Analysis of the Impact of Weather on PM₂.₅ Variability Using Machine Learning-Based Predictive Method

Agnieszka Głowacka; Aleksander Bies; Natalia Malcherczyk-Bies; Paulina Trybek

doi:10.31449/upinf.290

Authors

Agnieszka Głowacka University of Silesia, Faculty of Science and Technology
Aleksander Bies University of Silesia in Katowice, Faculty of Science and Technology
Natalia Malcherczyk-Bies University of Silesia in Katowice, Faculty of Natural Sciences
Paulina Trybek University of Silesia in Katowice, Faculty of Science and Technology

DOI:

https://doi.org/10.31449/upinf.290

Keywords:

air quality modeling, fine particulate matter, machine learning, meteorological drivers, PM₂.₅

Abstract

Fine particulate matter smaller than 2.5 micrometers significantly impacts human health, yet its concentrations are governed by complex interactions between emissions and meteorological conditions. This study investigates the seasonal drivers of PM₂.₅ in the Silesian Voivodeship and assesses the predictive performance of machine learning models, including Light Gradient Boosting Machine, Random Forest, and Support Vector Regression. Meteorological variables, temporal factors, and gaseous pollutants were analyzed to identify key determinants of pollution episodes. During the heating season, elevated residential heating emissions combined with stagnant atmospheric conditions dominate PM₂.₅ variability, while in the non-heating season, atmospheric dynamics and seasonal timing play a stronger role. Light Gradient Boosting Machine consistently achieved the highest predictive accuracy, capturing nonlinear interactions and providing interpretable insights into feature importance. These results highlight the potential of machine learning approaches for accurate air quality modeling and inform strategies for pollution mitigation and public health protection.

Author Biographies

Agnieszka Głowacka, University of Silesia, Faculty of Science and Technology

Agnieszka Głowacka is a second-year Master’s student of Micro and Nanotechnology at the University of Silesia in Katowice. Her interests focus on the application of modern technologies in the analysis and processing of scientific data. She is particularly interested in interdisciplinary approaches combining elements of physics, chemistry, and informatics, as well as practical applications of advanced technologies in science and industry.
Aleksander Bies, University of Silesia in Katowice, Faculty of Science and Technology

Aleksander Bies is a third-year PhD student in Computer Science at the University of Silesia in Katowice. His research interests focus on machine learning and the application of artificial intelligence methods to time-series analysis. He is particularly interested in developing predictive models, identifying temporal patterns, and applying advanced computational techniques to complex sequential data.
Natalia Malcherczyk-Bies, University of Silesia in Katowice, Faculty of Natural Sciences

Natalia Malcherczyk-Bies is a PhD student at the Institute of Biology, University of Silesia in Katowice. Her research interests focus on plant biology, the effects of environmental factors on plant development, and the use of plants from novel ecosystems as resources for creating green areas to support urban adaptation to climate change.
Paulina Trybek, University of Silesia in Katowice, Faculty of Science and Technology

Paulina Trybek is an assistant professor at the University of Silesia in Katowice and works at the Institute of Physics, where she specializes in complex systems and biomedical time-series analysis. She is actively involved in numerous student projects, supporting the development of data analysis competencies. She is also the coordinator of the project “Developing Talents in Artificial Intelligence to Solve Disruptive Environmental Problems”.