Investigating the relationship between PM10 and climatic parameters using GAM model: a case study of Skopje, Kumanovo, Bitola, North Macedonia

This study applies Generalized Additive Models (GAMs) to investigate the complex non-linear relationship between PM10 concentrations and climate factors in Skopje, Kumanovo and Bitola from 2012 to 2023. The three largest Macedonian cities have continuous challenges with the polluted air, which results from a combination of natural and socio-economic factors. Researches for the climate influences on air pollution are crucial for understanding when and why pollution episodes occur, which in turn informs more effective forecasting, public health warnings, and mitigation strategies. GAMs allow for a flexible, non-linear assessment of how temperature, relative humidity, wind speed, and precipitation influence air pollution concentrations. The analysis showed a consistent and significant negative relationship between temperature and PM10 concentrations, indicating higher pollution levels under colder conditions. Precipitation was generally associated with lower PM10 concentrations, whereas relative humidity was mostly positively associated with PM10, although, these relationships were not statistically significant for all observations. Wind speed showed variable associations with PM10 concentrations, with significant negative relationships observed only at some monitoring locations. These insights emphasize the role of climate forecasting in air quality management and highlight the need for targeted pollution mitigation strategies in North Macedonia.