Comparing land use regression and dispersion modelling to assess residential exposure to ambient air pollution for epidemiological studies

A comparison of PM2.5 exposure estimates from different estimation methods and their associations with cognitive testing and brain MRI outcomes

BACKGROUND

Reported associations between particulate matter with aerodynamic diameter ≤2.5 μm (PM2.5) and cognitive outcomes remain mixed. Differences in exposure estimation method may contribute to this heterogeneity.

OBJECTIVES

To assess agreement between PM2.5 exposure concentrations across 11 exposure estimation methods and to compare resulting associations between PM2.5 and cognitive or MRI outcomes.

METHODS

We used Visit 5 (2011-2013) cognitive testing and brain MRI data from the Atherosclerosis Risk in Communities (ARIC) Study. We derived address-linked average 2000-2007 PM2.5 exposure concentrations in areas immediately surrounding the four ARIC recruitment sites (Forsyth County, NC; Jackson, MS; suburbs of Minneapolis, MN; Washington County, MD) using 11 estimation methods. We assessed agreement between method-specific PM2.5 concentrations using descriptive statistics and plots, overall and by site. We used adjusted linear regression to estimate associations of method-specific PM2.5 exposure estimates with cognitive scores (n = 4678) and MRI outcomes (n = 1518) stratified by study site and combined site-specific estimates using meta-analyses to derive overall estimates. We explored the potential impact of unmeasured confounding by spatially patterned factors.

RESULTS

Exposure estimates from most methods had high agreement across sites, but low agreement within sites. Within-site exposure variation was limited for some methods. Consistently null findings for the PM2.5-cognitive outcome associations regardless of method precluded empirical conclusions about the potential impact of method on study findings in contexts where positive associations are observed. Not accounting for study site led to consistent, adverse associations, regardless of exposure estimation method, suggesting the potential for substantial bias due to residual confounding by spatially patterned factors.

DISCUSSION

PM2.5 estimation methods agreed across sites but not within sites. Choice of estimation method may impact findings when participants are concentrated in small geographic areas. Understanding unmeasured confounding by factors that are spatially patterned may be particularly important in studies of air pollution and cognitive or brain health.

High-Resolution Spatiotemporal Modeling for PM2.5 Major Components in the Pearl River Delta and Its Implications for Epidemiological Studies

Distinguishing the effects of different fine particulate matter components (PMCs) is crucial for mitigating their effects on human health. However, the sparse distribution of locations where PM is collected for component analysis makes it challenging to investigate the relevant health effects. This study aimed to investigate the agreement between data-fusion-enhanced exposure assessment and site monitoring data in estimating the effects of PMCs on gestational diabetes mellitus (GDM). We first improved the spatial resolution and accuracy of exposure assessment for five major PMCs (EC, OM, NO3-, NH4+, and SO42-) in the Pearl River Delta region by a data fusion model that combined inputs from multiple sources using a random forest model (10-fold cross-validation R2: 0.52 to 0.61; root mean square error: 0.55 to 2.26 μg/m3). Next, we compared the associations between exposures to PMCs during pregnancy and GDM in a hospital-based cohort of 1148 pregnant women in Heshan, China, using both site monitoring data and data-fusion model estimates. The comparative analysis showed that the data-fusion-based exposure generated stronger estimates of identifying statistical disparities. This study suggests that data-fusion-enhanced estimates can improve exposure assessment and potentially mitigate the misclassification of population exposure arising from the utilization of site monitoring data.

Modelling spatial & temporal variability of air pollution in an area of unconventional natural gas operations

Despite the growing unconventional natural gas production industry in northeastern British Columbia, Canada, few studies have explored the air quality implications on human health in nearby communities. Researchers who have worked with pregnant women in this area have found higher levels of volatile organic compounds (VOCs) in the indoor air of their homes associated with higher density and closer proximity to gas wells. To inform ongoing exposure assessments, this study develops land use regression (LUR) models to predict ambient air pollution at the homes of pregnant women by using natural gas production activities as predictor variables. Using the existing monitoring network, the models were developed for three temporal scales for 12 air pollutants. The models predicting monthly, bi-annual, and annual mean concentrations explained 23%-94%, 54%-94%, and 73%-91% of the variability in air pollutant concentrations, respectively. These models can be used to investigate associations between prenatal exposure to air pollutants associated with natural gas production and adverse health outcomes in northeastern British Columbia.

A machine learning-based ensemble model for estimating diurnal variations of nitrogen oxide concentrations in Taiwan

Air pollution is inextricable from human activity patterns. This is especially true for nitrogen oxide (NOx), a pollutant that exists naturally and also as a result of anthropogenic factors. Assessing exposure by considering diurnal variation is a challenge that has not been widely studied. Incorporating 27 years of data, we attempted to estimate diurnal variations in NOx across Taiwan. We developed a machine learning-based ensemble model that integrated hybrid kriging-LUR, machine-learning, and an ensemble learning approach. Hybrid kriging-LUR was performed to select the most influential predictors, and machine-learning algorithms were applied to improve model performance. The three best machine-learning algorithms were suited and reassessed to develop ensemble learning that was designed to improve model performance. Our ensemble model resulted in estimates of daytime, nighttime, and daily NOx with high explanatory powers (Adj-R2) of 0.93, 0.98, and 0.94, respectively. These explanatory powers increased from the initial model that used only hybrid kriging-LUR. Additionally, the results depicted the temporal variation of NOx, with concentrations higher during the daytime than the nighttime. Regarding spatial variation, the highest NOx concentrations were identified in northern and western Taiwan. Model evaluations confirmed the reliability of the models. This study could serve as a reference for regional planning supporting emission control for environmental and human health.

The methodology of quantitative risk assessment studies

Once an external factor has been deemed likely to influence human health and a dose response function is available, an assessment of its health impact or that of policies aimed at influencing this and possibly other factors in a specific population can be obtained through a quantitative risk assessment, or health impact assessment (HIA) study. The health impact is usually expressed as a number of disease cases or disability-adjusted life-years (DALYs) attributable to or expected from the exposure or policy. We review the methodology of quantitative risk assessment studies based on human data. The main steps of such studies include definition of counterfactual scenarios related to the exposure or policy, exposure(s) assessment, quantification of risks (usually relying on literature-based dose response functions), possibly economic assessment, followed by uncertainty analyses. We discuss issues and make recommendations relative to the accuracy and geographic scale at which factors are assessed, which can strongly influence the study results. If several factors are considered simultaneously, then correlation, mutual influences and possibly synergy between them should be taken into account. Gaps or issues in the methodology of quantitative risk assessment studies include 1) proposing a formal approach to the quantitative handling of the level of evidence regarding each exposure-health pair (essential to consider emerging factors); 2) contrasting risk assessment based on human dose-response functions with that relying on toxicological data; 3) clarification of terminology of health impact assessment and human-based risk assessment studies, which are actually very similar, and 4) other technical issues related to the simultaneous consideration of several factors, in particular when they are causally linked.

Comparison of PM2.5 Air Pollution Exposures and Health Effects Associations Using 11 Different Modeling Approaches in the Women's Health Initiative Memory Study (WHIMS)

BACKGROUND

Many approaches to quantifying air pollution exposures have been developed. However, the impact of choice of approach on air pollution estimates and health-effects associations remains unclear.

OBJECTIVES

Our objective is to compare particulate matter with aerodynamic diameter ≤ 2.5 μ m (PM 2.5 ) concentrations and resulting health effects associations using multiple estimation approaches previously used in epidemiologic analyses.

METHODS

We assigned annual PM 2.5 exposure estimates from 1999 to 2004 derived from 11 different approaches to Women's Health Initiative Memory Study (WHIMS) participant addresses within the contiguous US. Approaches included geostatistical interpolation approaches, land-use regression or spatiotemporal models, satellite-derived approaches, air dispersion and chemical transport models, and hybrid models. We used descriptive statistics and plots to assess relative and absolute agreement among exposure estimates and examined the impact of approach on associations between PM 2.5 and death due to natural causes, cardiovascular disease (CVD) mortality, and incident CVD events, adjusting for individual-level covariates and climate-based region.

RESULTS

With a few exceptions, relative agreement of approach-specific PM 2.5 exposure estimates was high for PM 2.5 concentrations across the contiguous US. Agreement among approach-specific exposure estimates was stronger near PM 2.5 monitors, in certain regions of the country, and in 2004 vs. 1999. Collectively, our results suggest but do not quantify lower agreement at local spatial scales for PM 2.5 . There was no evidence of large differences in health effects associations with PM 2.5 among estimation approaches in analyses adjusted for climate region.

CONCLUSIONS

Different estimation approaches produced similar spatial patterns of PM 2.5 concentrations across the contiguous US and in areas with dense monitoring data, and PM 2.5 -health effects associations were similar among estimation approaches. PM 2.5 estimates and PM 2.5 -health effects associations may differ more in samples drawn from smaller areas or areas without substantial monitoring data, or in analyses with finer adjustment for participant location. Our results can inform decisions about PM 2.5 estimation approach in epidemiologic studies, as investigators balance concerns about bias, efficiency, and resource allocation. Future work is needed to understand whether these conclusions also apply in the context of other air pollutants of interest. https://doi.org/10.1289/EHP12995.

Intra-urban variability of long-term exposure to PM2.5 and NO2 in five cities in Colombia

Rapidly urbanizing cities in Latin America experience high levels of air pollution which are known risk factors for population health. However, the estimates of long-term exposure to air pollution are scarce in the region. We developed intraurban land use regression (LUR) models to map long-term exposure to fine particulate matter (PM2.5) and nitrogen dioxide (NO2) in the five largest cities in Colombia. We conducted air pollution measurement campaigns using gravimetric PM2.5 and passive NO2 sensors for 2 weeks during both the dry and rainy seasons in 2021 in the cities of Barranquilla, Bucaramanga, Bogotá, Cali, and Medellín, and combined these data with geospatial and meteorological variables. Annual models were developed using multivariable spatial regression models. The city annual PM2.5 mean concentrations measured ranged between 12.32 and 15.99 µg/m3 while NO2 concentrations ranged between 24.92 and 49.15 µg/m3. The PM2.5 annual models explained 82% of the variance (R2) in Medellín, 77% in Bucaramanga, 73% in Barranquilla, 70% in Cali, and 44% in Bogotá. The NO2 models explained 65% of the variance in Bucaramanga, 57% in Medellín, 44% in Cali, 40% in Bogotá, and 30% in Barranquilla. Most of the predictor variables included in the models were a combination of specific land use characteristics and roadway variables. Cross-validation suggests that PM2.5 outperformed NO2 models. The developed models can be used as exposure estimate in epidemiological studies, as input in hybrid models to improve personal exposure assessment, and for policy evaluation.

PM2.5 and Dementia in a Low Exposure Setting: The Influence of Odor Identification Ability and APOE

BACKGROUND

Growing evidence show that long term exposure to air pollution increases the risk of dementia.

OBJECTIVE

The aim of this study was to investigate associations between PM2.5 exposure and dementia in a low exposure area, and to investigate the role of olfaction and the APOE ɛ4 allele in these associations.

METHODS

Data were drawn from the Betula project, a longitudinal study on aging, memory, and dementia in Sweden. Odor identification ability was assessed using the Scandinavian Odor Identification Test (SOIT). Annual mean PM2.5 concentrations were obtained from a dispersion-model and matched at the participants' residential address. Proportional hazard regression was used to calculate hazard ratios.

RESULTS

Of 1,846 participants, 348 developed dementia during the 21-year follow-up period. The average annual mean PM2.5 exposure at baseline was 6.77μg/m3, which is 1.77μg/m3 above the WHO definition of clean air. In a fully adjusted model (adjusted for age, sex, APOE, SOIT, cardiovascular diseases and risk factors, and education) each 1μg/m3 difference in annual mean PM2.5-concentration was associated with a hazard ratio of 1.23 for dementia (95% CI: 1.01-1.50). Analyses stratified by APOE status (ɛ4 carriers versus non-carriers), and odor identification ability (high versus low), showed associations only for ɛ4 carriers, and for low performance on odor identification ability.

CONCLUSION

PM2.5 was associated with an increased risk of dementia in this low pollution setting. The associations between PM2.5 and dementia seemed stronger in APOE carriers and those with below average odor identification ability.

Traffic-Related Air Pollution and Breast Cancer Risk: A Systematic Review and Meta-Analysis of Observational Studies

Current evidence of an association of breast cancer (BC) risk with air pollution exposure, in particular from traffic exhaust, remains inconclusive, and the exposure assessment methodologies are heterogeneous. This study aimed to conduct a systematic review and meta-analysis on the association between traffic-related air pollution (TRAP) and BC incidence (PROSPERO CRD42021286774). We systematically reviewed observational studies assessing exposure to TRAP and BC risk published until June 2022, available on Medline/PubMed and Web of Science databases. Studies using models for assessing exposure to traffic-related air pollutants or using exposure proxies (including traffic density, distance to road, etc.) were eligible for inclusion. A random-effects meta-analysis of studies investigating the association between NO₂/NO_x exposure and BC risk was conducted. Overall, 21 studies meeting the inclusion criteria were included (seven case-control, one nested case-control, 13 cohort studies); 13 studies (five case-control, eight cohort) provided data for inclusion in the meta-analyses. Individual studies provided little evidence of an association between TRAP and BC risk; exposure assessment methods and time periods of traffic emissions were different. The meta-estimate on NO₂ exposure indicated a positive association (pooled relative risk per 10 µg/m³ of NO₂: 1.015; 95% confidence interval, CI: 1.003; 1.028). No association between NO_x exposure and BC was found (three studies). Although there was limited evidence of an association for TRAP estimated with proxies, the meta-analysis showed a significant association between NO₂ exposure, a common TRAP pollutant marker, and BC risk, yet with a small effect size. Our findings provide additional support for air pollution carcinogenicity.

Locally weighted total least-squares variance component estimation for modeling urban air pollution

Land use regression (LUR) models are one of the standard methods for estimating air pollution concentration in urban areas. These models are usually low accurate due to inappropriate stochastic models (weight matrix). Furthermore, the measurement or modeling of dependent and independent variables used in LUR models is affected by various errors, which indicates the need to use an efficient stochastic and functional model to achieve the best estimation. This study proposes a locally weighted total least-squares variance component estimation (LW-TLS-VCE) for modeling urban air pollution. In the proposed method, in the first step, a locally weighted total least-squares (LW-TLS) regression is developed to simultaneously considers the non-stationary effects and errors of dependent and independent variables. In the second step, the variance components of the stochastic model are estimated to achieve the best linear unbiased estimation of unknowns. The efficiency of the proposed method is evaluated by modeling PM_2.5 concentrations via meteorological, land use, and traffic variables in Isfahan, Iran. The benefits provided by the proposed method, including considering non-stationary effects and random errors of all variables, besides estimating the actual variance of observations, are evaluated by comparing four consecutive methods. The obtained results demonstrate that using a suitable stochastic and functional model will significantly increase the proposed method's efficiency in PM_2.5 modeling.

Development of transferable neighborhood land use regression models for predicting intra-urban ambient nitrogen dioxide (NO2) spatial variations

Land use regression (LUR) models have been extensively used to predict air pollution exposure in epidemiological and environmental studies. The lack of dense routine monitoring networks in big cities places increased emphasis on the need for LUR models to be developed using purpose-designed neighborhood-scale monitoring data. However, the unsatisfactory model transferability limits these neighborhood LUR models to be then applied to other intra-urban areas in predicting air pollution exposure. In this study, we tackled this issue by proposing a method to develop transferable neighborhood NO₂ LUR models with comparable predictive power based on only micro-scale predictor variables for modeling intra-urban ambient air pollution exposure. Taking Auckland metropolis, New Zealand, as a case study, the proposed method was applied to three neighborhoods (urban, central business district, and dominion road) and compared with the corresponding counterpart models developed using pools of (a) only macro-scale predictor variables and (b) a mixture of both micro- and macro-scale predictor variables (traditional method). The results showed that the models using only macro-scale variables achieved the lowest accuracy (R²: 0.388-0.484) and had the worst direct (R²: 0.0001-0.349) and indirect transferability (R²: 0.07-0.352). Those models using the traditional method had the highest model fitting R² (0.629-0.966) with lower cross-validation R² (0.495-0.941) and slightly better direct transferability (R²: 0.0003-0.386) but suffered poor model interpretability when indirectly transferred to new locations. Our proposed models had comparable model fitting R² (0.601-0.966) and the best cross-validation R² (0.514-0.941). They also had the strongest direct transferability (R²: 0.006-0.590) and moderate-to-good indirect transferability (R²: 0.072-0.850) with much better model interpretability. This study advances our knowledge of developing transferable LUR models for the very first time from the perspective of the scale of the predictor variables used in the model development and will significantly benefit the wider application of LUR approaches in epidemiological and environmental studies.

Development and Evaluation of Spatio-Temporal Air Pollution Exposure Models and Their Combinations in the Greater London Area, UK

Land use regression (LUR) and dispersion/chemical transport models (D/CTMs) are frequently applied to predict exposure to air pollution concentrations at a fine scale for use in epidemiological studies. Moreover, the use of satellite aerosol optical depth data has been a key predictor especially for particulate matter pollution and when studying large populations. Within the STEAM project we present a hybrid spatio-temporal modeling framework by (a) incorporating predictions from dispersion modeling of nitrogen dioxide (NO₂), ozone (O₃) and particulate matter with an aerodynamic diameter equal or less than 10 μm (PM10) and less than 2.5 μm (PM2.5) into a spatio-temporal LUR model; and (b) combining the predictions LUR and dispersion modeling and additionally, only for PM2.5, from an ensemble machine learning approach using a generalized additive model (GAM). We used air pollution measurements from 2009 to 2013 from 62 fixed monitoring sites for O3, 115 for particles and up to 130 for NO₂, obtained from the dense network in the Greater London Area, UK. We assessed all models following a 10-fold cross validation (10-fold CV) procedure. The hybrid models performed better compared to separate LUR models. Incorporation of the dispersion estimates in the LUR models as a predictor, improved the LUR model fit: CV-R² increased to 0.76 from 0.71 for NO₂, to 0.79 from 0.57 for PM10, to 0.81 to 0.66 for PM2.5 and to 0.75 from 0.62 for O₃. The CV-R² obtained from the hybrid GAM framework was also increased compared to separate LUR models (CV-R² = 0.80 for NO₂, 0.76 for PM10, 0.79 for PM2.5 and 0.75 for O₃). Our study supports the combined use of different air pollution exposure assessment methods in a single modeling framework to improve the accuracy of spatio-temporal predictions for subsequent use in epidemiological studies.

Sampling Low Air Pollution Concentrations at a Neighborhood Scale in a Desert U.S. Metropolis with Volatile Weather Patterns

Background: Neighborhood-scale air pollution sampling methods have been used in a range of settings but not in low air pollution airsheds with extreme weather events such as volatile precipitation patterns and extreme summer heat and aridity—all of which will become increasingly common with climate change. The desert U.S. metropolis of Tucson, AZ, has historically low air pollution and a climate marked by volatile weather, presenting a unique opportunity. Methods: We adapted neighborhood-scale air pollution sampling methods to measure ambient NO₂, NO_x, and PM_2.5 and PM₁₀ in Tucson, AZ. Results: The air pollution concentrations in this location were well below regulatory guidelines and those of other locations using the same methods. While NO₂ and NO_x were reliably measured, PM_2.5 measurements were moderately correlated with those from a collocated reference monitor (r = 0.41, p = 0.13), potentially because of a combination of differences in inlet heights, oversampling of acutely high PM_2.5 events, and/or pump operation beyond temperature specifications. Conclusion: As the climate changes, sampling methods should be reevaluated for accuracy and precision, especially those that do not operate continuously. This is even more critical for low-pollution airsheds, as studies on low air pollution concentrations will help determine how such ambient exposures relate to health outcomes.

Characterization of ambient carbon monoxide and PM 2.5 effects on fetus development, liver enzymes and TSH in Isfahan City, central Iran

Ambient carbon monoxide (CO) and particulate matters (PMs) are two important air pollutants in urban areas with known impacts on fetuses. Hence, this study measured some biochemistry factors of 200 neonates with birth dates from January 19 to October 12, 2020, including the birth weight and height and the serum levels of ALT, AST, ALP, GGT, and TSH. The Support Vector Machine-fitted land-use regression approach was used to predict the spatio-temporal variability of intra-urban PM 2.5 and CO concentrations by month during the pregnancy period of the cases employing 5 variables of Digital Elevation Model (DEM), slope, and distance from Compressed Natural Gas (CNG) stations, Bus Rapid Transit (BRT) stations, and mines and industries. Spearman correlation analysis (p < 0.05) was performed between the neonate indices and mean monthly PM 2.5 and CO concentrations at the exact residential address of maternal cases and their nearby areas in 250, 500, 1000, 1500, and 2000 m-radius buffer rings. All modeling efforts succeeded in predicting CO and PM 2.5 levels with acceptable adjusted r² values. Northern Isfahan had relatively higher CO and PM 2.5 concentrations due to its adjacency to low-vegetated open lands and its high traffic load as compared to southern areas. The correlation results between the neonate biochemistry indices and mean PM 2.5 and CO concentrations were mostly positive in most buffer rings, especially in the >500 m-radius buffer rings for PM 2.5 and in the 2000 m-radius rings for CO. Although the correlation results of PM 2.5 followed a detectable trend in the buffer rings, the associations between CO and the neonate biochemistry indices differed significantly between the buffer rings. Results showed that increasing mean monthly concentration of CO and PM 2.5 may stimulate further production of liver enzymes while decreasing the birth weight and height.

Estimating PM2.5 concentration using the machine learning GA-SVM method to improve the land use regression model in Shaanxi, China

With rapid economic growth, urbanization and industrialization, fine particulate matter with aerodynamic diameters ≤ 2.5 µm (PM_2.5) has become a major pollutant and shows adverse effects on both human health and the atmospheric environment. Many studies on estimating PM_2.5 concentrations have been performed using statistical regression models and satellite remote sensing. However, the accuracy of PM_2.5 concentration estimates is limited by traditional regression models; machine learning methods have high predictive power, but fewer studies have been performed on the complementary advantages of different approaches. This study estimates PM_2.5 concentrations from satellite remote sensing-derived aerosol optical depth (AOD) products, meteorological data, terrain data and other predictors in 2015 in Shaanxi, China, using a combined genetic algorithm-support vector machine (GA-SVM) method, after which the spatial clustering pattern was explored at the season and year levels. The results indicated that temperature (r = -0.684), precipitation (r = -0.602) and normalized difference vegetation index (NDVI) (r = -0.523) were significantly negatively correlated with the PM_2.5 concentration, while AOD (r = 0.337) was significantly positively correlated with the PM_2.5 concentration. Compared to conventional land use regression (LUR) and SVM models and previous related studies, the GA-SVM method demonstrated a significantly better prediction accuracy of PM_2.5 concentration, with a higher 10-fold cross-validation coefficient of determination (R²) of 0.84 and lower root mean square error (RMSE) and mean absolute error (MAE) of 12.1 μg/m³ and 10.07 μg/m³, respectively. Y-scrambling test shows that the models have no chance correlation. The central and southern parts of Shaanxi have high PM_2.5 concentrations, which are mainly due to the pollutant emissions and meteorological and topographical conditions in those areas. There was a positive spatial agglomeration characteristic of regional PM_2.5 pollution, and the spatial spillover effect of PM_2.5 pollution for seasonal and annual variations does exist. In general, the GA-SVM method is robust and accurately estimates PM_2.5 concentrations via a novel modeling framework application and high-quality spatiotemporal information. It also has great significance for the exploration of PM_2.5 pollution estimation and high-precision mapping methods, especially early warning in high-risk areas. Finally, the prevention and control of atmospheric pollution should take pollution control measures from major cities and surrounding cities, and focus on the joint pollution control measures for plain cities.

Self-adaptive bandwidth eigenvector spatial filtering model for estimating PM2.5 concentrations in the Yangtze River Delta region of China

PM_2.5 concentrations are commonly estimated using geographically weighted regression (GWR) models, but these models may suffer from multi-collinearity and over-focus on local feature problems. To overcome these shortcomings, a self-adaptive bandwidth eigenvector spatial filtering (SA-ESF) model utilizing the golden section search (GO-ESF) and genetic algorithm (GA-ESF) was proposed. The SA-ESF model was applied to estimate ground PM_2.5 concentrations in the Yangtze River Delta (YRD) region of China both seasonally and annually from December 2015 to November 2016 using remotely sensing data, factory locations, and road networks. The results of the original eigenvector spatial filtering (ESF), GO-ESF, GA-ESF, and GWR models show that the GA-ESF model offers better performance and exhibits a better average adjusted R² which is 26.6%, 15.3%, and 10.8% higher than for the ESF, GO-ESF, and GWR models, respectively. We next calculated stochastic site indicators that can describe characteristics of regional concentration from interpolated concentration maps derived from the GA-ESF and GWR models. The concentration maps and stochastic site indicators point to major differences in the PM_2.5 concentrations in mountainous areas. There are notably high concentrations in those areas using the GWR model, in contrast with the GA-ESF results, indicating that there may be overfitting problems using the GWR model. Overall, the proposed SA-ESF model with the genetic algorithm technique can capture both global and local features and achieve promising results.

Application of land use regression modelling to describe atmospheric levels of semivolatile organic compounds on a national scale

Despite the success of passive sampler-based monitoring networks in capturing global atmospheric distributions of semivolatile organic compounds (SVOCs), their limited spatial resolution remains a challenge. Adequate spatial coverage is necessary to better characterize concentration gradients, identify point sources, estimate human exposure, and evaluate the effectiveness of chemical regulations such as the Stockholm Convention on Persistent Organic Pollutants. Land use regression (LUR) modelling can be used to integrate land use characteristics and other predictor variables (industrial emissions, traffic intensity, demographics, etc.) to describe or predict the distribution of air concentrations at unmeasured locations across a region or country. While LUR models are frequently applied to data-rich conventional air pollutants such as particulate matter, ozone, and nitrogen oxides, they are rarely applied to SVOCs. The MONET passive air sampling network (RECETOX, Masaryk University) continuously measures atmospheric SVOC levels across Czechia in monthly intervals. Using monitoring data from 29 MONET sites over a two-year period (2015-2017) and a variety of predictor variables, we developed LUR models to describe atmospheric levels and identify sources of polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyls (PCBs) and DDT across the country. Strong and statistically significant (R² > 0.6; p < 0.05) models were derived for PAH and PCB levels on a national scale. The PAH model retained three predictor variables - heating emissions represented by domestic fuel consumption, industrial PAH point sources, and the hill:valley index, a measure of site topography. The PCB model retained two predictor variables - site elevation, and secondary sources of PCBs represented by soil concentrations. These models were then applied to Czechia as a whole, highlighting the spatial variability of atmospheric SVOC levels, and providing a tool that can be used for further optimization of sampling network design, as well as evaluating potential human and environmental chemical exposures.

A review on the role of dispersion and receptor models in asthma research

There is substantial evidence that air pollution exposure is associated with asthma prevalence that affects millions of people worldwide. Air pollutant exposure can be determined using dispersion models and refined with receptor models. Dispersion models offer the advantage of giving spatially distributed outdoor pollutants concentration while the receptor models offer the source apportionment of specific chemical species. However, the use of dispersion and/or receptor models in asthma research requires a multidisciplinary approach, involving experts on air quality and respiratory diseases. Here, we provide a literature review on the role of dispersion and receptor models in air pollution and asthma research, their limitations, gaps and the way forward. We found that the methodologies used to incorporate atmospheric dispersion and receptor models in human health studies may vary considerably, and several of the studies overlook features such as indoor air pollution, model validation and subject pathway between indoor spaces. Studies also show contrasting results of relative risk or odds ratio for a health outcome, even using similar methodologies. Dispersion models are mostly used to estimate air pollution levels outside the subject's home, school or workplace; however, very few studies addressed the subject's routines or indoor/outdoor relationships. Conversely, receptor models are employed in regions where asthma incidence/prevalence is high or where a dispersion model has been previously used for this assessment. Road traffic (vehicle exhaust) and NOx are found to be the most targeted source and pollutant, respectively. Other key findings were the absence of a standard indicator, shortage of studies addressing VOC and UFP, and the shift toward chemical speciation of exposure.

Predicting Within-City Spatial Variations in Outdoor Ultrafine Particle and Black Carbon Concentrations in Bucaramanga, Colombia: A Hybrid Approach Using Open-Source Geographic Data and Digital Images

Outdoor ultrafine particles (UFP, <0.1 μm) and black carbon (BC) vary greatly within cities and may have adverse impacts on human health. In this study, we used a hybrid approach to develop new models to estimate within-city spatial variations in outdoor UFP and BC concentrations across Bucaramanga, Colombia. We conducted a mobile monitoring campaign over 20 days in 2019. Regression models were trained on land use data and combined with predictions from convolutional neural networks (CNN) trained to predict UFP and BC concentrations using satellite and street-level images. The combined UFP model (R² = 0.54) outperformed the CNN (R² = 0.47) and land use regression (LUR) models (R² = 0.47) on their own. Similarly, the combined BC model also outperformed the CNN and LUR BC models (R² = 0.51 vs 0.43 and 0.45, respectively). Spatial variations in model performance were more stable for the CNN and combined models compared to the LUR models, suggesting that the combined approach may be less likely to contribute to differential exposure measurement error in epidemiological studies. In general, our findings demonstrated that satellite and street-level images can be combined with a traditional LUR modeling approach to improve predictions of within-city spatial variations in outdoor UFP and BC concentrations.

Analysis of spatial factors, time-activity and infiltration on outdoor generated PM2.5 exposures of school children in five European cities

Atmospheric particles are a major environmental health risk. Assessments of air pollution related health burden are often based on outdoor concentrations estimated at residential locations, ignoring spatial mobility, time-activity patterns, and indoor exposures. The aim of this work is to quantify impacts of these factors on outdoor-originated fine particle exposures of school children. We apply nested WRF-CAMx modelling of PM_2.5 concentrations, gridded population, and school location data. Infiltration and enrichment factors were collected and applied to Athens, Kuopio, Lisbon, Porto, and Treviso. Exposures of school children were calculated for residential and school outdoor and indoor, other indoor, and traffic microenvironments. Combined with time-activity patterns six exposure models were created. Model complexity was increased incrementally starting from residential and school outdoor exposures. Even though levels in traffic and outdoors were considerably higher, 80-84% of the exposure to outdoor particles occurred in indoor environments. The simplest and also commonly used approach of using residential outdoor concentrations as population exposure descriptor (model 1), led on average to 26% higher estimates (15.7 μg/m³) compared with the most complex model (# 6) including home and school outdoor and indoor, other indoor and traffic microenvironments (12.5 μg/m³). These results emphasize the importance of including spatial mobility, time-activity and infiltration to reduce bias in exposure estimates.

Land use regression modelling of community noise in São Paulo, Brazil

Noise pollution has negative health consequences, which becomes increasingly relevant with rapid urbanization. In low- and middle-income countries research on health effects of noise is hampered by scarce exposure data and noise maps. In this study, we developed land use regression (LUR) models to assess spatial variability of community noise in the Western Region of São Paulo, Brazil.We measured outdoor noise levels continuously at 42 homes once or twice for one week in the summer and the winter season. These measurements were integrated with various geographic information system variables to develop LUR models for predicting average A-weighted (dB(A)) day-evening-night equivalent sound levels (L_den) and night sound levels (L_night). A supervised mixed linear regression analysis was conducted to test potential noise predictors for various buffer sizes and distances between home and noise source. Noise exposure levels in the study area were high with a site average L_den of 69.3 dB(A) ranging from 60.3 to 82.3 dB(A), and a site average L_night of 59.9 dB(A) ranging from 50.7 to 76.6 dB(A). LUR models had a good fit with a R² of 0.56 for L_den and 0.63 for L_night in a leave-one-site-out cross validation. Main predictors of noise were the inverse distance to medium roads, count of educational facilities within a 400 m buffer, mean Normalized Difference Vegetation Index (NDVI) within a 100 m buffer, residential areas within a 50 m (L_den) or 25 m (L_night) buffer and slum areas within a 400 m buffer. Our study suggests that LUR modelling with geographic predictor data is a promising and efficient approach for noise exposure assessment in low- and middle-income countries, where noise maps are not available.

The application of land use regression model to investigate spatiotemporal variations of PM2.5 in Guangzhou, China: Implications for the public health benefits of PM2.5 reduction

Understanding the intra-city variation of PM_2.5 is important for air quality management and exposure assessment. In this study, to investigate the spatiotemporal variation of PM_2.5 in Guangzhou, we developed land use regression (LUR) models using data from 49 routine air quality monitoring stations. The R², adjust R² and 10-fold cross validation R² for the annual PM_2.5 LUR model were 0.78, 0.72 and 0.66, respectively, indicating the robustness of the model. In all the LUR models, traffic variables (e.g., length of main road and the distance to nearest ancillary) were the most common variables in the LUR models, suggesting vehicle emission was the most important contributor to PM_2.5 and controlling vehicle emissions would be an effective way to reduce PM_2.5. The predicted PM_2.5 exhibited significant variations with different land uses, with the highest value for impervious surfaces, followed by green land, cropland, forest and water areas. Guangzhou as the third largest city that PM_2.5 concentration has achieved CAAQS Grade II guideline in China, it represents a useful case study city to examine the health and economic benefits of further reduction of PM_2.5 to the lower concentration ranges. So, the health and economic benefits of reducing PM_2.5 in Guangzhou was further estimated using the BenMAP model, based on the annual PM_2.5 concentration predicted by the LUR model. The results showed that the avoided all cause mortalities were 992 cases (95% CI: 221-2140) and the corresponding economic benefits were 1478 million CNY (95% CI: 257-2524) (willingness to pay approach) if the annual PM_2.5 concentration can be reduced to the annual CAAQS Grade I guideline value of 15 μg/m³. Our results are expected to provide valuable information for further air pollution control strategies in China.

A hybrid air pollution / land use regression model for predicting air pollution concentrations in Durban, South Africa

The objective of this paper was to incorporate source-meteorological interaction information from two commonly employed atmospheric dispersion models into the land use regression technique for predicting ambient nitrogen dioxide (NO₂), sulphur dioxide (SO₂), and particulate matter (PM₁₀). The study was undertaken across two regions in Durban, South Africa, one with a high industrial profile and a nearby harbour, and the other with a primarily commercial and residential profile. Multiple hybrid models were developed by integrating air pollution dispersion modelling predictions for source specific NO₂, SO₂, and PM₁₀ concentrations into LUR models following the European Study of Cohorts for Air Pollution Effects (ESCAPE) methodology to characterise exposure, in Durban. Industrial point sources, ship emissions, domestic fuel burning, and vehicle emissions were key emission sources. Standard linear regression was used to develop annual, summer and winter hybrid models to predict air pollutant concentrations. Higher levels of NO₂ and SO₂ were predicted in south Durban as compared to north Durban as these are industrial related pollutants. Slightly higher levels of PM₁₀ were predicted in north Durban as compared to south Durban and can be attributed to either traffic, bush burning or domestic fuel burning. The hybrid NO₂ models for annual, summer and winter explained 60%, 58% and 63%, respectively, of the variance with traffic, population and harbour being identified as important predictors. The SO₂ models were less robust with lower R² annual (44%), summer (53%) and winter (46%), in which industrial and traffic variables emerged as important predictors. The R² for PM₁₀ models ranged from 80% to 85% with population and urban land use type emerging as predictor variables.

A hybrid approach to predict daily NO2 concentrations at city block scale

Estimating the ambient concentration of nitrogen dioxide (NO2) is challenging because NO2 generated by local fossil fuel combustion varies greatly in concentration across space and time. This study demonstrates an integrated hybrid approach combining dispersion modeling and land use regression (LUR) to predict daily NO2 concentrations at a high spatial resolution (e.g., 50 m) in the New York tri-state area. The daily concentration of traffic-related NO2 was estimated at the Environmental Protection Agency's NO2 monitoring sites in the study area for the years 2015-2017, using the Research LINE source (R-LINE) model with inputs of traffic data provided by the Highway Performance and Management System and meteorological data provided by the NOAA Integrated Surface Database. We used the R-LINE-predicted daily concentrations of NO2 to build mixed-effects regression models, including additional variables representing land use features, geographic characteristics, weather, and other predictors. The mixed model was selected by the Elastic Net method. Each model's performance was evaluated using the out-of-sample coefficient of determination (R2) and the square root of mean squared error (RMSE) from ten-fold cross-validation (CV). The mixed model showed a good prediction performance (CV R2: 0.75-0.79, RMSE: 3.9-4.0 ppb). R-LINE outputs improved the overall, spatial, and temporal CV R2 by 10.0%, 18.9% and 7.7% respectively. Given the output of R-LINE is point-based and has a flexible spatial resolution, this hybrid approach allows prediction of daily NO2 at an extremely high spatial resolution such as city blocks.

Comparison of associations between mortality and air pollution exposure estimated with a hybrid, a land-use regression and a dispersion model

INTRODUCTION

To characterize air pollution exposure at a fine spatial scale, different exposure assessment methods have been applied. Comparison of associations with health from different exposure methods are scarce. The aim of this study was to evaluate associations of air pollution based on hybrid, land-use regression (LUR) and dispersion models with natural cause and cause-specific mortality.

METHODS

We followed a Dutch national cohort of approximately 10.5 million adults aged 29+ years from 2008 until 2012. We used Cox proportional hazard models with age as underlying time scale and adjusted for several potential individual and area-level socio-economic status confounders to evaluate associations of annual average residential NO2, PM2.5 and BC exposure estimates based on two stochastic models (Dutch LUR, European-wide hybrid) and deterministic Dutch dispersion models.

RESULTS

Spatial variability of PM2.5 and BC exposure was smaller for LUR compared to hybrid and dispersion models. NO2 exposure variability was similar for the three methods. Pearson correlations between hybrid, LUR and dispersion modeled NO2 and BC ranged from 0.72 to 0.83; correlations for PM2.5 were slightly lower (0.61-0.72). In general, all three models showed stronger associations of air pollutants with respiratory disease and lung cancer mortality than with natural cause and cardiovascular disease mortality. The strength of the associations differed between the three exposure models. Associations of air pollutants estimated by LUR were generally weaker compared to associations of air pollutants estimated by hybrid and dispersion models. For natural cause mortality, we found a hazard ratio (HR) of 1.030 (95% confidence interval (CI): 1.019, 1.041) per 10 µg/m3 for hybrid modeled NO2, a HR of 1.003 (95% CI: 0.993, 1.013) per 10 µg/m3 for LUR modeled NO2 and a HR of 1.015 (95% CI: 1.005, 1.024) per 10 µg/m3 for dispersion modeled NO2.

CONCLUSION

Air pollution was positively associated with natural cause and cause-specific mortality, but the strength of the associations differed between the three exposure models. Our study documents that the selected exposure model may contribute to heterogeneity in effect estimates of associations between air pollution and health.

Impact of different exposure models and spatial resolution on the long-term effects of air pollution

Long-term exposure to air pollution has been related to mortality in several epidemiological studies. The investigations have assessed exposure using various methods achieving different accuracy in predicting air pollutants concentrations. The comparison of the health effects estimates are therefore challenging. This paper aims to compare the effect estimates of the long-term effects of air pollutants (particulate matter with aerodynamic diameter less than 10 μm, PM₁₀, and nitrogen dioxide, NO₂) on cause-specific mortality in the Rome Longitudinal Study, using exposure estimates obtained with different models and spatial resolutions. Annual averages of NO₂ and PM₁₀ were estimated for the year 2015 in a large portion of the Rome urban area (12 × 12 km²) applying three modelling techniques available at increasing spatial resolution: 1) a chemical transport model (CTM) at 1km resolution; 2) a land-use random forest (LURF) approach at 200m resolution; 3) a micro-scale Lagrangian particle dispersion model (PMSS) taking into account the effect of buildings structure at 4 m resolution with results post processed at different buffer sizes (12, 24, 52, 100 and 200 m). All the exposures were assigned at the residential addresses of 482,259 citizens of Rome 30+ years of age who were enrolled on 2001 and followed-up till 2015. The association between annual exposures and natural-cause, cardiovascular (CVD) and respiratory (RESP) mortality were estimated using Cox proportional hazards models adjusted for individual and area-level confounders. We found different distributions of both NO₂ and PM₁₀ concentrations, across models and spatial resolutions. Natural cause and CVD mortality outcomes were all positively associated with NO₂ and PM₁₀ regardless of the model and spatial resolution when using a relative scale of the exposure such as the interquartile range (IQR): adjusted Hazard Ratios (HR), and 95% confidence intervals (CI), of natural cause mortality, per IQR increments in the two pollutants, ranged between 1.012 (1.004, 1.021) and 1.018 (1.007, 1.028) for the different NO₂ estimates, and between 1.010 (1.000, 1.020) and 1.020 (1.008, 1.031) for PM₁₀, with a tendency of larger effect for lower resolution exposures. The latter was even stronger when a fixed value of 10 μg/m³ is used to calculate HRs. Long-term effects of air pollution on mortality in Rome were consistent across different models for exposure assessment, and different spatial resolutions.

Air Pollution and Progression of Atherosclerosis in Different Vessel Beds-Results from a Prospective Cohort Study in the Ruhr Area, Germany

OBJECTIVES

Due to inconsistent epidemiological evidence on health effects of air pollution on progression of atherosclerosis, we investigated several air pollutants and their effects on progression of atherosclerosis, using carotid intima media thickness (cIMT), coronary calcification (CAC), and thoracic aortic calcification (TAC).

METHODS

We used baseline (2000-2003) and 5-y follow-up (2006-2008) data from the German Heinz Nixdorf Recall cohort study, including 4,814 middle-aged adults. Residence-based long-term air pollution exposure, including particulate matter (PM) with aerodynamic diameter ≤2.5μm (PM2.5), (PM10), and nitrogen dioxide (NO2) was assessed using chemistry transport and land use regression (LUR) models. cIMT was quantified as side-specific median IMT assessed from standardized ultrasound images. CAC and TAC were quantified by computed tomography using the Agatston score. Development (yes/no) and progression of atherosclerosis (change in cIMT and annual growth rate for CAC/TAC) were analyzed with logistic and linear regression models, adjusting for age, sex, lifestyle variables, socioeconomic status, and traffic noise.

RESULTS

While no clear associations were observed in the full study sample (mean age 59.1 (±7.6) y; 53% female), most air pollutants were marginally associated with progression of atherosclerosis in participants with no or low baseline atherosclerotic burden. Most consistently for CAC, e.g., a 1.5 μg/m3 higher exposure to PM2.5 (LUR) yielded an estimated odds ratio of 1.19 [95% confidence interval (CI): 1.03, 1.39] for progression of CAC and an increased annual growth rate of 2% (95% CI: 1%, 4%).

CONCLUSION

Our study suggests that development and progression of subclinical atherosclerosis is associated with long-term air pollution in middle-aged participants with no or minor atherosclerotic burden at baseline, while overall no consistent associations are observed. https://doi.org/10.1289/EHP7077.

Drivers of seasonal and annual air pollution exposure in a complex urban environment with multiple source contributions

Outdoor air pollution is a global health concern, but detailed exposure information is still limited for many parts of the world. In this study, high-resolution exposure surfaces were generated for annual and seasonal fine particulate matter (PM_2.5), coarse particulate matter (PM₁₀), and carbon monoxide (CO) for the Greater Beirut Area (GBA), Lebanon, an urban zone with a complex topography and multiple source contributions. Land use regression models (LUR) were calibrated and validated with monthly data collected from 58 locations between March 2017 and March 2018. The annual mean (±1 SD) concentrations of PM_2.5, PM₁₀, and CO across the monitoring locations were 68.1 (±15.7) μg/m³, 83.5 (±19.5) μg/m³, and 2.48 (±1.12) ppm, respectively. The coefficients of determination for LUR models ranged from 56 to 67% for PM_2.5, 44 to 63% for the PM₁₀ models, and 50 to 60% for the CO. LUR model structures varied significantly by season for both PM_2.5 and PM₁₀ but not for CO. Traffic emissions were consistently the main source of CO emissions throughout the year. The relative importance of industrial emissions and power generation sources towards predicted PM levels increased during the hot season while the contribution of the international airport diminished. Moreover, the complex topography of the study area along with the seasonal changes in the predominant wind directions affected the spatial predicted concentrations of all three pollutants. Overall, the predicted exposure surfaces were able to conserve the inter-pollution correlations determined from the field monitoring campaign, with the exception of the cold season. Our pollution surfaces suggest that the entire population of Beirut is regularly exposed to concentrations exceeding the World Health Organization (WHO) air quality standards for both PM_2.5 and PM₁₀.

Particle concentrations, dispersion modelling and evaluation in southern Sweden

Health impact assessments of differential air pollution rely on epidemiologically established relationships between concentration levels where people are exposed and adverse health outcomes. To assess air pollution concentrations, land use regression is commonly used. However, an alternative tool is dispersion modelling, where a detailed inventory of pollution sources together with meteorological data drives calculations of compound dispersion. With this, both spatial and temporal variation can be assessed. In this study, we evaluated results of a Gaussian dispersion model applied to an emissions inventory for Scania, the southernmost county in Sweden. The dispersion considered was particulate matter of aerodynamic diameter < 10 µm (PM10), particulate matter of aerodynamic diameter < 2.5 µm (PM2.5) and black carbon (BC) during an 11-year period (2000–2011). Mean concentrations and 95th percentiles expressed in µg/m3 ranged from 10.1 to 12.6 and 16.6 to 20.7 for PM2.5 and from 14.0 to 18.8 and 22.6 to 27.0 for PM10, respectively. Seven monitoring stations were used for evaluation. Correlations (R2) ranged from 0.44 to 0.86 for PM2.5 (mean bias from − 9.0 to 0.1 µg/m3) and from 0.46 to 0.83 for PM10 (mean bias − 6.1 to 3.5 µg/m3). An evaluated database of PM and BC concentrations for Scania is now available for future exposure assessment projects. Calculations were based on a well-known dispersion model with detailed emission data as input. The evaluation showed correlation coefficients for PM in line with previous literature. The data on PM10, PM2.5 and BC concentrations will, therefore, be used in subsequent studies, epidemiological as well as health impact assessments.

Spatial analysis and risk assessment of urban BTEX compounds in Urmia, Iran

Land Use Regression models (LUR) are the most common tools to estimate intra-urban air pollutant exposure in epidemiological studies. However, number of available and published models in developing and middle up income countries is still scarce. Here, we developed seasonal and overall LUR models for the spatial distribution of benzene, toluene, ethylbenzene and xylene (BTEX) based on 20 monitoring stations and 166 potentially predictive variables (PPVs) in Urmia, Iran. Carcinogenic and non-carcinogenic risks of exposure to BTEX and its sensitivity analysis were assessed using a probabilistic approach. The mean and standard deviation (in brackets) of overall benzene, toluene, ethylbenzene and xylene were 12.83 (16.19), 27.03 (32.00), 4.72 (4.15) and 27.35 (29.36) μg/m³, respectively. In all models the R² value of LUR models of benzene, toluene, ethylbenzene, xylene and total BTEX ranged from 0.66 to 0.85, 0.61, 0.88, 0.72 to 0.94, 0.75 to 0.84 and 0.67 to 0.93. The root mean square error (RMSE) for leave-one-out cross-validations (LOOCV) for benzene, toluene, ethylbenzene and xylene ranged from 7.48 to 10.31, 23.0 to 30.0, 3.40 to 6.90, 16.27 to 24.49, 36.10-50.0 μg/m³, respectively. The estimated lifetime carcinogenic risk (LTCR) indicated that ambient concentration of benzene is at a risk level for Urmia inhabitants (LTCR >10^-6). Sensitivity analysis for LTCR model indicated that concentration of benzene (C) was the most effective variable in increasing the carcinogenic risk (correlation coefficient ranged from 0.97 to 0.98 for all models).

Estimated health impacts from maritime transport in the Mediterranean region and benefits from the use of cleaner fuels

Ship traffic emissions degrade air quality in coastal areas and contribute to climate impacts globally. The estimated health burden of exposure to shipping emissions in coastal areas may inform policy makers as they seek to reduce exposure and associated potential health impacts. This work estimates the PM2.5-attributable impacts in the form of premature mortality and cardiovascular and respiratory hospital admissions, from long-term exposure to shipping emissions. Health impact assessment (HIA) was performed in 8 Mediterranean coastal cities, using a baseline conditions from the literature and a policy case accounting for the MARPOL Annex VI rules requiring cleaner fuels in 2020. Input data were (a) shipping contributions to ambient PM2.5 concentrations based on receptor modelling studies found in the literature, (b) population and health incidence data from national statistical registries, and (c) geographically-relevant concentration-response functions from the literature. Long-term exposure to ship-sourced PM2.5 accounted for 430 (95% CI: 220-650) premature deaths per year, in the 8 cities, distributed between groups of cities: Barcelona and Athens, with >100 premature deaths/year, and Nicosia, Brindisi, Genoa, Venice, Msida and Melilla, with tens of premature deaths/year. The more stringent standards in 2020 would reduce the number of PM2.5-attributable premature deaths by 15% on average. HIA provided a comparative assessment of the health burden of shipping emissions across Mediterranean coastal cities, which may provide decision support for urban planning with a special focus on harbour areas, and in view of the reduction in sulphur content of marine fuels due to MARPOL Annex VI in 2020.

A Random Forest Approach to Estimate Daily Particulate Matter, Nitrogen Dioxide, and Ozone at Fine Spatial Resolution in Sweden

Air pollution is one of the leading causes of mortality worldwide. An accurate assessment of its spatial and temporal distribution is mandatory to conduct epidemiological studies able to estimate long-term (e.g., annual) and short-term (e.g., daily) health effects. While spatiotemporal models for particulate matter (PM) have been developed in several countries, estimates of daily nitrogen dioxide (NO2) and ozone (O3) concentrations at high spatial resolution are lacking, and no such models have been developed in Sweden. We collected data on daily air pollutant concentrations from routine monitoring networks over the period 2005–2016 and matched them with satellite data, dispersion models, meteorological parameters, and land-use variables. We developed a machine-learning approach, the random forest (RF), to estimate daily concentrations of PM10 (PM<10 microns), PM2.5 (PM<2.5 microns), PM2.5–10 (PM between 2.5 and 10 microns), NO2, and O3 for each squared kilometer of Sweden over the period 2005–2016. Our models were able to describe between 64% (PM10) and 78% (O3) of air pollutant variability in held-out observations, and between 37% (NO2) and 61% (O3) in held-out monitors, with no major differences across years and seasons and better performance in larger cities such as Stockholm. These estimates will allow to investigate air pollution effects across the whole of Sweden, including suburban and rural areas, previously neglected by epidemiological investigations.

An approach to predict population exposure to ambient air PM2.5 concentrations and its dependence on population activity for the megacity London

A comprehensive modelling approach has been developed to predict population exposure to the ambient air PM_2.5 concentrations in different microenvironments in London. The modelling approach integrates air pollution dispersion and exposure assessment, including treatment of the locations and time activity of the population in three microenvironments, namely, residential, work and transport, based on national demographic information. The approach also includes differences between urban centre and suburban areas of London by taking account of the population movements and the infiltration of PM_2.5 from outdoor to indoor. The approach is tested comprehensively by modelling ambient air concentrations of PM_2.5 at street scale for the year 2008, including both regional and urban contributions. Model analysis of the exposure in the three microenvironments shows that most of the total exposure, 85%, occurred at home and work microenvironments and 15% in the transport microenvironment. However, the annual population weighted mean (PWM) concentrations of PM_2.5 for London in transport microenvironments were almost twice as high (corresponding to 13-20 μg/m³) as those for home and work environments (7-12 μg/m³). Analysis has shown that the PWM PM_2.5 concentrations in central London were almost 20% higher than in the surrounding suburban areas. Moreover, the population exposure in the central London per unit area was almost three times higher than that in suburban regions. The exposure resulting from all activities, including outdoor to indoor infiltration, was about 20% higher, when compared with the corresponding value obtained assuming inside home exposure for all times. The exposure assessment methodology used in this study predicted approximately over one quarter (-28%) lower population exposure, compared with using simply outdoor concentrations at residential locations. An important implication of this study is that for estimating population exposure, one needs to consider the population movements, and the infiltration of pollution from outdoors to indoors.

Endotoxin and particulate matter emitted by livestock farms and respiratory health effects in neighboring residents

BACKGROUND

Living in livestock-dense areas has been associated with health effects, suggesting airborne exposures to livestock farm emissions to be relevant for public health. Livestock farm emissions involve complex mixtures of various gases and particles. Endotoxin, a pro-inflammatory agent of microbial origin, is a constituent of livestock farm emitted particulate matter (PM) that is potentially related to the observed health effects. Quantification of livestock associated endotoxin exposure at residential addresses in relation to health outcomes has not been performed earlier.

OBJECTIVES

We aimed to assess exposure-response relations for a range of respiratory endpoints and atopic sensitization in relation to livestock farm associated PM₁₀ and endotoxin levels.

METHODS

Self-reported respiratory symptoms of 12,117 persons participating in a population-based cross-sectional study were analyzed. For 2494 persons, data on lung function (spirometry) and serologically assessed atopic sensitization was additionally available. Annual-average PM₁₀ and endotoxin concentrations at home addresses were predicted by dispersion modelling and land-use regression (LUR) modelling. Exposure-response relations were analyzed with generalized additive models.

RESULTS

Health outcomes were generally more strongly associated with exposure to livestock farm emitted endotoxin compared to PM₁₀. An inverse association was observed for dispersion modelled exposure with atopic sensitization (endotoxin: p = .004, PM₁₀: p = .07) and asthma (endotoxin: p = .029, PM₁₀: p = .022). Prevalence of respiratory symptoms decreased with increasing endotoxin concentration at the lower range, while at the higher range prevalence increased with increasing concentration (p < .05). Associations between lung function parameters with exposure to PM₁₀ and endotoxin were not statistically significant (p > .05).

CONCLUSIONS

Exposure to livestock farm emitted particulate matter is associated with respiratory health effects and atopic sensitization in non-farming residents. Results indicate endotoxin to be a potentially plausible etiologic agent, suggesting non-infectious aspects of microbial emissions from livestock farms to be important with respect to public health.

Cancer and mortality in relation to traffic-related air pollution among coronary patients: Using an ensemble of exposure estimates to identify high-risk individuals

BACKGROUND

Moderate correlations were previously observed between individual estimates of traffic-related air pollution (TRAP) produced by different exposure modeling approaches. This induces exposure misclassification for a substantial fraction of subjects.

AIM

We used an ensemble of well-established modeling approaches to increase certainty of exposure classification and reevaluated the association with cancers previously linked to TRAP (lung, breast and prostate), other cancers, and all-cause mortality in a cohort of coronary patients.

METHODS

Patients undergoing percutaneous coronary interventions in a major Israeli medical center from 2004 to 2014 (n = 10,627) were followed for cancer (through 2015) and mortality (through 2017) via national registries. Residential exposure to nitrogen oxides (NO_x) -a proxy for TRAP- was estimated by optimized dispersion model (ODM) and land use regression (LUR) (r_Pearson = 0.50). Mutually exclusive groups of subjects classified as exposed by none of the methods (high-certainty low-exposed), ODM alone, LUR alone, or both methods (high-certainty high-exposed) were created. Associations were examined using Cox regression models.

RESULTS

During follow-up, 741 incident cancer cases were diagnosed and 3051 deaths occurred. Using a ≥25 ppb cutoff, compared with high-certainty low exposed, the multivariable-adjusted hazard ratios (95% confidence intervals) for lung, breast and prostate cancer were 1.56 (1.13-2.15) in high-certainty exposed, 1.27 (0.86-1.86) in LUR-exposed alone, and 1.13 (0.77-1.65) in ODM-exposed alone. The association of the former category was strengthened using more extreme NO_x cutoffs. A similar pattern, albeit less strong, was observed for mortality, whereas no association was shown for cancers not previously linked to TRAP.

CONCLUSIONS

Use of an ensemble of TRAP exposure estimates may improve classification, resulting in a stronger association with outcomes.

Air pollution, lung function and COPD: results from the population-based UK Biobank study

Ambient air pollution increases the risk of respiratory mortality, but evidence for impacts on lung function and chronic obstructive pulmonary disease (COPD) is less well established. The aim was to evaluate whether ambient air pollution is associated with lung function and COPD, and explore potential vulnerability factors.We used UK Biobank data on 303 887 individuals aged 40-69 years, with complete covariate data and valid lung function measures. Cross-sectional analyses examined associations of land use regression-based estimates of particulate matter (particles with a 50% cut-off aerodynamic diameter of 2.5 and 10 µm: PM_2.5 and PM₁₀, respectively; and coarse particles with diameter between 2.5 μm and 10 μm: PM_coarse) and nitrogen dioxide (NO₂) concentrations with forced expiratory volume in 1 s (FEV₁), forced vital capacity (FVC), the FEV₁/FVC ratio and COPD (FEV₁/FVC <lower limit of normal). Effect modification was investigated for sex, age, obesity, smoking status, household income, asthma status and occupations previously linked to COPD.Higher exposures to each pollutant were significantly associated with lower lung function. A 5 µg·m^-3 increase in PM_2.5 concentration was associated with lower FEV₁ (-83.13 mL, 95% CI -92.50- -73.75 mL) and FVC (-62.62 mL, 95% CI -73.91- -51.32 mL). COPD prevalence was associated with higher concentrations of PM_2.5 (OR 1.52, 95% CI 1.42-1.62, per 5 µg·m^-3), PM₁₀ (OR 1.08, 95% CI 1.00-1.16, per 5 µg·m^-3) and NO₂ (OR 1.12, 95% CI 1.10-1.14, per 10 µg·m^-3), but not with PM_coarse Stronger lung function associations were seen for males, individuals from lower income households, and "at-risk" occupations, and higher COPD associations were seen for obese, lower income, and non-asthmatic participants.Ambient air pollution was associated with lower lung function and increased COPD prevalence in this large study.

Analytical Study of Fuel Switching from Heavy Fuel Oil to Natural Gas in clay brick factories at Arab Abu Saed, Greater Cairo

Arab Abu Saed area in Giza governorate, south to Cairo contains more than 228 clay brick kilns represent the largest cluster of brickworks in Egypt. Burning of Heavy Fuel Oil (HFO) in such kilns is the main source of air pollution in the surrounding locations. In this study, investigation of switching the fuel used in brick kilns from (HFO) to Natural Gas (NG) is carried out and the pollution loads are assessed in both cases. In addition, two Gaussian dispersion plume models are employed to estimate the concentration of primary pollutants; PM₁₀, SO₂, and NO₂ at seven locations in the vicinity of Arab Abu Saed to determine the most adversely affected locations. Statistical analysis is applied to evaluate the correlation and conformity of the results of both models. Results show that using of NG leads to a significant reduction of pollution loads of PM₁₀, SO₂ and NO₂ reaches 96%, 72%, and 24% respectively. In addition, the reduction of naturally occurring radionuclides in air is analyzed. Activity concentrations of Ra-226, Th-232 and K-40 in Bq/l for HFO were measured using HPGe detector for six HFO samples. Exposure due to air submersion of naturally occurring radionuclides in the study area leads to annual equivalent dose ranged between 2.16 mSv/y (received by Uterus) and 14 mSv/y (received by skin), and average effective dose 2.65 mSv/y which represent valuable exposure.

Comparison of model estimates from an intra-city land use regression model with a national satellite-LUR and a regional Bayesian Maximum Entropy model, in estimating NO2 for a birth cohort in Sydney, Australia

BACKGROUND

Methods for estimating air pollutant exposures for epidemiological studies are becoming more complex in an effort to minimise exposure error and its associated bias. While land use regression (LUR) modelling is now an established method, there has been little comparison between LUR and other recent, more complex estimation methods. Our aim was to develop a LUR model to estimate intra-city exposures to nitrogen dioxide (NO₂) for a Sydney cohort, and to compare those with estimates from a national satellite-based LUR model (Sat-LUR) and a regional Bayesian Maximum Entropy (BME) model.

METHODS

Satellite-based LUR and BME estimates were obtained using existing models. We used methods consistent with the European Study of Cohorts for Air Pollution Effects (ESCAPE) methodology to develop LUR models for NO₂ and NOx. We deployed 46 Ogawa passive samplers across western Sydney during 2013/2014 and acquired data on land use, population density, and traffic volumes for the study area. Annual average NO₂ concentrations for 2013 were estimated for 947 addresses in the study area using the three models: standard LUR, Sat-LUR and a BME model. Agreement between the estimates from the three models was assessed using interclass correlation coefficient (ICC), Bland-Altman methods and correlation analysis (CC).

RESULTS

The NO₂ LUR model predicted 84% of spatial variability in annual mean NO₂ (RMSE: 1.2 ppb; cross-validated R²: 0.82) with predictors of major roads, population and dwelling density, heavy traffic and commercial land use. A separate model was developed that captured 92% of variability in NOx (RMSE 2.3 ppb; cross-validated R²: 0.90). The annual average NO₂ concentrations were 7.31 ppb (SD: 1.91), 7.01 ppb (SD: 1.92) and 7.90 ppb (SD: 1.85), for the LUR, Sat-LUR and BME models respectively. Comparing the standard LUR with Sat-LUR NO₂ cohort estimates, the mean estimates from the LUR were 4% higher than the Sat-LUR estimates, and the ICC was 0.73. The Pearson's correlation coefficients (CC) for the LUR vs Sat-LUR values were r = 0.73 (log-transformed data) and r = 0.69 (untransformed data). Comparison of the NO₂ cohort estimates from the LUR model with the BME blended model indicated that the LUR mean estimates were 8% lower than the BME estimates. The ICC for the LUR vs BME estimates was 0.73. The CC for the logged LUR vs BME estimates was r = 0.73 and for the unlogged estimates was r = 0.69.

CONCLUSIONS

Our LUR models explained a high degree of spatial variability in annual mean NO₂ and NOx in western Sydney. The results indicate very good agreement between the intra-city LUR, national-scale sat-LUR, and regional BME models for estimating NO₂ for a cohort of children residing in Sydney, despite the different data inputs and differences in spatial scales of the models, providing confidence in their use in epidemiological studies.

Considerations for evaluating green infrastructure impacts in microscale and macroscale air pollution dispersion models

Green infrastructure (GI) in urban areas may be adopted as a passive control system to reduce air pollutant concentrations. However, current dispersion models offer limited modelling options to evaluate its impact on ambient pollutant concentrations. The scope of this review revolves around the following question: how can GI be considered in readily available dispersion models to allow evaluation of its impacts on pollutant concentrations and health risk assessment? We examined the published literature on the parameterisation of deposition velocities and datasets for both particulate matter and gaseous pollutants that are required for deposition schemes. We evaluated the limitations of different air pollution dispersion models at two spatial scales - microscale (i.e. 10-500 m) and macroscale (i.e. 5-100 km) - in considering the effects of GI on air pollutant concentrations and exposure alteration. We conclude that the deposition schemes that represent GI impacts in detail are complex, resource-intensive, and involve an abundant volume of input data. An appropriate handling of GI characteristics (such as aerodynamic effect, deposition of air pollutants and surface roughness) in dispersion models is necessary for understanding the mechanism of air pollutant concentrations simulation in presence of GI at different spatial scales. The impacts of GI on air pollutant concentrations and health risk assessment (e.g., mortality, morbidity) are partly explored. The i-Tree tool with the BenMap model has been used to estimate the health outcomes of annually-averaged air pollutant removed by deposition over GI canopies at the macroscale. However, studies relating air pollution health risk assessments due to GI-related changes in short-term exposure, via pollutant concentrations redistribution at the microscale and enhanced atmospheric pollutant dilution by increased surface roughness at the macroscale, along with deposition, are rare. Suitable treatments of all physical and chemical processes in coupled dispersion-deposition models and assessments against real-world scenarios are vital for health risk assessments.

Estimate annual and seasonal PM1, PM2.5 and PM10 concentrations using land use regression model

Exposure to ambient particulate matter (PM) can increase mortality and morbidity in urban area. In this study, annual and seasonal spatial pattern of PM₁, PM_2.5 and PM₁₀ pollutants were assessed using land use regression (LUR) models in Sabzevar, Iran. The studied pollutants were measured at 26 monitoring stations of different microenvironments in the study area. Sampling was conducted during four campaigns from April 2017 to February 2018. LUR models were developed based on 104 potentially predictive variables (PPVs) subdivided in six categories and 22 sub-categories. The annual mean (standard deviation) of PM₁, PM_2.5 and PM₁₀ were 36.46 (8.56), 39.62 (10.55) and 51.99 (16.25) μg/m³, respectively. The R² values and root mean square error for leave-one-out cross validations (RMSE for LOOCV) of PM₁ models ranged from 0.23 to 0.79 and 3.43-22.5, respectively. Further, R² and RMSE for LOOCV of PM_2.5 models ranged from 0.56 to 0.93 and 3.66-28.3, respectively. For PM₁₀ models the R² ranged from 0.31 to 0.82 and the RMSE for LOOCV ranged from 9.16 to 33.9. The generated models can be useful for population based epidemiologic studies and to estimate these pollutants in different parts of the study area for scientific decision making.

Spatial Analysis of Built Environment Risk for Respiratory Health and Its Implication for Urban Planning: A Case Study of Shanghai

Urban planning has been proven and is expected to promote public health by improving the built environment. With a focus on respiratory health, this paper explores the impact of the built environment on the incidence of lung cancer and its planning implications. While the occurrence of lung cancer is a complicated and cumulative process, it would be valuable to discover the potential risks of the built environment. Based on the data of 52,009 lung cancer cases in Shanghai, China from 2009 to 2013, this paper adopts spatial analytical methods to unravel the spatial distribution of lung cancer cases. With the assistance of geographic information system and Geo-Detector, this paper identifies certain built environments that are correlated with the distribution pattern of lung cancer cases in Shanghai, including the percentage of industrial land (which explains 28% of the cases), location factors (11%), and the percentages of cultivated land and green space (6% and 5%, respectively). Based on the quantitative study, this paper facilitates additional consideration and planning intervention measures for respiratory health such as green buffering. It is an ecological study to illustrate correlation that provides approaches for further study to unravel the causality of disease incidence and the built environment.

Assessing the transferability of landuse regression models for ultrafine particles across two Canadian cities

Land use regression (LUR) models have been increasingly used to predict intra-city variations in the concentrations of different air pollutants. However, limited research assessing the transferability of these models between cities has been published to date. In this study, LUR models were generated for Ultra-Fine Particles (UFP) (<0.1 um) using data collected from mobile monitoring campaigns in two Canadian cities, Montreal and Toronto. City-specific models were first generated for each city before the models were transferred to the second city with and without recalibration. The calibrated transferred models showed only a slight decrease in performance, with the coefficient of determination (R²), dropping from 0.49 to 0.36 for Toronto and from 0.41 to 0.38 for Montreal. Transferring models between cities with no calibration resulted in low R²; 0.11 in Toronto and 0.18 in Montreal. Moreover, two additional models were generated by combining data from the two cities. The first combined model (CM1) assumed a spatially invariant effect of the predictors, while the second (CM2) relaxed the assumption of spatial invariance for some of the model coefficients. The performance of both combined models (R² ranged between 0.41 for CM1 and 0.43 for CM2; root mean squared error (RMSE) ranged between 0.34 for CM1 and 0.33 for CM2) was found to be on par with the Toronto city-specific model and outperformed the Montreal model. The results of this study highlight that the UFP LUR models appear to support transferability of model structures between cities with similar geographical characteristics, with a minor drop in model fit and predictive skill.

Using Low-Cost Air Quality Sensor Networks to Improve the Spatial and Temporal Resolution of Concentration Maps

We present an approach to analyzing fine particulate matter (PM2.5) data from a network of "low cost air quality monitors" (LCAQM) to obtain a finely resolved concentration map. In the approach, based on a dispersion model, we first identify the probable locations of the sources, and then estimate the magnitudes of the emissions from these sources by fitting model estimates of concentrations to corresponding measurements. The emissions are then used to estimate concentrations on a grid covering the domain of interest. The residuals between model estimates at the monitor locations and the measured concentrations are then interpolated to the grid points using Kriging. We illustrate this approach by applying it to a network of 20 LCAQMs located in the Imperial Valley of Southern California. Estimating the underlying mean concentration field with a dispersion model provides a more realistic estimate of the spatial distribution of PM2.5 concentrations than that from the Kriging observations directly.

Development and performance evaluation of a GIS-based metric to assess exposure to airborne pollutant emissions from industrial sources

BACKGROUND

Dioxins are environmental and persistent organic carcinogens with endocrine disrupting properties. A positive association with several cancers, including risk of breast cancer has been suggested.

OBJECTIVES

This study aimed to develop and assess performances of an exposure metric based on a Geographic Information System (GIS) through comparison with a validated dispersion model to estimate historical industrial dioxin exposure for its use in a case-control study nested within a prospective cohort.

METHODS

Industrial dioxin sources were inventoried over the whole French territory (n > 2500) and annual average releases were estimated between 1990 and 2008. In three selected areas (rural, urban and urban-costal), dioxin dispersion was modelled using SIRANE, an urban Gaussian model and exposure of the French E3N cohort participants was estimated. The GIS-based metric was developed, calibrated and compared to SIRANE results using a set of parameters (local meteorological data, characteristics of industrial sources, e.g. emission intensity and stack height), by calculating weighted kappa statistics (wκ) and coefficient of determination (R2). Furthermore, as performance evaluation, the final GIS-based metric was tested to assess atmospheric exposure to cadmium.

RESULTS

The concordance between the GIS-based metric and the dispersion model for dioxin exposure estimate was strong (wκ median = 0.78 (1st quintile = 0.72, 3rd quintile =0.82) and R2 median = 0.82 (1st quintile = 0.71, 3rd quintile = 0.87)). We observed similar performance for cadmium.

CONCLUSIONS

Our study demonstrated the ability of the GIS-based metric to reliably characterize long-term environmental dioxin and cadmium exposures as well as the pertinence of using dispersion modelling to construct and calibrate the GIS-based metric.

Noise estimation model development using high-resolution transportation and land use regression

Noise pollution is a common phenomenon of the 21st century. Noise prediction models tend to estimate noise levels mainly from road traffic sources (such as cars, public transportation etc.). This paper describes the adoption of land use regression (LUR) modeling methodology to assess noise pollution in two periods of the day (rush hour and off-peak), in two major cities in Israel (Tel Aviv and Beer Sheva). For both rush hour and off-peak times, 20 min short term measurements were used to develop a LUR noise estimation model. We used GIS-based predictors alongside commonly used traffic predictors. The findings show good fits for our model, with rush hour "out of sample" ten folds cross-validated R² of 0.79 (Tel Aviv) and 0.52 (Beer Sheva). The Tel Aviv model performance was also tested with independent monitoring data in an adjacent city (Bat Yam), presenting a good performance as well (R² of 0.93). The findings demonstrate the viability of using a LUR approach for applying high-resolution spatial data to estimate and map noise pollution for environmental noise assessment.

Estimated health benefits of exhaust free transport in the city of Malmö, Southern Sweden

Air pollution is responsible for one in eight premature deaths worldwide, and thereby a major threat to human health. Health impact assessments of hypothetic changes in air pollution concentrations can be used as a mean of assessing the health impacts of policy, plans and projects, and support decision-makers in choices to prevent disease. The aim of this study was to estimate health impacts attributable to a hypothetical decrease in air pollution concentrations in the city of Malmö in Southern Sweden corresponding to a policy on-road transportations without tail-pipe emissions in the municipality. We used air pollution data modelled for each of the 326,092 inhabitants in Malmö by a Gaussian dispersion model combined with an emission database with >40,000 sources. The dispersion model calculates Nitrogen Oxides (NO_x) (later transformed into Nitrogen Dioxide (NO₂)) and particulate matter with an aerodynamic diameter < 2.5 μg/m³ (PM_2.5) with high spatial and temporal resolution (85 m and 1 h, respectively). The average individual reduction was 5.1 (ranging from 0.6 to 11.8) μg/m³ in NO_2, which would prevent 55 (2% of all deaths) to 93 (4%) deaths annually, depending on dose-response function used. Furthermore, we estimate that the NO₂ reduction would result in 21 (6%) fewer cases of incident asthma in children, 95 (10%) fewer children with bronchitis every year, 30 (1%) fewer hospital admissions for respiratory disease, 87(4%) fewer dementia cases, and 11(11%) fewer cases of preeclampsia every year. The average reduction in PM_2.5 of 0.6 (ranging from 0.1 till 1.7) μg/m³ would mean that 2729 (0.3%) work days would not be lost due to sick-days and that there would be 16,472 fewer restricted activity days (0.3%) that year had all on-road transportations been without tail-pipe emissions. Even though the estimates are sensitive to the dose-response functions used and to exposure misclassification errors, even the most conservative estimate of the number of prevented deaths is 7 times larger than the annual traffic fatalities in Malmö, indicating a substantial possibility to reduce the health burden attributed to tail-pipe emissions in the study area.

Modelling local uncertainty in relations between birth weight and air quality within an urban area: combining geographically weighted regression with geostatistical simulation

In this study, we combine known methods to present a new approach to assess local distributions of estimated parameters measuring associations between air quality and birth weight in the urban area of Sines (Portugal). To model exposure and capture short-distance variations in air quality, we use a Regression Kriging estimator combining air quality point data with land use auxiliary data. To assess uncertainty of exposure, the Kriging estimator is incorporated in a sequential Gaussian simulation algorithm (sGs) providing a set of simulated exposure maps with similar spatial structural dependence and statistical properties of observed data. Following the completion of the simulation runs, we fit a geographically weighted generalized linear model (GWGLM) for each mother's place of residence, using observed health data and simulated exposure data, and repeat this procedure for each simulated map. Once the fit of GWGLM with all exposure maps is finished, we take the distribution of local estimated parameters measuring associations between exposure and birth weight, thus providing a measure of uncertainty in the local estimates. Results reveal that the distribution of local parameters did not vary substantially. Combining both methods (GWGLM and sGs), however, we are able to incorporate local uncertainty on the estimated associations providing an additional tool for analysis of the impacts of place in health.

Spatiotemporal modeling of PM2.5 concentrations at the national scale combining land use regression and Bayesian maximum entropy in China

Concentrations of particulate matter with aerodynamic diameter <2.5 μm (PM_2.5) are relatively high in China. Estimation of PM_2.5 exposure is complex because PM_2.5 exhibits complex spatiotemporal patterns. To improve the validity of exposure predictions, several methods have been developed and applied worldwide. A hybrid approach combining a land use regression (LUR) model and Bayesian Maximum Entropy (BME) interpolation of the LUR space-time residuals were developed to estimate the PM_2.5 concentrations on a national scale in China. This hybrid model could potentially provide more valid predictions than a commonly-used LUR model. The LUR/BME model had good performance characteristics, with R² = 0.82 and root mean square error (RMSE) of 4.6 μg/m³. Prediction errors of the LUR/BME model were reduced by incorporating soft data accounting for data uncertainty, with the R² increasing by 6%. The performance of LUR/BME is better than OK/BME. The LUR/BME model is the most accurate fine spatial scale PM_2.5 model developed to date for China.

Full-chain health impact assessment of traffic-related air pollution and childhood asthma

BACKGROUND

Asthma is the most common chronic disease in children. Traffic-related air pollution (TRAP) may be an important exposure contributing to its development. In the UK, Bradford is a deprived city suffering from childhood asthma rates higher than national and regional averages and TRAP is of particular concern to the local communities.

AIMS

We estimated the burden of childhood asthma attributable to air pollution and specifically TRAP in Bradford. Air pollution exposures were estimated using a newly developed full-chain exposure assessment model and an existing land-use regression model (LUR).

METHODS

We estimated childhood population exposure to NO_x and, by conversion, NO₂ at the smallest census area level using a newly developed full-chain model knitting together distinct traffic (SATURN), vehicle emission (COPERT) and atmospheric dispersion (ADMS-Urban) models. We compared these estimates with measurements and estimates from ESCAPE's LUR model. Using the UK incidence rate for childhood asthma, meta-analytical exposure-response functions, and estimates from the two exposure models, we estimated annual number of asthma cases attributable to NO₂ and NO_x in Bradford, and annual number of asthma cases specifically attributable to traffic.

RESULTS

The annual average census tract levels of NO₂ and NO_x estimated using the full-chain model were 15.41 and 25.68 μg/m³, respectively. On average, 2.75 μg/m³ NO₂ and 4.59 μg/m³ NO_x were specifically contributed by traffic, without minor roads and cold starts. The annual average census tract levels of NO₂ and NO_x estimated using the LUR model were 21.93 and 35.60 μg/m³, respectively. The results indicated that up to 687 (or 38% of all) annual childhood asthma cases in Bradford may be attributable to air pollution. Up to 109 cases (6%) and 219 cases (12%) may be specifically attributable to TRAP, with and without minor roads and cold starts, respectively.

CONCLUSIONS

This is the first study undertaking full-chain health impact assessment of TRAP and childhood asthma in a disadvantaged population with public concern about TRAP. It further adds to scarce literature exploring the impact of different exposure assessments. In conservative estimates, air pollution and TRAP are estimated to cause a large, but largely preventable, childhood asthma burden. Future progress with childhood asthma requires a move beyond the prevalent disease control-based approach toward asthma prevention.

Local- and regional-scale air pollution modelling (PM10) and exposure assessment for pregnancy trimesters, infancy, and childhood to age 15 years: Avon Longitudinal Study of Parents And Children (ALSPAC)

We established air pollution modelling to study particle (PM₁₀) exposures during pregnancy and infancy (1990-1993) through childhood and adolescence up to age ~15 years (1991-2008) for the Avon Longitudinal Study of Parents And Children (ALSPAC) birth cohort. For pregnancy trimesters and infancy (birth to 6 months; 7 to 12 months) we used local (ADMS-Urban) and regional/long-range (NAME-III) air pollution models, with a model constant for local, non-anthropogenic sources. For longer exposure periods (annually and the average of birth to age ~8 and to age ~15 years to coincide with relevant follow-up clinics) we assessed spatial contrasts in local sources of PM₁₀ with a yearly-varying concentration for all background sources. We modelled PM₁₀ (μg/m³) for 36,986 address locations over 19 years and then accounted for changes in address in calculating exposures for different periods: trimesters/infancy (n = 11,929); each year of life to age ~15 (n = 10,383). Intra-subject exposure contrasts were largest between pregnancy trimesters (5^th to 95^th centile: 24.4-37.3 μg/m³) and mostly related to temporal variability in regional/long-range PM₁₀. PM₁₀ exposures fell on average by 11.6 μg/m³ from first year of life (mean concentration = 31.2 μg/m³) to age ~15 (mean = 19.6 μg/m³), and 5.4 μg/m³ between follow-up clinics (age ~8 to age ~15). Spatial contrasts in 8-year average PM₁₀ exposures (5^th to 95^th centile) were relatively low: 25.4-30.0 μg/m³ to age ~8 years and 20.7-23.9 μg/m³ from age ~8 to age ~15 years. The contribution of local sources to total PM₁₀ was 18.5%-19.5% during pregnancy and infancy, and 14.4%-17.0% for periods leading up to follow-up clinics. Main roads within the study area contributed on average ~3.0% to total PM₁₀ exposures in all periods; 9.5% of address locations were within 50 m of a main road. Exposure estimates will be used in a number of planned epidemiological studies.