Regulatory Monitoring Sites Research Articles

Changes in ambient air pollutants in New York State from 2005 to 2019: Effects of policy implementations and economic and technological changes

Over the past 20 years, a number of regulatory efforts have been applied to improve air quality in the United States and specifically in New York State. These measures generally focused on mobile emissions through emissions controls and improved fuel quality, and controls on electricity generation to reduce emissions from older, uncontrolled electricity generation units (EGUs). In addition, economic drivers such as the major recession in 2007–2009 and the change in the relative costs of natural gas and coal also drove changes in the mixture of EGU technologies. To assess the effects of these changes and to define the baseline for future changes as the economy further decarbonizes through renewable electricity generation and electric vehicles, the concentrations of all pollutants measured at all regulatory monitoring sites in New York State were assessed for their trends. Trends were examined using seasonal-trend decomposition with local regression smoothing (STL), Mann-Kendall trend analysis with the Theil-Sen nonparametric slope estimation, and piecewise regression analysis to identify breakpoints in the slopes of the time series data. The concentrations of primary gaseous pollutants, CO, NO2, and SO2 have decreased substantially in step with the declining emissions. PM2.5 has substantially declined largely due to the reductions in particulate sulfate. However, in recent years, the rate of decline has diminished due to relatively constant or increasing particulate nitrate and secondary organic aerosol. O3 has also generally increased at the urban sites likely as a result of reduced NOx emissions, while it declined or remained constant at the rural sites. Thus, the promulgated regulations assisted by the economic drivers have improved air quality, but additional actions will be needed to further reduce urban O3 and PM2.5.

Human Serum Albumin Cys34 Adducts in Newborn Dried Blood Spots: Associations With Air Pollution Exposure During Pregnancy.

Background: Increasing evidence suggests that exposure to air pollution during pregnancy is associated with adverse pregnancy outcomes. However, biomarkers associated with air pollution exposure are widely lacking and often transient. In addition, ascertaining biospecimens during pregnacy to assess the prenatal environment remains largely infeasible.Objectives: To address these challenges, we investigated relationships between air pollution exposure during pregnancy and human serum albumin Cys34 (HSA-Cys34) adducts in newborn dried blood spots (DBS) samples, which captures an integration of perinatal exposures to small reactive molecules in circulating blood.Methods: Newborn DBS were obtained from a state archive for a cohort of 120 children born at one Kaiser Permanente Southern California (KPSC) hospitals in 2007. These children were selected to maximize the range of residential air pollution exposure during the entire pregnancy to PM2.5, PM10, NO2, O3, based on monthly estimates interpolated from regulatory monitoring sites. HSA-Cys34 adducts were selected based on previously reported relationships with air pollution exposure and oxidative stress.Results: Six adducts measured in newborn DBS samples were associated with air pollution exposures during pregnancy; these included direct oxidation products, adducts formed with small thiol compounds, and adducts formed with reactive aldehydes. Two general trends were identified: Exposure to air pollution late in pregnancy (i.e., in the last 30 days) was associated with increased oxidative stress, and exposure to air pollution earlier in pregnancy (i.e., not in the last 30 days) was associated with decreased oxidative stress around the time of birth.Discussion: Air pollution exposure occurring during pregnancy can alter biology and leave measurable impacts on the developing infant captured in the newborn DBS adductome, which represents a promising tool for investigating adverse birth outcomes in population-based studies.

Application of nonlinear land use regression models for ambient air pollutants and air quality index

Air pollution is a major global environmental problem that affects health. In view of this, it is important to improve the prediction method of air pollutant concentrations to obtain accurate exposure estimates. Currently, land use regression (LUR) models are widely used to predict the fine-scale spatial variation of air pollutants. However, most of previous studies used linear regression methods such as generalized linear models (GLM) with less applicability to fit LUR models. Considering the potential nonlinear relationship between predictor variables, this study adopted generalized additive models (GAM) to derive LUR models of air pollutants (including PM2.5, PM10, CO, NO2, SO2, and O3) and air quality index (AQI) in Beijing with annual resolution. These models were based on routine monitoring data from 35 national regulatory monitoring sites and combined with a set of predictor variables such as land-use type, traffic and industrial emissions, population density, and meteorological factors. Results indicated that compared with traditional methods, the GAM approach significantly improved the performance of LUR models with explanatory power adjusted R2 levels ranging from 70 to 90%, and the cross-validation analysis also showed high prediction accuracy of the GAM approach. Besides, this approach emphasized the importance of meteorology in predicting air pollutant concentrations and AQI values. Generally, this study provides a feasible way to determine exposure assessment in heavily polluted cities and future support for long-term environmental epidemiological studies.

Associations between Vehicle Miles Traveled and Ambient Air Pollution at Regulatory Monitoring Sites: A Natural Experiment using the COVID-19 Shutdown

BACKGROUND AND AIM: Transportation planners calculate vehicle miles traveled (VMT) to forecast roadway usage and inform policy decisions. However, the relationship between VMT and ambient air pollution is not fully understood. In this analysis, we use proprietary smartphone sensing data with VMT calculations to examine the impact of VMT on ambient air pollution during the decline in vehicle usage from the COVID-19 shutdown. We aim to determine how VMT change contributes to change in NO2, O3, and PM2.5 concentrations and at what spatial scale these changes occur. METHODS: We obtained access to a daily national database with VMT calculations for January to July 2020, which corresponds to various stay at home orders that limited travel needs. We linked VMT measures at different buffer distances (100, 300, 500, 1000, 2000 and 3000m) to ambient air pollution monitor from the Environmental Protection Agency (EPA). We implement multi-level linear regression models to examine how NO2, O3, and PM2.5 changed with VMT reductions in different distance buffers around monitoring locations. RESULTS:Preliminary results show consistent associations between VMT and O3 and NO2 at varying buffer distances, but these associations are less consistent for PM2.5. For O3, we observe a 0.18 ppb (95% CI: -0.21, -0.15) reduction per 1,000 VMT at the 100m buffer but this reduction increases to 0.47 ppb (95% CI: 0.52, 0.43) at the 3,000m buffer. For NO2, we observe a 2.21 ppb (95% CI: 2.13, 2.30) increase per 1,000 VMT at the 100m buffer and results are similar out to the 3,000m buffer. When we split out VMT by road type, we find that models for VMT from expressways yield smaller magnitudes of association than the models for VMT from non-expressways. CONCLUSIONS:These results will be used to inform future analyses on how congestion reduction projects impact local air pollution and population health. KEYWORDS: Traffic-related, Policy research, Exposure assessment-air pollution, Oxides of nitrogen, Particle components, Ozone

Application and Evaluation of a Low-cost PM Sensor and Data Fusion with CMAQ Simulations to Quantify the Impacts of Prescribed Burning on Air Quality in Southwestern Georgia, USA

ABSTRACT Prescribed burning (PB) is a prominent source of PM2.5 in the southeastern US and exposure to PB smoke is a health risk. As demand for burning increases and stricter controls are implemented for other anthropogenic sources, PB emissions tend to be responsible for an increasing fraction of PM2.5 concentrations. Here, to quantify the effect of PB on air quality, low-cost sensors are used to measure PM2.5 concentrations in Southwestern Georgia. The feasibility of using low-cost sensors as a supplemental measurement tool is evaluated by comparing them with reference instruments. A chemical transport model, CMAQ, is also used to simulate the contribution of PB to PM2.5 concentrations. Simulated PM2.5 concentrations are compared to observations from both low-cost sensors and reference monitors. Finally, a data fusion method is applied to generate hourly spatiotemporal exposure fields by fusing PM2.5 concentrations from the CMAQ model and all observations. The results show that the severe impact of PB on local air quality and public health may be missed due to the dearth of regulatory monitoring sites. In Southwestern Georgia PM2.5 concentrations are highly non-homogeneous and the spatial variation is not captured even with a 4-km horizontal resolution in model simulations. Low-cost PM sensors can improve the detection of PB impacts and provide useful spatial and temporal information for integration with air quality models. R2 of regression with observations increases from an average of 0.09 to 0.40 when data fusion is applied to model simulation withholding the observations at the evaluation site. Adding observations from low-cost sensors reduces the underestimation of nighttime PM2.5 concentrations and reproduces the peaks that are missed by the simulations. In the future, observations from a dense network of low-cost sensors could be fused with the model simulated PM2.5 fields to provide better estimates of hourly exposures to smoke from PB. Prescribed burning emissions are a major cause of elevated PM2.5 concentrations, posing a risk to human health. However, their impact cannot be quantified properly due to a dearth of regulatory monitoring sites in certain regions of the United States such as Southwestern Georgia. Low-cost PM sensors can be used as a supplemental measurement tool and provide useful spatial and temporal information for integration with air quality model simulations. In the future, data from a dense network of low-cost sensors could be fused with model simulated PM2.5 fields to provide improved estimates of hourly exposures to smoke from prescribed burning. Impliation statement

Impact of the exclusive median bus lane system on air pollution concentrations in Seoul, Korea

Since many previous studies reported the health effect of air pollution and indicated traffic as a major pollution source, significant policy efforts have been made to control traffic to reduce air pollution. However, there have been few studies that evaluated such policy implementation. In Seoul, Korea, the exclusive median bus lane system was implemented in 2004, and the metropolitan government applied air pollution reduction policies such as conversion of diesel buses to compressed natural gas buses and installation of emission control devices. This paper aimed to investigate the impact of the exclusive median bus lane system on air pollution reduction. Using hourly concentrations of particulate matter (PM10) and nitrogen dioxide (NO2) measured at 131 regulatory monitoring sites in Seoul and Gyeonggi-do for 2001-2014, we calculated annual and daily average concentrations at each site. We assessed the impact of the policy using differences-in-differences analysis by annual and daily average models after adjusting for geographic and/or meteorological variables. This method divides population into treatment and control groups with and without policy application, and compares the difference between the two time periods before and after the policy implementation in the treatment group with the difference in the control group. We classified all monitoring sites into treatment and control groups using two definitions: 1) Seoul vs. Gyeonggi; 2) within vs. outside 300 meters from the median bus lane. Pre- and post-policy periods were defined as 2001-2005 and 2006-2014, and 2004 and 2014 in the annual and daily models, respectively. The decrease in PM10 concentrations between the two periods across monitoring sites in the treatment group was larger by 1.73-5.88 μg/m3 than in the control group. NO2 also showed the decrease without statistical significance. Our findings suggest that an efficient public transport policy combined with pollution abatement policies can contribute to reduction in air pollution.

Impacts of Aviation Emissions on Near-Airport Residential Air Quality.

Impacts of aviation emissions on air quality in and around residences near airports remain underexamined. We measured gases (CO, CO2, NO, and NO2) and particles (black carbon, particle-bound aromatic hydrocarbons, fine particulate matter (PM2.5), and ultrafine particles (reported using particle number concentrations (PNC) as a proxy)) continuously for 1 month at a residence near the Logan International Airport, Boston. The residence was located under a flight trajectory of the most utilized runway configuration. We found that when the residence was downwind of the airport, the concentrations of all gaseous and particulate pollutants (except PM2.5) were 1.1- to 4.8-fold higher than when the residence was not downwind of the airport. Controlling for runway usage and meteorology, the impacts were highest during overhead landing operations: average PNC was 7.5-fold higher from overhead landings versus takeoffs on the closest runway. Infiltration of aviation-origin emissions resulted in indoor PNC that were comparable to ambient concentrations measured locally on roadways and near highways. In addition, ambient NO2 concentrations at the residence exceeded those measured at regulatory monitoring sites in the area including near-road monitors. Our results highlight the need for further characterization of outdoor and indoor impacts of aviation emissions at the neighborhood scale to more accurately estimate residential exposures.

Improving Correlations between Land Use and Air Pollutant Concentrations Using Wavelet Analysis: Insights from a Low-cost Sensor Network

ABSTRACT City-wide air pollution assessments have typically relied on a small number of widely separated regulatory monitoring sites or land use regression (LUR) models built using time-integrated samples to assess annual average population-scale exposure. However, air pollutant concentrations often exhibit significant spatial and temporal variability depending on local sources and features of the built environment. In 2016, the Center for Air, Climate, and Energy Solutions (CACES) Air Quality Observatory was launched at Carnegie Mellon University to better understand urban spatial and temporal pollution gradients on the 8 h) above the regional background. Compared to the non-decomposed total pollutant signal, the short-lived or persistent enhancement pollutant signals, which should come from local sources, were better correlated with covariates used in LUR model construction. For example, Pearson r between total vehicle counts in a 100 m buffer and NO2 increased from 0.57 using the total pollutant signal to 0.83 using the persistent enhancement only. The findings from this study support building more accurate and higher time resolution (e.g., daily, hourly) LURs using low-cost sensors.

Evaluating and improving the reliability of gas-phase sensor system calibrations across new locations for ambient measurements and personal exposure monitoring

Abstract. Advances in ambient environmental monitoring technologies are enabling concerned communities and citizens to collect data to better understand their local environment and potential exposures. These mobile, low-cost tools make it possible to collect data with increased temporal and spatial resolution, providing data on a large scale with unprecedented levels of detail. This type of data has the potential to empower people to make personal decisions about their exposure and support the development of local strategies for reducing pollution and improving health outcomes. However, calibration of these low-cost instruments has been a challenge. Often, a sensor package is calibrated via field calibration. This involves colocating the sensor package with a high-quality reference instrument for an extended period and then applying machine learning or other model fitting technique such as multiple linear regression to develop a calibration model for converting raw sensor signals to pollutant concentrations. Although this method helps to correct for the effects of ambient conditions (e.g., temperature) and cross sensitivities with nontarget pollutants, there is a growing body of evidence that calibration models can overfit to a given location or set of environmental conditions on account of the incidental correlation between pollutant levels and environmental conditions, including diurnal cycles. As a result, a sensor package trained at a field site may provide less reliable data when moved, or transferred, to a different location. This is a potential concern for applications seeking to perform monitoring away from regulatory monitoring sites, such as personal mobile monitoring or high-resolution monitoring of a neighborhood. We performed experiments confirming that transferability is indeed a problem and show that it can be improved by collecting data from multiple regulatory sites and building a calibration model that leverages data from a more diverse data set. We deployed three sensor packages to each of three sites with reference monitors (nine packages total) and then rotated the sensor packages through the sites over time. Two sites were in San Diego, CA, with a third outside of Bakersfield, CA, offering varying environmental conditions, general air quality composition, and pollutant concentrations. When compared to prior single-site calibration, the multisite approach exhibits better model transferability for a range of modeling approaches. Our experiments also reveal that random forest is especially prone to overfitting and confirm prior results that transfer is a significant source of both bias and standard error. Linear regression, on the other hand, although it exhibits relatively high error, does not degrade much in transfer. Bias dominated in our experiments, suggesting that transferability might be easily increased by detecting and correcting for bias. Also, given that many monitoring applications involve the deployment of many sensor packages based on the same sensing technology, there is an opportunity to leverage the availability of multiple sensors at multiple sites during calibration to lower the cost of training and better tolerate transfer. We contribute a new neural network architecture model termed split-NN that splits the model into two stages, in which the first stage corrects for sensor-to-sensor variation and the second stage uses the combined data of all the sensors to build a model for a single sensor package. The split-NN modeling approach outperforms multiple linear regression, traditional two- and four-layer neural networks, and random forest models. Depending on the training configuration, compared to random forest the split-NN method reduced error 0 %–11 % for NO2 and 6 %–13 % for O3.

Examining spatiotemporal variability of urban particulate matter and application of high-time resolution data from a network of low-cost air pollution sensors

Traditional air monitoring approaches using regulatory monitors have historically been used to assess regional-scale trends in air pollutants across large geographical areas. Recent advances in air pollution sensor technologies could provide additional information about nearby sources, support the siting of regulatory monitoring stations, and improve our knowledge of finer-scale spatiotemporal variation of ambient air pollutants and their associated health effects. Sensors are now being developed that are much smaller and lower cost than traditional ambient air monitoring systems and are capable of being deployed as a network to provide greater coverage of a given area. The CitySpace project conducted by the US EPA and the Shelby County Health Department included the deployment of a network of 17 sensor pods using Alphasense OPC-N2 particulate matter (PM) sensors integrated with meteorological sensors in Memphis, TN for six months. Sensor pods were collocated with a federal equivalent method (FEM) tapered element oscillating microbalance (TEOM) monitor both before and after the primary study period. Six of the sensor pods were found to meet the data quality objective (DQO) of coefficient of determination (R2) greater than 0.5 when collocated with the TEOM. Seven pods were decommissioned before the end of the study due to mechanical failure. The six pods meeting the DQO were used to examine the spatiotemporal variability of fine PM (PM2.5) across the Memphis area. One site was found to have higher relative PM2.5 concentrations when compared to the other sites in the network. The 1-min data from this sensor pod were evaluated to quantify the regional urban background and local-scale contributions to PM2.5 at that monitoring location. This method found that approximately 20% of the PM2.5 was attributed to local sources at this location, compared to 9% at a local regulatory monitoring site. Additionally, the 1-min data were combined with 1-min wind speed and wind direction data to examine potential sources in the area using the nonparametric trajectory analysis (NTA) technique. This method geographically identified local source areas that contributed to the measured concentrations at the high reading sensor location throughout the course of the study.

Prediction of local particle pollution level based on artificial neural network

Citizens eager to know the local pollution level to prevent from air pollution. The real-time measurement for everywhere is a very expensive way, a statistical model based on artificial neural network is applied in this research. This model can estimate particle pollution level with some influencing factors, including background pollution level, weather conditions, urban morphology and local pollution sources. The monitoring from regulatory monitoring sites is considered as the background level. The field measurements of 20 locations are conducted to feed the output layer of ANN model. The average relative error of prediction compared with measurement is 9.24% for PM10 and 18.90% for PM2.5.

Performance of NO, NO<sub>2</sub> low cost sensors and three calibration approaches within a real world application

Abstract. Low cost sensors for measuring atmospheric pollutants are experiencing an increase in popularity worldwide among practitioners, academia and environmental agencies, and a large amount of data by these devices are being delivered to the public. Notwithstanding their behaviour, performance and reliability are not yet fully investigated and understood. In the present study we investigate the medium term performance of a set of NO and NO2 electrochemical sensors in Switzerland using three different regression algorithms within a field calibration approach. In order to mimic a realistic application of these devices, the sensors were initially co-located at a rural regulatory monitoring site for a 4-month calibration period, and subsequently deployed for 4 months at two distant regulatory urban sites in traffic and urban background conditions, where the performance of the calibration algorithms was explored. The applied algorithms were Multivariate Linear Regression, Support Vector Regression and Random Forest; these were tested, along with the sensors, in terms of generalisability, selectivity, drift, uncertainty, bias, noise and suitability for spatial mapping intra-urban pollution gradients with hourly resolution. Results from the deployment at the urban sites show a better performance of the non-linear algorithms (Support Vector Regression and Random Forest) achieving RMSE < 5 ppb, R2 between 0.74 and 0.95 and MAE between 2 and 4 ppb. The combined use of both NO and NO2 sensor output in the estimate of each pollutant showed some contribution by NO sensor to NO2 estimate and vice-versa. All algorithms exhibited a drift ranging between 5 and 10 ppb for Random Forest and 15 ppb for Multivariate Linear Regression at the end of the deployment. The lowest concentration correctly estimated, with a 25 % relative expanded uncertainty, resulted in ca. 15–20 ppb and was provided by the non-linear algorithms. As an assessment for the suitability of the tested sensors for a targeted application, the probability of resolving hourly concentration difference in cities was investigated. It was found that NO concentration differences of 5–10 ppb (8–10 for NO2) can reliably be detected (90 % confidence), depending on the air pollution level. The findings of this study, although derived from a specific sensor type and sensor model, are based on a flexible methodology and have extensive potential for exploring the performance of other low cost sensors, that are different in their target pollutant and sensing technology.

Association between ambient air pollution and pregnancy rate in women who underwent IVF.

Are the concentrations of five criteria air pollutants associated with probabilities of biochemical pregnancy loss and intrauterine pregnancy in women? Increased concentrations of ambient particulate matter (PM10), nitrogen dioxide (NO2), carbon monoxide (CO) during controlled ovarian stimulation (COS) and after embryo transfer were associated with a decreased probability of intrauterine pregnancy. Exposure to high ambient air pollution was suggested to be associated with low fertility and high early pregnancy loss in women. Using a retrospective cohort study design, we analysed 6621 cycles of 4581 patients who underwent one or more fresh IVF cycles at a fertility centre from January 2006 to December 2014, and lived in Seoul at the time of IVF treatment. To estimate patients' individual exposure to air pollution, we computed averages of hourly concentrations of five air pollutants including PM10, NO2, CO, sulphur dioxide (SO2) and ozone (O3) measured at 40 regulatory monitoring sites in Seoul for each of the four exposure periods: period 1 (start of COS to oocyte retrieval), period 2 (oocyte retrieval to embryo transfer), period 3 (embryo transfer to hCG test), and period 4 (start of COS to hCG test). Hazard ratios (HRs) from the time-varying Cox-proportional hazards model were used to estimate probabilities of biochemical pregnancy loss and intrauterine pregnancy for an interquartile range (IQR) increase in each air pollutant concentration during each period, after adjusting for individual characteristics. We tested the robustness of the result using generalised linear mixed model, accounting for within-woman correlation. Mean age of the women was 35 years. Average BMI was 20.9 kg/m2 and the study population underwent 1.4 IVF cycles on average. Cumulative pregnancy rate in multiple IVF cycles was 51.3% per person. Survival analysis showed that air pollution during periods 1 and 3 was generally associated with IVF outcomes. Increased NO2 (adjusted HR = 0.93, 95% CI: 0.87, 0.99) and CO (0.94, 95% CI: 0.89, 1.00) during period 1 were associated with decreased probability of intrauterine pregnancy. PM10 (0.92, 95% CI: 0.85, 0.99), NO2 (0.93, 95% CI = 0.86, 1.00) and CO (0.93, 95% CI: 0.87, 1.00) levels during period 3 were also inversely associated with intrauterine pregnancy. Both PM10 (1.17, 95% CI: 1.04 1.33) and NO2 (1.18, 95% CI: 1.03, 1.34) during period 3 showed positive associations with biochemical pregnancy loss. The district-specific ambient air pollution treated as an individual exposure may not represent the actual level of each woman's exposure to air pollution. Smoking, working status, parity or gravidity of women, and semen analysis data were not included in the analysis. This study provided evidence of an association between increased ambient concentrations of PM10, NO2 and CO and reduced probabilities for achieving intrauterine pregnancy using multiple IVF cycle data. Specifically, our results indicated that lower intrauterine pregnancy rates in IVF cycles may be linked to ambient air pollution during COS and the post-transfer period. This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2013 R1A6A3A04059017, 2016 R1D1A1B03933410 and 2018 R1A2B6004608) and the National Cancer Center of Korea (NCC-1810220-01). The authors report no conflicts of interest. N/A.

A panel study of airborne particulate matter concentration and impaired cardiopulmonary function in young adults by two different exposure measurement

This study sought to clarify the correlation of individual exposure measurements and PM2.5 measurements collected at regulatory monitoring sites in short-term panel study settings. To achieve this goal, 30 young, healthy adult participants were assigned to three groups with 4 samplers in each group to collect individual exposures during four weekends in March 2016. Participants also completed cardiopulmonary function tests during the same periods. For comparison, ambient air pollution data were obtained from the Air Pollution Surveillance Network in Guangzhou, China. The 8-h ambient pollutant averages and group sampler concentrations were used as separate indicators of air pollution exposure. Results showed that the 8-h mean concentration of personal PM2.5 exposure was 65.09 ± 22.18 μg/m3, which was 24.34 μg/m3 statistically higher than the ambient concentrations over the same period (p < 0.05). However, these concentrations were strongly correlated (Spearman's r = 0.937, p < 0.01). Separate mixed-effect models were fit for ambient and personal exposures to estimate their associations with cardiopulmonary outcomes. Higher PM2.5 and PM10 exposures were related to lower lung function of maximal mid-expiratory flow (MMEF). A 10 μg/m3 higher PM was associated with 0.11 L/S to 0.52 L/S lower MMEF. No effects on cardiovascular function were found. In conclusion, personal PM2.5 exposure might be higher than ambient concentrations. Young, healthy adults in urban areas may experience reduced lung function (lower MMEF), even after just 8 h of exposure to PM2.5 and PM10.

Air Pollution Monitoring Design for Epidemiological Application in a Densely Populated City.

Introduction: Many studies have reported the association between air pollution and human health based on regulatory air pollution monitoring data. However, because regulatory monitoring networks were not designed for epidemiological studies, the collected data may not provide sufficient spatial contrasts for assessing such associations. Our goal was to develop a monitoring design supplementary to the regulatory monitoring network in Seoul, Korea. This design focused on the selection of 20 new monitoring sites to represent the variability in PM2.5 across people’s residences for cohort studies. Methods: We obtained hourly measurements of PM2.5 at 37 regulatory monitoring sites in 2010 in Seoul, and computed the annual average at each site. We also computed 313 geographic variables representing various pollution sources at the regulatory monitoring sites, 31,097 children’s homes from the Atopy Free School survey, and 412 community service centers in Seoul. These three types of locations represented current, subject, and candidate locations. Using the regulatory monitoring data, we performed forward variable selection and chose five variables most related to PM2.5. Then, k-means clustering was applied to categorize all locations into several groups representing a diversity in the spatial variability of the five selected variables. Finally, we computed the proportion of current to subject location in each cluster, and randomly selected new monitoring sites from candidate sites in the cluster with the minimum proportion until 20 sites were selected. Results: The five selected geographic variables were related to traffic or urbanicity with a cross-validated R2 value of 0.69. Clustering analysis categorized all locations into nine clusters. Finally, one to eight new monitoring sites were selected from five clusters. Discussion: The proposed monitoring design will help future studies determine the locations of new monitoring sites representing spatial variability across residences for epidemiological analyses.

National-scale exposure prediction for long-term concentrations of particulate matter and nitrogen dioxide in South Korea

The limited spatial coverage of the air pollution data available from regulatory air quality monitoring networks hampers national-scale epidemiological studies of air pollution. The present study aimed to develop a national-scale exposure prediction model for estimating annual average concentrations of PM10 and NO2 at residences in South Korea using regulatory monitoring data for 2010. Using hourly measurements of PM10 and NO2 at 277 regulatory monitoring sites, we calculated the annual average concentrations at each site. We also computed 322 geographic variables in order to represent plausible local and regional pollution sources. Using these data, we developed universal kriging models, including three summary predictors estimated by partial least squares (PLS). The model performance was evaluated with fivefold cross-validation. In sensitivity analyses, we compared our approach with two alternative approaches, which added regional interactions and replaced the PLS predictors with up to ten selected variables. Finally, we predicted the annual average concentrations of PM10 and NO2 at 83,463 centroids of residential census output areas in South Korea to investigate the population exposure to these pollutants and to compare the exposure levels between monitored and unmonitored areas. The means of the annual average concentrations of PM10 and NO2 for 2010, across regulatory monitoring sites in South Korea, were 51.63 μg/m3 (SD = 8.58) and 25.64 ppb (11.05), respectively. The universal kriging exposure prediction models yielded cross-validated R2s of 0.45 and 0.82 for PM10 and NO2, respectively. Compared to our model, the two alternative approaches gave consistent or worse performances. Population exposure levels in unmonitored areas were lower than in monitored areas. This is the first study that focused on developing a national-scale point wise exposure prediction approach in South Korea, which will allow national exposure assessments and epidemiological research to answer policy-related questions and to draw comparisons among different countries.

Air pollution monitoring design for epidemiological application in a metropolitan area

Introduction: The association between long-term exposures to air pollution and health in cohort studies was assessed by relying on spatial contrasts in exposures and outcomes. However, existing regulatory monitoring networks may not represent sufficient spatial contrasts of air pollution exposures across people’s residences. We aimed to develop an air pollution monitoring design to supplement regulatory monitoring networks in Seoul, Korea. This design specifically focused on the selection of 20 new monitoring sites for representing spatial variability of PM2.5 and NO2 across Seoul residents for the application to epidemiological studies. Methods: We computed annual averages of PM2.5 and NO2 at 37 regulatory monitoring sites during 2010 in Seoul, Korea. About 300 geographic variables representing various pollution sources were computed at regulatory monitoring sites, 38,201 participants’ homes of the Atopy Free School survey, and 422 community service centers in Seoul. The design included three procedures. We selected five geographic variables most related to each of PM2.5 and NO2 from regulatory monitoring data. Then, k-means clustering was used for categorizing all locations into groups representing contrasts in five selected variables. Finally, using community service centers as candidate sites, we randomly selected ten new sites from clusters with low ratios of regulatory monitoring sites to participant homes. Results: Five selected variables generally included variables relevant to traffic and urbanicity with cross-validated R2s of 0.69 and 0.58 for PM2.5 and NO2, respectively. K-means clustering gave eight and night clusters for PM2.5 and NO2, respectively, representing different geographic characteristics. One to three new monitoring sites were selected from four to six clusters for PM2.5 and NO2. Conclusions: Our monitoring design provides a useful tool applicable to metropolitan areas for assessing air pollution levels of people and resulting health effects.

Computation of geographic variables for air pollution prediction models in South Korea.

Recent cohort studies have relied on exposure prediction models to estimate individuallevel air pollution concentrations because individual air pollution measurements are not available for cohort locations. For such prediction models, geographic variables related to pollution sources are important inputs. We demonstrated the computation process of geographic variables mostly recorded in 2010 at regulatory air pollution monitoring sites in South Korea. On the basis of previous studies, we finalized a list of 313 geographic variables related to air pollution sources in eight categories including traffic, demographic characteristics, land use, transportation facilities, physical geography, emissions, vegetation, and altitude. We then obtained data from different sources such as the Statistics Geographic Information Service and Korean Transport Database. After integrating all available data to a single database by matching coordinate systems and converting non-spatial data to spatial data, we computed geographic variables at 294 regulatory monitoring sites in South Korea. The data integration and variable computation were performed by using ArcGIS version 10.2 (ESRI Inc., Redlands, CA, USA). For traffic, we computed the distances to the nearest roads and the sums of road lengths within different sizes of circular buffers. In addition, we calculated the numbers of residents, households, housing buildings, companies, and employees within the buffers. The percentages of areas for different types of land use compared to total areas were calculated within the buffers. For transportation facilities and physical geography, we computed the distances to the closest public transportation depots and the boundary lines. The vegetation index and altitude were estimated at a given location by using satellite data. The summary statistics of geographic variables in Seoul across monitoring sites showed different patterns between urban background and urban roadside sites. This study provided practical knowledge on the computation process of geographic variables in South Korea, which will improve air pollution prediction models and contribute to subsequent health analyses.

Exposure Prediction Approaches For Estimating Individual-Level Long-Term Particulate Matter Concentrations In Seoul, Korea

Introduction: Cohort studies have developed exposure prediction approaches to estimate individual-level long-term exposures to particulate matter (PM) air pollution. However, these approaches were largely based on North American and European cities. This study developed spatial prediction approaches for individual-level long-term PM10 and PM22.5 concentrations in Seoul, Korea, with the goal of application to cohort studies. Methods: We obtained hourly PM10 and PM2.5 concentrations measured at 36 regulatory monitoring sites in Seoul during 2010 from the Ministry of Environment and computed the annual average at each site. One or two monitoring sites are located in each of the 25 districts of Seoul (about 600 km2). We also computed 315 geographic variables related to potential pollution sources such as traffic, population, emission, vegetation index, geography, and altitude at each monitoring site. Given the PM and geographic data, we developed two prediction approaches: land use regression (LUR) and universal kriging (UK). LUR included three geographic predictors into a regression framework, whereas UK additionally incorporated a spatial correlation structure. Three predictors were estimated as a linear combination of 315 geographic variables by partial least squares. For the model evaluation, we performed 5-fold cross-validation and computed cross-validation statistics. Results: The means of annual average PM10 and PM2.5 concentrations for 2010 in Seoul were 51.27 (standard deviation = 5.02) and 27.22 (3.59) µg/m3, respectively. The cross-validated R2s in LUR were 0.80 and 0.75 for PM10 and PM2.5, respectively. UK produced consistent R2s reflecting little contribution of spatial correlation in addition to geographic predictors in Seoul. Discussion: This study showed good model performance of our spatial prediction approaches for PM10 and PM2.5 in Seoul, Korea and will allow us to assess associations with various health outcomes based on individual PM estimates.

Field calibrations of a low-cost aerosol sensor at a regulatory monitoring site in California

Abstract. Health effects attributed to ambient fine particulate matter (PM2.5) now rank it among the risk factors with the highest health burdens in the world, but existing monitoring infrastructure cannot adequately characterize spatial and temporal variability in urban PM2.5 concentrations, nor in human population exposures. The development and evaluation of more portable and affordable monitoring instruments based on low-cost sensors may offer a means to supplement and extend existing infrastructure, increasing the density and coverage of empirical measurements and thereby improving exposure science and control. Here, we report on field calibrations of a custom-built, battery-operated aerosol monitoring instrument we developed using low-cost, off-the-shelf optical aerosol sensors. We calibrated our instruments using 1 h and 24 h PM2.5 data from a class III US EPA Federal Equivalent Method (FEM) PM2.5 β-attenuation monitor in continuous operation at a regulatory monitoring site in Oakland, California. We observed negligible associations with ambient humidity and temperature; linear corrections were sufficient to explain 60% of the variance in 1 h reference PM2.5 data and 72% of the variance in 24 h data. Performance at 1 h integration times was comparable to commercially available optical instruments costing considerably more. These findings warrant further exploration of the circumstances under which this class of aerosol sensors may profitably be deployed to generate improved PM2.5 data sets.