The implementation of different chemical mechanisms in air quality models can exert a significant impact on simulations of ozone and its precursors in the troposphere. In the present study, we applied three different chemical mechanisms (RADM2, SAPRC99 and CBMZ) in a three-dimensional mesoscale air quality model, WRF-Chem, to capture the spatiotemporal variations of the surface ozone and its precursors such as NO_x and VOCs in Shanghai, China in May 2016, so that the influences on simulations exerted by using different chemical mechanisms can be clarified. Observational data from surface monitoring stations were also adopted for the comparison of the model simulations. By comparing the model predictions with the observational data, we found that  the model with different chemical mechanisms can generally capture the temporal change of ozone in Shanghai during the simulated time period. However, it was found that SAPRC99 predicts a higher ozone than the other two mechanisms during a high-ozone period, while RADM2 gives a higher ozone level in other time periods. Moreover, in simulations using RADM2 and SAPRC99, NO₂ is overestimated. In contrast, NO₂ predicted by CBMZ is closer to the observational data. A comparison between simulations using different mechanisms also shows that SAPRC99 predicts higher OH, HO₂ and NO_x than the other two mechanisms. In contrast, RADM2 predicts the highest HNO₃ and ozone among these three mechanisms. In addition, we also found that these simulations using different mechanisms all indicate that  Shanghai city is VOC-controlled. However, VOCs and NO_x concentrations predicted by CBMZ were found to be mostly lower than those predicted by the other two mechanisms under a same condition.
 

Ozone；WRF-Chem；Chemical mechanisms