Case Study on Pollutant Trend Analysis: Ozone
EPA analyses show that over the past 20 years, national 1-hr average O3 concentrations have decreased by about 18%; most major metropolitan areas showed downward tends, and the Northeast and West showed the greatest improvements. In contrast, the South and North-central regions of the country and many national parks show increased O3 concentrations over the past 10 years. However, numerous uncertainties are associated with those O3 trend estimates. The formation and accumulation of O3 are strongly affected by prevailing meteorological conditions. Therefore, O3 trends, driven by emission-control policies, can be easily masked by year-to-year meteorological variability, and trend analyses can reach very different conclusions based on the methods used to factor out meteorological influences. O3 trends have been routinely examined using the second highest daily maximum 1-hr average concentration in a given year, because that value is used to establish compliance with the 1-hr O3 NAAQS; however, that value tends to fluctuate with a larger amplitude than a more robust statistic, such as the 95th percentile. As a result, subtle trends can be masked by the fluctuations of the extreme values.
Several research groups have questioned EPA’s reported O3 trends and carried out independent studies using data from national monitoring networks. A comprehensive review of such studies is presented by Wolff et al. (2001). Some locations, such as Los Angeles, have shown consistently decreasing O3 trends, but national-scale studies paint a more complicated picture. Milanchus et al. (1998) examined daily maximum 1-hr O3 concentrations over the period 1984–1995 and found that most locations exhibited downward trends during that time, and a few isolated sites showed upward trends. Fiore et al. (1998) examined median summer afternoon O3 concentrations for the years 1980–1995 and found that trends were insignificant over most of the continental United States, although O3 concentrations decreased in New York, Los Angeles, and Chicago. A follow-up study examined trends in the number of exceedances of the O3 standards (Lin et al. 2001). Over the period 1980 to 1998, strong downward trends in the number of days in exceedance of either O3 standard were found along the northeastern and southwestern coasts, and significant upward trends were found in various locations elsewhere in the country. In most locations, the O3 air quality improvements seen in the 1980s leveled off in the 1990s. The results differ among those and other studies because they are based on different yearly ranges, time periods within each year, pollution metrics, and methods to account for meteorological effects. Thus, it is very difficult to compare and evaluate the validity of reported O3 trends systematically based on past analyses.