Modelling climatic variable impacts on ground-level ozone in Malaysia using backward trajectory and Generative Additive Models

Climate change has a recognized global effect on ozone concentration, yet its impact varies across regions and countries. Local studies are imperative for precisely evaluating the accurate, robust, and up-to-date relationship between climatic variables and ozone concentration at regional scale. In this work, we elucidate the spatiotemporal and seasonal variability of ground-level ozone (O₃) in Malaysia using backward trajectory and Generative Additive Model. Concentrations of O₃ and other air pollutants (NO₂, CO, SO₂ and PM_2.5) from a total of 43 air quality stations across the country from 2107 to 2020 have been analyzed along with the meteorological auxiliary data. Ozone pollution is susceptible in the Central, Northern and Southern of Peninsular Malaysia, and occurs at different times subject to the monsoon variability. In the Central zone, 60% of days during March and April had unhealthy ozone levels with a maximum daily averaged O₃ 73.5 ± 9.3 ppb. The backward trajectory analysis indicates that ozone pollution in the Central zone is strongly affected by northeasterly transboundary air pollution from Indochina and East China. The Generative Additive Model analysis highlights O₃ variability in the Central zone is possibly modulated by stratospheric air intrusion and PM_2.5 inhibitory effect that suppressed surface solar radiation and weakened O₃ production. Overall, the work advances the understanding of O₃ variability in Malaysia, provides valuable insights into complex interplay between O₃ concentrations and climatic variables, and offers a framework for future research in air quality modeling.