Variations in the water vapor distribution and the associated effects on fog and haze events over Xi'an based on Raman lidar data and back trajectories.

A combination of more than two years of water vapor lidar data with back trajectory analysis using the hybrid single-particle Lagrangian integrated trajectory (HYSPLIT) model was used to study the long-range transport of air masses and the water vapor distribution characteristics and variations over Xi'an, China (34.233° N, 108.911° E), which is a typical city in Northwest China. High-quality profiles of the water vapor density were derived from a multifunction Raman lidar system built in Xi'an, and more than 2000 sets of profiles with >400 nighttime observations from October 2013 to July 2016 were collected and used for statistical and quantitative analyses. The vertical variations in the water vapor content were discussed. A mutation height of the water vapor exists at 2-4 km with a high occurrence rate of ∼60% during the autumn and winter seasons. This height reflects a distinct stratification in the water vapor content. Additionally, the atmospheric water vapor content was mainly concentrated in the lower troposphere, and the proportion of the water vapor content at 0.5-5 km accounted for 80%-90% of the total water vapor below 10 km. Obvious seasonal variations were observed, including large water vapor content during the spring and summer and small content during the autumn and winter. Combined with back trajectory analysis, the results showed that markedly different water vapor transport pathways contribute to seasonal variations in the water vapor content. South and southeast airflows dominated during the summer, with 30% of the 84 trajectories originating from these areas; however, the air masses during the winter originated from the north and local regions (64.3%) and from the northwest (27%). In addition, we discussed variations in the water vapor during fog and haze weather conditions during the winter. A considerable enhancement in the mean water vapor density at 0.5-3 km exhibited a clear positive correlation (correlation coefficient >0.8) with the PM2.5 and PM10 concentrations. The results indicate that local airflow trajectories mainly affect water vapor transport below the boundary layer, and that these flows are closely related to the formation of fog and haze events in the Xi'an area.