Thirty years of air quality change in North Dakota: Seasonal trends and machine-learning forecasts of NO<sub>2</sub> and SO<sub>2</sub>

Faisal Ahmed; Md Nasifur Rahman Sakib; Manna Khan; Sabrina Karim; Faisal Ahmed; Md Nasifur Rahman Sakib; Manna Khan; Sabrina Karim

doi:10.3934/environsci.2026024

AIMS Environmental Science

2026, Volume 13, Issue 3: 579-597. doi: 10.3934/environsci.2026024

Previous Article Next Article

Research article Special Issues

Thirty years of air quality change in North Dakota: Seasonal trends and machine-learning forecasts of NO₂ and SO₂

1.
GIScience and Geoenvironment, Western Illinois University, Macomb, IL 61455, USA
2.
Department of Geography and Geographic Information Science, University of North Dakota, Grand Forks, ND 58201, USA
3.
Department of Earth System Science and Policy, University of North Dakota, Grand Forks, ND 58201, USA
4.
Department of Geosciences, Auburn University, Auburn, AL 36830, USA

Received: 16 March 2026 Revised: 29 June 2026 Accepted: 13 August 2026 Published: 18 August 2026

Long-term air quality patterns in the Northern Great Plains remain insufficiently characterized, particularly for gaseous pollutants that are sensitive to seasonal atmospheric processes and regional emission changes. This study provides a multidecadal analysis of nitrogen dioxide (NO₂) and sulfur dioxide (SO₂) concentrations from 1995 to 2025 at two monitoring stations in Fargo and Bismarck, North Dakota, using seasonally averaged (summer and winter) observations. Historical trends show substantial declines in both pollutants over the 30-year period, with winter concentrations consistently exceeding summer values. To evaluate future air quality trajectories, four machine-learning models (i.e., random forest, extreme gradient boosting, support vector machine, and a feedforward neural network) were trained using engineered lag features and standardized predictors. Model validation with independent 2020–2024 data demonstrated that the neural network achieved the highest predictive accuracy across all scenarios, with coefficients of determination exceeding 0.96 and minimal forecast error. Using a recursive forecasting framework, pollutant levels were projected from 2025 to 2050. The forecasts indicate continued improvement in some cases, stabilization in others, and slight upward tendencies for specific pollutant-season combinations, particularly winter NO₂ in Bismarck and summer SO₂ in both cities. These results highlight that while historical emission reductions have been highly effective, future progress may not be uniform across pollutants or seasons. The findings underscore the importance of sustained monitoring and proactive regulatory planning to preserve North Dakota's historically clean air.
- nitrogen dioxide,
- sulfur dioxide,
- machine learning,
- seasonal trends,
- long-term prediction,
- North Dakota
Citation: Faisal Ahmed, Md Nasifur Rahman Sakib, Manna Khan, Sabrina Karim. Thirty years of air quality change in North Dakota: Seasonal trends and machine-learning forecasts of NO₂ and SO₂[J]. AIMS Environmental Science, 2026, 13(3): 579-597. doi: 10.3934/environsci.2026024

Related Papers:

Abstract

Long-term air quality patterns in the Northern Great Plains remain insufficiently characterized, particularly for gaseous pollutants that are sensitive to seasonal atmospheric processes and regional emission changes. This study provides a multidecadal analysis of nitrogen dioxide (NO₂) and sulfur dioxide (SO₂) concentrations from 1995 to 2025 at two monitoring stations in Fargo and Bismarck, North Dakota, using seasonally averaged (summer and winter) observations. Historical trends show substantial declines in both pollutants over the 30-year period, with winter concentrations consistently exceeding summer values. To evaluate future air quality trajectories, four machine-learning models (i.e., random forest, extreme gradient boosting, support vector machine, and a feedforward neural network) were trained using engineered lag features and standardized predictors. Model validation with independent 2020–2024 data demonstrated that the neural network achieved the highest predictive accuracy across all scenarios, with coefficients of determination exceeding 0.96 and minimal forecast error. Using a recursive forecasting framework, pollutant levels were projected from 2025 to 2050. The forecasts indicate continued improvement in some cases, stabilization in others, and slight upward tendencies for specific pollutant-season combinations, particularly winter NO₂ in Bismarck and summer SO₂ in both cities. These results highlight that while historical emission reductions have been highly effective, future progress may not be uniform across pollutants or seasons. The findings underscore the importance of sustained monitoring and proactive regulatory planning to preserve North Dakota's historically clean air.

References

[1]	Mustafa ARA, Shokr MS, Alharbi T, et al. (2025) Integration of Google Earth Engine and aggregated air quality index for monitoring and mapping the spatio-temporal air quality to improve environmental sustainability in arid regions. Sustainability 17: 3450. https://doi.org/10.3390/su17083450 doi: 10.3390/su17083450
[2]	Badh A, Akyuz A, Vocke G, et al. (2009) Impact of climate change on the growing seasons in select cities of North Dakota, United States of America. Int J Clim Change Impacts Responses 1: 105–118. https://doi.org/10.18848/1835-7156/CGP/v01i01/37130 doi: 10.18848/1835-7156/CGP/v01i01/37130
[3]	Dhukuchhu A, Kaya O, Hatterman-Valenti H (2025) Planting date and cultivar selection effects on cauliflower growth, physiology, and yield performance in North Dakota growing conditions. Horticulturae 11: 1314. https://doi.org/10.3390/horticulturae11111314 doi: 10.3390/horticulturae11111314
[4]	Karim S, Ahmed F, Mollick T, et al. (2026) Climate-driven forecasting of agricultural land cover change in North Dakota's Red River Valley using remote sensing and machine learning. Earth Syst Environ https://doi.org/10.1007/s41748-026-01205-4 doi: 10.1007/s41748-026-01205-4
[5]	Biagi B, Brattich E, Cintolesi C, et al. (2025) Dynamical and chemical impacts of urban green areas on air pollution in a city environment. Urban Clim 60: 102343. https://doi.org/10.1016/j.uclim.2025.102343 doi: 10.1016/j.uclim.2025.102343
[6]	Zhou M, Xie Y, Wang C, et al. (2024) Impacts of current and climate-induced changes in atmospheric stagnation on Indian surface PM2.5 pollution. Nat Commun 15: 7448. https://doi.org/10.1038/s41467-024-51462-y doi: 10.1038/s41467-024-51462-y
[7]	Clean Air Task Force (2017) Health risks in North Dakota from oil and gas air pollution. Fact Sheet. Available from: https://www.catf.us/resource/health-risks-in-north-dakota/
[8]	Allison E, Mandler B (2018) Air quality impacts of oil and gas emissions from production, processing, refining, and use. American Geosciences Institute, Petroleum and the Environment Report. Available from: https://profession.americangeosciences.org/reports/petroleum-environment-2018/air-quality-impacts-oil-gas/
[9]	Prenni AJ, Day DE, Evanoski-Cole AR, et al. (2016) Oil and gas impacts on air quality in federal lands in the Bakken region: An overview of the Bakken Air Quality Study and first results. Atmos Chem Phys 16: 1401–1416. https://doi.org/10.5194/acp-16-1401-2016 doi: 10.5194/acp-16-1401-2016
[10]	Evanoski-Cole AR, Gebhart KA, Sive BC, et al. (2017) Composition and sources of winter haze in the Bakken oil and gas extraction region. Atmos Environ 156: 77–87. https://doi.org/10.1016/j.atmosenv.2017.02.019 doi: 10.1016/j.atmosenv.2017.02.019
[11]	Gebhart KA, Day DE, Prenni AJ, et al. (2018) Visibility impacts at Class Ⅰ areas near the Bakken oil and gas development. J Air Waste Manag Assoc 68: 477–493. https://doi.org/10.1080/10962247.2018.1429334 doi: 10.1080/10962247.2018.1429334
[12]	Schwarz JP, Holloway JS, Katich JM, et al. (2015) Black carbon emissions from the Bakken oil and gas development region. Environ Sci Technol Lett 2: 281–285. https://doi.org/10.1021/acs.estlett.5b00225 doi: 10.1021/acs.estlett.5b00225
[13]	Jeon JW, Park SW, Han YJ, et al. (2024) Nitrate formation mechanisms causing high concentration of PM2.5 in a residential city with low anthropogenic emissions during cold season. Environ Pollut 352: 124141. https://doi.org/10.1016/j.envpol.2024.12414 doi: 10.1016/j.envpol.2024.12414
[14]	CIRES (2019) A new view of wintertime air pollution: Study could help improve air quality in cities across the U. S. West. Cooperative Institute for Research in Environmental Sciences. Available from: https://cires.colorado.edu/news/new-view-wintertime-air-pollution
[15]	Munia MMFJ, Jannat JNJ, Mia MY, et al. (2023) A critical review and prospect of NO₂ and SO₂ pollution over Asia: Hotspots, trends, and sources. Sci Total Environ 876: 162851. https://doi.org/10.1016/j.scitotenv.2023.162851 doi: 10.1016/j.scitotenv.2023.162851
[16]	Elisephane I, Török Z (2025) Environmental determinants of ambient air pollution in urban East Africa: Unveiling critical insights from Kigali, Rwanda. Sci Afr 30: e03055. https://doi.org/10.1016/j.sciaf.2025.e03055 doi: 10.1016/j.sciaf.2025.e03055
[17]	Cohen AJ, Brauer M, Burnett R, et al. (2017) Estimates and 25-year trends of the global burden of disease attributable to ambient air pollution. Lancet 389: 1907–1918. https://doi.org/10.1016/S0140-6736(17)30505-6 doi: 10.1016/S0140-6736(17)30505-6
[18]	Landrigan PJ, Fuller R, Acosta NJR, et al. (2018) The Lancet Commission on pollution and health. Lancet 391: 462–512. https://doi.org/10.1016/S0140-6736(17)32345-0 doi: 10.1016/S0140-6736(17)32345-0
[19]	Olawade DB, Wada OZ, Ige AO, et al. (2024) Artificial intelligence in environmental monitoring: Advancements, challenges, and future directions. Hyg Environ Health Adv 12: 100114. https://doi.org/10.1016/j.heha.2024.100114 doi: 10.1016/j.heha.2024.100114
[20]	Wickham H, Averick M, Bryan J, et al. (2019) Welcome to the tidyverse. J Open Source Softw 4: 1686. https://doi.org/10.21105/joss.01686 doi: 10.21105/joss.01686
[21]	Grolemund G, Wickham H (2011) Dates and times made easy with lubridate. J Stat Softw 40: 1–25. https://doi.org/10.18637/jss.v040.i03 doi: 10.18637/jss.v040.i03
[22]	Kuhn M (2008) Building predictive models in R using the caret package. J Stat Softw 28: 1–26. https://doi.org/10.18637/jss.v028.i05 doi: 10.18637/jss.v028.i05
[23]	Liaw A, Wiener M (2002) Classification and regression by randomForest. R News 2: 18–22.
[24]	Chen TQ, Guestrin C (2016) XGBoost: A scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, 785–794. https://doi.org/10.1145/2939672.2939785
[25]	Meyer D, Dimitriadou E, Hornik K, et al. (2024) e1071: Misc functions of the Department of Statistics, Probability Theory Group (Version 1.7-16). R package.
[26]	Venables WN, Ripley BD (2002) Modern Applied Statistics with S. 4 Eds., New York: Springer. https://doi.org/10.1007/978-0-387-21706-2
[27]	Schauberger P, Walker A (2025) openxlsx: Read, write and edit xlsx files (Version 4.2. 8). R package. Available from: https://ycphs.github.io/openxlsx/index.html
[28]	US EPA (2024) Air pollution trends continue to link clean air and strong economy. Available from: https://www.epa.gov/newsreleases/air-pollution-trends-continue-link-clean-air-and-strong-economy
[29]	US EPA (n. d.) Clean Air Act overview. Available from: https://www.epa.gov/clean-air-act-overview/progress-cleaning-air-and-improving-peoples-health
[30]	Tırınk S (2025) Machine learning-based forecasting of air quality index under long-term environmental patterns: A comparative approach with XGBoost, LightGBM, and SVM. PLoS One 20: e0334252. https://doi.org/10.1371/journal.pone.0334252 doi: 10.1371/journal.pone.0334252
[31]	Seo JH, Cho J, Hong E, et al. (2026) Seasonally adaptive data-driven ozone prediction in megacity environments. Environ Res 293: 123741. https://doi.org/10.1016/j.envres.2026.123741 doi: 10.1016/j.envres.2026.123741
[32]	Lee J, Barquilla CAM, Park K, et al. (2024) Urban form and seasonal PM2.5 dynamics: Enhancing air quality prediction using interpretable machine learning and IoT sensor data. Sustain Cities Soc 117: 105976. https://doi.org/10.1016/j.scs.2024.105976 doi: 10.1016/j.scs.2024.105976
[33]	Abatzoglou JT, Williams AP (2016) Impact of anthropogenic climate change on wildfire across western US forests. Proc Natl Acad Sci USA 113: 11770–11775. https://doi.org/10.1073/pnas.1607171113 doi: 10.1073/pnas.1607171113
[34]	IPCC (2023) Climate Change 2023: Synthesis Report. Core Writing Team, Lee H, Romero J, Eds., Geneva: Intergovernmental Panel on Climate Change, 35–115. https://doi.org/10.59327/IPCC/AR6-9789291691647
[35]	Jaffe DA, O'Neill SM, Larkin NK, et al. (2020) Wildfire and prescribed burning impacts on air quality in the United States. J Air Waste Manag Assoc 70: 583–615. https://doi.org/10.1080/10962247.2020.1749731 doi: 10.1080/10962247.2020.1749731
[36]	Crippa M, Janssens-Maenhout G, Dentener F, et al. (2016) Forty years of improvements in European air quality: Regional policy-industry interactions with global impacts. Atmos Chem Phys 16: 3825–3841. https://doi.org/10.5194/acp-16-3825-2016 doi: 10.5194/acp-16-3825-2016
[37]	McLinden CA, Fioletov V, Shephard MW, et al. (2016) Space-based detection of missing sulfur dioxide sources of global air pollution. Nat Geosci 9: 496–500. https://doi.org/10.1038/ngeo2724 doi: 10.1038/ngeo2724
[38]	Lin M, Horowitz LW, Payton R, et al. (2017) US surface ozone trends and extremes from 1980 to 2014: Quantifying the roles of rising Asian emissions, domestic controls, wildfires, and climate. Atmos Chem Phys 17: 2943–2970. https://doi.org/10.5194/acp-17-2943-2017 doi: 10.5194/acp-17-2943-2017

Reader Comments

Your name:*

Email:*
© 2026 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)