An event of heavy pollution in Chengdu during 1-6 January 2017 was analyzed in this study by using the data of micro-pulse lidar, the concentration of PM2.5, surface meteorological observations and vertical soundings. The results show that the extinction coefficient calculated by the lidar data is highly consistent with the change of PM2.5 concentration. When PM2.5 concentration increases, the near-surface extinction coefficient increases. Conversely, the near-surface extinction coefficient decreases. For this particular case, when there is no influence of cold air, diurnal variations of mixing layer height and relative humidity have obvious influence on the profile of extinction coefficient. The decrease of mixing layer height and the increase of relative humidity result in the increase of extinction coefficient and the aggravation of the pollution. The sky condition has a significant effect on the vertical distribution of aerosol. In sunny or cloudy days, the water vapor and a large number of aerosols concentrate in the area below the top of inversion layer because of the strong inversion in the morning. As the aerosol layer thickens, the near-surface extinction coefficient decreases significantly due to the evenly mixing of aerosols in the mixed layer with the development of mixed layer at noon. When the previous day was sunny or cloudy and the very day is overcast, the aerosol is obviously divided into two layers in the morning, one is near the ground and the other one is near the top of the residual layer. Part of residual layer aerosol is mixed downward into the mixed layer because the vertical turbulence at noon makes the extinction coefficient in the mixed layer increase significantly. The strong inversion layer near the surface, the decrease of the height of the mixed layer, the downward mixing of the residual aerosol layer and the increase of relative humidity are all responsible for the aggravation of the pollution.