Air Quality Modeling and Decisions for Ozone Reduction Strategies

Despite the widespread application of photochemical air quality models (AQMs) in U.S. state implementation planning (SIP) for attainment of the ambient ozone standard, documentation for the reliability of projections has remained highly subjective. An "idealized" evaluation framework is proposed that provides a means for assessing reliability. Applied to 18 cases of regulatory modeling in the early 1990s in North America, a comparative review of these applications is reported. The intercomparisons suggest that more than two thirds of these AQM applications suffered from having inadequate air quality and meteorological databases. Emissions representations often were unreliable; uncertainties were too high. More than two thirds of the performance evaluation efforts were judged to be substandard compared with idealized goals. Meteorological conditions chosen according regulatory guidelines were limited to one or two cases and tended to be similar, thus limiting the extent to which public policy makers could be confident that the emission controls adopted would yield attainment for a broad range of adverse atmospheric conditions. More than half of the studies reviewed did not give sufficient attention to addressing the potential for compensating errors. Corroborative analyses were conducted in only one of the 18 studies reviewed. Insufficient attention was given to the estimation of model and/or input database errors, uncertainties, or variability in all of the cases examined. However, recent SIP and policy-related regional modeling provides evidence of substantial improvements in the underlying science and available modeling systems used for regulatory decision making. Nevertheless, the availability of suitable databases to support increasingly sophisticated modeling continues to be a concern for many locations. Thus, AQM results may still be subject to significant uncertainties. The evaluative process used here provides a framework for modelers and public policy makers to assess the adequacy of contemporary and future modeling work.