Multivariate Regression Analysis for Ground-Level Ozone Modeling in Kurupelit, Samsun, Turkey

Abstract

In this study, the multivariate linear regression was described by the hourly frequency distributions of continuous measurements using meteorological variables and ozone levels. The ambient O<inf>3</inf> concentration of was monitored continuously in the town of Kurupelit, Samsun, Turkey for one year (2005-2006). The meteorological variables were temperature, relative humidity, wind speed, solar radiation, and ambient pressure. A negative association was observed between O<inf>3</inf> and temperature a weak relationship (r2=0,289), and between O<inf>3</inf> and solar radiation a weak relationship (r2=0,102). The data from the continuous monitoring of O3 shows a negative correlation with relative humidity and atmospheric pressure.