Circular Statistical Analysis Of Wind Direction And Wind Speed On Ozone Concentration

• المؤلف الرئيسي: Zinol, Noorain
• التنسيق: أطروحة
• اللغة: الإنجليزية
• منشور في: 2017
• الموضوعات: T Technology; TA1-2040 Engineering (General). Civil engineering (General)
• الوصول للمادة أونلاين: http://eprints.usm.my/46469/

______________________________________________________________ The rapid development along with increased population and transportation usage has deteriorated air quality status in Malaysia. The purpose of this research is to predict and modelling the casual relationship between circular and linear variables using probability density functions and regression analysis. The observations were made at Shah Alam, Prai, Pasir Gudang and Jerantut station over ten years period from 2004 until 2013 respectively. Mean values for O3 concentrations at all locations did not exceed MAAQG level limit. Circular density and wind rose plot were used to describe the characteristics of wind direction including mean direction, mean resultant length and concentration. This study suggested the presence of prominent wind that were blowing from north and south direction of Peninsular Malaysia due to southwest monsoon and northeast monsoon. Four distributions function were fitted to the wind direction data and it was found that wrapped Cauchy fit the data very well based on mean chord length. The analysis was further conducted with circular-linear correlation, circular-linear regression and multi linear regression model. Most of positive correlations were found between wind direction and O3 concentration in all stations during daytime and nightime. From the circular – linear regression, O3 concentration were predicted from wind direction and the results showed that few values in Shah Alam station exceeded the permissible limit by MAAQG (0.06 ppm). Prediction model were continued with multi linear regression by including the sine and cosine of wind direction as independent variables as well as wind speed variable. It was found that nighttime models provide better accuracy than daytime.