Abstract

Background and Objective: The analysis of the behavior of daily PM10 occurrence is becoming important nowadays and the results obtained may be useful for the prediction and decision making purposes. This study considered the behavior of PM10 concentration that related with its dependency nature. Therefore, this study is attempted to determine the sequences of polluted and non-polluted days affected by PM10 concentration based on the optimum order of a Markov chain model. Methodology: Twelve years of monitoring records which is from 2002-2013 and have been analyzed for this purpose. The PM10 concentration data that possess Markov chain properties show that the successive event is dependent on the previous event and is suited for further analysis using this model. Results: The optimum order of the Markov chain model for Shah Alam monitoring station shows that the order of two and three are optimum for threshold values less than 120 μg m–3 and a simple order is optimum for a threshold value of 150 μg m–3. The results mean that the occurrence of the polluted or non-polluted days affected by PM10 is dependent on the 2 or 3 days before the observed day for threshold value less than 120 μg m–3. For a threshold value of 150 μg m–3, the occurrence depends only on a day before the observed day. Besides that, the distribution of polluted events is well fitted based on the optimum order for each threshold value used. Conclusion: The information of polluted (non-polluted) occurrences is important in monitoring the PM10 concentrations which can be used for predicting related future events and helpful in providing the necessary precautionary measures to public and protect their health.