Design and testing of a low-cost sensor and sampling platform for indoor air quality

Optimal Methodology to Generate Road Traffic Emissions for Air Quality Modeling

In India, the sparse air quality monitoring infrastructure beyond urban centers limits our understanding of fine particulate matter (PM2.5) spatiotemporal variability, exposure dynamics, and their health and ecological impacts. To address this challenge, the Ambient Air Quality Monitoring in Rural Areas using Indigenous Technology (AMRIT) project was launched, deploying 1,400 low-cost air quality sensors (LCS) across diverse demographics and land-use settings in the states of Uttar Pradesh and Bihar, within the Indo-Gangetic plains (IGP). Additionally, the Dynamic Hyper-local Source Apportionment (DHSA) project was initiated to explore the potential of LCS for real-time source apportionment, facilitating pollution source identification and providing policymakers with critical data for effective air pollution control. Together, these initiatives aim to capture regional pollution dynamics and impacts beyond urban centers for the first time in India.Our investigation showed that the AMRIT-LCS network effectively captured fine-scale pollution patterns across urban and rural areas of Bihar, with a mean concentration of 135 µg/m³. Rural areas exhibited 6-8% higher pollution levels than urban settings. Within urban areas, pollution levels were notably higher in suburban and small-town regions, highlighting the need for targeted pollution management in these emerging cities. Leveraging machine learning (ML) frameworks, the village-level exposure assessment revealed significant spatiotemporal heterogeneity in PM2.5 exposure, with consistently high levels in the northwestern districts and a mortality rate (per lakh) of 50 (95% CI: 33–64) in the region. Data-driven techniques were applied to delineate air sheds for future air quality management. Five distinct air-sheds were identified, with three (in the north) and two (in the south) of the Ganges River, flowing west to east and dividing Bihar into two unequal parts. Consistent with the exposure model, the air-sheds over the northwestern and north-central regions were more polluted. Additionally, the LCS network was used to identify pollution hotspots by leveraging satellite-driven datasets within a deep learning framework, across various temporal scales.These findings underscore the critical need for high-density monitoring networks to capture local variability and generate spatially resolved estimates. Under the DHSA project, ML techniques enabled real-time, accurate predictions of pollution source contributions using LCS datasets, with a mean absolute error of less than 5%. This approach strengthens the potential for deploying widespread LCS networks capable of high-resolution source apportionment with spatiotemporal precision. The data-driven approach provides a solid foundation for designing decentralized mitigation strategies and comprehensive policy frameworks aimed at improving air quality and public health in the IGP. Thus, the portable LCS technology offers a promising alternative to conventional air quality monitoring methods, presenting a viable solution not only for India but also for data-sparse regions globally, contributing to the achievement of sustainable development goals.Keywords: PM2.5, Low-Cost Sensors, Indo-Gangetic Plains, Machine Learning, Exposure Assessment, Urban-Rural Disparity, Airsheds, Hotspot Analysis, Source Apportionment

The first case of COVID-19 in Brazil was registered in the city of São Paulo on February 26, 2020; however, restrictive measures and social distancing were only determined in the city on March 17, 2020. A partial lockdown aimed to mitigate the advance of the virus by raising the social isolation rates, by limiting the operation of several services and the mobility of the population. Thus, this study aims to analyze the relationship between the social isolation index in the city of São Paulo and the emission levels of the main air pollutants (PM10, PM2.5, NOx, NO, NO2, SO2 and CO), as well as air temperature. We analyzed the data collected from three urban air quality monitoring stations located in the city center of São Paulo from March 16, 2020 to July 20, 2020. The data for 2020 were compared with those of the previous period in 2020 and the same period in the previous 5 years (2015–2019), and also to the city’s official indices of social isolation. The relationships between pollutant concentrations and the social isolation index showed that the decrease in mobility influenced the reduction in air pollution. Pollutants NO2, NOx, NO and CO had the strongest negative associations (Pearson’s correlation = − 0.582; 0.481; − 0.433 and − 0.367, respectively). Our results showed that the partial lockdown (from March 17, 2020, to July 20, 2020) had a positive impact on air quality, with a reduction in the emission of pollutants NO (31.75%), NO2 (20.60%), NOx (27.21%) and CO (29.95%). The greatest reductions in the emission of pollutants were observed when the social isolation index reached an average of 52.20%. Small negative fluctuations in the social isolation index broke the most significant reductions observed at the beginning of social isolation.Graphic abstract

Fine chemical industrial park (FCIP) is a major source of atmospheric pollutants in China. A long-term high spatial resolution monitoring campaign on air pollutants had been firstly conducted in a major FCIP in Yangtze River Delta (YRD) from December 2019 to November 2020. The grid-based monitoring platform consisting of 30 miniature air quality monitoring stations (AQMSs) provided comprehensive coverage of a FCIP, and long-term monitoring studies solved the problem of lack of clarity about pollution sources in industrial parks. Overall, NO2 pollution was particularly high in the pharmaceutical industry, while TVOCs and O3 pollution were most serious in the textile dyeing industry, with PM pollution much higher in the metal smelting industry than in other industries, and in the leather industry, O3 pollution was relatively severe. The spatial and temporal variations of air pollutants showed that higher PM, CO and NO2 concentrations were revealed in winter while lower in summer due to better meteorological diffusion conditions. TVOCs concentrations were higher with an average of 1954 ppb in summer possibly due to their increased volatilization from their sources at higher ambient temperatures. O3 concentrations were at their peaks in spring (88.8 μg m−3) and early fall (78.5 μg m−3). The daily trends of O3 precursors (TVOCs and NO2) were clearly negatively correlated with O3, and they showed bimodal peaks due to anthropogenic activities, plant emissions, lowering of the mixed boundary layer, etc. The O3 formed in FCIP was judged to be NO2-limited during the monitoring period based on the ratios of NO2 to TVOCs. Therefore, the effective strategy to reduce O3 formation in FCIP is to decrease the ambient NO2 concentration. Based on Pearson correlation coefficients, it appeared that WS promoted O3 formation through long-term transport and that high air temperatures also contributed to O3 formation in the environment. It was also stated in the study that the closer the residential area is to the industrial sources, the more significant the correlation. Thus, the results of this study will also be helpful for policymakers to design pollutant control strategies for different industries to mitigate the impact of pollutants on human health.

Jakarta’s rapid population growth and industrialization have severely impacted air quality, with limited monitoring stations and increasingly complex pollutant sources influenced by meteorological factors. Source apportionment studies are critical for addressing urban air quality issues because knowing the source of pollution is an important step in overcoming urban air quality problems. This study uses the Geographically Weighted Regression (GWR) and Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) models to analyze the distribution and sources of NO2, SO2, and CO in Jakarta and surrounding areas. Data includes spatial information from road networks, industries, and the energy sector during the wet (December) and dry (June) months of 2023. GWR results show that network density significantly contributes to pollutant concentrations, particularly during the dry month, while industrial and energy sectors display seasonal variations. The HYSPLIT model indicates that in December, air masses carrying pollutants mainly originate from the northwest and north, and in June, from the southeast and southwest. These patterns underscore the regional influence of transboundary air pollution from industrial and energy plants. The combined GWR and HYSPLIT findings provide a comprehensive understanding of air pollution dynamics and the critical role of meteorological factors in pollutant distribution.

(Barely) living in smog: China and air pollution

Ground-level ozone (O3) is a secondary air pollutant and one of the major air pollutants that shape the atmospheric chemistry and influence many chemical reactions in the atmosphere. O3 is the second most significant air pollutant after particulate matter contributing mortality world-wide. To explore the O3 (a criteria pollutant) concentration variability influenced by primary air pollutants, especially criteria air pollutants such as volatile organic compounds (VOC), carbon monoxide (CO), nitrogen dioxide (NO2), and sulfur dioxide (SO2), we conducted a field campaign at Kharagpur city in India on April 2024. Sample air was measured by the USEPA approved Serinus 10 ozone analyser, Serinus 40 NOx analyser, Serius 500 Portable Air Quality Monitor (with swappable sensor heads) to get O3, NOx (NO+NO2), total VOC (TVOC), CO, and SO2 concentrations, respectively. The measurement was carried out on National Highway 49 (NH-49) (22.379128°N, 87.361647°E) in Kharagpur. Results showed a strong negative correlation between O3 and NO (r = -0.82), a weak positive correlation with NO2 (r = 0.19), moderate negative correlations with TVOC (r = -0.54) and CO (r = -0.49), and a very weak positive correlation with SO2 (r = 0.11). All correlations are statistically significant at the p < 0.01 level. We applied Quantile Regression Model (QRM) to explore a robust framework for analyzing the relationships between dependent (O3) and independent variables (NO, NO2, TVOC, CO, SO2) across different points of the data distribution by capturing conditional quantiles. Analysis revealed nonlinear distribution of O3 concentration across all the quantiles (τ) with a strong performance at the median quantile (τ = 0.5) that explained 76.06% of variability in O3 concentration (R1(τ)0.5) = 0.7606) with high accuracy and low predicting errors (MAE = 14.19, RMSE = 17.86). The local measure of goodness of fit, R1(τ) were diminished at lower (below τ = 0.1) and higher quantiles (above τ = 0.95) and the Quantile Loss of 7.10 confirms effective handling of O3 variability. The standardized coefficients for NO were negative across all quantiles and became less negative at higher quantiles (0.8-1.0) that indicated a weaker adverse effect as O3 increased. NO2 showed positive coefficients that peaked at the 0.4 quantile and declined at higher quantiles and suggesting a stronger influence at moderate O3 levels. TVOC consistently exhibited negative coefficients with a stronger effect at lower quantiles (0.1-0.2) that stabilizes at higher quantiles. CO and SO2 coefficients fluctuate around zero and shows minimal and inconsistent influence. Monte Carlo simulation of health risk assessment showed that O3 could significantly pose the development of non-carcinogenic health risks (Hazard Quotient (HQ) > 1). Sensitivity analysis revealed the variance in the O3 Hazard Index (HI) where NO significantly contributed the most (60.4%), followed by NO2 (23.3%), TVOC (12.1%), SO2 (4.0%), and CO (0.2%). The Air Quality Index (AQI) analysis categorized Kharagpur as a ‘Moderately Polluted’ region. Overall, NOx and TVOC are the two types of major gaseous pollutants that contributed majorly in O3 concentration variability, thus O3 pollution levels. Targeted policies to reduce VOC and NOx emissions are essential.

Air pollution has become an integral part of modern life. The main source of air pollution can be considered combustion processes associated with energy-intensive corporate activities. Energy companies consume about one-third of the fuel produced and are a significant source of air pollution [1]. State and public air quality monitoring networks were created to monitor the situation. Public monitoring networks are cheaper and have more coverage than government ones. Although the state monitoring system shows more accurate data, an inexpensive network is sufficient to inform the public about the presence or absence of pollution (air quality). In order to inform the public, the idea arose to test the possibility of detecting types of pollution using data from cheap air quality monitoring sensors. In general, to use a cheap sensor for measurements, it must first be calibrated (corrected) by comparing its readings with a reference device. Various mathematical methods can be used for this. One of such method is neural network training, which has proven itself well for correcting PM particle readings due to relative humidity impact [2].The idea of using a neural network to improve data quality is not new, but it is quite promising, as the authors showed in [3]. The main problem to implement this method is connected with a reliable dataset for training the network. For this, it is necessary to register sensor readings for relatively clean air and for artificially generated or known sources of pollution. Training the neural network on the basis of collected data can be used to determine (classify) types of air: with pollution (pollutant) or without. For this, an experiment was set up in the "ReLab" co-working space at the Taras Shevchenko National University of Kyiv. The sensors were placed in a closed box, in which airflow ventilation is provided. The ZPHS01B [4] sensor module was used for inbox measurements, as well as, calibrated sensors PMS7003 [5] and BME280 [6]. Additionally, IPS 7100 [7] and SPS30 [8] were added to enrich the database for ML training. A platform based on HiLink 7688 was used for data collecting, processing, and transmission.Data was measured every two seconds, independently from each sensor. Before each experiment, the room was ventilated to avoid influence on the next series of experiments.

Air quality directly affects the health of humans. The health implications of poor air quality are recognized by professionals and the public alike and these concerns have driven both the proliferation of formal sensor networks, but also low-cost sensors which can be used in the home. The advancement of technology in recent years has also led to the rapid development of low-cost sensors. Given that citizens are concerned about the air quality of the environment in which they live, they are turning to the supply of low-cost sensors, as they are affordable. The question of the reliability of measurements from low-cost sensors remains an area of research. In this research work, the optimization of ozone (O3) and nitrogen dioxide (NO2) measurements of low-cost electrochemical air quality sensors is investigated by applying nonlinear regression, using a second-order polynomial equation as a correction factor. The proposed correction method is implementable in IoT devices, as it does not require high computational resources. The results show that the measurements are susceptible to correction, with the effect that the corrected values are close to the actual values obtained by the reference instruments of the Department of Environmental Pollution Control Project of Athens (PERPA), a service of the Greek Ministry of the Environment and Energy.

Background: The air quality of any area depends upon the various PMs (particulate matter) and hazardous gases present in the air. Low-cost PM sensors and gas sensors are present in different target places to monitor the air quality, read the environmental data, and transmit it to local servers through the IoT device. The low-cost sensor is not reliable due to its low sensing capacity; therefore, the read data is calibrated with the meteorological data presented by the nearby meteorological Centre of that particular area. The calibrated reading data sent to the server could be analyzed through some Machine Learning [ML] models. The ML models help to predict the risk of asthma in a particular area. The risk of asthma is directly related to the air quality of the surroundings. It is observed that the air quality of the industrial area is much worse than the non-industrial belt. Air quality monitoring of industrial areas is always a challenging task due to the ununiformed pollution in some segregated places around the industry, emitting pollutants mostly from chimneys. The air quality of any area depends upon the PM (PM), i.e., PM2.5 and PM10.0, as well as the gasses like NO2(Nitrogen Dioxide), NH3 (Ammonia), SO2(Sulfur dioxide), CO(Carbon monoxide), O3(Ozone) and Benzene. These are the most hazardous gases generally emitted by common heavy industries like iron and steel. In this article, the researchers considered the industrial belt of the Asansol- Durgapur region of West Bengal, India, and predicted the risk of asthma attacks for the test dataset. The experiment was carried out on 10 different supervised machine learning [SML] models as well as semi-supervised machine learning (SSML) models. The SML models have been further refined through hyper-parameter tuning, and better results have been obtained in the case of some ML models. The result has been compared with the existing literature considering the same external environment from where the meteorological data was collected, and similar ML models have been used. The research outperformed the existing literature, which is depicted in the result and analysis section of the article. Methods: The study evaluated ML models, both supervised and semi-supervised, to assess pollution levels. Relevant features were selected while less relevant ones were discarded. Accuracy levels of different ML algorithms werecompared in the results. The most effective model for an IoT system was chosen to maximize accuracy. In semi-supervised learning, feature selection followed supervised learning, but testing was akin to unsupervised learning. Results were compared with supervised learning data, enhancing reliability. Results: The result employing various classifiers werepresented across tables after the independent parameter Ozone was removed. Following the output of several classifiers, the results were verified using the k-fold validation method, with k being set to 5 or 10, accordingly. Tables display the best outcome, which is indicated in bold characters. method: In this research work the researcher considered 9 different ML models and used them as supervised as well as semi supervised model to determine the pollution level of the certain area. In this research work the researcher also selected the most relevant features and discarded the less relevant features. In case of SML algorithm, the accuracy level of the different ML algorithm has been determined and depicted in the result analysis section. The most effective ML model has been chosen for the proposed embedded system so that accuracy could be achieved at most. In case of semi supervised algorithm the feature selection is done as per the supervised algorithm. In this case the training is done same as the SML algorithm, but the testing phase is done like unsupervised machine learning algorithm where the decision parameter is predicted and ultimately matched with the previously achieved data of SML algorithm. The reliability of this approach is much more effective than simple SML algorithm. Conclusion: This study focused on predicting asthma risk in the Asansol-Durgapur industrial belt, India, using low-cost PM and gas sensors. Data calibration with meteorological inputs enhanced accuracy. ML models predicted risk and were refined through hyper-parameter tuning. Comparative analysis showed superior performance, emphasizing the importance of precise air quality monitoring. While offering a robust framework for future research, the study’s limitation lies in its area-specific dataset.

Air pollution is one of the leading risk factors for poor health in Africa, resulting in millions of premature deaths and economic losses. Of particular interest is exposure to fine particulate matter (PM2.5) which is the driver for a majority of deaths across the continent. However, PM monitoring, and by extension, ground-level data on PM2.5 is very limited; this limits our understanding of the widespread societal and health impacts linked to PM pollution. The robustness of low-cost PM sensors and their ability to report in situ data in tropical environments via internet-based platforms as well as relative affordability has created the opportunity to employ low-cost sensors (LCS) for air quality monitoring but calibration methodologies and the usefulness of the high-temporal resolution data for source identification remain a challenge. Increasingly, local governments in African countries are also turning to low-cost sensors to monitor air quality. In this study, two Airnote PM monitors were colocated with reference-grade Teledyne PM mass monitor T640 for ~4 weeks at the University of Ghana, Accra to establish their performance using a simplified data correction methodology - multiple linear regression (MLR) model. A split ratio of 80% and 20% was used to train and test the populated Airnote PM2.5 data respectively based on measurements from Teledyne T640 with temperature and relative humidity values from the Airnote monitor. Sectoral and calendar analysis with wind component data were used to triangulate the sources of PM2.5. We observed a high consistency between the two Airnote monitors. Hourly and 24-hour average PM2.5 values ranged from 25 to 95 μg/m3, and 29 to 54 μg/m3 respectively, and in most cases, were significantly higher than the WHO Air Quality Guideline. MLR using Pearson’s correlation analysis improved the out-of-the-box quality of low-cost Airnote PM2.5 data; the R2 improved from 0.69 to 0.84 and the mean absolute error from 11.75 to 4.20 μg/m3 respectively. Also, the MLR correction model was found to improve the Airnote PM2.5 data quality for higher relative humidity (between 50 and 90%) but not lower. PM2.5 pollution was local and from N, NE and SW winds for the raw, corrected and Teledyne PM mass monitor T640 measurements. Together, these results indicate that with appropriate corrections, low-cost PM sensors can generate the much needed data for air pollution research and mitigation in areas with limited air quality monitoring and data.

Background: Air pollution is a major global health risk, causing seven million deaths each year. In The Gambia, the use of firewood and charcoal for household energy, along with harmattan dust, significantly increases exposure to harmful particulate matter (PM2.5). Women and children are particularly vulnerable to these health risks.To address the air quality data gaps in The Gambia, we launched the Clean Air Initiative in 2023. Our goal was to set up a nationwide ambient air quality monitoring network using low-cost sensors. Led by the Permian Health Lung Institute (PHLI) and the National Environment Agency (NEA) in collaboration with various stakeholders, the initiative aimed to systematically tackle air pollution challenges.Methodology: We began with stakeholder mapping to identify the multifaceted challenges of air pollution. Key participants included government ministries of health and environment, private construction companies, and academic institutions. The process involved installing air quality sensors, conducting education workshops, and integrating scientific data with actionable plans.Results: Over 12 months, we deployed 17 low-cost air quality sensors (IQAir AVO) across all seven regions of The Gambia. This network helped identify air quality trends and assess their health impacts. Challenges such as sensor security and cellular network reliability were addressed through community engagement and solar panel installations. Our focus on the densely populated Greater Banjul Area enabled crucial observations and set the stage for nationwide expansion.Data from our initial four sensors showed a 12-month average PM2.5 concentration of 36.9 µg/m³ (95% Confidence Interval: 31.9 - 42.0 µg/m³), which is seven times higher than the World Health Organization (WHO) guideline of 5 µg/m³. Seasonal variations were significant, with dry season levels (November to May) averaging 44.9 µg/m³ (95% CI: 39.7 - 50.2 µg/m³), compared to 25.8 µg/m³ (95% CI: 18.6 - 32.9 µg/m³) during the rainy season.Discussion: A key outcome of the initiative was The Gambia's first Air Quality State Report, bolstered by continuous PM2.5 monitoring and capacity building at the National Environment Agency through international partnerships. Funding from the University of Chicago Energy Policy Institute was crucial for our success. Collaborations with regional entities like Makerere University are improving data accessibility across Africa. Despite these advances, challenges such as connectivity issues with remote sensors and environmental impacts on equipment remain, especially during seasonal transitions.Conclusion: The Clean Air Initiative marks a significant leap in managing air quality in The Gambia. By establishing a comprehensive monitoring network and creating foundational policies, we are paving the way for improved public health and governance. Continued collaboration and overcoming operational hurdles are essential for sustained progress. Future plans for expansion and regional data integration offer promising pathways for effective air quality management in the region.Keywords: Air pollution, PM2.5, The Gambia

In the beginning, even before the desert skyline was punctuated by stacks from the first LNG plants in the northeast corner of Qatar, air quality models were being run to estimate potential air quality impacts (AQIA) due to the proposed developments. AQIAs were required as part of the Environmental Impact Assessment (EIA) obligation to SCENR (now Ministry of Environment). With time, Ras Laffan (RLC) has grown into one of the premier energy hubs in the world with contemporaneous improvement in the pedigree of air modelling. This paper provides a genesis of air quality modelling in Qatar with initial screening estimates graduating to Industrial Source Complex (ISCST3) and now AERMOD as the mainstay model. The RLC coastline with its dense mass of industrial growth presents a daunting challenge in terms of predictions of air quality impacts due to a mix of uncertainties in emission inventories, coastal phenomena etc., which cannot be accounted for by the existing raft of air dispersion models. This paper makes a case for moving to an established scalable puff model (CALPUFF) operated in near real-time with dynamic Weather Research and Forecasting (WRF) data as the regulatory model of choice. The proposed platform will vastly improve RLC's Air Quality Management Plan (AQMP) by improving allowable increment planning, handle complex coastline conditions local to RLC, addressing the lacunae of plume models and photochemistry. Applications: Assess and mitigate air quality impacts from large gas and industrial projects through air dispersion modellingAQMP to assess future industrial growth using real-time AQIA Results and Conclusions: Results from preliminary one-year RLC WRF-CALPUFF compared against other platformsInnovative technological approach by leveraging dynamic computation platforms to improve air quality impact predictions Technical Contributions: Innovative air quality dispersion modelling using WRF-CALPUFF for assessing near real-time air quality impacts

Global warming and climate change have become universal issues recently. One of the leading sources of climate change is automobiles. Automobiles are the prime source of air pollution in urban areas globally. This has resulted in a problematic and chaotic state in the development of an automatic traffic management system for capturing and monitoring vehicles’ hourly and daily passage. With the significant advancement of sensor technology, atmospheric information such as air pollu-tion, meteorological, and motor vehicle data can be harvested and stored in databases. However, due to the complexity and non-linear associations between air quality, meteorological, and traffic variables, it is difficult for the traditional statistical and mathematical models to analyze them. Recently, machine learning algorithms in the field of traffic emissions prediction have become a popular tool. Meteorological and traffic variables influence the variation and the trend of the traffic pollutants. In this paper, an optimized artificial neural network (OANN) was developed to enhance the existing artificial neural network (ANN) model by updating the initial weights in the network using a Genetic Algorithm (GA). The OANN model was implemented to predict the concentration of CO, NO, NO2, and NOx pollutants produced by motor vehicles in Kuala Lumpur, Malaysia. OANN was compared with Artificial Neural Network (ANN), Random Forest (RF), and Decision Tree (DT) models. The results show that the developed OANN model performed better than the ANN, RF, and DT models with the lowest MSE values of 0.0247 for CO, 0.0365 for NO, 0.0542 NO2, and 0.1128 for NOx. It can be concluded that the developed OANN model is a better choice in predicting traffic emission concentrations. The developed OANN model can help environmental agencies monitor traffic-related air pollution levels efficiently and take necessary measures to ensure the effectiveness of traffic management policy. The OANN model can also help decision-makers mitigate traffic emissions to protect citizens living in the neighborhood of highways.

Workers of São Paulo city, Brazil, exposed to air pollution: Assessment of genotoxicity