Evaluation of the SAPRC-07 mechanism against CSIRO smog chamber data

The updated SAPRC-07 mechanism was evaluated against data from experiments performed in the CSIRO smog chamber. The mechanism predictions have been compared to experimental results as well as predictions by SAPRC-99.Experiments were performed using either toluene or m-xylene in the presence of NO_x at sub-0.1 ppmv concentrations. For the majority of m-xylene experiments, the modelled Δ(O₃–NO) concentration was within 20% of observed values for both SAPRC mechanisms. However during the oxidation of toluene the production of radicals was poorly predicted, with final Δ(O₃–NO) concentration under-predicted by up to 60%. The predictions of major oxidants from isoprene oxidation were in good agreement with observed values. For the NO_x-limited conditions however, the ozone concentration predicted by both mechanisms were under-predicted by approximately 20% in the five experiments tested.The performance of the SAPRC-07 mechanism was also evaluated against twelve evaporated fuel experiments. Two types of evaporative mode experiments were performed: headspace evaporated fuel and wholly evaporated fuel. The major difference was a significantly higher concentration of aromatic hydrocarbons and larger alkane products in wholly evaporated fuels. For headspace evaporated fuel experiments both SAPRC mechanisms were in good agreement with experimental results. For wholly evaporated experiments the average Δ(O₃–NO) model error was ?25% with SAPRC-07 compared to less than ?5% for SAPRC-99. Updates to the photolysis data for dicarbonyls, the light source used and the experimental conditions under which these experiments were performed are possible causes for the discrepancy between SAPRC-99 and -07 predictions for wholly evaporated experiments.