Environmental Protection Agency's Monitoring Research Articles

Higher mold levels found in in the deteriorated housing in the Sun Valley neighborhood of Denver, Colorado compared to other Denver housing and higher rates of health insurance claims for some diseases documented for the Sun Valley residents compared to other Denver residents.

The Sun Valley Homes public housing in Denver, Colorado (CO) will be replaced because of its deteriorated condition. Our goal was to document the mold contamination and particulate matter (PM2.5) concentrations in Sun Valley homes and the circulatory and respiratory health of Sun Valley compared to all Denver residents (total 2761 and 1,049,046, respectively) based on insurance claims data for 2015 to 2019. Mold contamination in Sun Valley homes (n=49) was quantified using the Environmental Relative Moldiness Index (ERMI) scale. Indoor PM2.5 concentrations were measured in Sun Valley homes (n=11) using time-integrated, filter-based samples and quantified using gravimetric analysis. Outdoor PM2.5 concentrations data were obtained from a near-by United States Environmental Protection Agency monitoring station. In Sun Valley homes, the average ERMI value was 5.25 compared to -1.25 for other Denver homes. The PM2.5 median concentration inside Sun Valley homes was 7.6μg/m3 (interquartile range - 6.4μg/m3). The ratio of indoor to outdoor concentrations of PM2.5 was 2.3 (interquartile range - 1.5). In the last five years, ischemic heart disease was significantly more likely for Denver compared to Sun Valley residents. However, acute upper respiratory infections, chronic lower respiratory diseases and asthma were all significantly more likely for Sun Valley than Denver residents. Since the process of replacing and occupying the new housing will take several years, the next phase of the study will not occur until that process is complete.

Impact of Smoke Plumes Transport on Air Quality in Sydney during Extensive Bushfires (2019) in New South Wales, Australia Using Remote Sensing and Ground Data

Smoke aerosol dispersion and transport have a significant impact on air quality levels and can be examined by environmental monitoring and modelling techniques. The purpose of this study is to determine the characteristics of the smoke aerosols and the level of air quality during November and December 2019 under the influence of extensive bushfires in the Sydney area, New South Wales (NSW), Australia. To achieve this goal, air quality and meteorological data were analysed in combination with remote sensing satellite measurements. Meteorological and air quality data were obtained from the Bureau of Meteorology (BOM) and Environmental Protection Agency monitoring sites in NSW. In Richmond the daily maximum average hourly concentration of particulate matter (PM10) was 848.9 μg/m3 at 07:00 UTC on 26 November 2019 and 785 μg/m3 at 07:00 UTC on 12 December 2019. On 10 December 2019, the highest PM10 recorded in the Sydney region was 961.5 μg/m3 in St Marys at 01:00 UTC, while the highest PM2.5 concentration was 714.6 μg/m3 in Oakdale in southwest Sydney at 18:00 UTC. These values all decreased again to the standard level (<50 μg/m3) in a few days. The potential sources of smoke aerosols originated from bushfires to the northwest of Sydney (Blue Mountains) as well as from southwest and northwest NSW. The smoke plumes were revealed by the combined AOD values from Aqua and Terra sensors on the MODIS satellite. In each case, the smoke travelled towards the east coast of Australia and out over the Pacific Ocean. The NAAPS model displays the existence of smoke at ground level, while the CALIPSO satellite data showed that the plumes extended 14 km up into the stratosphere layer. Backward trajectories obtained from the HYSPLIT model agree well with the movement of smoke plumes observed in the MODIS satellite images.

Prenatal and early life exposure to air pollution and the incidence of Kawasaki disease

Kawasaki disease (KD) is the most common form of acquired pediatric cardiac disease in the developed world. However, its etiology is still unclear. Epidemiological studies have shown that air pollution is a plausible risk factor in stimulating oxidative stress, inducing inflammation and causing autoimmune diseases. This study aims to assess the connections between prenatal and early life air pollution exposure to the incidence of KD. The main data source of this nationwide longitudinal study was the National Health Insurance Research Database (NHIRD) of Taiwan. NHIRD was linked with Taiwan Maternal and Child Health Database to establish the link between mothers and children. In total, 4192 KD cases involving children under 6 years of age were identified between January 2004 and December 2010. Children in the control group were randomly selected at a 1:4 ratio and matched using their age and index year. Integrated data for the air pollutants were obtained from 71 Environmental Protection Agency monitoring stations across Taiwan. Patients who had main admission diagnosis of KD and subsequently received intravenous immunoglobulin treatment were defined as incidence cases. Ambient exposure, including pollutant standards index (PSI), carbon monoxide (CO), nitric oxide (NO), nitric dioxide (NO2), and nitrogen oxide (NOx) during pregnancy were all positively associated with KD incidence. Conversely, ozone (O3) exposure had a negative correlation. Exposure to CO, NO, NO2, and NOx after childbirth remained consistent with regards to having a positive association with KD incidence. Exposure to PSI and O3 after delivery displayed no significant association with KD. Both prenatal and postnatal cumulative CO, NO, NO2, and NOx exposure had a dose dependent effect towards increasing KD incidence. Certain prenatal and early life air pollutant exposure may increase the incidence of KD.

A new approach for determining optimal placement of PM2.5 air quality sensors: case study for the contiguous United States

Considerable financial resources are allocated for measuring ambient air pollution in the United States, yet the locations for these monitoring sites may not be optimized to capture the full extent of current pollution variability. Prior research on best sensor placement for monitoring fine particulate matter (PM2.5) pollution is scarce: most studies do not span areas larger than a medium-sized city or examine timescales longer than 1 week. Here we present a pilot study using multiresolution dynamic mode decomposition (mrDMD) to identify the optimal placement of PM2.5 sensors from 2000 to 2016 over the contiguous United States. This novel approach incorporates the variation of PM2.5 on timescales ranging from 1 d to over a decade to capture air pollution variability. We find that the mrDMD algorithm identifies more high-priority sensor locations in the western United States than those expected along the eastern coast, where a large number of Environmental Protection Agency (EPA) PM2.5 monitors currently reside. Specifically, 53% of mrDMD optimized sensor locations are west of the 100th meridian, compared to only 32% in the current EPA network. The mrDMD sensor locations can capture PM2.5 from wildfires and high pollution events, with particularly high skill in the west. These results suggest significant gaps in the current EPA monitoring network in the San Joaquin Valley in California, northern California, and in the Pacific Northwest (Idaho, and Eastern Washington and Oregon). Our framework diagnoses where to place air quality sensors so that they can best monitor smoke from wildfires. Our framework may also be applied to urban areas for equitable placement of PM2.5 monitors.

Characterizing the Relationship between Social Determinants of Health and Air Quality

Background: Much of health is determined by peoples’ social determinants of health (SDH) and environments. Understanding them gives a more realistic model for risk stratification and the true complexity of patient populations, which may lead to more effective patient, family, and community engagement. For example, in the current COVID-19 pandemic African Americans are experiencing worse outcomes due to existing medical issues, less access to healthcare, and a higher likelihood to work unstable jobs. At the same time, there is likely an increased severity in COVID-19 cases following long-term exposure to air pollution. The objective of this study was to find relationships between SDH and environmental factors independent of a particular disease.Methods: Using the Factors Affecting Communities and Enabling Targeted Services model, we compiled SDH data for Utah census tracts from the American Community Survey data for the years 2012-2016. We assigned daily air quality measurements (AQ) for each of these census tracts based on the closest Environmental Protection Agency monitoring station. We performed clustering on the cleaned dataset using the Python package scikit-learn and visualized the obtained clusters with Tableau. We used the elbow method to obtain an optimum number of clusters and used the values for each variable in the clusters’ centroid to discover relationships between SDH and environment.Results: We obtained 10 distinct clusters of SDH and AQ. Clusters of census tracts with bad air tended to have higher proportions of individuals who were non-white, non-college educated, lower income, and lacking insurance.Conclusion: This analysis suggests a tangible relationship between poor AQ and SDH. Next steps include using modeled AQ data for finer spatial granularity and expanding timeframe and locations. The relationship between SDH and AQ irrespective of the disease is worth exploring further. The whole is more than the sum of its parts.

A Multifactorial Evaluation of the Effects of Air Pollution and Meteorological Factors on Asthma Exacerbation.

In the real world, dynamic changes in air pollutants and meteorological factors coexist simultaneously. Studies identifying the effects of individual pollutants on acute exacerbation (AE) of asthma may overlook the health effects of the overall combination. A comprehensive study examining the influence of air pollution and meteorological factors is required. Asthma AE data from emergency room visits were collected from the Taiwan National Health Insurance Research Database. Complete monitoring data for air pollutants (SO2; NO2; O3; CO; PM2.5; PM10) and meteorological factors were collected from the Environmental Protection Agency monitoring stations. A bi-directional case-crossover analysis was used to investigate the effects of air pollution and meteorological factors on asthma AE. Among age group divisions, a 1 °C temperature increase was a protective factor for asthma ER visits with OR = 0.981 (95% CI, 0.971–0.991) and 0.985 (95% CI, 0.975–0.994) for pediatric and adult patients, respectively. Children, especially younger females, are more susceptible to asthma AE due to the effects of outdoor air pollution than adults. Meteorological factors are important modulators for asthma AE in both asthmatic children and adults. When studying the effects of air pollution on asthma AE, meteorological factors should be considered.

Evaluation of El Niño-Southern oscillation influence on ozone exceedances along the United States Gulf Coast

An investigation into climate-scale influences on tropospheric ozone concentrations in cities along the United States Gulf Coast was performed using a thirty-year database of observed surface ozone data. Hourly surface ozone data were obtained from eight surface Environmental Protection Agency monitoring stations from Texas to Florida. Eight-hour peak concentrations were computed for each day in the record and a database created for further study. To account for regulatory changes, each time history was detrended over the thirty-year period. Days with precipitation, which tend to have low surface ozone concentrations due to low solar insolation and wet deposition of ozone precursors on surfaces, were also removed. The resulting database was stratified by Oceanic Niño Index, which is a measure of the state of the El Niño-Southern Oscillation (ENSO) based on water temperatures in the eastern Pacific Ocean. A common seasonal dependency was identified for all eight stations, with peaks in ozone concentrations in spring and autumn regardless of ENSO phase. The summer ozone minimum was more pronounced for the western Gulf of Mexico than the east. This extension of work done for the city of Houston, Texas confirmed that the influence of ENSO on surface air quality applies to the entire coast of the Gulf of Mexico. Statistically significant enhancements of 5–10 ppbv were found during the months of March and April for El Niño years relative to neutral years. Distributions fit to each city's dataset of 8-h peak ozone values for spring and fall confirmed the difference between the behavior of ozone concentrations in spring and fall for different ENSO phases. El Niño spring data possessed a broader distribution more centered at a higher mean value than the neutral or La Niña datasets for each city. Taken together, the higher mean and broader distribution of 8-h ozone concentrations lead to 30% more exceedances of the National Ambient Air Quality Standard of 70 ppbv for El Niño years relative to neutral years for the composite Gulf Coast dataset. These changes in the ozone climate of the Gulf Coast during El Niño years can be attributed in part to an area of anomalously low pressure in the eastern Gulf of Mexico which influences meridional and zonal flow along the coast. This feature leads to stagnation and a persistent continental airmass along the northern Gulf coast. Smaller but still statistically significant differences between El Niño and neutral autumn were also identified for some stations and were attributed to increased frontal activity during El Niño autumn which leads to an earlier, longer ozone season in previous research. This work extends and improves understanding of the influence of the El Niño-Southern Oscillation on air quality in a large and economically important region of the United States.

Chronic PM2.5 Exposure and Lipids/Lipoproteins among Midlife Women

Fine particles (PM2.5) are known to increase risks of cardiovascular diseases, but very few studies have examined impact of long-term PM2.5 exposure on plasma lipid. In this study, we examined the association between PM2.5 exposure and lipid/lipoprotein among 2289 midlife women from six sites across U.S. between 1999 and 2005 (average 49±3 years old at the first visit included), and explored potential vulnerable subgroups. PM2.5 data were obtained from the U.S. Environmental Protection Agency monitoring network. The average prior one-year exposure was calculated and assigned to each woman based on proximity of the monitors to her residential address. Mixed effects models were used to account for repeated measures for each woman, adjusted for site, race/ethnicity, education, and time-varying variables including age at the visit, visit number, menopausal status, body mass index (BMI), smoking status, alcohol use in the past 24 hours, hormone use, and non-work activity level. The changes in lipids/lipoproteins presented different patterns by dyslipidemia status, menopausal status, and BMI categories. With increased PM2.5 exposure, increased atherogenic lipoproteins, such as, Apolipoprotein (Apo) B and lipoprotein(a), were observed in women without dyslipidemia and women with BMI less than 25, e.g., 5.0% (95%th CIs: 0.8%, 9.4%) and 6.1% (1.1%, 11.3%) increases in ApoB in these two groups, respectively, per 10 µg/m3 increase of one-year average PM2.5 exposure. Reduced protective lipoproteins, including ApoA1 and lipoprotein A1, were observed in women with dyslipidemia and overweight (25≤BMI<30) and obese (BMI≥30) women. Both trends were observed in peri-menopausal women (had a period within the past 11 months but became irregular). These results suggested that chronic PM2.5 exposure was adversely associated with lipid levels through different mechanisms in midlife women, and all lead to increased cardiovascular risks.

Ozone response to emission reductions in the southeastern United States

Abstract. Ozone (O3) formation in the southeastern US is studied in relation to nitrogen oxide (NOx) emissions using long-term (1990s–2015) surface measurements of the Southeastern Aerosol Research and Characterization (SEARCH) network, U.S. Environmental Protection Agency (EPA) O3 measurements, and EPA Clean Air Status and Trends Network (CASTNET) nitrate deposition data. Annual fourth-highest daily peak 8 h O3 mixing ratios at EPA monitoring sites in Georgia, Alabama, and Mississippi exhibit statistically significant (p < 0.0001) linear correlations with annual NOx emissions in those states between 1996 and 2015. The annual fourth-highest daily peak 8 h O3 mixing ratios declined toward values of ∼ 45–50 ppbv and monthly O3 maxima decreased at rates averaging ∼ 1–1.5 ppbv yr−1. Mean annual total oxidized nitrogen (NOy) mixing ratios at SEARCH sites declined in proportion to NOx emission reductions. CASTNET data show declining wet and dry nitrate deposition since the late 1990s, with total (wet plus dry) nitrate deposition fluxes decreasing linearly in proportion to reductions of NOx emissions by ∼ 60 % in Alabama and Georgia. Annual nitrate deposition rates at Georgia and Alabama CASTNET sites correspond to 30 % of Georgia emission rates and 36 % of Alabama emission rates, respectively. The fraction of NOx emissions lost to deposition has not changed. SEARCH and CASTNET sites exhibit downward trends in mean annual nitric acid (HNO3) concentrations. Observed relationships of O3 to NOz (NOy–NOx) support past model predictions of increases in cycling of NO and increasing responsiveness of O3 to NOx. The study data provide a long-term record that can be used to examine the accuracy of process relationships embedded in modeling efforts. Quantifying observed O3 trends and relating them to reductions in ambient NOy species concentrations offers key insights into processes of general relevance to air quality management and provides important information supporting strategies for reducing O3 mixing ratios.

Impact of California Fires on Local and Regional Air Quality: The Role of a Low-Cost Sensor Network and Satellite Observations.

PM2.5, or fine particulate matter, is a category of air pollutant consisting of particles with effective aerodynamic diameter equal to or less than 2.5 μm. These particles have been linked to human health impacts as well as regional haze, visibility, and climate change issues. Due to cost and space restrictions, the U.S. Environmental Protection Agency monitoring network remains spatially sparse. To increase the spatial resolution of monitoring, previous studies have used satellite data to estimate ground‐level PM concentrations, despite these estimates being associated with moderate to large uncertainties when relating a column measure of aerosol (aerosol optical depth) with surface measurements. To this end, we discuss a low‐cost air quality monitor (LCAQM) network deployed in California. In this study, we present an application of LCAQM and satellite data for quantifying the impact of wildfires in California during October 2017. The impacts of fires on PM2.5 concentration at varying temporal (hourly, daily, and weekly) and spatial (local to regional) scales have been evaluated. Comparison between low‐cost air quality sensors and reference‐grade air quality instruments shows expected performance with moderate to high uncertainties. The LCAQM measurements, in the absence of federal equivalent method data, were also found to be very useful in developing statistical models to convert aerosol optical depth into PM2.5 with performance of satellite‐derived PM2.5, similar to that obtained using the federal equivalent method data. This paper also highlights challenges associated with both LCAQM and satellite‐based PM2.5 measurements, which require further investigation and research.

Formaldehyde (HCHO) As a Hazardous Air Pollutant: Mapping Surface Air Concentrations from Satellite and Inferring Cancer Risks in the United States.

Formaldehyde (HCHO) is the most important carcinogen in outdoor air among the 187 hazardous air pollutants (HAPs) identified by the U.S. Environmental Protection Agency (EPA), not including ozone and particulate matter. However, surface observations of HCHO are sparse and the EPA monitoring network could be prone to positive interferences. Here we use 2005-2016 summertime HCHO column data from the OMI satellite instrument, validated with high-quality aircraft data and oversampled on a 5 × 5 km2 grid, to map surface air HCHO concentrations across the contiguous U.S. OMI-derived summertime HCHO values are converted to annual averages using the GEOS-Chem chemical transport model. Results are in good agreement with high-quality summertime observations from urban sites (-2% bias, r = 0.95) but a factor of 1.9 lower than annual means from the EPA network. We thus estimate that up to 6600-12 500 people in the U.S. will develop cancer over their lifetimes by exposure to outdoor HCHO. The main HCHO source in the U.S. is atmospheric oxidation of biogenic isoprene, but the corresponding HCHO yield decreases as the concentration of nitrogen oxides (NOx ≡ NO + NO2) decreases. A GEOS-Chem sensitivity simulation indicates that HCHO levels would decrease by 20-30% in the absence of U.S. anthropogenic NOx emissions. Thus, NOx emission controls to improve ozone air quality have a significant cobenefit in reducing HCHO-related cancer risks.

Adult atopic dermatitis and exposure to air pollutants—a nationwide population-based study

BackgroundThere is a trend toward an increased worldwide prevalence of allergic diseases. It is speculated that industrialization with resultant air pollution plays a role. However, there are sparse epidemiologic data on the relation between air pollution and atopic dermatitis (AD) in adults. ObjectiveTo investigate the relation between exposure to air pollutants and adult AD in a cross-sectional study based on data from the National Health Insurance Research Database in Taiwan. MethodsWe identified 1,023 adult patients with AD and 4,092 age- and sex-matched controls without allergic diseases in 2011. Using data from 71 Environmental Protection Agency monitoring stations across Taiwan, levels of exposure to air pollutants were determined by the location of a subject's place of residence. Multivariate logistic regression analysis, adjusted for age, sex, levels of urbanization, and family income, was performed. ResultsWe found an association between particulate matter <2.5 μm in diameter or the Pollutant Standards Index (the highest sub-index of the concentrations of 5 main air pollutants after transformation) and the development of adult AD. The adjusted odds ratios were 1.05 (95% confidence interval 1.02–1.08) and 1.02 (95% confidence interval 1.01–1.03), respectively. ConclusionThese results demonstrated that air pollution, represented by particulate matter <2.5 μm in diameter or the Pollutant Standards Index, was modestly associated with the development of AD in adults.

Spatiotemporal calibration and resolution refinement of output from deterministic models.

Spatiotemporal calibration of output from deterministic models is an increasingly popular tool to more accurately and efficiently estimate the true distribution of spatial and temporal processes. Current calibration techniques have focused on a single source of data on observed measurements of the process of interest that are both temporally and spatially dense. Additionally, these methods often calibrate deterministic models available in grid-cell format with pixel sizes small enough that the centroid of the pixel closely approximates the measurement for other points within the pixel. We develop a modeling strategy that allows us to simultaneously incorporate information from two sources of data on observed measurements of the process (that differ in their spatial and temporal resolutions) to calibrate estimates from a deterministic model available on a regular grid. This method not only improves estimates of the pollutant at the grid centroids but also refines the spatial resolution of the grid data. The modeling strategy is illustrated by calibrating and spatially refining daily estimates of ambient nitrogen dioxide concentration over Connecticut for 1994 from the Community Multiscale Air Quality model (temporally dense grid-cell estimates on a large pixel size) using observations from an epidemiologic study (spatially dense and temporally sparse) and Environmental Protection Agency monitoring stations (temporally dense and spatially sparse). Copyright © 2016 John Wiley & Sons, Ltd.

Measuring Pedestrian Exposure to PM-2.5: Case of the Seattle, Washington, International District

Traffic-related air pollution is dangerous to human health. Although transportation and land use planning policies often focus on making walking more attractive, there is a lack of research on pedestrian exposure to air pollution levels. This research focused on pedestrian exposure to particulate matter with a diameter of 2.5 µm or less (PM-2.5) in the International District (ID) adjacent to downtown Seattle, Washington. Several types of equipment were used: ( a) a portable nephelometer (Radiance Research M903) mounted on a backpack (arranged by the Puget Sound Clean Air Agency); ( b) an AirCasting mobile application (by Habitmap) in a cell phone to record the researcher’s location and exposure levels while walking; and ( c) a GoPro Hero camera to record visual images of the surrounding built environment, traffic volume, and other activities. The field data were collected three times a day (morning, midday, and evening) for one week in winter (December 31, 2014–January 9, 2015) and one week in spring (March 21–30, 2015) on selected routes in the ID. The study found seasonal and time-of-day variability of exposure levels: there were higher PM-2.5 concentration levels during the winter (57.77 µg/m3) than in the spring (6.99 µg/m3), and higher levels in the morning (25 µg/m3) than in the evening (17 µg/m3). Also, the average PM-2.5 levels of ID data were slightly higher (20.7 µg/m3) than those at the nearest U.S. Environmental Protection Agency monitoring station (19.0 µg/m3). The researchers concluded that the key contributors of pedestrian exposure to air pollution are traffic, construction activities, and smokers on sidewalks.

Forecasting fish biomasses, densities, productions, and bioaccumulation potentials of mid-atlantic wadeable streams.

Regional fishery conditions of Mid-Atlantic wadeable streams in the eastern United States are estimated using the Bioaccumulation and Aquatic System Simulator (BASS) bioaccumulation and fish community model and data collected by the US Environmental Protection Agency's Environmental Monitoring and Assessment Program (EMAP). Average annual biomasses and population densities and annual productions are estimated for 352 randomly selected streams. Realized bioaccumulation factors (BAF) and biomagnification factors (BMF), which are dependent on these forecasted biomasses, population densities, and productions, are also estimated by assuming constant water exposures to methylmercury and tetra-, penta-, hexa-, and hepta-chlorinated biphenyls. Using observed biomasses, observed densities, and estimated annual productions of total fish from 3 regions assumed to support healthy fisheries as benchmarks (eastern Tennessee and Catskill Mountain trout streams and Ozark Mountains smallmouth bass streams), 58% of the region's wadeable streams are estimated to be in marginal or poor condition (i.e., not healthy). Using simulated BAFs and EMAP Hg fish concentrations, we also estimate that approximately 24% of the game fish and subsistence fishing species that are found in streams having detectable Hg concentrations would exceed an acceptable human consumption criterion of 0.185 μg/g wet wt. Importantly, such streams have been estimated to represent 78.2% to 84.4% of the Mid-Atlantic's wadeable stream lengths. Our results demonstrate how a dynamic simulation model can support regional assessment and trends analysis for fisheries.

Chlorate Challenges for Water Systems

Chlorate, currently included in the US Environmental Protection Agency's monitoring of unregulated contaminants and on the contaminant candidate list, could potentially receive a regulatory determination in the near future. This article, using available literature along with past and current monitoring data, assesses the presence of chlorate in drinking water and the potential impact of its regulation. The article gives specific attention to the variety of threshold concentrations that appear most often in the literature—210, 700, and 840 μg/L—and evaluates the effect a regulatory requirement at each of these values would have on utilities. The research indicated that potential regulatory thresholds &gt; 700 μg/L would be violated by only &lt; 10% of utilities. The effects of regional conditions and type of disinfection used depend greatly on the adopted thresholds. Utilities in the southern region of the United States and those using chloramination are most at risk if a low maximum contaminant level is adopted.

Reduced-Rank Spatio-Temporal Modeling of Air Pollution Concentrations in the Multi-Ethnic Study of Atherosclerosis and Air Pollution.

There is growing evidence in the epidemiologic literature of the relationship between air pollution and adverse health outcomes. Prediction of individual air pollution exposure in the Environmental Protection Agency (EPA) funded Multi-Ethnic Study of Atheroscelerosis and Air Pollution (MESA Air) study relies on a flexible spatio-temporal prediction model that integrates land-use regression with kriging to account for spatial dependence in pollutant concentrations. Temporal variability is captured using temporal trends estimated via modified singular value decomposition and temporally varying spatial residuals. This model utilizes monitoring data from existing regulatory networks and supplementary MESA Air monitoring data to predict concentrations for individual cohort members. In general, spatio-temporal models are limited in their efficacy for large data sets due to computational intractability. We develop reduced-rank versions of the MESA Air spatio-temporal model. To do so, we apply low-rank kriging to account for spatial variation in the mean process and discuss the limitations of this approach. As an alternative, we represent spatial variation using thin plate regression splines. We compare the performance of the outlined models using EPA and MESA Air monitoring data for predicting concentrations of oxides of nitrogen (NO x )-a pollutant of primary interest in MESA Air-in the Los Angeles metropolitan area via cross-validated R2. Our findings suggest that use of reduced-rank models can improve computational efficiency in certain cases. Low-rank kriging and thin plate regression splines were competitive across the formulations considered, although TPRS appeared to be more robust in some settings.

Estimating daily nitrogen dioxide level: application of a longitudinal model with spatially-correlated random effects

Typically available nitrogen dioxide (\({\hbox {NO}}_{2}\)) measurements are either spatially representative but temporally sparse or temporally representative but spatially sparse. This article illustrates an approach for making use of the strengths of these two types of data to obtain pollution estimates that are representative both spatially and temporally. The model uses three data sources: a Connecticut epidemiological study with average monthly \({\hbox {NO}}_{2}\) measurements at 651 sites in each season for 1 year; hourly measurements at four Environmental Protection Agency (EPA) sites; and, geospatial data on traffic volume, population density, elevation, and land use. We start by establishing a relationship between observations from the epidemiological study and those at the EPA sites at the monthly level. This relationship is assumed to hold for daily \({\hbox {NO}}_{2}\) levels, enabling us to estimate average daily \({\hbox {NO}}_{2}\) levels using EPA data. Validation study confirmed that this assumption is sound and that the method performs best among three alternatives: a linear model, a mixed-effects model, and the proposed mix-effects model that takes into account spatial correlation. The model can provide predictions of daily \({\hbox {NO}}_{2}\) levels at both study sites and random sites that have no measurements at all. Application of this approach provides pollution estimates that make it possible to study short term relationships between \({\hbox {NO}}_{2}\) exposure and respiratory symptoms such as asthma severity. The approach also offers significant cost reduction for future studies of higher temporal/spatial resolution of air pollution levels by making effective use of readily available EPA monitoring site measurements and observations at spatially dispersed sites from epidemiological studies. The model is implemented under the Bayes framework with a Gibbs sampler.

PM2.5 Exposure and Birth Outcomes

Air pollution may be related to adverse birth outcomes. Exposure information from land-based monitoring stations often suffers from limited spatial coverage. Satellite data offer an alternative data source for exposure assessment. We used birth certificate data for births in Connecticut and Massachusetts, United States (2000-2006). Gestational exposure to PM2.5 was estimated from US Environmental Protection Agency monitoring data and from satellite data. Satellite data were processed and modeled by using two methods-denoted satellite (1) and satellite (2)-before exposure assessment. Regression models related PM2.5 exposure to birth outcomes while controlling for several confounders. Birth outcomes were mean birth weight at term birth, low birth weight at term (<2500 g), small for gestational age (SGA, <10th percentile for gestational age and sex), and preterm birth (<37 weeks). Overall, the exposure assessment method modified the magnitude of the effect estimates of PM2.5 on birth outcomes. Change in birth weight per interquartile range (2.41 μg/m) increase in PM2.5 was -6 g (95% confidence interval = -8 to -5), -16 g (-21 to -11), and -19 g (-23 to -15), using the monitor, satellite (1), and satellite (2) methods, respectively. Adjusted odds ratios, based on the same three exposure methods, for term low birth weight were 1.01 (0.98-1.04), 1.06 (0.97-1.16), and 1.08 (1.01-1.16); for SGA, 1.03 (1.01-1.04), 1.06 (1.03-1.10), and 1.08 (1.04-1.11); and for preterm birth, 1.00 (0.99-1.02), 0.98 (0.94-1.03), and 0.99 (0.95-1.03). Under exposure assessment methods, we found associations between PM2.5 exposure and adverse birth outcomes particularly for birth weight among term births and for SGA. These results add to the growing concerns that air pollution adversely affects infant health and suggest that analysis of health consequences based on satellite-based exposure assessment can provide additional useful information.

Estimating daily nitrogen dioxide level: Exploring traffic effects

Data used to assess acute health effects from air pollution typically have good temporal but poor spatial resolution or the opposite. A modified longitudinal model was developed that sought to improve resolution in both domains by bringing together data from three sources to estimate daily levels of nitrogen dioxide (NO2) at a geographic location. Monthly NO2 measurements at 316 sites were made available by the Study of Traffic, Air quality and Respiratory health (STAR). Four US Environmental Protection Agency monitoring stations have hourly measurements of NO2. Finally, the Connecticut Department of Transportation provides data on traffic density on major roadways, a primary contributor to NO2 pollution. Inclusion of a traffic variable improved performance of the model, and it provides a method for estimating exposure at points that do not have direct measurements of the outcome. This approach can be used to estimate daily variation in levels of NO2 over a region.