Impact of PM10 concentrations: a Markov Chain approach to examining wind speed effects in Petaling Jaya

Air pollution, particularly PM10 concentration, poses significant health and environmental challenges in urban areas such as Petaling Jaya. Wind speed is known to influence pollutant dispersion, however, the dynamic relationship between wind speed and PM10 levels over time has not been adequately modelled using probabilistic approaches. This study aims to investigate the influence of wind speed on PM10 concentrations in Petaling Jaya from 2013 to 2023. It seeks to model the temporal evolution and dependency between wind speed and PM10 concentrations using a Markov Chain framework and to identify the most suitable statistical models for accurate prediction and assessment. A first order Markov Chain model was developed using Transition Probability Matrices (TPMs— mathematical frameworks that capture the likelihood of transitioning between different environmental states over time) constructed through the Count Method. The model’s assumptions, which include periodicity, irreducibility, and state classification, were validated through statistical tests. Additionally, a Partial Proportional Odds Model (PPOM), an Ordered Logit Model (OLM), and an Ordered Probit Model (OPM) were applied and compared with the Count Method. Monte Carlo simulations were used to assess the models’ performance under varying environmental conditions. The findings reveal that higher wind speeds significantly enhance the dispersion of PM10, whereas lower wind speeds lead to the accumulation of pollutants. Among the models, the OPM best captures the distribution of wind speed, whilst the PPOM demonstrates the highest accuracy in predicting PM10 concentrations. These results provide valuable insights into air quality management and environmental policymaking.