A study of spatial distribution of ozone and its correlation with meteorological parameters and precursors

Scope and method of study. The existing status of Oklahoma is complete attainment with the 1-hr and 8-hr ozone standards set by the NAAQS. However, Skiatook a suburban site in Tulsa county showed higher exceedances of ozone standards in the years 1998 to 2000. Temporal analysis was performed to study ozone exceedance patterns and geostatistical analysis was performed to study diurnal variation of ozone in Oklahoma and specifically in Skiatook. An investigation was performed on high ozone episodes in Skiatook to study the effects of temperature, relative humidity, wind speed, solar radiation and pressure on ozone concentrations. Multi-linear regression analysis was performed to forecast next day 1-hr ozone, 8-hr ozone, and daily maximum 1-hr ozone in Skiatook. The data from 10:00 a.m. to 5:00 p.m. for the months of May to September from 1998--2000 was used for the analysis. The forecasted values obtained from these models were tested to be within +/- 0.005 ppm of the actual ozone values. The geostatistical analysis was performed using software GS+ and the regression analysis was performed using MINITAB 14.; Findings and conclusions. Temperature was the only meteorological parameter found to have noticeable effect on high ozone concentrations in Skiatook. The model using the meteorological parameters and nitrogen dioxide as predictors showed the highest percentage (35.88%) of satisfying the criterion of forecasted values being within +/- 0.005 ppm of the actual ozone values. It was found that using the previous day ozone concentration as a predictor helped achieve a higher R2 value for ozone forecast. The unavailability of actual nitrogen dioxide concentrations in Skiatook resulted in poor forecasts using the models developed.