Investigation of Voc Reactivity Effects Using Existing Regional Air Quality Models

Results to date on an ongoing project to assess VOC reactivity effects in the Eastern United States using the CRC-NARSTO database representing the July-5, 1995 episode and the Carbon Bond mechanism are described. The Direct Decoupled Method (DDM) as implemented in the CAMx model was used to calculate the incremental ozone impacts of 8 Carbon Bond VOC species, CO and ethane, and various methods were used to quantify ozone impacts and derive regional reactivity metrics based on them. The results indicated that relative ozone impacts of the modeled species varied significantly throughout the domain, but for most model species except for TOL reasonably consistent results are obtained if “minimum substitution error” or “regional maximum incremental reactivity” metrics are used. There were no large differences in regional relative reactivities derived based on maximum 8-hour ozone averages compared to those based on 1-hour averages. Metrics based on the regional maximum ozone concentration were found not to give consistent results. The regional reactivity results are similar to the reactivity scales derived using the Carbon Bond mechanism with the scenarios and methods of Carter (1994a) but some differences are observed. Although the large scale reactivity-based substitution calculations called for this project have not yet been conducted, a preliminary analysis of calculations where all anthropogenic VOCs are removed or replaced by ethane indicate that with a few exceptions their results are reasonably consistent with what is predicted by incremental reactivity analysis based on these DDM results. Preliminary conclusions obtained from the project thus far are summarized.