Numerical study of the effects of initial conditions and emissions on PM<sub>2.5</sub> concentration simulations with CAMx v6.1: a Xi'an case study

Abstract. A series of model sensitivity experiments was designed to explore the effects different initial conditions and emissions in Xi'an December 2016; is a major city Fenwei Plain, which key area with respect air pollution control China. Three methods were applied for condition tests: clean simulation, restart continuous simulation. In simulation test, C00, C06, C12, C18, C24 conducted effect intercepted time periods used. The results these showed that fine particulate matter (PM2.5) performance better when start delayed. For C00 C24, absolute mean bias (MB) decreased from 51.07 3.72 µg m?3, index agreement (IOA) increased 0.49 0.86, illustrates much than C00. R1120 R1124 used and, turn, date first day While times simulations different, nearly consistent after spin-up period, revealed approximately 27 h. CT12 CT24 conducted. same, almost identical. Based on results, best all experiments, correlation coefficient (R), MB, IOA reaching 0.81, 6.29 0.90 respectively. emission tests, an updated local inventory construction fugitive dust added compared original inventory. PM2.5 modelling. Therefore, combining method can satisfactorily improve Xi'an: MB 35.16 reached 0.90.