Comparisons of the response of simulated ozone concentrations to changes in emissions using the Carbon Bonb IV and SAPRC99 chemical mechanisms

There is a trend in regional air quality modeling studies towards more detailed chemical mechanisms. In California, the Carbon Bond IV (CB4) chemical mechanism (32 chemical species -- 92 reactions) is giving way to the SAPRC99f (fixed) mechanism (76 chemical species -- 217 reactions). However, aside from comparisons from environmental chamber studies, there have been few reports describing the impact of changing the chemical mechanism on ozone concentrations simulated for regional air quality studies.

In this study, we compared air quality simulations using the CB4 and SAPRC99f chemical mechanisms for two modeling domains in California. The CALGRID model was used to simulate the August, 1997 episode of the South California Ozone Study (SCOS) in the Los Angeles area and the CAMx model was used to simulate the July/August, 2000 episode of the Central California Ozone Study (CCOS) in the San Joaquin Valley. For both episodes, air quality simulations were run using base-year anthropogenic emissions of reactive organic gases (ROG) and oxides of nitrogen (NOx) reduced by 20, 40, 60, and 80%. For the SCOS episode, the peak ozone concentration from each simulation was extracted from the subregion of the domain where the peak ozone concentration was observed. For the CCOS episode, peak simulated ozone concentrations were extracted from 3 such subregions. The relative reduction in simulated ozone concentrations resulting from the emissions reductions, for each chemical mechanism, was determined.

The results of this analysis showed that ozone concentrations simulated using the SAPRC99 chemical mechanism were generally more responsive to ROG emissions reductions than when using the CB4 chemical mechanism. The response to NOx emissions reductions were similar for both mechanisms above a certain threshold of emissions. This threshold appeared to be correlated with the ROG/NOx emissions ratio; however, this was not conclusive.