Modelling distribution function of surface ozone concentration for selected suburban areas in Malaysia

Abstract

Ozone is known as an important secondary pollutant in the atmosphere. The aim of this study is to find the best fit distribution for calculating exceedance and return period of ozone based on suburban areas; Perak (AMS1) and Pulau Pinang (AMS2). Three distributions namely Gamma, Rayleigh and Laplace were used to fit 2 years ozone data (2010 and 2011). The parameters were estimated by using Maximum Likelihood Estimation (MLE) in order to plot probability distribution function (PDF) and cumulative distribution function (CDF). Four performance indicators were used to find the best distribution namely, normalized absolute error (NAE), prediction accuracy (PA), coefficient of determination (R2) and root mean square error (RMSE). The best distribution to represent ozone concentration at both sites in 2010 and 2011 is Gamma distribution with the smallest error measure (NAE and RMSE) and the highest adequacy measure (PA and R2). For the 2010 data, AMS1 was predicted to exceed 0.1 ppm for 2 days in 2011 with a return period of one occurrence per 179 days. However, AMS2 do not exceed MAAQG limit (0.1 ppm) based on both 2010 and 2011. From the results, the exceedance and return period can be used as a guidance to overcome future air pollution problem