[Scale Dependence Between PM_2.5 and Meteorological Factors and Its Influencing Factors in "2+26" Cities]

Based on the PM_2.5 concentration and meteorological data of "2+26" cities, the variations in PM_2.5 time series were analyzed by the continuous wavelet transform(CWT) and discrete wavelet transform(DWT). Wavelet coherence(WTC) and multiple wavelet coherence(MWC) were used to quantify the response relationship between PM_2.5 and single/multiple meteorological factors in the time-frequency domain. Partial wavelet coherence(PWC) was used to quantitatively evaluate the influence of atmospheric teleconnection factors on the response relationship. The results showed that:① the concentration of PM_2.5 in the "2+26" cities had the spatial distribution characteristics of high in the middle area and low in the peripheral area. The PM_2.5 mutation events were mainly concentrated before 2018 and mostly occurred in winter when the meteorological conditions were stable. The annual scale period of 256-512 d was relatively stable, and it was also the dominant period of the PM_2.5 time series. ② The coherences between PM_2.5 and meteorological factors depended on the time-frequency scale and variable combination. At all time-frequency scales, PM_2.5 had strong coherences with relative humidity and temperature. At small and medium time-frequency scales, PM_2.5 had strong coherences with wind speed. At large scales, PM_2.5 had strong coherences with temperature. The combination of precipitation, temperature, and relative humidity could explain the variation in PM_2.5 at all time-frequency scales. ③ At different time-frequency scales, the enhancement/weakening effects of atmospheric teleconnection factors on the response relationship were not the same. At all time-frequency scales, the El Niño-Southern Oscillation(ENSO) had a greater impact on the response relationship between PM_2.5 and precipitation/temperature, and the Pacific decadal oscillation(PDO) had a greater impact on the response relationship between PM_2.5 and relative humidity/wind speed. These results provide reference for regional air pollution control.