Demonstrating attainment of the air quality standards: Integration of observations and model predictions into the probabilistic framework

This paper introduces an integrated observational-modeling approach to tran sform the deterministic nature of attainment demonstrations of the National Ambient Air Quality Standard (NAAQS) into the probabilistic framework. Whi le the methods presented here can be used to address any air quality standa rd that is based on extreme values, this paper focuses on the application t o the 1-hr and 8-hr NAAQS for ozone. Extreme value statistics and resamplin g techniques are applied to estimate the probability of exceeding the NAAQS for both 1-hr and 8-hr ozone concentrations. Within the integrated observa tion-modeling analysis approach, we show that the model-to-model difference s in the predicted responses to emission reductions are smaller than the mo del-to-model differences in predicted absolute ozone concentrations. We ill ustrate that the emission reductions stemming from a real-world emission co ntrol strategy would substantially reduce the probability of exceeding the NAAQS over a large portion of the eastern United States, especially for the 8-hr average ozone concentrations.