Modelling of PM 10 Concentration in Industrialized Area in Malaysia: A Case Study in Nilai Norazian Mohamed Noor 1 , Cheng Yau Tan 1 , Mohd Mustafa Al Bakri Abdullah 2 , Nor Azam Ramli 3 and Ahmad Shukri Yahaya 3 1 School of Environmental Engineering, Universiti Malaysia Perlis, Pejabat Pos Besar 01007 Kangar, Perlis 2 School of Material Engineering, Universiti Malaysia Perlis, Pejabat Pos Besar 01007 Kangar, Perlis 3 School of Civil Engineering, Universiti Sains Malaysia, Engineering Campus 14300 Nibong Tebal, Pulau Pinang Abstract. Three distributions, namely Weibull, log-normal and gamma were chosen to model the PM 10 observations at selected industrial area i.e. Nilai, Negeri Sembilan. One-year period hourly average data for 2006 and 2007 was used for this research. For parameters estimation, method of maximum likelihood estimation (MLE) was selected. Four performance indicators that are mean absolute error (MAE), root mean squared error (RMSE), coefficient of determination (R 2 ) and prediction accuracy (PA), were applied to determine the goodness-of-fit criteria of the distributions. The best distribution that fits with the PM 10 observations in Nilai was found to be gamma distribution. The probabilities of the exceedences concentration were calculated and the return period for the coming year was predicted from the cumulative density function (cdf) obtained from the best-fit distributions. For the 2006 data, Nilai was predicted to exceed 150 μg/m 3 for 2.7 days in 2007 with a return period of one occurrence per 137 days. Keywords: Particulate matter, statistical analysis, probability distributions, performance indicators, exceedences, return period. 1. Introduction Exponential development of science and technology nowadays has lead to the rapid growing industrialization which is the major sources of various environmental pollutions, especially air pollution. Air pollutants, specifically particulate matter (PM) smaller than about 10 micrometers, referred as PM 10 , have received extensive attention, due to its capability to settle in the bronchi and lungs and cause health problems. Malaysian Ambient Air Quality Guidelines (MAAQG) were issued and target values for annual and daily mean mass concentrations for various air pollutant were established to control and reduce air pollutant levels in the atmosphere. Monitoring data and studies on ambient air quality show that some of the air pollutants in several large cities are increasing with time and are not always at acceptable levels according to the MAAQG. There are very limited data and case studies on air pollution in our country. While the application is almost non-existent in our country, it is an attractive analytical option as it can reasonably predict the return period and exceedences in the succeeding period to meet the evolving information needs of environmental quality management [1]. Many types of probability distributions have been used to fit air pollutant concentrations including Weibull distribution [2], lognormal distribution [3], gamma distribution [4] and Rayleigh distribution [5]. Lu [6] and Chen et al. [7] have studied the goodness-of-fit for selected probability distributions by using several performance indicators such as mean absolute error (MAE), root means error (RMSE), index of agreement (d 2 ), bias (B), normalized absolute error (NAE), prediction accuracy (PA) and coefficient of determination (R 2 ). 18 2011 International Conference on Environment and Industrial Innovation IPCBEE vol.12 (2011) © (2011) IACSIT Press, Singapore