**2. Materials and Methods**

## *2.1. Data*

The National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) daily reanalysis data of the zonal wind and vertical velocity from 1989 to 2018 are used with the horizontal resolution of 2.5◦ × 2.5◦ (the number of grid points is 144 × 73) and the vertical resolution of 17 layers.

The surface pollutant data and the surface meteorological variables including the relative humidity, temperature, surface meridional wind and surface zonal wind data are derived from the second version of the high-resolution air pollution reanalysis dataset in China in the Science Data Bank during 2013–2018. The dataset mainly contains two parts. The first part is the surface concentration reanalysis data of six conventional air pollutants (PM2.5, PM10, SO2, NO2, CO and O3) in China from 2013 to 2018. These data are obtained by assimilating the surface observation data provided by the China National Environmental Monitoring Center by using the ensemble Kalman filter and the Nested Air Quality Prediction Modeling System. The second part is the Weather Research and Forecasting model simulation data of surface meteorological variables including the wind speed, temperature, air pressure and relative humidity during the same period. The spatiotemporal resolution of the dataset is high with the temporal resolution of 1 h and the spatial resolution of 15 km. By using cross-validation, independent data verification and comparing with similar data at home and abroad, it was found that the dataset is highly accurate [22]. In this study, four surface pollutants of the NO2, PM10, PM2.5 and O3 are selected as the main research objects, and the O3 data are processed into the format of the maximum concentration in 8 h per day.
