition and analysis of data sources that can be used to derive indicators known to influence community exposure to transportation activities. Several such indicators include proximity of residential locations to transportation infrastructure (and related activity) as well as some measure of socio-economic status. Increased proximity to transportation activity is thought to increase health risk posed by transportation while decreased socio economic status is thought to increase risk of health impacts to effected populations. Several types of geographic information are required for calculation of these risk indicators. First, one must account for the location of residential locations as well as how they are to be

3

represented in the analysis. Residential locations can be represented as either the sites of individual buildings (provided the availability of this data) or as some aggregation of residential locations such as a zip code area, census block group or a transportation analysis zone (TAZ). In this application, a geographic information system (GIS) database containing building footprints for all areas within the city of Baltimore was obtained from (2000). To compliment the building footprints, another GIS dataset detailing the location, extent, and land use for land parcels in the city was also obtained from (2000). Second, geospatial information on the location and usage of transportation infrastructure is needed. In this case, the regional planning transportation network from (2005) was used to facilitate this task. This planning network was developed to model and assign transportation activity between TAZs in the region and as such, various types of inter-TAZ traffic for the year 2005 had been already assigned to road segments in the network. That is, the volume of cars and trucks using each road segment in the network had already been estimated and attributed to each segment by (2005). Finally, socio-economic information (e.g., race, median income, education, etc.) on the study area’s population must be known to characterize socio-economic status. Here, year 2000 median income for each U.S. Census block group was used as a proxy for socioeconomic status. Block group data was selected since it is the smallest spatial unit for which census tabulations of household income data are available. Methodology Two indicator variables are used to characterize community health risk due to traffic exposure: 1) proximity and 2) socio-economic status. One way of assessing proximity to traffic is to compute the total vehicle miles (# vehicles on road segment*length of road segment) within a threshold distance (e.g., 300 feet) of a residential structure or within a tabulation area (e.g., TAZ). As discussed earlier, socio-economic status can be represented in a variety of ways (e.g., median household income). Thus, using these two basic indicators of health risk associated with traffic exposure, a simple index (RIi) can be computed representing the level of risk associated with each residence or a tabulation area i.

RI i  TVMTi  SES i

(1)

Where, i = unit of analysis (e.g., building or tabulation area) RIi = Risk index for unit i TVMTi = Total Vehicle Miles Traveled (# of vehicles * length of road segment) within some proximity threshold of unit i SESi = Socio-economic status for unit i Obviously, an issue with this index is that both variables are of a ratio nature, thus it is necessary to convert them to ordinal measurements to facilitate their integration and comparison. One way to accomplish this is to consider the range associated with each variable and split the range into bins of equal intervals. Thus, both TVMT and SES variables can be split into 10 intervals of equal size indexed 1 through 10, such that a

4

value of 1 indicates the least risk and a value of 10 indicates the greatest level of risk. For example, those residences with the lowest levels of TVMT would be assigned a value of 1 while those with the highest TVMT would be assigned a value of 10. Similarly, those residences with the highest levels of SES would be assigned a value of 1 while those with the lowest SES would be assigned a value of 10. Therefore, computing RIi using the transformed ordinal variables results in a RIi with values ranging between 2 (lowest TVMT and highest income) and 20 (highest TVMT and lowest income). TransCAD, a GIS specifically oriented toward the analysis of transportation data was used to facilitate analyses of the datasets discussed above to compute the components of the risk index. Although TransCAD is primarily oriented toward transportation analysis, it is also well suited as a general purpose GIS and, hence, has proven useful in addressing broad and diverse research questions such as those involved in deriving meaningful discoveries related to traffic, health, and environmental justice. First, TransCAD was used to derive a variable indicating proximity to transportation activity. Proximity for each residential building is defined here as all transportation activity falling within 200 feet of the building. Proximity to transportation was derived using the following GIS methodology: a. Select buildings falling within city parcels denoted as residential. b. Calculate vehicle miles traveled (VMT) for each road segment in the transportation network. Here, VMT relates to daily vehicle (cars and trucks) use of a road segment. c. Generate a 200 foot buffer for each residential building polygon. d. Overlay buffer polygons with transportation network to compute the total VMT (TVMT) falling within 200 feet of each residential building. Again, it should be noted that while residential buildings were used in this analysis, larger spatial units of analysis could also be used as well. For instance, one could use TAZs and compute an overlay with the transportation planning network as is done above with each building and compute total VMT for each TAZ. Next, TransCAD was used to attribute each residential building in the building footprint dataset with the median income of the census block group the building falls within. This measure of socio-economic status was integrated with the building data by first attributing each residential building polygon with the median income of the census block group in which the building’s centroid is located. Results The environmental justice risk modeling was applied to four Baltimore communities (Cherry Hill, Federal Hill, Kirk Avenue, and Highway to Nowhere) to exemplify it as a tool as a part of the Baltimore Region Environmental Justice in Transportation Project (BREJTP).

5

Figure 1. Risk index applied to Cherry Hill.

6

Figure 2. Risk index applied to Federal Hill.

7

Figure 3. Risk index applied to Kirk Ave.

8

Figure 4. Risk Index applied to "Hightway to Nowhere.

These maps identify building-level “hot spots” within communities where it would be reasonable to hypothesize that individuals are at risk due to the combined influence of low SES and proximity to high levels of traffic. This index is effective in differentiating neighborhoods that are both socio-economically disadvantaged and in close proximity to busy roadways. As would be expected, it can be observed that the highest index values are associated with homes in close proximity to highways and busy urban arterials (Figure 4), however, this influence and risk can be offset by blocks with high income (Figure 2). We recognize that SES and roadway proximity are not independent. The scatter plot in Figure 5 illustrates the relationship between median income and proximity to transportation activity for households in Baltimore. As hypothesized, the plotted relationship between these two variables does appear to indicate a general trend whereby traffic exposure is decreasing with increasing income.

9

Figure 5. Scatter plot of vehicle miles within 200 feet of Baltimore residences in relation to median household income

Now that TVMT and SES have been generated for each building, the ratio data can be converted to ordinal measurements and the risk index RIi for each building can then be computed. Figure 6 shows a scatterplot of the relationship between the new ordinal measurements of median household income and proximity to transportation activity. In this case, smaller values for income represent groups of households with higher median income levels. Smaller vehicle mile values indicate households with lower proximity to transportation activity. Thus, those households in higher vehicle mile intervals and in higher median income intervals are those at greatest risk for transportation related health impacts.

10

Figure 6 Scatter plot of ordinal transformation of transportation and socio economic status indicators for Baltimore households Discussion Although there is strong evidence that proximity to heavily trafficked roadways and SES conspire to place communities at risk for health threats that range from cardiovascular disease to cancer, research is needed to fully elucidate the nature of the risk and develop strategies mitigation. The current analysis provides a strategy for identifying communities at risk. This strategy relies on publicly available data combined using a GIS platform that accommodates assessment of both community-level traffic and demographics. Using this platform, we developed a risk index that accounts for factors known to place communities at risk including level of traffic, roadway proximity, and SES. The identification of such communities using this approach is a necessary first step to fully evaluating the risk and developing strategies for mitigating that risk. Furthermore, this analysis tool can be used to empower communities. References Apelberg, B. J.; Buckley, T. J.; White, R. H. Socioeconomic and racial disparities in cancer risk from air toxics in Maryland; Environ. Health Perspect. 2005, 113, 693-699. Bigert, C.; Gustavsson, P.; Hallqvist, J.; Hogstedt, C.; Lewne, M.; Plato, N.; Reuterwall, C.; Scheele, P. Myocardial infarction among professional drivers; Epidemiology 2003, 14, 333-339.

11

Brown, P. 1995. Race, Class, and Environmental-Health - A Review and Systematization of the Literature. Environmental Research 69:15-30. Brunekreef, B.; Janssen, N. A.; de Hartog, J.; Harssema, H.; Knape, M.; van Vliet, P. Air pollution from truck traffic and lung function in children living near motorways; Epidemiology 1997, 8, 298-303. Caldwell, JC, Woodruff, TJ, Morello-Frosch, R, Axelrad, DA. 1998. Application of health information to hazardous air pollutants modeled in EPA's Cumulative Exposure Project. Toxicology and Industrial Health 14:429-454. Clinton,WJ. 1994. Executive Order 12898: Federal actions to address environmental justice in minority population and low income populations. Fed Reg 59(32). Friedman, M. S.; Powell, K. E.; Hutwagner, L.; Graham, L. M.; Teague, W. G. Impact of changes in transportation and commuting behaviors during the 1996 Summer Olympic Games in Atlanta on air quality and childhood asthma; JAMA 2001, 285, 897-905. Green, RS, Smorodinsky, S, Kim, JJ, McLaughlin, R, Ostro, B. 2004. Proximity of California public schools to busy roads. Environmental Health Perspectives 112:61-66. Gunier, RB, Hertz, A, Von Behren, J, Reynolds, P. 2003. Traffic density in California: Socioeconomic and ethnic differences among potentially exposed children. Journal of Exposure Analysis and Environmental Epidemiology 13:240-246. Institute of Medicine, eds. 1999. Toward environmental justice: research, education, and health policy needs. Washington, D.C: National Academies Press. Knox, E. G. Oil combustion and childhood cancers; J. Epidemiol. Community Health 2005, 59, 755-760. Lopez, R. 2002. Segregation and Black/White differences in exposure to air toxics in 1990. Environmental Health Perspectives 110:289-295. Morello-Frosch, R, Pastor, M, Porras, C, Sadd, J. 2002. Environmental justice and regional inequality in southern California: Implications for future research. Environmental Health Perspectives 110:149-154. Morello-Frosch, RA, Woodruff, TJ, Axelrad, DA, Caldwell, JC. 2000. Air toxics and health risks in California: The public health implications of outdoor concentrations. Risk Analysis 20:273-291. Pastor, M, Sadd, J, Hipp, J. 2001. Which came first? Toxic facilities, minority move-in, and environmental justice. Journal of Urban Affairs 23:1-21.

12

Pearson, R. L.; Wachtel, H.; Ebi, K. L. Distance-weighted traffic density in proximity to a home is a risk factor for leukemia and other childhood cancers; J. Air Waste Manag. Assoc. 2000, 50, 175-180. Perlin, SA, Wong, D, Sexton, K. 2001. Residential proximity to industrial sources of air pollution: interrelationships among race, poverty, and age. J Air Waste Manag Assoc 51:406-421. Samet, JM, Dearry, A, Eggleston, PA, Ford, J, Froines, J, Gelobter, M et al. 2001. Urban air pollution and health inequities: A workshop report. Environmental Health Perspectives 109:357-374. Schweitzer, L and Valenzuela, A. 2004. Environmental injustice and transportation: The claims and the evidence. Journal of Planning Literature 18:383-398. U.S.EPA. 2004a. Environmental Justice. Available: http://www.epa.gov/compliance/environmentaljustice/ [accessed: July 2004a]. Weiland, S. K.; Mundt, K. A.; Ruckmann, A.; Keil, U. Self-reported wheezing and allergic rhinitis in children and traffic density on street of residence; Ann. Epidemiol. 1994, 4, 243-247. Wjst, M.; Reitmeir, P.; Dold, S.; Wulff, A.; Nicolai, T.; von Loeffelholz-Colberg, E. F.; von Mutius, E. Road traffic and adverse effects on respiratory health in children; BMJ 1993, 307, 596-600. Woodruff, TJ, Axelrad, DA, Caldwell, J, Morello-Frosch, R, Rosenbaum, A. 1998, Public health implications of 1990 air toxics concentrations across the United States. Environmental Health Perspectives 106:245-251. Woodruff, TJ, Caldwell, J, Cogliano, VJ, Axelrad, DA. 2000. Estimating cancer risk from outdoor concentrations of hazardous air pollutants in 1990. Environmental Research 82:194-206.

13