Field-Level Examination of Air Quality in a Financially Challenged and Demographically Diverse region of Virginia (USA)

Field-Level Examination of Air Quality in a Financially Challenged and Demographically Diverse region of Virginia (USA)

Tim O. Moore II, P. E., Ph D, Peerawat Charuwat, EIT

Department of Civil & Environmental Engineering Virginia Military Institute

Lexington, Virginia (USA)

Abstract Concentrations of ambient NOx and PM2.5 were measured in an urban area of Virginia using a mobile emissions measurement lab. Concentrations were correlated with demographic and socioeconomic information using GIS to detect instances of adverse air pollution exposure by disadvantaged populations. Race comparison results showed that both minority and mixed populations experienced NOx and PM2.5 concentration levels as high as 89 ppb and 19 g/m3, respectively. However, in all cases of adverse air quality exposure, income level was a factor. For example, low income populations, regardless of race, were exposed to average NOx concentrations ranging from 54 to 89 ppb with 30 minute average concentrations as high as 130 to 137 ppb. For PM2.5, mixed race, low income populations experienced average concentrations of 19 g/m3 with 30 minute sustained concentrations as high as 42 g/m3, 23 95% higher than the NAAQS limit. On the contrary, high-income neighborhoods with median household incomes (MHIs) ranging from $42,600

$59,800 experienced much lower NOx concentrations between 22

• 26 ppb, 70 109% lower than high minority, low income sites. Comparative studies reveal that low income, minority populations tended to experience cancer risks 3-12 times higher than high-income populations. A DPM risk analysis was also conducted. Low-income populations in the Norfolk area, regardless of race, were experiencing DPM concentration ranging between 0.2-3.2 g/m3. Using EPA DPM risk analysis methods, results showed an increase of 183-1029 extra cancers per one million people at various low income sample locations, which is 9-53 times higher than the high-income populations in the same urban area.

  KeywordsEnvironmental justice, air pollution, ambient air quality, NOx emissions, PM2.5 emissions, socioeconomic air quality factors, adverse air quality exposure

  1. INTRODUCTION

     1. Air Quality and Environmental Justice

        Environmental justice issues involve the adverse health and economic effects of environmental hazards when disproportionately suffered by minority and low-income communities. It is well known that air pollution adversely affects the health of disadvantaged populations [1-5]. These populations often live in urban settings, have low socioeconomic status and more often than not include a large number of ethnic minorities [5]. Many studies have been conducted that link adverse health effects with the demographic characteristics of particular urban areas. One study conducted on British air quality concluded that there

        were positive relationships between poverty and known respiratory diseases (i.e. asthma) in the London area [6]. Another similar study conducted in Leeds, United Kingdom determined that there were strong positive correlations between social deprivation and respiratory health [7]. In 2001, a workshop report published by the American Lung Association confirmed that higher air pollution levels have been directly linked to the prevalence of asthma in children and adults, and in the United States, most deaths from asthma occur in the urban areas [5]. The report also determined that the asthma mortality rates among African Americans was 2.5 times higher than among Caucasians, and that analysis of those deaths showed further correlation with high poverty rates and air pollution. Similar studies conducted in New York City have shown that asthma mortality rates associated with urban air pollution were four times the citywide rate in the predominantly African American neighborhood of East Harlem [5]. A 2007 study from Canada reported higher mortality rates and increased cardiovascular disease factors associated with poor environmental conditions such as air pollution [2]. More interestingly, the study looked at environmental inequalities from a neighborhood perspective, better defining the necessity for scientists to further study the adverse effects of poor air quality on a micro level.

        Correlations between urban industrial air pollution and disadvantaged areas have been recognized by researchers as well. A study conducted in Hamilton, Ontario, Canada linked mortality and cardio-respiratory issues with exposure of minority and lower income neighborhoods to PM2.5 generated from the manufacturing of steel [8]. In 1999, researchers examined the sociodemographic characteristics of people living near industrial sources of air pollution in Kanawha Valley, West Virginia, Baton Rouge, Louisiana and Baltimore, Maryland. Results of the study determined that higher instances of diminished health quality existed in lower socioeconomic status areas and areas consisting of high minority concentrations. Adverse health effects associated with vehicle related air pollution are also of concern. A study of southern California vehicle related air pollution exposure found that minority and high poverty neighborhoods bear more than two times the level of traffic density than the rest of southern California. Furthermore, research showed increased exposure to vehicle related air pollutants in those areas [9]. Another study conducted in Boston, Massachusetts

        estimated the exposures of 413 children within a disadvantaged neighborhood to traffic related pollution. The study found an association between air pollution and asthma, and went even further by determining that children exposed to violence are more susceptible to air pollution and asthma[10]. A recent study, also in Massachusetts, found that disadvantaged neighborhoods were disproportionately exposed to diesel vehicle particulate matter emissions which were linked to increased incidences of lung cancer and asthma in those neighborhoods [11]. In New Zealand, researchers revealed that there are approximately 400 cases of premature mortality per year due to exposure to particulates emitted from vehicles; most adverse health outcomes related to poor air quality are increasingly associated with areas of low socioeconomic status and higher social deprivation [12, 13].

     2. GIS and Air Pollution in Disadvantaged Areas

        Geographical Information Systems (GIS) has been used extensively with reported air quality information to draw conclusions about air emissions and their effects on disadvantaged populations; however, very few GIS based studies use actual measured neighborhood scale emissions to determine exposure. Many studies locate populations by geocoding (assigning mapping coordinates) addresses and then using proximity analysis of a contaminant source as a surrogate for exposure. Environmental monitoring data is then integrated into the analysis to predict scenario based health outcomes[14]. Studies have shown that this method has bias and errors associated with it. One study in Orange County, Florida, compared four different geocoding methods with proximity analysis to determine the effect of positional error associated with these techniques on the analysis of exposure to traffic related air pollution of children at various school locations. Results of this study determined that the 95% root mean square error, statistical magnitude of various quantities, was greater than 300 meters in some cases, which could indicate positional inaccuracies on the data sources[15]. Other studies have used GIS to graphically provide information on the demographic characteristics of neighborhoods and correate these results with industrial pollutant releases from emissions inventories such as the Toxic Release Inventory (TRI) maintained by the EPA[16, 17]. GIS has the ability to provide a graphical database capable of providing health officials with the information necessary to properly direct programs for environmental clean-up and disease prevention. GIS can also be used as an environmental justice indicator by relating air quality risks with various sociodemographic characteristics[18]. This research uses the exploration capabilities of GIS with actual measured neighborhood scale emissions to provide an exposure analysis of various disadvantaged neighborhoods within the Norfolk, Virginia area to harmful air pollutant concentrations. By using GIS capabilities with actual measurements, achieving a more accurate exposure footprint is possible and thereby can provide public health officials with more comprehensive information on where to target remediation programs.

     3. Direct Emissions Measurements in Disadvantaged Areas

     Much research has been conducted using proximity analysis and EPA reported air quality data (i.e. – NATA reports, Air Emissions Inventories, etc.) to provide information about the exposure of disadvantaged populations to polluted air emissions; however, very little work has been done that provides actual neighborhood scale air pollutant concentration estimates within disadvantaged neighborhoods and how those measured concentrations adversely affect the residents of that neighborhood. This research focuses on nitrogen oxide (NOx) and fine particulate matter (PM2.5) concentrations in the Norfolk, Virginia and surrounding areas measured during the summer of 2008. A comparative study is used to analyze the adverse effects of NOx and PM2.5 versus other published data and a diesel particulate matter (DPM) risk analysis is conducted on all sites. The potential harmful effects of these pollutant concentrations on varying demographic and socioeconomic population characteristics are explored. The objective of the research is to compare perceived risks associated with each location and identify instances of possible environmental inequity. GIS and local health information are used in conjunction with recorded demographic characteristics of the measured areas to determine their relationship. Areas of both low and high socioeconomic status and ethnic diversity are located within the measurement footprints. Measurements are obtained using the Flux Lab for the Atmospheric Measurement of Emissions (FLAME). The FLAME is a uniquely mobile air quality measurement system capable of taking pollutant concentration measurements at any location and has been extensively described in previous publications[19]. The analyses use the concentration measurements to estimate risk within particular neighborhoods. Measurements were taken at 16 locations (SL 1-16) within a 12 square kilometer area and include parts of Norfolk, Chesapeake, Portsmouth and Virginia Beach, Virginia (Figure 1). Due to instrument malfunction at SL 9, SL 11 and SL 12, concentration measurements at those locations are not included in this analysis.

  2. METHODS

     1. Site

        A 12 square kilometer area within the city of Norfolk, Virginia and its surrounding areas (Chesapeake, Portsmouth and Virginia Beach) was the focal point for this measurement campaign. Norfolk is located within the Greater Tidewater area of Virginia and has a population of approximately 250,000. Norfolk is home to a significant amount of industries to include coal processing, rail yard activities, shipping industry, power generation and much more. The minority (i.e.

        – African American, Asian, American Indian/Pacific Islander, etc.) population in Norfolk makes up approximately 48% of the residents. According to 2000 census data, the median household income in the Norfolk area was $32,000. Approximately 20% of the population of Norfolk lives below the poverty level and about 45% of those residents are single mother households[20]. Norfolk and its surrounding areas are designated as one and eight hour ozone non-attainment areas, and based on EPA records, during the hot summer months, often experiences high levels of particulate matter. In 2006,

        the Virginia Department of Environmental Quality reported annual point source emissions of 3,600 tons in the City of Norfolk alone. Figure 1 shows the 12 representative sample locations chosen for measurement in the Norfolk area. Locations were chosen based on their proximity to various sources of anthropogenic area emissions sources as well as their varied demographic characteristics. Most locations are within close proximity to neighborhoods of varying economic status. Air quality in each location was sampled during normal weekly operations for 10 hours beginning at 7:00 AM and ending at 5:00 PM.

        Figure 1. Sample locations within Norfolk, Virginia and surrounding areas

     2. Equipment

        The FLAME is a customized television news van with an extendable mast that rises to 15.5 m. A sonic anemometer (Applied Technologies SATI-3K) and sample tubing are mounted on a rotating platform on top of the mast. A pump draws air at 20 L min-1 through 0.5-inch PTFE conductive tubing (TELEFLEX T1618-08) down to ground level, and gas and particle analyzers subsample the air through a custom designed Teflon manifold. Analyzers inside the van measure NOx (Eco Physics CLD 88Y, 1-s response time) and PM2.5 (DustTrak 8520, 1-s response time) concentrations. A data logger (National Instruments Compact FieldPoint 2110) records the measurements at 10 Hz. The equipment is powered using a 4500 W gasoline generator (Onan GENSET 4500 Series).

     3. Quality Control and Post Processing of Data

        Quality assurance and control measures included calibration of the NOx analyzer before and during the field campaign and testing for sampling line losses. Losses of NOx were 0.57%, and water vapor losses were eclipsed by humidity variations in the atmosphere during the test periods. A slight loss in PM2.5 (8% ± 5%) was also noted. Gravimetric filter samples of PM2.5 were collected during the field campaign for calibration of the DustTrak, an aerosol photometer whose response is dependent on particles optical properties. The DustTraks average concentrations were 14 ± 0.3% higher than the filter-based ones, and because filters may also be subject to sampling artifacts, we have elected to report the factory-calibrated DustTrak PM2.5 values rather than correct them to match the filters. Standard post-processing of

        the measurements included hard spike removal, soft spike removal and application of a low pass filter to ensure valid concentration measurements.[21-23]

     4. GIS Methods

     GIS was used to analyze and validate demographic comparisons of various neighborhood populations and their exposure to harmful levels of NOx and PM2.5. Using GIS, we combined information regarding demographics with socioeconomic data retrieved from the 2000 census compiled by the US Census Bureau. Combining the statistical information obtained from GIS with measured NOx and PM2.5 concentrations from the FLAME, we were able to identify instances of disproportional air pollutant exposure by various demographic and socioeconomic groups within Norfolk and the surrounding areas. 13 site locations within the 12 square kilometer Norfolk sample area were chosen for analysis. Sample location statistics within the reference areas were compiled using GIS based files from the US Census Bureau, US Geological Survey (USGS), and the Earth Resources Observation and Science (EROS). Using the compiled data, we were able to discern population demographic and financial information. The census data identifies population characteristics such as minority populations, income levels, age and gender. The research focuses on the demographic disproportion of ambient air quality as it relates to various demographic and socioeconomic characteristics within each sampling area.

  3. RESULTS ND DISCUSSION

     Figures 2 and 3 show the median household incomes and demographic composition at each of the 13 measurement locations as determined by 2000 census data. As shown in the figure, inner city areas, where higher levels of industrial activity were noted, contain higher percentages of minorities. The figure also shows that a high percentage of the predominantly minority inner city sites are also financially challenged and living below the U.S. Department of Health and Human Services (HSS) Poverty Line of $22,000 for the average 4 person family. Sites were located within a varied range of socioeconomic and demographic characteristics.

     Figure 2. Median household income of 13 sample locations in Norfolk and surrounding area

     Figure 3. Demographic characteristics at 13 sample locations in Norfolk and surrounding area

     Minority populations surrounding the sample locations ranged from 9% to 98%. Median household incomes (MHI) ranged from $20,268 to $59,779. Table 1 shows NOx and PM2.5 concentrations at each site compared with sample location demographic and financial characteristics. SL 10 maintained the highest average daily NOx concentration at 89 ppb (170 g/m3) with SL 4 reporting the highest average PM2.5 concentrations at 19 g/m3. Population characteristics at SL 10 show that the area is 98% minority with a MHI just below the HHS poverty line at $21,131. SL 1, which also had high minority percentages at 85.9%, experienced high NOx concentrations at 54 ppb with high average PM2.5 concentrations of 9.6 g/m3. The MHI at SL 1 was $20,268, 8.2% below the HHS poverty line. SL 14 and SL 16, which had the smallest minority populations, experienced the lowest NOx concentrations at 22 ppb. PM2.5 concentrations at SL 14 and SL 16 were 6.9 g/m3 and 1.7 g/m3 respectively. MHIs

     at SL 14 and SL16 were $42,563 and $59,779 respectively, which are 64% and 92% higher than the HHS poverty line.

     Other demographic characteristics such as age and predominant gender are also shown in Table 1 for each site location. Gender comparisons of each site show that most sites contained a female majority, however, on average, the male and female percentages were very close to 50% at all sites. Average age at the sample locations ranged from 20 to 48 years with SL 2 having the oldest population range and SL 4 having the youngest. SL 14 and SL 16 had some of the oldest residents at 42 and 43 years respectively. The oldest residents were in SL 2 and 6 at 48 and 47 years respectively. MHIs at SL 2 and SL 6 were in the mid-range at $35,833 and

     $34,583 respectively, approximately 47% higher than the HHS poverty line.

     A further review of the available GIS data revealed that SL 2 had the highest percentages of elderly (above 65 years old) at 34.9% and some of the lowest levels of NOx but highest PM2.5 concentrations measured during the campaign. SL 1, which had a relatively high NOx concentration of 54 ppb, had the lowest median age at 24 years and approximately 30% of the population at SL 1 was under the age of 17. Elderly populations at SL 1 were 8.6%, which is lower than the average percentages of elderly at all sites of 15.4%. SL 4, which had the lowest median age at 20 years old, experienced the highest PM2.5 concentrations at 19 g/m3. SL 4 also had the lowest percentages of the elderly in the areas. High PM2.5 concentrations at SL 4 can likely be attributed to ongoing construction in the area. For SL 10, which experienced the highest NOx concentration, both the underage and elderly percentages were higher than most sites at 26.8% and 17.4% respectively.

     Table 1. NOx and PM2.5 concentrations at 12 sample locations with demographic and financial characteristics

     Date	Site	NOx(ppb)	PM2.5(µg/m3)	Financial Characteristics	Demographic Characteristics
     					Race (%)		Gender (%)		Median Age
     				Median Household Income($)	Caucasian	Minority	Male	Female	(years)
     06/02/08	SL 1	54±75	9.6±1.8	20,268	14	86	50	50	24
     06/03/08	SL 2	24±23	8.6±4.9	35,833	29	71	46	54	48
     06/04/08	SL 3	26±64	1.1±0.4	41,346	55	45	50	50	37
     06/05/08	SL 4	24±18	19±23	24,091	71	29	49	51	20
     06/09/08	SL 5	31±60	1.0±0.3	38,846	20	80	48	52	31
     06/10/08	SL 6	36±48	1.0±0.5	34,583	79	21	45	55	47
     06/12/08	SL 8	34±33	16±15	22,829	38	62	48	52	35
     06/17/08	SL 10	89±48a,b	3.3±0.3b	21,131	2	98	46	54	38
     06/23/08	SL 13	46±17	6.4±3.8	35,223	81	19	50	50	35
     06/24/08	SL 14	22±14	6.9±1.1	42,563	91	9	48	52	42
     06/25/08	SL 15	26±24	2.3±2.7	46,250	49	51	52	48	33
     06/26/08	SL 16	22±7	1.7±0.6	59,779	64	36	47	53	43

     a Lower bound since concentrations exceeded the analyzers maximum range of 5000 ppb for ~40 s in 8.5 hours.

     b Excludes the first 90 min of measurements, when concentrations exceeded analyzers upper limits.

     1. Effects of NOx and PM2.5 Concentrations (Comparative Study)

        As shown in Table 2, significant research has been conducted providing relationships between NOx and PM2.5 exposure and increased incidences of asthma, asthma related symptoms, and mortality due to lung cancer, respiratory issues and poor cardiovascular health [10, 24-34]. For example, a study conducted in 31 cities around China determined that NOx as a single pollutant corresponded to a 1.5%, 2.3%, 2.6%, and 2.7% increase of total mortality, cardiovascular, respiratory mortality, and lung cancer respectively for every 10 g/m3 increase in concentration [24]. The average concentration measured in the China study was approximately 50 g/m3. When compared with the Norfolk data, concentrations at six of the 12 sample locations were higher than the average NOx concentrations measured in China, which based on the study would significantly increase the probability of adverse health effects associated with NOx exposure. For example, at SL10, NOx concentrations were 89 ppb (170 g/m3) corresponding to risk increases of 5.5%, 39.1%, 44.2%, and 45.9% for total mortality, cardiovascular, respiratory mortality, and lung cancer respectively. On average, the population surrounding SL10 is experiencing a risk of mortality that is 17 times higher than the 31 cities studied in China. A further review of the GIS census data reveals that at SL 10, these adverse conditions are being experienced by a low income, high minority population. Another similar study conducted in Denmark found that NOx concentrations between 17.2 and 30 g/m3 could be associated with an average 9% increase in cancer risk with a 30% increase in cancer risk for every 10 g/m3 increase in concentration above 30 g/m3[32]. With this logic, the residents at SL 1, 8 and 10 would experience a 215%, 101% and 413% increase in cancer risk respectively, which is approximately 3.7, 1.2 and 8.2 times higher than the other sample locations. Also, when compared with GIS census data, SL 1, 8 and 10 are all high minority, low income neighborhoods being adversely subjected to higher levels of NOx pollution. On the contrary, SL 2, 3, 14, 15 and 16 are all high income, low minority or mixed neighborhoods that are

        subjected to much lower levels of NOx pollution indicating possible exposure inequities. For example, at SL 14 and 16, residents are subjected to NOx concentrations of approximately 22 ppb (41 g/m3). Based on the China study, this equates to a 34% increase in cancer risk, versus the 100 400% increase experienced by residents around SL 1, 8, and 10.

        From Table 2, adverse health effects attributable to PM2.5 exposure were noted at concentrations ranging between 17.7 and 94 g/m3. For example, one report conducted in six US cities found that premature death can be associated with PM2.5 concentrations as low as 2 g/m3 [35]. SL 1, 2, 4, 8, 13 and 14 experienced concentrations that were between 25% and 117% higher than those measured in the study indicating a greater risk of premature death at these sites. Another study conducted using data obtained through the American Cancer Society (ACS) found that every 10 g/m3 increase in PM2.5 concentration above 17 g/m3 corresponded to 6 and 8% increases in the risk of cardiovascular and respiratory mortality respectively [31]. According to this study, all of the sample locations experienced no increase in risk due to PM2.5 exposure except SL 4. Based on an average PM2.5 concentration of 19 g/m3, SL 4 experienced increased cardiovascular and respiratory mortality risks of 6.78 and 9.04% respectively. This increase was likely due to noted instances of increased construction activity in the area, contributing to a temporary increase in PM2.5 emissions. At SL 10, PM2.5 data was excluded from the measurements during the hours of 7:30 to 10:30 due to a sustained exceedance of the analyzers upper measurement range of 0.1 mg/m3. Many studies provided similar results but with varying average concentrations. For example, a study conducted in 27 communities around the US reported a 1.2%, 1.0%, and 1.8% increase in total, cardiovascular, and respiratory mortality for every 10 g/m3 increase in PM2.5 concentration above 15.7 g/m3[26]. In this instance, the population at SL 4 would be experiencing a 1.6%, 1.3%, and 2.4% increase in the risk of total, cardiovascular, and respiratory mortality respectively.

        Table 2. Comparative NOx and PM2.5 Studies

        Authors	Pollutants				Health Effects
        	NOx concentration (µg/m3)	Increment (µg/m3)	PM2.5 concentration (µg/m3)	Increment (µg/m3)	NOx	PM2.5
        Nielsen et. al.[32]	17.2 29.7	10	—	—	Increased risk of lung cancer	—
        Cao et al[24]	50	10	94	10	Increased risk of mortality from cardiovascular or respiratory complications and lung cancer	Increased risk of mortality from cardiovascular or respiratory complications and lung cancer
        Pope et al (ACS)[3 1]	—	—	17.7	10	—	Increased risk of mortality from cardiovascular or respiratory complications and lung cancer
        Franklin et al[26]	—	—	15.7	10	—	Increased risk of mortality from cardiovascular or respiratory complications and lung cancer

     2. Correlation of NOx and PM2.5 Concentrations to Characteristics and the NAAQS

        The EPA regulates six criteria pollutants for the protection of public health and the environment. Among these pollutants are NO2 and PM2.5. Knowing that NOx consists of NO2 and NO, and using the EPA recommended ambient ratio method (ARM) of 0.75 NO2 to NOx, we can determine the theoretical amount of NO2 at each sample location (Table 2). Of the 12 sites, only SL 10 exceeds the NAAQS for NO2; however, based on previous studies, concentrations as low as 7 ppb have caused adverse respiratory and cardiopulmonary reactions in children and elderly adults. For example, a study conducted in Perth, Australia [36], concluded that children experienced adverse health effects from average NO2 concentrations of 7 ppb, with a range of 0 24 ppb, which in the upper range is between 8% and 115% lower than concentrations experienced at eight of the 12 sample locations.

        From Table 3, demographics at the sample locations varied with median age ranges of 24 years up to 48 years. Some sites consisted of high percentages of children under the age of 17 and elderly populations over the age of 65. For example, SL 5 consisted of 32.4% of the population being under the age of 17 and SL 2 had an elderly population of 34.9%. Based on many epidemiological studies and Integrated Science Assessments (ISA), child and elderly populations face much higher risks of developing long term respiratory symptoms and infections such as asthma from poor air quality and NO2 concentrations ranging from 3 50 ppb. Also, a number of studies reported an increase of between 1% and 13% for children and elderly adults (<65 years) exposed to adverse air quality [37].

        When conducting a race and financial based comparison, some striking correlations are noted. For example, the populations at SL 1, 8 and 10 were mostly minority at 86%,

        62% and 98% respectively. These sites also had some of the highest concentrations of NO2 at 41, 26 and 67 ppb respectively. Populations at SL 14 and 16 were low minority areas and had much lower concentrations of NO2 at 17 ppb for both sites. Of all characteristics however, financial characteristics proved to have the highest degree of correlation when compared with concentration data. In all cases, higher concentrations of NOx and PM2.5 were associated with populations living just above, at or below the HHS poverty line of $22,000. For example, at SL 8 NOx concentrations were very high in the morning hours between 07:30-0830. During this time, the 1-hour NOx concentration was measured at 112 ppb. 62.3% of the population surrounding this site is minority with a MHI of $22,829, which is 33% lower than the overall MHI of the Norfolk area. At SL 10, average NOx concentrations were approximately 89 ppb, which is 68% higher than NAAQS NO2 standards and 23% higher using the EPA ARM method. Populations surrounding SL 10 experienced high NOx concentrations at most times during the day; especially, during the rush hour, when the 1-hour average NOx concentrations were between 130-150 ppb. 98% of the population at SL 10 is minority with a median household income of $21,131, 41% lower than the Norfolk average. SL 10 was located close to major highways (I-264, I-464, Highway 460, and Highway 337) and surrouding the residential area at SL 10 was an industrial ship painting facility, ship repairing industries, on-going construction, a port authority shipping operation, and fuel storage facilities. Also adjacent to SL 10 (< 1 km) were three industries required to report emissions releases to the EPA toxic release inventory (TRI). On the contrary, at SL 14 and 16, residents only experienced NOx concentrations of 22 ppb and NO2 concentrations of 17 ppb, 68% lower than the EPA NAAQS. The majority of the population at SL 14 and 16 was Caucasian with MHIs of $42,563 and $59,779 respectively, which is 33% and 87% higher than the average Norfolk area MHI.

        Table 3. Sample location concentration comparisons to EPA NAAQS

        Date	Site	NOx(ppb)	EPA ARM Method	PM2.5(µg/m3)	NAAQS
        			NO2(ppb)		NO2	PM2.5
        06/02/08	SL 1	54±75	41	9.6±1.8
        06/03/08	SL 2	24±23	18	8.6±4.9
        06/04/08	SL 3	26±64	20	1.1±0.4
        06/05/08	SL 4	24±18	18	19±23
        06/09/08	SL 5	31±60	23	1.0±0.3
        06/10/08	SL 6	36±48	27	1.0±0.5	53 ppb (Annual)	15 µg/m3 (Annual)
        06/12/08	SL 8	34±33	26	16±15	100 ppb (1 hr)	35 µg/m3 (24 hr)
        06/17/08	SL 10	89±48a,b	67	3.3±0.3b
        06/23/08	SL 13	46±17	35	6.4±3.8
        06/24/08	SL 14	22±14	17	6.9±1.1
        06/25/08	SL 15	26±24	20	2.3±2.7
        06/26/08	SL 16	22±7	17	1.7±0.6

        a Lower bound since concentrations exceeded the analyzers maximum range of 5000 ppb for ~40 s in 8.5 hours.

        b Excludes the first 90 min of measurements, when concentrations exceeded analyzers upper limits.

     3. Correlation of NOx and PM2.5 Concentrations to Characteristics and the NAAQS

     According to the California EPA Air Resources Board (ARB), diesel particulate matter (DPM) emissions make up approximately 6% to 17% of all rural and urban particulate emissions respectively. DPM emissions cause health effects from both short term or acute exposures and also long term chronic exposures. Acute exposure to DPM may cause irritation to the eyes, nose, throat and lungs as well as some neurological effects such as lightheadedness. There is also considerable evidence that DPM is a likely carcinogen [38]. As a final comparison to this analysis and to further study the possible harmful effects of the ambient PM2.5 emissions experienced at the 12 sites, an EPA based risk analysis was conducted utilizing EPA based methodology [39]. A risk analysis helps to identify probabilities of adverse health effects to populations at the various sample locations and how the risk is proportioned based on demographic characteristics at each location. Utilizing the PM2.5 concentrations reported in Table 1 above and correcting for DPM percentages of 6 and 17%, a range of DPM exposures can be calculated using Equation 1.

     1.1 mg/kg-d, the cancer risk of DPM at each sample location can be calculated using Equation 2.

     = × (2)

     DPM cancer risk assessments are calculated in Table 4 for the 12 sample locations using ARB derived nationwide and urban estimates of 6% and 17% respectively. Based on the California ARB study, DPM corresponds to more than 70% of all adverse health risks with an estimated 540 cancers per one million people nationwide. From Table 4, SL 1 and 8, which were high minority sites with MHIs at or below the poverty line, experience cancer risks consistent with ARB estimates with SL 8 being 60% higher than ARB estimates in the upper range. PM2.5 data from SL 10, which recorded the highest NOx concentrations during the campaign, was excluded from this analysis due to PM2.5 concentrations exceeding the analyzers maximum range of 100 g/m3 during the first 90 minutes of measurements. Contrary to SL 1 and 8, SL4 had the highest range of DPM related cancer risk at 363 1,029 cancers per one million people, 91% higher than the ARB estimates. SL 4 consisted of a low minority, low income population. Residents of SL 1, 4, 8, and

     = × × × ×

     (1)

     10 experienced significantly higher risks than the other eight sample locations, which had MHI ranges between 57% and 171% above the poverty line and experienced 13% to 96%

     Ca is the concentration of DPM in the air (mg/m3), BR is the breathing rate (302 L/kg-day), EF is the exposure frequency (350 days/year), ED is the exposure duration (70 years), CF is the conversion factor (1000 L/m3), and AT is the averaging time (25,550 days). Using the calculated DPM exposure values and an EPA derived DPM cancer slope factor (CSF) of

     lower risk than the ARB estimate. Results indicate that adverse health effects are not race specific but rather income specific suggesting that residents of poorer neighborhoods in Norfolk, regardless of race, are subjected to cancer risks that are 9 53 times higher than residents within the high-income sample locations.

     Table 4. Risk Assessment for DPM within the 12 sample locations using 6% and 17% of ambient PM2.5 as DPM

     Date	Site	DPM (mg/m3)	Exposure (mg/kg-d)	Risk	One per million	Minority (%)	Income ($)
     06/02/08	SL 1	5.8-16.3E-04	1.7-4.7E-04	1.8-5.2E-04	183-520	86	20,268
     06/03/08	SL 2	5.2-14.6E-04	1.5-4.2E-04	1.64-4.6E-04	164-466	71	35,833
     06/04/08	SL 3	6.6-18.7E-05	1.9-5.4E-05	2.1-5.9E-05	21-60	45	41,346
     06/05/08	SL 4	1.1-3.2E-03	3.3-9.3E-04	3.6-10.3E-04	363-1029	29	24,091
     06/09/08	SL 5	6.0-17.0E-05	1.7-4.9E-05	1.9-5.4E-05	19-54	80	38,846
     06/10/08	SL 6	6.0-17.0E-05	1.7-4.9E-05	1.9-5.4E-05	19-54	21	34,583
     06/12/08	SL 8	9.6-27.2E-04	2.8-7.8E-04	3.1-8.6E-04	306-866	62	22,829
     06/17/08	SL 10	2.0-5.6E-04	5.7-16.2E-05	6.3-1.7E-05	63-179/p>	98	21,131
     06/23/08	SL 13	3.8-10.9E-04	1.1-3.1E-04	1.2-3.4E-04	122-347	19	35,223
     06/24/08	SL 14	4.1-11.7E-04	1.2-3.4E-04	1.3-3.4E-04	132-374	9	42,563
     06/25/08	SL 15	1.4-3.9E-04	4.0-11.3E-05	4.4-12.5E-05	44-125	51	46,250
     06/26/08	SL 16	1.0-2.8E-04	3.0-8.3E-05	3.2-9.2E-05	32-92	36	59,779

     a Not included in correlation since construction activities occurring during sampling campaign

  4. CONCLUSION

Ambient NOx and PM2.5 concentrations measured in the Norfolk area clearly indicate a relationship between demographics and exposure to harmful levels of air pollution. Results revealed that both high minority and mixed neighborhoods were experiencing NOx and PM2.5 concentration levels as high as 89 ppb and 19 g/m3, respectively. The strongest relationship existed between socioeconomic characteristics and pollution exposure levels. For example, high minority (86 98%), low income (MHI:

$20,200 $21,200) populations at SL 1 and 10 were exposed to average NOx concentrations ranging from 54 to 89 ppb with 30 minute sustained concentrations in those areas as high as 130 and 137 ppb, respectively which clearly exceed the NAAQS NO2 limit of 53 ppb. SL 8, a low minority (38%), low income (MHI: $22,300) neighborhood experienced average NOx concentrations of 34 ppb with 30 minute sustained concentrations as high as 67 ppb. For PM2.5, SL 4, a low minority (29%) low income (MHI:

$24,000) neighborhood experienced an average

concentration of 19 g/m3 with 30 minute sustained concentrations as high as 42 g/m3, 23 95% higher than the NAAQS limit of 15 g/m3. On the contrary, high income neighborhoods such as those at SL 14, 15 and 16 (MHI: $42,600 – $59,800) experienced much lower NOx concentrations between 22 26 ppb, 70 109% lower than concentrations experienced at SL 1 and 10. A comparative literature analysis on the adverse health effects of NOx and PM2.5 indicates that minority and low income populations in the Norfolk area may be experiencing cancer risks that are 3

12 times higher than the high income populations within the same urban area. Results of PM2.5 DPM risk comparisons clearly demonstrated that, regardless of race, low income populations in the Norfolk area were exposed to higher DPM concentrations ranging between 0.2 3.2 g/m3, which according to EPA DPM risk analysis procedures likely results in 183 1029 extra cancers per one million people per year at these locations which is 9 53 times higher than risks associated with the high income population areas.

ACKNOWLEDGMENT

This research was supported by a National Science Foundation (NSF) CAREER award (CBET-0547107), NSF Research Experience for Undergraduates grant (CBET- 0715162), and a Virginia Tech NSF Advance seed grant. Summer undergraduate research was funded by a VMI Summer Undergraduate Research Institute (SURI) grant. We thank Michael Klapmeyer and Linsey Marr of the Virginia Tech Environmental and Water Resources Department for their training and guidance. We also thank the Virginia Tech Transportation Institute for use of the van and the City of Norfolk for their support during the measurement campaign.

REFERENCES

1. Apelberg, B.J., T.J. Buckley, and R.H. White, Socioeconomic and racial disparities in cancer risk from air toxics in Maryland. Environmental Health Perspectives, 2005. 113(6): p. 693-699.

2. Bernard, P., et al., Health inequalities and place: A theoretical conception of neighbourhood. Social Science & Medicine, 2007. 65(9): p. 1839-1852.

3. Downey, L. and B. Hawkins, Single-mother families and air pollution: A national study. Social Science Quarterly, 2008. 89(2): p. 523-536.

4. Mirabelli, M.C., et al., Race, poverty, and potential exposure of middle-school students to air emissions from confined swine feeding operations. Environmental Health Perspectives, 2006. 114(4): p. 591-596.

5. Samet, J.M., et al., Urban air pollution and health inequities: A workshop report. Environmental Health Perspectives, 2001. 109: p. 357-374.

6. Mitchell, G. and D. Dorling, An environmental justice analysis of British air quality. Environment and Planning A, 2003. 35(5): p. 909-929.

7. Namdeo, A. and C. Stringer, Investigating the relationship between air pollution, health and social deprivation in Leeds, UK. Environment International, 2008. 34(5): p. 585- 591.

8. Elliott, S.J., et al., The power of perception: Health risk attributed to air pollution in an urban industrial neighbourhood. Risk Analysis, 1999. 19(4): p. 621-634.

9. Houston, D., et al., Structural disparities of urban traffic in Southern California: Implications for vehicle-related air pollution exposure in minority and high-poverty neighborhoods. Journal of Urban Affairs, 2004. 26(5): p. 565-592.

10. Clougherty, J.E., et al., Synergistic effects of traffic-related air pollution and exposure to violence on urban asthma etiology. Environmental Health Perspectives, 2007. 115(8): p. 1140-1146.

11. McEntee, J.C. and Y. Ogneva-Himmelberger, Diesel particulate matter, lung cancer, and asthma incidences along major traffic corridors in MA, USA: A GIS analysis. Health & Place, 2008. 14(4): p. 817-828.

12. Fisher, J.B., M. Kelly, and J. Romm, Scales of environmental justice: Combining GIS and spatial analysis for air toxics in West Oakland, California. Health & Place, 2006. 12(4): p. 701-714.

13. Kingham, S., J. Pearce, and P. Zawar-Reza, Driven to injustice? Environmental justice and vehicle pollution in Christchurch, New Zealand. Transportation Research Part D- Transport and Environment, 2007. 12(4): p. 254-263.

14. Nuckols, J.R., M.H. Ward, and L. Jarup, Using geographic information systems for exposure assessment in environmental epidemiology studies. Environmental Health Perspectives, 2004. 112(9): p. 1007-1015.

15. Zandbergen, P.A. and J.W. Green, Error and bias in determining exposure potential of children at school locations using proximity-based GIS techniques. Environmental Health Perspectives, 2007. 115(9): p. 1363- 1370.

16. Perlin, S.A., K. Sexton, and D.W.S. Wong, An examination of race and poverty for populations living near industrial sources of air pollution. Journal of Exposure Analysis and Environmental Epidemiology, 1999. 9(1): p. 29-48.

17. Perlin, S.A., D. Wong, and K. Sexton, Residential proximity to industrial sources of air pollution: Interrelationships among race, poverty, and age. Journal of the Air & Waste Management Association, 2001. 51(3): p. 406-421.

18. Jerrett, M., et al., A GIS – environmental justice analysis of particulate air pollution in Hamilton, Canada. Environment and Planning A, 2001. 33(6): p. 955-973.

19. Moore, T.O., D.C. Doughty, and L.C. Marr, Demonstration of a mobile Flux Laboratory for the Atmospheric Measurement of Emissions (FLAME) to assess emissions inventories. Journal of Environmental Monitoring, 2009. 11(2): p. 259-268.

20. City of Norfolk, Community Profile, in Community Profile,

    E.I.S. Division, Editor 2008, Virginia Employment Commission: Norfolk, Virginia. p. 38.

21. Aubinet, M., et al., Estimates of the annual net carbon and water exchange of forests: The EUROFLUX methodology, in Advances in Ecological Research, Vol 302000. p. 113-175.

22. Grimmond, C.S.B., et al., Local-scale fluxes of carbon dioxide in urban environments: Methdological challenges and results from Chicago. Environmental Pollution, 2002. 116(SUPPL 1): p. 243-254.

23. Velasco, E., et al., Measurements of CO2 fluxes from the Mexico City urban landscape. Atmospheric Environment, 2005. 39(38): p. 7433-7446.

24. Cao, J., et al., Association between long-term exposure to outdoor air pollution and mortality in China: A cohort study. Journal of Hazardous Materials, 2011. 186(2-3): p. 1594- 1600.

25. Delfino, R.J., et al., Asthma symptoms in Hispanic children and daily ambient exposures to toxic and criteria air pollutants. Environmental Health Perspectives, 2003. 111(4): p. 647-656.

26. Franklin, M., A. Zeka, and J. Schwartz, Association between PM2.5 and all-cause and specific-cause mortality in 27 US communities. Journal of Exposure Science and Environmental Epidemiology, 2007. 17(3): p. 279-287.

27. Gauderman, W.J., et al., Childhood asthma and exposure to traffic and nitrogen dioxide. Epidemiology, 2005. 16(6): p. 737-743.

28. Ma, Y.J., et al., Fine particulate air pollution and daily mortality in Shenyang, China. Science of the Total Environment, 2011. 409(13): p. 2473-2477.

29. Nafstad, P., et al., Urban air pollution and mortality in a cohort of Norwegian men. Environmental Health Perspectives, 2004. 112(5): p. 610-615.

30. Ostro, B., et al., Fine particulate air pollution and mortality in nine California counties: Results from CALFINE. Environmental Health Perspectives, 2006. 114(1): p. 29-33.

31. Pope, C.A., et al., Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. Jama-Journal of the American Medical Association, 2002. 287(9): p. 1132-1141.

32. Raaschou-Nielsen, O., et al., Lung Cancer Incidence and Long-Term Exposure to Air Pollution from Traffic. Environmental Health Perspectives, 2011. 119(6): p. 860- 865.

33. Studnicka, M., et al., Traffic-related NO2 and the prevalence of asthma and respiratory symptoms in seven year olds. European Respiratory Journal, 1997. 10(10): p. 2275-2278.

34. Villeneuve, P.J., et al., Fine particulate air pollution and all- cause mortality within the Harvard Six-Cities study: Variations in risk by period of exposure. Annals of Epidemiology, 2002. 12(8): p. 568-576.

35. Schwartz, J., F. Laden, and A. Zanobetti, The concentration- response relation between PM2.5 and daily deaths. Environmental Health Perspectives, 2002. 110(10): p. 1025- 1029.

36. Rodriguez, C., et al., The relationship between outdoor air quality and respiratory symptoms in young children. Epidemiology, 2006. 17(6): p. S107-S107.

37. EPA, ISA for Oxides of Nitrogen-Health Criteria, 2008, National Center for Environmental Assessment, Research Triangle Park, NC.

38. Ris, C., US EPA health assessment for diesel engine exhaust: A review. Inhalation Toxicology, 2007. 19: p. 229-239.

39. ARB, Methodology for Estimating Premature Deaths Associated with Long-term Exposures to Fine Airborne Particulate Matter in California, in Draft Staff Report2008, California Environmental Protection Agency, Air Resources Board.