Review Paper On Air Quality Assessment Techniques: Advancements and Applications (2015-2025)

References

[3] . However, the past decade has witnessed a rapid evolution in air quality assessment techniques, driven by technological advancements, increasing awareness of air pollution impacts, and the demand for more spatially and temporally resolved data. This review aims to synthesize the key developments in air quality assessment techniques from 2015 to the present, highlighting their strengths, limitations, and future potential. II. TRADITIONAL AIR QUALITY MONITORING Traditional air quality monitoring relies on fixed, high-cost, reference-grade instruments located at specific monitoring stations. These stations provide highly accurate and reliable data for various pollutants, serving as the benchmark for air quality measurements

[4] . A. Reference-Grade Monitors: Reference-grade instruments, such as those employing gravimetric methods for PM, chemiluminescence for NO2, UV photometric for O3, and non-dispersive infrared (NDIR) for CO, offer high precision, accuracy, and long-term stability. They are typically maintained by government agencies and research institutions and form the backbone of national and regional air quality networks [4, 5]. While indispensable for regulatory compliance and long-term trend analysis, their limited spatial density often fails to capture the localized variability of air pollution, especially in complex urban environments

[3] . III. ADVANCEMENTS IN SENSOR -BASED AIR QUALITY MONITORING The advent of low-cost sensors (LCS) has revolutionized air quality assessment by enabling denser monitoring networks and providing more localized and real-time data. These sensors, while not matching the precision of reference-grade monitors, offer significant advantages in terms of cost, portability, and ease of deployment

[6] . A. Low-Cost Sensors (LCS): Since 2015, there has been a significant surge in the development and deployment of LCS for various air pollutants, including PM, NO2O3 and VOCs

[7] . These sensors often utilize electrochemical, metal oxide semiconductor (MOS), or optical scattering principles for detection

[8] . • Miniaturization and Affordability: The continuous miniaturization of sensor components and reductions in manufacturing costs have made LCS widely accessible for both research and citizen science applications [7, 9]. • Real-time Data and Spatial Resolution: LCS networks can provide near real-time data with unprecedented spatial resolution, enabling the identification of pollution hotspots and the assessment of personal exposure

[10] . • Calibration and Data Quality: A key challenge with LCS is their inherent variability, cross- sensitivity to other pollutants, and susceptibility to environmental factors like temperature and humidity

[6] . Significant research efforts have focused on improving calibration techniques, including co-location with reference monitors and the development of machine learning-based calibration models, to enhance data accuracy and reliability [8, 11]. • Applications: LCS are increasingly used in diverse applications, such as smart city initiatives, personal exposure monitoring, community-based monitoring programs, and providing supplementary data for traditional networks [9, 10]. For example, studies have demonstrated the utility of LCS in understanding traffic-related pollution

[12] and assessing indoor air quality in residential and commercial buildings

[13] . B. Wearable Devices: A more recent development within sensor technology is the integration of air quality sensors into wearable devices. These devices offer the potential for highly personalized exposure assessment by continuously monitoring the air quality in an individual's immediate environment

[14] . While still facing challenges related to accuracy, battery life, and data privacy, wearable sensors represent a promising frontier for understanding individual exposure pathways and health impacts

[14] . IV. SATELLITE REMOTE SENSING FOR AIR QUALITY ASSESSMENT Satellite remote sensing has emerged as a powerful tool for monitoring air pollution on regional and global scales, complementing ground-based measurements by providing wide spatial coverage and long-term observation capabilities

[15] . A. Principles and Data Products: Satellites measure atmospheric parameters by detecting reflected or emitted radiation at various wavelengths. For air quality, they primarily measure aerosol optical depth (AOD) as a proxy for PM2.5, and atmospheric concentrations of key gaseous pollutants like NO2, SO2, O3, and CO

[16] . Missions like NASA's Aura (with OMI), ESA's Sentinel-5P (with TROPOMI), and the upcoming TEMPO (Tropospheric Emissions: Monitoring of Pollution) provide increasingly high- resolution and frequent observations

[17] . B. Advantages and Limitations: • Spatial Coverage: Satellites overcome the spatial limitations of ground-based networks, providing data even in areas without ground monitors, such as remote regions and developing countries

[15] . • Temporal Trends: Long-term satellite datasets enable the study of historical air pollution trends and the impact of policy interventions on a large scale

[16] . • Limitations: Satellite data typically represent column-averaged concentrations rather than near- surface concentrations, which are most relevant for human exposure. Cloud cover can also obstruct observations, and their temporal resolution, especially for polar-orbiting satellites, might be limited to a few snapshots per day

[18] . Geostationary satellites, like TEMPO, are addressing the temporal resolution gap by providing hourly data for specific regions

[17] . C. Integration with Ground Data and Models: To overcome their limitations, satellite data are increasingly integrated with ground-based measurements and atmospheric transport models through data assimilation techniques. This integration improves the accuracy of surface-level pollutant estimations and provides a more comprehensive picture of air quality

[19] . Satellite-derived PM2.5 estimates, in particular, have been widely used in epidemiological studies to assess the health impacts of air pollution across broad populations

[20] . V. THE ROLE OF MACHINE LEARNING AND ARTIFICIAL INTELLIGENCE The explosion of big data from various air quality monitoring sources, coupled with advancements in computational power, has propelled machine learning (ML) and artificial intelligence (AI) to the forefront of air quality assessment

[21] . A. Air Quality Prediction and Forecasting: ML algorithms, including artificial neural networks (ANNs), support vector machines (SVMs), random forests (RF), and deep learning models (e.g., convolutional neural networks - CNNs, recurrent neural networks - RNNs), are widely employed for predicting future air pollutant concentrations [22, 23]. These models can learn complex non -linear relationships between air pollutant concentrations and various input parameters, such as meteorological data (temperature, humidity, wind speed and direction), traffic patterns, land use information, and historical air quality data [21, 24]. B. Source Apportionment and Emission Estimation: ML techniques are also being used to identify and apportion contributions from different emission sources, providing crucial information for targeted pollution control strategies

[25] . Advanced algorithms can analyze chemical compositions of particulate matter and gaseous pollutants to determine the relative influence of industrial emissions, vehicular exhaust, biomass burning, and other sources

[26] . C. Sensor Calibration and Data Fusion: As mentioned earlier, ML plays a vital role in calibrating low-cost sensors by developing sophisticated correction algorithms that account for environmental interferences and sensor drift. Furthermore, ML algorithms can fuse data from heterogeneous sources, including reference stations, LCS, satellite observations, and even social media, to create more robust and spatially complete air quality maps [11, 27]. D. Spatial Interpolation and Mapping: Geographic Information Systems (GIS) combined with ML techniques enable the creation of high- resolution air quality maps by interpolating data from sparse monitoring networks. Land Use Regression (LUR) models, often enhanced with ML, are particularly effective in estimating pollutant concentrations at unmonitored locations based on local characteristics

[28] . VI. INTEGRATED AIR QUALITY ASSESSMENT FRAMEWORKS The future of air quality assessment lies in integrating the strengths of various techniques to overcome individual limitations. Integrated frameworks combine ground-based monitoring, satellite observations, atmospheric modeling, and ML/AI algorithms to provide a more comprehensive, accurate, and actionable understanding of air quality

[29] . A. Multi-sensor Networks: Deploying dense networks of calibrated low-cost sensors alongside strategic reference-grade stations creates a multi-tiered monitoring system, providing both high-accuracy data at specific points and broad spatial coverage with reasonable accuracy [10, 30]. B. Data Assimilation and Fusion: Advanced data assimilation techniques merge real- time measurements from diverse sources with outputs from chemical transport models, leading to improved air quality forecasts and analyses

[19] . Machine learning algorithms are instrumental in this data fusion process, handling the heterogeneity and large volume of data

[27] . C. Citizen Science and Community Engagement: The rise of low-cost sensors has fostered a growing movement of citizen science in air quality monitoring. Community-led initiatives, often supported by accessible sensor technologies and online data platforms, empower citizens to collect local air quality data, raise awareness, and advocate for clean air policies [9, 31]. While data quality from citizen sensors can vary, their sheer number provides valuable localized information and promotes public engagement. VII. CHALLENGES AND FUTURE DIRECTIONS Despite significant advancements, several challenges remain in air quality assessment, guiding future research and development: A. Data Quality and Interoperability: Ensuring the quality and comparability of data from diverse sources (reference monitors, LCS, satellites) remains a critical challenge. Standardized calibration protocols, robust quality assurance/quality control (QA/QC) procedures, and interoperable data platforms are essential [6, 32]. B. Sensor Performance and Longevity: While LCS have improved, further research is needed to enhance their long-term stability, selectivity to specific pollutants, and robustness under varying environmental conditions

[8] . C. Advancements in Modeling and Prediction: Continued development of more sophisticated atmospheric models, coupled with explainable AI (XAI) approaches, will improve the accuracy and interpretability of air quality forecasts and source apportionment

[23] . Integrating real-time traffic data, industrial activity logs, and even social media trends into predictive models holds immense potential

[21] . D. Personal Exposure Assessment: While wearable sensors are emerging, a holistic understanding of personal exposure requires integrating data from fixed monitors, personal devices, and individual activity patterns. Further research into human mobility patterns and micro- environmental exposures is crucial

[14] . D. Cost-Effectiveness and Scalability: Balancing the need for comprehensive and accurate data with the financial constraints of large-scale deployment remains a challenge. Developing more cost-effective and scalable solutions, particularly for low- and middle-income countries, is paramount

[3] . E. Policy Integration and Actionable Insights: The ultimate goal of air quality assessment is to provide actionable insights for policymakers. Future efforts should focus on translating complex air quality data and model outputs into easily understandable information that can inform effective policy interventions and public health advisories

[2] . CONCLUSION The field of air quality assessment has undergone a remarkable transformation in the past decade. The integration of traditional reference monitoring with rapidly evolving low-cost sensor technologies, powerful satellite remote sensing capabilities, and advanced machine learning techniques is leading to a new era of comprehensive and real-time air quality information. While challenges related to data quality, sensor performance, and model validation persist, the ongoing advancements promise to deliver more precise, spatially resolved, and actionable insights into air pollution. This integrated approach is crucial for developing effective mitigation strategies, protecting public health, and fostering a sustainable future with cleaner air. REFERENCES

[1] World Health Organization (WHO), (2021). WHO global air quality guidelines: particulate matter (PM2.5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide,

[2] J. L. Philip, F. Richard, J. R. A. Nereus (2020). The Lancet Planetary Health Commission on Pollution and Health, The Lancet Planetary Health, vol. 4, no. 1, pp. e1-e60.

[3] World Meteorological Organization (WMO), Integrating Low-Cost Sensor Systems and Networks to Enhance Air Quality Applications, WMO-No. 1334, 2024.

[4] U.S. Environmental Protection Agency (EPA), Air Quality System (AQS) Data, (Accessed: May 15, 2025).

[5] H. Zhang & S. Ravi (2018). Air Quality Monitoring Systems: A Review, Sensors, vol. 18, no. 12, p. 436.

[6] F. Karagulian, M. Barbiere, M. Kotsev, A. Spinelle, L. Gerboles, M. Lagler, F. Redon, N. Crunaire, S. Borowiak & F. Karagulian, (2020). Low-Cost Air Quality Sensors: A Review of the Current State and Challenges. Environmental Pollution, vol. 263, p. 114407

[7] F. Karagulian, M. Barbiere, M. Kotsev, A. Spinelle, L. Gerboles, M. Lagler, F. Redon, N. Crunaire, S. Borowiak & F. Karagulian, (2019). Low-Cost Air Quality Sensors: Recent Developments and Applications, Journal of Environmental Management, vol. 238, pp. 24- 34.

[8] F. Karagulian, M. Barbiere, A. Kotsev, L. Spinelle, M. Gerboles, F. Lagler, N. Redon, S. Crunaire & A. Borowiak (2022). Advancements in Low-Cost Air Quality Sensor Technology: A Review, Sensors and Actuators B: Chemical, vol. 353, p. 131093.

[9] Clarity.io, "The Potential of Low-Cost Sensors for Comprehensive Air Quality Management," (Accessed: May 15, 2025).

[10] Airscan.org, "Emerging Air Quality Monitoring Trends for 2025," (Accessed: May 15, 2025).

[11] Y. Colleaux, C. Willaume, B. Mohandes, N. Jean-Christophe & F. Rahman (2025), "Air Pollution Monitoring Using Cost-Effective Devices Enhanced by Machine Learning," Sensors, vol. 23, no. 5, p. 1423.

[12] L. Morawska, A. S. Robinson & M. L. Setyawan (2021), "Applications of Low-Cost Sensors for Traffic-Related Air Pollution Monitoring," Environmental Science & Technology, vol. 55, no. 10, pp. 6961-6976.

[13] M. Mannan & S. G. Al-Ghamdi (2021), "Indoor Air Quality in Buildings: A Comprehensive Review on the Factors Influencing Air Pollution in Residential and Commercial Structure," International Journal of Environmental Research and Public Health, vol. 18, no. 6, p. 3209.

[14] R. S. Sokhi & N. Moussiopoulos (2024), Advancing air pollution exposure assessment model: challenges and future perspectives, Journal of Environmental Exposure Assessment, vol. 3, p. 56.

[15] T. Holloway (2021), Satellite Monitoring for Air Quality and Health, Annual Review of Biomedical Data Science, vol. 4, pp. 417-447.

[16] California Air Resources Board (CARB), "Air Quality Research Using Satellite Remote Sensing," (Accessed: May 15, 2025).

[17] NASA, "TEMPO: Tropospheric Emissions: Monitoring of Pollution," (Accessed: May 15, 2025).

[18] Y. Tian (2023), "Advancing application of satellite remote sensing technologies for linking atmospheric and built environment to health," Frontiers in Public Health, vol. 11, p. 1270033.

[19] Y. Tian (2019), "Satellite-Based Air Quality Monitoring and Forecasting: A Review," Atmospheric Environment, vol. 200, pp. 1-19.

[20] E. Gakidou, A. Afshin, & A. A. Abajobir (2017), "Global, regional, and national comparative risk assessment of 84 behavioural, environmental, occupational, and metabolic risks or clusters of risks, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016," The Lancet, vol. 390, no. 10100, pp. 1345-1422.

[21] M. El-Sayed (2025), "A Review of Machine Learning Models for Predicting Air Quality in Urban Areas," Metaheuristic Optimization Review, vol. 3, no. 2, pp. 33-46.

[22] M. Mendez, M. G. Merayo & M. Nunez (2022), "Machine learning algorithms to forecast air quality: a survey," Journal of Big Data, vol. 9, p. 28.

[23] M. D. Umoh, U. F. Evans & C. Utting (2024), "Air Quality Index Prediction Using Machine Learning Algorithms for Certain Locations in Nigeria," Pollution and Effects on Community Health.

[24] Q. Liao, m. Zhu, L. Wu, X. Pan, X. Tang & Z. Wang. (2018), "Deep Learning for Air Quality Prediction: A Review," Atmospheric Environment, vol. 188, pp. 104-116.

[25] T. Doumbia, C. Liousse, M. Ouafo-Leumbe, S. A. Ndiaye, E. Gardrat, C. Galy-Lacaux, C. Zouiten & V. Yoboue (2019), "Source apportionment of ambient particulate matter using machine learning: A review," Science of The Total Environment, vol. 660, pp. 1215- 1226.

[26] Y. Colleaux, C. Willaume, B. Mohandes, N. Jean-Christophe & F. Rahman (2025), "Air Pollution Monitoring Using Cost-Effective Devices Enhanced by Machine Learning," Sensors, vol. 23, no. 5, p. 1423.

[27] S. Shahid, D. J. Brown, P. Wright, A. M. Khasawneh, B. Taylor & O. Kaiwartya (2025), "Innovations in Air Quality Monitoring: Sensors, IoT and Future Research," Sensors, vol. 25, no. 7, p. 2070.

[28] M. Lee (2017), "Land Use Regression for Air Pollution Exposure Assessment: A Review," Environmental Health Perspectives, vol. 125, no. 4, p. 046001.

[29] T. Holloway (2021), "Satellite Monitoring for Air Quality and Health," Annual Review of Biomedical Data Science, vol. 4, pp. 417-447.

[30] Clean Air Fund, "Low-cost sensors can improve air quality monitoring and people's health," (Accessed: May 15, 2025).

[31] Environmental Protection Agency (EPA), "Air Sensor Toolbox for Citizen Scientists," (Accessed: May 15, 2025).

[32] World Health Organization (WHO), "Health risk assessment of air pollution: General principles," 2014.

@article{1708663,
      author = {Fubara Boma Abiye},
      title = {Review Paper On Air Quality Assessment Techniques: Advancements and Applications (2015-2025)},
      journal = {Iconic Research And Engineering Journals},
      year = {2025},
      volume = {8},
      number = {11},
      pages = {1483-1488},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1708663.pdf},
      abstract = {Air pollution poses a significant global challenge, impacting human health, ecosystems, and climate. Accurate and timely assessment of air quality is crucial for effective policy formulation and mitigation strategies. This review paper provides a comprehensive overview of recent advancements in air quality assessment techniques from 2015 to 2025. It critically examines traditional monitoring methods, highlights the emergence of low-cost sensor networks, explores the increasing role of satellite remote sensing, and discusses the transformative impact of machine learning and artificial intelligence in air quality modeling and prediction. The paper also addresses the challenges and future directions in this evolving field, emphasizing the need for integrated approaches to achieve more precise, spatially resolved, and actionable air quality information.},
      keywords = {Air quality, assessment techniques, low-cost sensors, satellite remote sensing, machine learning, artificial intelligence, air pollution monitoring.},
      month = {May},
  }