Critical Review of Methods to Estimate PM2.5 Concentrations within Specified Research Region

Do Machine Learning Methods Improve Prediction of Ambient Air Pollutants with High Spatial Contrast? A Systematic Review

BACKGROUND & OBJECTIVE

The use of machine learning for air pollution modelling is rapidly increasing. We conducted a systematic review of studies comparing statistical and machine learning models predicting the spatiotemporal variation of ambient nitrogen dioxide (NO2), ultrafine particles (UFPs) and black carbon (BC) to determine whether and in which scenarios machine learning generates more accurate predictions.

METHODS

Web of Science and Scopus were searched up to June 13, 2024. All records were screened by two independent reviewers. Differences in the coefficient of determination (R2) and Root Mean Square Error (RMSE) between best statistical and machine learning methods were compared across categories of methodological elements.

RESULTS

A total of 38 studies with 46 model comparisons (30 for NO2, 8 for UFPs and 8 for BC) were included. Linear non-regularized methods and Random Forest were most frequently used. Machine learning outperformed statistical models in 34 comparisons. Mean differences (95% confidence intervals) in R2 and RMSE between best machine learning and statistical models were 0.12 (0.08, 0.17) and 20% (11%, 29%) respectively. Tree-based methods performed best in 12 of 17 multi-model comparisons. Nonlinear or regularization regression methods were used in only 12 comparisons and provided similar performance to machine learning methods.

CONCLUSION

This systematic review suggests that machine learning methods, especially tree-based methods, may be superior to linear non-regularized methods for predicting ambient concentrations of NO2, UFPs and BC. Additional comparison studies using nonlinear, regularized and a wider array of machine learning methods are needed to confirm their relative performance. Future air pollution studies would also benefit from more explicit and standardized reporting of methodologies and results.

Air Pollution (PM2.5) Negatively Affects Urban Livability in South Korea and China

This study investigated the effect of the concentration of ambient fine particulate matter (PM2.5), a transboundary air pollutant, on the livability of neighboring areas of China and South Korea with the aim of informing common policy development. Grey relational analysis (GRA) and panel regression analysis were performed to examine the effect of PM2.5 concentration on various livability indicators. The results revealed that urban living infrastructure was an indicator of effect in both South Korea and China. Based on the high correlation between urban living infrastructure and PM2.5 concentration, it can be seen that PM2.5 clearly affects livability, shown by panel regression analysis. Other key livability indicators were traffic safety, culture and leisure, and climate indicators. Spatial analysis of the livability index revealed that from 2015 to 2019, livability improved in both South Korea and China, but there was a clear difference in the spatial distribution in China. High-vulnerability areas showed potential risks that can reduce livability in the long run. In South Korea and China, areas surrounding large cities were found to be highly vulnerable. The findings of this research can guide the establishment of policies grading PM2.5 pollution at the regional or city macro-level.

Effects of Urban Vibrancy on an Urban Eco-Environment: Case Study on Wuhan City

In the context of rapid urbanisation and an emerging need for a healthy urban environment, revitalising urban spaces and its effects on the urban eco-environment in Chinese cities have attracted widespread attention. This study assessed urban vibrancy from the dimensions of density, accessibility, liveability, diversity, and human activity, with various indicators using an adjusted spatial TOPSIS (technique for order preference by similarity to an ideal solution) method. The study also explored the effects of urban vibrancy on the urban eco-environment by interpreting PM 2.5 and land surface temperature using “big” and “dynamic” data, such as those from mobile and social network data. Thereafter, spatial modelling was performed to investigate the influence of urban vibrancy on air pollution and temperature with inverted and extracted remote sensing data. This process identified spatial heterogeneity and spatial autocorrelation. The majority of the dimensions, such as density, accessibility, liveability, and diversity, are negatively correlated with PM 2.5, thereby indicating that the advancement of urban vibrancy in these dimensions potentially improves air quality. Conversely, improved accessibility increases the surface temperature in most of the districts, and large-scale infrastructure construction generally contributes to the increase. Diversity and human activity appear to have a cooling effect. In the future, applying spatial heterogeneity is advised to assess urban vibrancy and its effect on the urban eco-environment, to provide valuable references for spatial urban planning, improve public health and human wellbeing, and ensure sustainable urban development.

Interpolation-Based Fusion of Sentinel-5P, SRTM, and Regulatory-Grade Ground Stations Data for Producing Spatially Continuous Maps of PM2.5 Concentrations Nationwide over Thailand

Atmospheric pollution has recently drawn significant attention due to its proven adverse effects on public health and the environment. This concern has been aggravated specifically in Southeast Asia due to increasing vehicular use, industrial activity, and agricultural burning practices. Consequently, elevated PM2.5 concentrations have become a matter of intervention for national authorities who have addressed the needs of monitoring air pollution by operating ground stations. However, their spatial coverage is limited and the installation and maintenance are costly. Therefore, alternative approaches are necessary at national and regional scales. In the current paper, we investigated interpolation models to fuse PM2.5 measurements from ground stations and satellite data in an attempt to produce spatially continuous maps of PM2.5 nationwide over Thailand. Four approaches are compared, namely the inverse distance weighted (IDW), ordinary kriging (OK), random forest (RF), and random forest combined with OK (RFK) leveraging on the NO2, SO2, CO, HCHO, AI, and O3 products from the Sentinel-5P satellite, regulatory-grade ground PM2.5 measurements, and topographic parameters. The results suggest that RFK is the most robust, especially when the pollution levels are moderate or extreme, achieving an RMSE value of 7.11 μg/m3 and an R2 value of 0.77 during a 10-day long period in February, and an RMSE of 10.77 μg/m3 and R2 and 0.91 during the entire month of March. The proposed approach can be adopted operationally and expanded by leveraging regulatory-grade stations, low-cost sensors, as well as upcoming satellite missions such as the GEMS and the Sentinel-5.

Correlation between Exposure to Fine Particulate Matter (PM2.5) during Pregnancy and Congenital Anomalies: Its Surgical Perspectives

BACKGROUND

Fine particulate matter (PM_2.5) can easily penetrate blood vessels and tissues through the human respiratory tract and cause various health problems. Some studies reported that particular matter (PM) exposure during pregnancy is associated with low birth weight or congenital cardiovascular anomalies. This study aimed to investigate the correlation between the degree of exposure to PM ≤ 2.5 μm (PM_2.5) during pregnancy and congenital anomalies relevant to the field of pediatric surgery.

METHODS

Mother-infant dyads with registered addresses in the Metropolitan City were selected during 3 years. The electronic medical records of mothers and neonates were retrospectively analyzed, with a focus on maternal age at delivery, date of delivery, gestation week, presence of diabetes mellitus (DM) or hypertension, parity, the residence of the mother and infant, infant sex, birth weight, Apgar score, and presence of congenital anomaly. The monthly PM_2.5 concentration from the first month of pregnancy to the delivery was computed based on the mothers' residences.

RESULTS

PM_2.5 exposure concentration in the second trimester was higher in the congenital anomaly group than in the non-congenital anomaly group (24.82 ± 4.78 µg/m³, P = 0.023). PM_2.5 exposure concentration did not affect the incidence of nervous, cardiovascular, and gastrointestinal anomalies. While statistically insignificant, the groups with nervous, cardiovascular, gastrointestinal, musculoskeletal, and other congenital anomalies were exposed to higher PM_2.5 concentrations in the first trimester compared with their respective counterparts. The effect of PM_2.5 concentration on the incidence of congenital anomalies was significant even after adjusting for the mother's age, presence of DM, hypertension, and parity. The incidence of congenital anomalies increased by 26.0% (95% confidence interval of 4.3% and 49.2%) per 7.23 µg/m³ elevation of PM_2.5 interquartile range in the second trimester.

CONCLUSIONS

The congenital anomaly group was exposed to a higher PM_2.5 concentration in the second trimester than the non-congenital anomaly group. The PM_2.5 exposure concentration level in the first trimester tended to be higher in groups with anomalies than those without anomalies. This suggests that continuous exposure to a high PM_2.5 concentration during pregnancy influences the incidence of neonatal anomalies in surgical respects.

Satellite Monitoring for Air Quality and Health

Data from satellite instruments provide estimates of gas and particle levels relevant to human health, even pollutants invisible to the human eye. However, the successful interpretation of satellite data requires an understanding of how satellites relate to other data sources, as well as factors affecting their application to health challenges. Drawing from the expertise and experience of the 2016-2020 NASA HAQAST (Health and Air Quality Applied Sciences Team), we present a review of satellite data for air quality and health applications. We include a discussion of satellite data for epidemiological studies and health impact assessments, as well as the use of satellite data to evaluate air quality trends, support air quality regulation, characterize smoke from wildfires, and quantify emission sources. The primary advantage of satellite data compared to in situ measurements, e.g., from air quality monitoring stations, is their spatial coverage. Satellite data can reveal where pollution levels are highest around the world, how levels have changed over daily to decadal periods, and where pollutants are transported from urban to global scales. To date, air quality and health applications have primarily utilized satellite observations and satellite-derived products relevant to near-surface particulate matter <2.5 μm in diameter (PM_2.5) and nitrogen dioxide (NO₂). Health and air quality communities have grown increasingly engaged in the use of satellite data, and this trend is expected to continue. From health researchers to air quality managers, and from global applications to community impacts, satellite data are transforming the way air pollution exposure is evaluated.

Spatial regression modelling of particulate pollution in Calgary, Canada

The study presents a spatial analysis of particulate pollution, which includes not only particulate matter, but also black carbon, a pollutant of growing concern for human health. We developed land use regression (LUR) models for two particulate matter size fractions, PM_2.5 and PM₁₀, and for δC, an index calculated from black carbon (BC)-a component of PM_2.5-which indicates the portion of organic versus elemental BC. LUR models were estimated over Calgary (Canada) for summer 2015 and winter 2016. As all samples exhibited significant spatial autocorrelation, spatial autoregressive lag (SARlag) and error (SARerr) models were computed. SARlag models were preferred for all pollutants in both seasons, and yielded goodness of fit aligned with or higher than values reported in the literature. LUR models yielded consistent sets of predictors, representing industrial activities, traffic, and elevation. The obtained model coefficients were then combined with local land use variables to compute fine-scale concentration predictions over the entire city. The predicted concentrations were slightly lower and less dispersed than the observed ones. Consistent with observed pollution records, prediction maps exhibited higher concentration over the road network, industrial areas, and the eastern quadrants of the city. Lastly, results of a corresponding study of PM in summer 2010 and winter 2011 were considered. While the small size of the 2010-2011 sample hampered a multi-temporal analysis, we cautiously note comparable seasonal patterns and consistent association with land use variables for both PM fine fractions over the 5-year interval.

A Framework to Predict High-Resolution Spatiotemporal PM2.5 Distributions Using a Deep-Learning Model: A Case Study of Shijiazhuang, China

Air-borne particulate matter, PM2.5 (PM having a diameter of less than 2.5 micrometers), has aroused widespread concern and is a core indicator of severe air pollution in many cities globally. In our study, we present a validated framework to predict the daily PM2.5 distributions, exemplified by a use case of Shijiazhuang City, China, based on daily aerosol optical depth (AOD) datasets. The framework involves obtaining the high-resolution spatiotemporal AOD distributions, estimation of the spatial distributions of PM2.5 and the prediction of these based on a convolutional long short-term memory (ConvLSTM) model. In the estimation part, the eXtreme gradient boosting (XGBoost) model has been determined as the estimation model with the lowest root mean square error (RMSE) of 32.86 µg/m3 and the highest coefficient of determination regression score function (R2) of 0.71, compared to other common models used as a baseline for comparison (linear, ridge, least absolute shrinkage and selection operator (LASSO) and cubist). For the prediction part, after validation and comparison with a seasonal autoregressive integrated moving average (SARIMA), which is a traditional time-series prediction model, in both time and space, the ConvLSTM gives a more accurate performance for the prediction, with a total average prediction RMSE of 14.94 µg/m3 compared to SARIMA’s 17.41 µg/m3. Furthermore, ConvLSTM is more stable and with less fluctuations for the prediction of PM2.5 in time, and it can also eliminate better the spatial predicted errors compared to SARIMA.

Potential Approach for Single-Peak Extinction Fitting of Aerosol Profiles Based on In Situ Measurements for the Improvement of Surface PM2.5 Retrieval from Satellite AOD Product

The vertical distribution of aerosols is important for accurate surface PM2.5 retrieval and initial modeling forecasts of air pollution, but the observation of aerosol profiles on the regional scale is usually limited. Therefore, in this study, an approach to aerosol extinction profile fitting is proposed to improve surface PM2.5 retrieval from satellite observations. Owing to the high similarity of the single-peak extinction profile in the distribution pattern, the log-normal distribution is explored for the fitting model based on a decadal dataset (3248 in total) from Micro Pulse LiDAR (MPL) measurements. The logarithmic mean, standard deviation, and the height of peak extinction near-surface (Mode) are manually derived as the references for model construction. Considering the seasonal impacts on the planetary boundary layer height (PBLH), Mode, and the height of the surface layer, the extinction profile is then constructed in terms of the planetary boundary layer height (PBLH) and the total column aerosol optical depth (AOD). A comparison between fitted profiles and in situ measurements showed a high level of consistency in terms of the correlation coefficient (0.8973) and root-mean-square error (0.0415). The satellite AOD is subsequently applied for three-dimensional aerosol extinction, and the good agreement of the extinction coefficient with the PM2.5 within the surface layer indicates the good performance of the proposed fitting approach and the potential of satellite observations for providing accurate PM2.5 data on a regional scale.

Application of an advanced spatiotemporal model for PM2.5 prediction in Jiangsu Province, China

Either long- or short-term of fine particle (PM_2.5) exposure is associated with adverse health effects especially for children. Primary school students spend much time in schools whereas PM_2.5 prediction for such fine-scale places remains a demanding task, let alone a combined prediction with high temporal resolution. The study aimed to estimate PM_2.5 levels of different time scales for primary schools in Jiangsu Province, China. Hourly PM_2.5 measurements within the academic year (Sept. 2016-June 2017) were collected from 72 routine monitoring sites. Together with PM_2.5 emission inventory and dozens of geographic variables, an advanced spatiotemporal land use regression (LUR) model was employed to estimate PM_2.5 concentrations of biweekly, seasonal and academic year levels in Jiangsu Province at 2457 primary school locations. 10-fold cross-validation verified high prediction ability with squared correlations R_CV² of 0.72 for temporal and 0.71 for spatial changes. PM_2.5 levels in primary schools in Nanjing and Nantong were >10% higher than that of the corresponding cities while pollution levels in primary schools in Xuzhou were >20% lower. For 10 out of the 13 cities in Jiangsu, PM_2.5 levels for primary schools surpassed 70 μg/m³ in winter. Schools in Lianyungang, Zhenjiang and Huai'an suffered the most. This study demonstrated the fine-scale prediction ability of the novel spatiotemporal LUR model, as well as the potential and necessity to apply it in epidemiological studies. It also verified the emergency of pollution control for primary schools from cities such as Lianyungang and Zhenjiang.

Differences in fine particle chemical composition on clear and cloudy days

Clouds are prevalent and alter PM_2.5 mass and chemical composition. Cloud-affected satellite retrievals are often removed from data products, hindering estimates of tropospheric chemical composition during cloudy times. We examine surface fine particulate matter (PM_2.5) chemical constituent concentrations in the Interagency Monitoring of PROtected Visual Environments network during Cloudy and Clear Sky times defined using Moderate Resolution Imaging Spectroradiometer (MODIS) cloud flags from 2010-2014 with a focus on differences in particle hygroscopicity and aerosol liquid water (ALW). Cloudy and Clear Sky periods exhibit significant differences in PM_2.5 and chemical composition that vary regionally and seasonally. In the eastern US, relative humidity alone cannot explain differences in ALW, suggesting emissions and in situ chemistry exert determining impacts. An implicit clear sky bias may hinder efforts to quantitatively to understand and improve model representation of aerosol-cloud interactions.

PM2.5 Pollutant in Asia-A Comparison of Metropolis Cities in Indonesia and Taiwan

Air pollution has emerged as a significant health, environmental, economic, and social problem all over the world. In this study, geospatial technologies coupled with a LUR (Land Use Regression) approach were applied to assess the spatial-temporal distribution of fine particulate (PM_2.5). In-situ observations of air pollutants from ground monitoring stations from 2016-2018 were used as dependent variables, while the land-use/land cover, a NDVI (Normalized Difference Vegetation Index) from a MODIS sensors, and meteorology data allocations surrounding the monitoring stations from 0.25-5 km buffer ranges were collected as spatial predictors from GIS and remote sensing databases. A linear regression method was developed for the LUR model and 10-fold cross-validation was used to assess the model robustness. The R² model obtained was 56% for DKI Jakarta, Indonesia, and 83% for Taipei Metropolis, Taiwan. According to the results of the PM_2.5 model, the essential predictors for DKI Jakarta were influenced by temperature, NDVI, humidity, and residential area, while those for the Taipei Metropolis region were influenced by PM₁₀, NO₂, SO₂, UV, rainfall, spring, main road, railroad, airport, proximity to airports, mining areas, and NDVI. The validation of the results of the estimated PM_2.5 distribution use 10-cross validation with indicated R² values of 0.62 for DKI Jakarta and 0.84 for Taipei Metropolis. The results of cross-validation show the strength of the model.

Methods, availability, and applications of PM2.5 exposure estimates derived from ground measurements, satellite, and atmospheric models

Fine particulate matter (PM2.5) is a well-established risk factor for public health. To support both health risk assessment and epidemiological studies, data are needed on spatial and temporal patterns of PM2.5 exposures. This review article surveys publicly available exposure datasets for surface PM2.5 mass concentrations over the contiguous U.S., summarizes their applications and limitations, and provides suggestions on future research needs. The complex landscape of satellite instruments, model capabilities, monitor networks, and data synthesis methods offers opportunities for research development, but would benefit from guidance for new users. Guidance is provided to access publicly available PM2.5 datasets, to explain and compare different approaches for dataset generation, and to identify sources of uncertainties associated with various types of datasets. Three main sources used to create PM2.5 exposure data are ground-based measurements (especially regulatory monitoring), satellite retrievals (especially aerosol optical depth, AOD), and atmospheric chemistry models. We find inconsistencies among several publicly available PM2.5 estimates, highlighting uncertainties in the exposure datasets that are often overlooked in health effects analyses. Major differences among PM2.5 estimates emerge from the choice of data (ground-based, satellite, and/or model), the spatiotemporal resolutions, and the algorithms used to fuse data sources.Implications: Fine particulate matter (PM2.5) has large impacts on human morbidity and mortality. Even though the methods for generating the PM2.5 exposure estimates have been significantly improved in recent years, there is a lack of review articles that document PM2.5 exposure datasets that are publicly available and easily accessible by the health and air quality communities. In this article, we discuss the main methods that generate PM2.5 data, compare several publicly available datasets, and show the applications of various data fusion approaches. Guidance to access and critique these datasets are provided for stakeholders in public health sectors.

Integration of Remote Sensing and Social Sensing Data in a Deep Learning Framework for Hourly Urban PM2.5 Mapping

Fine spatiotemporal mapping of PM2.5 concentration in urban areas is of great significance in epidemiologic research. However, both the diversity and the complex nonlinear relationships of PM2.5 influencing factors pose challenges for accurate mapping. To address these issues, we innovatively combined social sensing data with remote sensing data and other auxiliary variables, which can bring both natural and social factors into the modeling; meanwhile, we used a deep learning method to learn the nonlinear relationships. The geospatial analysis methods were applied to realize effective feature extraction of the social sensing data and a grid matching process was carried out to integrate the spatiotemporal multi-source heterogeneous data. Based on this research strategy, we finally generated hourly PM2.5 concentration data at a spatial resolution of 0.01°. This method was successfully applied to the central urban area of Wuhan in China, which the optimal result of the 10-fold cross-validation R2 was 0.832. Our work indicated that the real-time check-in and traffic index variables can improve both quantitative and mapping results. The mapping results could be potentially applied for urban environmental monitoring, pollution exposure assessment, and health risk research.

Large-Scale, Fine-Grained, Spatial, and Temporal Analysis, and Prediction of Mobile Phone Users' Distributions Based upon a Convolution Long Short-Term Model

Accurate and timely estimations of large-scale population distributions are a valuable input for social geography and economic research and for policy-making. The most popular large-scale method to calculate such estimations uses mobile phone data. We propose a novel method, firstly based upon using a kernel density estimation (KDE) to estimate dynamic mobile phone users' distributions at a two-hourly scale temporal resolution. Secondly, a convolutional long short-term memory (ConvLSTM) model was used in our study to predict mobile phone users' spatial and temporal distributions for the first time at such a fine-grained temporal resolution. The evaluation results show that the predicted people's mobility derived from the mobile phone users' density correlates much better with the actual density, both temporally and spatially, as compared to traditional methods such as time-series prediction, autoregressive moving average model (ARMA), and LSTM.