A Comparative Evaluation of LSTM and Bidirectional LSTM Architectures for Nitrogen Oxide Forecasting in Air Quality Monitoring
DOI:
https://doi.org/10.66279/9c0e5874Keywords:
LSTM, BiLSTM, Nitrogen Oxide Forecasting, Air Quality MonitoringAbstract
Reliable, real-time monitoring of nitrogen oxides (NOx ) is essential for air quality management, yet conventional monitoring networks are limited by sparse spatial coverage and inconsistent predictive reliability. This study compares Long Short-Term Memory (LSTM) and Bidirectional LSTM (BiLSTM) networks for forecasting NOx concentrations using the UCI Air Quality Dataset. To assess these architectures rigorously, we benchmark them against a persistence forecast and an ARIMA model, apply expanding-window temporal cross-validation to guard against data leakage, and test the significance of any performance gap with the Diebold–Mariano statistic. Both recurrent architectures attain high accuracy (R2 > 0.999), with BiLSTM showing a marginally higher mean R2 (0.99977 versus 0.99972 for LSTM); however, the Diebold–Mariano test indicates that this difference is not statistically significant (p = 0.384), so the two architectures should be regarded as practically equivalent for this task. Both deep learning models substantially outperform the statistical baselines, most notably ARIMA (R2 = 0.854), which supports the value of recurrent architectures for NOx forecasting while also underscoring the importance of baseline comparison and leakage-aware validation in this line of work.
Downloads
References
[1] A. P. Pribadi, A. U. Rauf, Y. M. R. Rahman, and Z. F. Haq, “Air quality and urban sustainable development-current issues and future directions,” in Sustainable Urban Environment and Waste Management: Theory and Practice, pp. 23–51, Springer, 2025. DOI: https://doi.org/10.1007/978-981-96-1140-9_2
[2] A. Manders and M. Ketzel, “Regional and urban air quality in europe,” in Handbook of Air Quality and Climate Change, pp. 1–21, Springer, 2023. DOI: https://doi.org/10.1007/978-981-15-2527-8_14-1
[3] S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural computation, vol. 9, no. 8, pp. 1735–1780, 1997. DOI: https://doi.org/10.1162/neco.1997.9.8.1735
[4] P. A. B. Andrade, P. E. C. Zamora, A. E. C. Lara, and J. P. P. Piedra, “A comprehensive evaluation of ai techniques for air quality index prediction: Rnns and transformers,” Ingenius, no. 33, pp. 60–75, 2025. DOI: https://doi.org/10.17163/ings.n33.2025.06
[5] B. Zhang, Y. Rong, R. Yong, D. Qin, M. Li, G. Zou, and J. Pan, “Deep learning for air pollutant concentration prediction: A review,” Atmospheric Environment, vol. 290, p. 119347, 2022. DOI: https://doi.org/10.1016/j.atmosenv.2022.119347
[6] L. Zhang, P. Liu, L. Zhao, G. Wang, W. Zhang, and J. Liu, “Air quality predictions with a semi-supervised bidirectional lstm neural network,” Atmospheric Pollution Research, vol. 12, no. 1, pp. 328–339, 2021. DOI: https://doi.org/10.1016/j.apr.2020.09.003
[7] G. Naresh and B. Indira, “Air pollution prediction using multivariate lstm deep learning model,” International Journal of Intelligent Systems and Applications in Engineering, vol. 12, pp. 211–220, 2024.
[8] K. Nagrecha, P. Muthukumar, E. Cocom, J. Holm, D. Comer, I. Burga, and M. Pourhomayoun, “Sensor-based air pollution prediction using deep cnn-lstm,” in 2020 International Conference on Computational Science and Computational Intelligence (CSCI), pp. 694–696, IEEE, 2020. DOI: https://doi.org/10.1109/CSCI51800.2020.00127
[9] S. Kabir, R. U. Islam, M. S. Hossain, and K. Andersson, “An integrated approach of belief rule base and deep learning to predict air pollution,” Sensors, vol. 20, no. 7, p. 1956, 2020. DOI: https://doi.org/10.3390/s20071956
[10] Z. Wu, C. Ma, X. Shi, L. Wu, Y. Dong, and M. Stojmenovic, “Imputing missing indoor air quality data with inverse mapping generative adversarial network,” Building and Environment, vol. 215, p. 108896, 2022. DOI: https://doi.org/10.1016/j.buildenv.2022.108896
[11] S. Gautam, A. S. Gautam, A. Awasthi, and R. N, “The nature of air pollution,” in Sustainable Air: Strategies for Cleaner Atmosphere and Healthier Communities, pp. 29–37, Springer, 2024. DOI: https://doi.org/10.1007/978-3-031-77057-9_5
[12] Á. M. Trivino, J. Palacios, P. Brassard, S. Godbout, and V. Raghavan, “Evolution of research on air emissions from agricultural activities: A comprehensive review,” Environmental Science and Pollution Research, vol. 31, no. 59, DOI: https://doi.org/10.1007/s11356-024-35635-2
pp. 66551–66567, 2024.
[13] R. Munsif, M. Zubair, A. Aziz, and M. N. Zafar, “Industrial air emission pollution: potential sources and sustainable mitigation,” in Environmental emissions, IntechOpen, 2021. DOI: https://doi.org/10.5772/intechopen.93104
[14] I. Manisalidis, E. Stavropoulou, A. Stavropoulos, and E. Bezirtzoglou, “Environmental and health impacts of air pollution: a review,” Frontiers in public health, vol. 8, p. 14, 2020. DOI: https://doi.org/10.3389/fpubh.2020.00014
[15] C. He, R. Kumar, W. Tang, G. Pfister, Y. Xu, Y. Qian, and G. Brasseur, “Air pollution interactions with weather and climate extremes: current knowledge, gaps, and future directions,” Current Pollution Reports, vol. 10, no. 3, pp. 430–442, 2024. DOI: https://doi.org/10.1007/s40726-024-00296-9
[16] S. Wang, R. Song, Z. Xu, M. Chen, G. L. Di Tanna, L. Downey, S. Jan, and L. Si, “The costs, health and economic impact of air pollution control strategies: a systematic review,” Global Health Research and Policy, vol. 9, no. 1, p. 30, 2024. DOI: https://doi.org/10.1186/s41256-024-00373-y
[17] F. Jelti, A. Allouhi, and K. A. Tabet Aoul, “Transition paths towards a sustainable transportation system: a literature review,” Sustainability, vol. 15, no. 21, p. 15457, 2023. DOI: https://doi.org/10.3390/su152115457
[18] S. Singh, M. J. Kulshrestha, N. Rani, K. Kumar, C. Sharma, and D. Aswal, “An overview of vehicular emission standards,” Mapan, vol. 38, no. 1, pp. 241–263, 2023. DOI: https://doi.org/10.1007/s12647-022-00555-4
[19] P. Sharma, P. Sharma, and N. Thakur, “Sustainable farming practices and soil health: A pathway to achieving sdgs and future prospects,” Discover Sustainability, vol. 5, no. 1, p. 250, 2024. DOI: https://doi.org/10.1007/s43621-024-00447-4
[20] S. Verghese and A. K. Nema, “Optimal design of air quality monitoring networks: A systematic review,” Stochastic Environmental Research and Risk Assessment, vol. 36, no. 10, pp. 2963–2978, 2022. DOI: https://doi.org/10.1007/s00477-022-02187-1
[21] A. R. Whitehill, M. Lunden, B. LaFranchi, S. Kaushik, and P. A. Solomon, “Mobile air quality monitoring and comparison to fixed monitoring sites for instrument performance assessment,” Atmospheric measurement techniques, vol. 17, no. 9, pp. 2991–3009, 2024. DOI: https://doi.org/10.5194/amt-17-2991-2024
[22] Y. Tang, J. Cape, and M. Sutton, “Development and types of passive samplers for monitoring atmospheric no2 and nh3 concentrations,” The Scientific World Journal, vol. 1, no. 2, pp. 513–529, 2001. DOI: https://doi.org/10.1100/tsw.2001.82
[23] L. Caretto, “Chemical analysis of air pollution sources,” in Combustion-Generated Air Pollution: A Short Course on Combustion-Generated Air Pollution held at the University of California, Berkeley September 22–26, 1969, pp. 165–204, Springer, 1971. DOI: https://doi.org/10.1007/978-1-4684-7574-6_7
[24] M. Rezaali, R. Fouladi-Fard, P. O’Shaughnessy, K. Naddafi, and A. Karimi, “Assessment of aermod and adms for nox dispersion modeling with a combination of line and point sources,” Stochastic Environmental Research and Risk Assessment, vol. 39, no. 2, pp. 813–827, 2025. DOI: https://doi.org/10.1007/s00477-024-02903-z
[25] M. Pantusheva, R. Mitkov, P. O. Hristov, and D. Petrova-Antonova, “Air pollution dispersion modelling in urban environment using cfd: a systematic review,” Atmosphere, vol. 13, no. 10, p. 1640, 2022. DOI: https://doi.org/10.3390/atmos13101640
[26] C. A. Onwudiegwu, L. Sylva, A. O. Aigberua, and M. Hait, “Biological monitoring of air pollutants,” in Air Pollutants in the Context of One Health: Fundamentals, Sources, and Impacts, pp. 457–484, Springer, 2024. DOI: https://doi.org/10.1007/698_2024_1139
[27]S. Hochheiser, Methods of measuring and monitoring atmospheric sulfur dioxide. US Department of Health, Education,
and Welfare, and Public Health Service, Division of Air Pollution, 1964.
[28] B. S. Freeman, G. Taylor, B. Gharabaghi, and J. Thé, “Forecasting air quality time series using deep learning,” Journal of the Air & Waste Management Association, vol. 68, no. 8, pp. 866–886, 2018. DOI: https://doi.org/10.1080/10962247.2018.1459956
[29] K. Rajesh and S. R. Kumar, “Deep reinforcement learning for urban air quality management: Multi-objective optimization of pollution mitigation booth placement in metropolitan environments,” IEEE Access, 2025. DOI: https://doi.org/10.1109/ACCESS.2025.3599541
[30] S. De Vito, E. Massera, M. Piga, L. Martinotto, and G. Di Francia, “On field calibration of an electronic nose for benzene estimation in an urban pollution monitoring scenario,” Sensors and Actuators B: Chemical, vol. 129, no. 2, pp. 750–757, 2008. DOI: https://doi.org/10.1016/j.snb.2007.09.060
[31] G. Box and G. Jenkins, “Analysis: Forecasting and control,” San francisco, vol. 10, 1976.
[32] A. Graves and J. Schmidhuber, “Framewise phoneme classification with bidirectional lstm and other neural network architectures,” Neural networks, vol. 18, no. 5-6, pp. 602–610, 2005. DOI: https://doi.org/10.1016/j.neunet.2005.06.042
[33] F. X. Diebold and R. S. Mariano, “Comparing predictive accuracy,” Journal of Business & economic statistics, vol. 20, no. 1, pp. 134–144, 2002. DOI: https://doi.org/10.1198/073500102753410444
Downloads
Published
Issue
Section
Categories
License
Copyright (c) 2026 Engineering Systems and Intelligent Technologies (ESIT)
This work is licensed under a Creative Commons Attribution 4.0 International License.
Engineering Systems and Intelligent Technologies (ESIT) content is published under a Creative Commons Attribution License (CCBY). This means that content is freely available to all readers upon publication, and content is published as soon as production is complete.
Engineering Systems and Intelligent Technologies (ESIT) seeks to publish the most influential papers that will significantly advance scientific understanding. Selected articles must present new and widely significant data, syntheses, or concepts. They should merit recognition by the wider scientific community and the general public through publication in a reputable scientific journal.
