Air Pollutant Regression Analysis (CO, O3 and PM10) and Meteorological Parameters, Case Study: Birjand-Iran

Abstract

Air pollution is a global threat to public and environmental health, especially in urban areas. Therefore, modeling is used to control and planing concentration of pollutants. In this paper, a model is proposed based on linear regression for short term forecasting of CO, PM10 and O3 based on meteorological parameters, and the results are presented. Data of meteorological parameters including humidity, pressure, minimum and maximum temperature, wind speed and wind direction (Birjand Meteorological Organization), and air pollution data (CO, PM10, and O3 concentrations) from the Birjand weather organization were prepared and used as daily average. SPSS16 software was used for linear regression modeling. The results showed that the highest correlation coefficient for CO pollutant with minimum temperature was 0.53 and the lowest correlation coefficient with the value of 0.166 was wtih the wind direction. The maximum correlation coefficient of PM10 contamination with wind speed was 0.33 and the lowest correlation coefficient of this pollutant with a pressure was 0.882. Finally, the highest correlation coefficient of O3 contamination with maximum temperature was 0.50 and also with regard to the regression coefficient obtained for carbon monoxide (R = 0.33) compared to the other two pollutants, has been obtained better.

Keywords

• modeling
• pollutants
• prediction
• linear regression
• meteorological parameters