AN EXAMINATION OF URBAN OZONE TRENDS IN THE NORTHEASTERN UNITED-STATES (1973-1983) USING A ROBUST STATISTICAL-METHOD

Stringent volatile organic compounds (VOC) control measures for both mobile and stationary sources have been introduced in the U.S. in an attempt to reduce ambient concentrations of ozone (O3). Despite the expense of billions of dollars to implement these measures, the impact on O3 air quality is still unknown, and most major metropolitan areas are still in violation of the daily 1-h National Ambient Air Quality Standard for O3 of 120 ppb. This study was initiated to examine the trends of O3 from 1973 to 1983 at eight major population centers in the northeastern U.S. using a robust statistical method. The effects of the late 1970s change in O3 calibration methodology and meteorology on these trends were also examined. The 75th percentile was shown to be a good statistic for determining trends and was used for analysis of the trends before and after the calibration change was applied. The surface temperature and upper air temperature variables were found to be the best predictors of O3 behavior. Two regression procedures, the fixed-range and fixed-value techniques, were performed to remove the variability of meteorological conditions conducive to high O3 (i.e. O3 > 80 ppb). Finally, a robust, nonparametric statistical procedure, which does not require specific statistical assumptions to be met, was used for each set of estimates derived above. The results of these analyses show there has been a decrease of a few ppb on a yearly basis for the majority of the sites.