Abstract
Ozone formation in California in year 2017 was simulated using three chemical mechanisms: the Statewide Air Pollution Research Center 2007 mechanism with updated toxics and isoprene chemistry (S07TIC); the Carbon Bond 6, revision 2 (CB6r2) mechanism; a modified CB6r2 mechanism with increased hydroxyl radical (OH) production from isoprene oxidation at low nitrogen oxide (NOx) concentrations (CB6r2OH). The simulations with a photochemical grid model (PGM) showed little difference in ozone between CB6r2 and CB6r2OH. Further analysis with a box model found little sensitivity of ozone to OH production in the isoprene chemistry of all three mechanisms under conditions representative of California. The S07TIC generally predicts greater ozone than the CB6r2, and the difference in the predicted concentrations exceeds the uncertainty estimated from uncertainties in the emissions inventory. The sensitivity of ozone to anthropogenic volatile organic compound (VOC), NOx and carbon monoxide (CO) emissions and to isoprene emissions was also calculated with the PGM. Accounting for uncertainties in the emissions inventory, the sensitivity to NOx emissions is significantly different between S07TIC and CB6r2 in central and northern California; the sensitivity to isoprene emissions is significantly different in southern California. All mechanisms give a negative sensitivity of ozone to isoprene emissions in areas of northern California where NOx emissions are small. The CB6r2 and S07TIC differ in the sensitivity of ozone to organic nitrate (ON) formation from isoprene oxidation, likely because all ONs recycle to NOx in the S07TIC mechanism but only first-generation ONs do so in the CB6r2. For current California conditions, uncertainties in OH production from isoprene have little impact on ozone, but the extent to which isoprene ONs recycle NOx does impact ozone concentration and ozone sensitivity to NOx emissions.