Total ozone over Ankara and its forecasting using regression models

Abstract

The present study is concerned with the total ozone variation in Ankara, Turkey (39°57′ N; 32°53′ E) and developing a total ozone prediction model by Total Ozone Mapping Spectrometer (TOMS) and single-site aerological data (1984–2006). The daily averages of the total column ozone over Ankara show a seasonal variation, with larger values in spring/winter and lower in autumn/summer and a large day-to-day variability. In this study, in order to forecast the total column ozone over Ankara, a multi-linear regression equation was performed. Predictors are selected by stepwise regression method. The ozone value on the previous day, temperature at the 100 hPa pressure level, previous day's temperature for the 700 hPa pressure level and temperature difference for 50 hPa and 500 hPa are the most important predictors. The success of monthly prognostic total column ozone models built up for Ankara is tested by using Brewer MK III Spectrophotometer data and satellite (OMI) data with the upper-air data in 2007. The Brewer MKIII has been in operation since November 2006 in Ankara. The ground-based one-year data show good agreement with the satellite OMI data.

Acknowledgements

This study has been part of the Project 105G032 supported by TUBITAK (Turkish Scientific and Technological Research Council).