Statistical distribution models for urban air quality management

The air quality data collected for four years (February 2005 to December 2008) from the continuous ambient air quality monitoring station set up by Tamilnadu Pollution Control Board (TNPCB) at Royapuram in Chennai city (India) has been analyzed. During peak hour traffic flows, it is found that carbon monoxide (CO), oxides of nitrogen (NOx) and suspended particulate matter (SPM) concentrations are violating the national ambient air quality standards (NAAQS) at the study region. The Emission inventory at the study region indicates that two wheelers are the dominating source of air pollution. The statistical distribution models have been fitted to the Royapuram air quality data using the SYSTAT software. Two parameter statistical distributions such as normal, lognormal, Weibull, chi square, gamma and exponential have been fitted using the method of moments and maximum likelihood methods presented in Jakeman et al. (1986). The goodness of fit test has been used to identify the suitable statistical distribution model. The result shows that the two parameter lognormal distribution model is best fitted to the Royapuram air quality data.