Estimating morning and evening commute period O3 concentration in Taiwan using a fine spatial-temporal resolution ensemble mixed spatial model with Geo-AI technology.

Elevated levels of ground-level ozone (O3) can have harmful effects on health. While previous studies have focused mainly on daily averages and daytime patterns, it's crucial to consider the effects of air pollution during daily commutes, as this can significantly contribute to overall exposure. This study is also the first to employ an ensemble mixed spatial model (EMSM) that integrates multiple machine learning algorithms and predictor variables selected using Shapley Additive exExplanations (SHAP) values to predict spatial-temporal fluctuations in O3 concentrations across the entire island of Taiwan. We utilized geospatial-artificial intelligence (Geo-AI), incorporating kriging, land use regression (LUR), machine learning (random forest (RF), categorical boosting (CatBoost), gradient boosting (GBM), extreme gradient boosting (XGBoost), and light gradient boosting (LightGBM)), and ensemble learning techniques to develop ensemble mixed spatial models (EMSMs) for morning and evening commute periods. The EMSMs were used to estimate long-term spatiotemporal variations of O3 levels, accounting for in-situ measurements, meteorological factors, geospatial predictors, and social and seasonal influences over a 26-year period. Compared to conventional LUR-based approaches, the EMSMs improved performance by 58% for both commute periods, with high explanatory power and an adjusted R2 of 0.91. Internal and external validation procedures and verification of O3 concentrations at the upper percentile ranges (in 1%, 5%, 10%, 15%, 20%, and 25%) and other conditions (including rain, no rain, weekday, weekend, festival, and no festival) have demonstrated that the models are stable and free from overfitting issues. Estimation maps were generated to examine changes in O3 levels before and during the implementation of COVID-19 restrictions. These findings provide accurate variations of O3 levels in commute period with high spatiotemporal resolution of daily and 50m * 50m grid, which can support control pollution efforts and aid in epidemiological studies.

Frequent occurrence of respiratory symptoms in children is associated with exposure to air pollution, land use types, and parental mental health in the Greater Taipei area.

Although studies have investigated the individual effects of air pollution, land use types, and parental mental health on children's respiratory health, few studies have examined the effects of these risk factors simultaneously in children aged <2 years. We investigated the effects of exposure to air pollution, land use types surrounding residences, and parental mental health on the frequent occurrence of respiratory symptoms in children aged <2 years in the Greater Taipei area. Participants were recruited from an ongoing Taiwanese birth cohort study. We analyzed the data of the participants who had been recruited from January 2011 to April 2014 and had responded to the follow-up questionnaires at 6, 12, and 24 months. Self-administered questionnaires were used to collect participants' sociodemographic background and health, such as respiratory symptoms, and parental mental health. Pre- and postnatal pollution levels were estimated using the spatial interpolation technique (ordinary kriging) at children's residential addresses. Land use types surrounding participants' homes were evaluated by performing buffer analysis. Multiple logistic regression analyses were conducted to examine the effects of risk factors on the frequent occurrence of child respiratory symptoms in children aged 6, 12, and 24 months. We included 228, 360, and 441 children aged 6, 12, and 24 months, respectively. Our results indicated that postnatal exposure to PM2.5 and O3 was positively associated with children's respiratory symptoms. Traffic-related land-use types, sports facilities, and commercial land surrounding homes exerted adverse effects on children's respiratory symptoms, whereas the presence of schools in the neighborhood was beneficial. Parental mental health was also associated with children's respiratory symptoms. Postnatal exposure to air pollution and land use types surrounding residences were associated with respiratory health in children aged <2 years. The residential environment is a critical factor affecting children's respiratory health of children aged <2 years.

Effects of air pollution, land-use type, and maternal mental health on child development in the first two years of life in the Greater Taipei area.

INTRODUCTION: Few studies have investigated the associations of child development with air pollution, land-use type, and maternal mental health simultaneously. Therefore, we evaluated the effect of exposure to air pollutants during several critical periods of life, with adjustment for land-use type and maternal mental status, on child development at 6, 12, and 24 months of age in the Greater Taipei area. METHODS: Participants were selected from an ongoing Taiwanese birth cohort study. We analyzed the data of the participants who had been recruited from January 2011 to April 2014. Self-administered standardized questionnaires were used to collect information on sociodemographic factors, infant development and health, maternal mental status, etc. Air pollution levels in pre- and postnatal periods were estimated using a spatial interpolation technique (ordinary kriging) at children's residential addresses. Land-use types around participants' homes were evaluated using buffer analysis. We used multiple logistic regression analysis to examine the relationships between child development delay and environmental factors. RESULTS: In total, 228, 361, and 441 families completed child development forms at 6, 12, and 24 months of age, respectively. Our results indicated that prenatal exposure to particulate matter with aerodynamic diameter ≤10 µm and O3 and postnatal exposure to NO2 were negatively associated with child development. Traffic-related land-use types, gas stations, and power generation areas around participants' homes were also adversely correlated with child development. Moreover, poor maternal mental health was associated with child development delay. CONCLUSION: Prenatal exposure and postnatal exposure to air pollution were associated with development delay in children under 2 years of age, specifically those under 1 year of age, even after adjustment for land-use type and maternal mental status. Living environment is critical for the development of children under 2 years of age.

Auramine O, an incense smoke ingredient, promotes lung cancer malignancy.

Burning incense to worship deities is a popular religious ritual in large parts of Asia, and is a popular custom affecting more than 1.5 billion adherents. Due to incomplete combustion, burning incense has been well recognized to generate airborne hazards to human health. However, the correlation between burning incense and lung cancer in epidemiological studies remains controversy. Therefore, we speculated that some unknown materials in incense smoke are involved in the initiation or progression of lung cancer. Based on this hypothesis, we identified a major compound auramine O (AuO) from the water-soluble fraction of incense burned condensate using mass spectrometry. AuO is commonly used in incense manufacture as a colorant. Due to thermostable, AuO released from burned incenses becomes an unexpected air pollutant. AuO is classified as a Group 2B chemical by the International Agency of Research on Cancer (IARC), however, the damage of AuO to the respiratory system remains elusive. Our study revealed that AuO has no apparent effect on malignant transformation; but, it dramatically promotes lung cancer malignancy. AuO accumulates in the nucleus and induces the autophagy activity in lung tumor cells. AuO significantly enhances migration and invasive abilities and the in vitro and in vivo stemness features of lung tumor cells through activating the expression of aldehyde dehydrogenase family 1 member A1 (ALDH1A1), and ALDH1A1 knockdown attenuates AuO-induced autophagy activity and blocks AuO-induced lung tumor malignancy. In conclusion, we found that AuO, an ingredient of incense smoke, significantly increases the metastatic abilities and stemness characters of lung tumor cells through the activation of ALDH1A1, which is known to be associated with poor outcome and progression of lung cancer. For public health, reducing or avoiding the use of AuO in incense is recommended.

Explainable geospatial-artificial intelligence models for the estimation of PM2.5 concentration variation during commuting rush hours in Taiwan.

PM2.5 concentrations are higher during rush hours at background stations compared to the average concentration across these stations. Few studies have investigated PM2.5 concentration and its spatial distribution during rush hours using machine learning models. This study employs a geospatial-artificial intelligence (Geo-AI) prediction model to estimate the spatial and temporal variations of PM2.5 concentrations during morning and dusk rush hours in Taiwan. Mean hourly PM2.5 measurements were collected from 2006 to 2020, and aggregated into morning (7 a.m.-9 a.m.) and dusk (4 p.m.-6 p.m.) rush-hour mean concentrations. The Geo-AI prediction model was generated by integrating kriging interpolation, land-use regression, machine learning, and a stacking ensemble approach. A forward stepwise variable selection method based on the SHapley Additive exPlanations (SHAP) index was used to identify the most influential variables. The performance of the Geo-AI models for morning and dusk rush hours had accuracy scores of 0.95 and 0.93, respectively and these results were validated, indicating robust model performance. Spatially, PM2.5 concentrations were higher in southwestern Taiwan for morning rush hours, and suburban areas for dusk rush hours. Key predictors included kriged PM2.5 values, SO2 concentrations, forest density, and the distance to incinerators for both morning and dusk rush hours. These PM2.5 estimates for morning and dusk rush hours can support the development of alternative commuting routes with lower concentrations.

Using a land use regression model with machine learning to estimate ground level PM2.5.

Ambient fine particulate matter (PM2.5) has been ranked as the sixth leading risk factor globally for death and disability. Modelling methods based on having access to a limited number of monitor stations are required for capturing PM2.5 spatial and temporal continuous variations with a sufficient resolution. This study utilized a land use regression (LUR) model with machine learning to assess the spatial-temporal variability of PM2.5. Daily average PM2.5 data was collected from 73 fixed air quality monitoring stations that belonged to the Taiwan EPA on the main island of Taiwan. Nearly 280,000 observations from 2006 to 2016 were used for the analysis. Several datasets were collected to determine spatial predictor variables, including the EPA environmental resources dataset, a meteorological dataset, a land-use inventory, a landmark dataset, a digital road network map, a digital terrain model, MODIS Normalized Difference Vegetation Index (NDVI) database, and a power plant distribution dataset. First, conventional LUR and Hybrid Kriging-LUR were utilized to identify the important predictor variables. Then, deep neural network, random forest, and XGBoost algorithms were used to fit the prediction model based on the variables selected by the LUR models. Data splitting, 10-fold cross validation, external data verification, and seasonal-based and county-based validation methods were used to verify the robustness of the developed models. The results demonstrated that the proposed conventional LUR and Hybrid Kriging-LUR models captured 58% and 89% of PM2.5 variations, respectively. When XGBoost algorithm was incorporated, the explanatory power of the models increased to 73% and 94%, respectively. The Hybrid Kriging-LUR with XGBoost algorithm outperformed the other integrated methods. This study demonstrates the value of combining Hybrid Kriging-LUR model and an XGBoost algorithm for estimating the spatial-temporal variability of PM2.5 exposures.

Micro-scale particle simulation and traffic-related particle exposure assessment in an Asian residential community.

Conducting studies on sharp particulate matter (PM) gradients in Asian residential communities is difficult due to their complex building arrangements and various emission sources, particularly road traffic. In this study, a synthetic methodology, combining numerical simulations and minor field observations, was set up to investigate the dispersion of traffic-related PM in a typical Asian residential community and its contribution to PM exposure. A Lagrangian particle model (GRAL) was applied to estimate the spatiotemporal variation of the traffic-related PM increments within the community. A detailed topography dataset with 5 m horizontal resolution was used to simulate a micro-scale flow field. The model performance was comprehensively validated using both in-situ and mobile observations. The coefficient of determination (R2) of the simulated vs. observed PM2.5 reached 0.81 by an artery road, and 0.85 in alleys without significant road traffic. The maximum increments of kerbside PM exposure concentration contributed by road traffic during rush hour were found to be 38% (PM10) and 40% (PM2.5). This synthetic method was used to assess the impact of synoptic wind and canyon orientation on residents' PM2.5 exposure related to traffic exhaust. Perfect exponential decay curves of traffic-related PM2.5 were found within canyons. The decrease of road-traffic PM2.5 on five different floor levels, compared with that on kerbside levels, ranged between 42% and 100%. The results demonstrated that in complex Asian communities, Lagrangian particle models such as GRAL can simulate the spatial distribution of PM10 and PM2.5 and assess the residents' outdoor exposure.

PM2.5 exposure of various microenvironments in a community: Characteristics and applications.

While the measurement of particulate matter (PM) with a diameter of less than 2.5 µm (PM2.5) has been conducted for personal exposure assessment, it remains unclear how models that integrate microenvironmental levels with resolved activity and location information predict personal exposure to PM. We comprehensively investigated PM2.5 concentrations in various microenvironments and estimated personal exposure stratified by the microenvironment. A variety of microenvironments (>200 places and locations, divided into 23 components according to indoor, outdoor, and transit modes) in a community were selected to characterize PM2.5 concentrations. Infiltration factors calculated from microenvironmental/central-site station (M/S) monitoring campaigns with time-activity patterns were used to estimate time-weighted exposure to PM2.5 for university students. We evaluated exposures using a four-stage modeling approach and quantified the performance of each component. It was found that the SidePak monitor overestimated the concentration by 3.5 times as compared with the filter-based measurements. Higher mean concentrations of PM2.5 were observed in the Taoist temple and night market microenvironments; in contrast, lower concentrations were observed in air-conditioned offices and car microenvironments. While the exposure model incorporating detailed time-location information and infiltration factors achieved the highest prediction (R2 = 0.49) of personal exposure to PM2.5, the use of indoor, outdoor, and transit components for modeling also generated a consistent result (R2 = 0.44).

A hybrid kriging/land-use regression model with Asian culture-specific sources to assess NO2 spatial-temporal variations.

Kriging interpolation and land use regression (LUR) have characterized the spatial variability of long-term nitrogen dioxide (NO2), but there has been little research on combining these two methods to capture small-scale spatial variation. Furthermore, studies predicting NO2 exposure are almost exclusively based on traffic-related variables, which may not be transferable to Taiwan, a typical Asian country with diverse local emission sources, where densely distributed temples and restaurants may be important for NO2 levels. To advance the exposure estimates in Taiwan, a hybrid kriging/LUR model incorporates culture-specific sources as potential predictors. Based on 14-year NO2 observations from 73 monitoring stations across Taiwan, a set of interpolated NO2 values were generated through a leave-one-out ordinary kriging algorithm, and this was included as an explanatory variable in the stepwise LUR procedures. Kriging interpolated NO2 and culture-specific predictors were entered in the final models, which captured 90% and 87% of NO2 variation in annual and monthly resolution, respectively. Results from 10-fold cross-validation and external data verification demonstrate robust performance of the developed models. This study demonstrates the value of incorporating the kriging-interpolated estimates and culture-specific emission sources into the traditional LUR model structure for predicting NO2, which can be particularly useful for Asian countries.

Variations of Cd/Pb and Zn/Pb ratios in Taipei aerosols reflecting long-range transport or local pollution emissions.

Along with windblown dust, large quantities of pollutants are annually brought out of continental China by the westerlies in winter and spring; thereafter, they are partly subjected to transport by northeastern monsoon winds to Taiwan. To characterize the heavy metal composition differences between long-range transported and local aerosols and to evaluate metal contributions from long-range transported aerosols during the northeastern monsoon season, both PM(10) and PM(2.5) aerosols collected from Taipei, Taiwan from February 2002 to March 2003 were analyzed for three selected heavy metals, namely Pb, Cd and Zn using ICP-MS. Monthly patterns show that Pb concentrations in winter (62 ng/m(3)) were over two times higher than those in the other seasons, which is attributed to long-range transport from areas under development in China. Low Cd/Pb (0.017) and Zn/Pb (1.82) ratios were measured in aerosols collected during the Asian dust period, in which the ambient aerosols consisted predominantly of long-range transported pollutants. By contrast, high Cd/Pb (0.030) and Zn/Pb (3.44) ratios were observed during the summer monsoon season, in which aerosols were dominated by local pollutant emissions. Cd/Pb and Zn/Pb ratios appear to be successfully applied to identify the pollutants originating principally from the long-range transport or from local emissions. In addition, by assuming that a significant fraction of heavy metals associated with coarse airborne dust have settled to the sea prior to reaching Taiwan in spring, a mechanism is suggested to explain why higher anthropogenic metal concentrations occurred in winter than those in dust-rich spring.