Regression analysis of the effect of meteorological parameters on air quality in three neighboring cities located on the Mediterranean coast of Turkey

Abstract

For the past couple of decades, the increasing health problems caused by air pollution have caught more and more attention. As a result, analysis and prediction of the air quality were widely studied. Several methodologies, both deterministic and statistical, have been proposed. In this study, we use the linear model to detect the relationship between the concentration of an air pollutant (PM10, i.e. particulate matter with an aerodynamic diameter of less than 10 µm) at a specific site and meteorological variables in three neighboring cities, namely Antalya, Isparta and Burdur, located on the Mediterranean region of Turkey. Daily average PM10 concentration values for the years 2014 and 2015 were used as the dependent variable. The PM10 concentrations in each city is measured at air quality stations operated by Ministry of Civilization and Environment. The PM10 concentrations were tried to be explained by using different meteorological parameters as independent variables. The meteorological parameters used in this study were hours of sunshine, wind speed, difference between maximum and minimum temperature, average daily temperature, actual pressure, precipitation and relative humidity. The results showed that in all three cities, increase in hours of sun shine and in wind speed, decreases the PM10 concentration in the ambient atmosphere. Furthermore, increase in pressure and in difference between maximum and minimum temperature, increases the PM10 levels. The R values for the cities varies from 0.62 and 0.68, which were statistically significant in 95% confidence level.For the past couple of decades, the increasing health problems caused by air pollution have caught more and more attention. As a result, analysis and prediction of the air quality were widely studied. Several methodologies, both deterministic and statistical, have been proposed. In this study, we use the linear model to detect the relationship between the concentration of an air pollutant (PM10, i.e. particulate matter with an aerodynamic diameter of less than 10 µm) at a specific site and meteorological variables in three neighboring cities, namely Antalya, Isparta and Burdur, located on the Mediterranean region of Turkey. Daily average PM10 concentration values for the years 2014 and 2015 were used as the dependent variable. The PM10 concentrations in each city is measured at air quality stations operated by Ministry of Civilization and Environment. The PM10 concentrations were tried to be explained by using different meteorological parameters as independent variables. The meteorological parameters used...