Low-level jet (LLJ) is a phenomenon where the horizontal wind speed appears to be extreme in the vertical profile. When the extreme value is within the atmospheric boundary layer, it is called boundary layer LLJ. The characteristics of LLJ in the vertical structure make it often appear in the research on heavy convective weather such as heavy rainfall, aviation safety, pollutant transport, and wind energy development and utilization. Main observation instruments for LLJ and vertical wind fields include radiosonde, wind profile radar, and Doppler wind lidar. Doppler lidar can obtain vertical wind field information with a high spatial and temporal resolution. Additionally, its higher vertical resolution and lower detection blind area give itself an obvious advantage in observing the structural features of LLJ. The higher detection accuracy of vertical wind speed enables us to better find the change of atmospheric vertical diffusion ability in the LLJ process, which is also a key mechanism of the impact of LLJ on pollutant concentration. In China, the LLJ research is mainly focused on weather-scale LLJ related to rainfall, while less research is on boundary layer LLJ. Due to the lack of observation data, the studies on structural characteristics of LLJ are rare. Juehua Island, an island about 10 km offshore in the Bohai Sea, has a distinctive LLJ structure due to the influence of land and sea, which makes our study meaningful.

The coherent Doppler wind lidar is operated in Juehua Island, Huludao City, Liaoning Province from September 21st, 2020 to May 8th, 2021. The vertical wind field lidar data is employed to find the atmospheric boundary layer LLJ in this area. Statistical characteristics are analyzed by combining the mean sea level pressure and 1000 hPa temperature data provided by the ERA5 reanalysis model. Combined with the PM_2.5 mass concentrations of Huludao City, the impact of LLJ on the change in PM_2.5 mass concentrations is analyzed. In previous studies, the criteria for LLJ are often selected according to the research purposes, instruments, and environmental characteristics of the observation area. Referring to Wu and Bass, our criteria for judging the LLJ are as follows. The extreme value of wind speed should be greater than or equal to 8 m?s^-1. The difference between the extreme wind speed and the minimum wind speed at higher altitudes shall be greater than or equal to half of the maximum wind speed.

Among effectively observed 24550 wind profiles, 2766 wind profiles are determined as LLJ wind profiles. During the observation period, the occurrence frequency of LLJ is 0.11, with obvious monthly changes (Fig. 2). The average jet wind speed is 13.2 m?s^-1, and the wind directions of the jets are mainly concentrated in the northeast and southwest. Statistically, the jet height is mainly below 500 m, and the height of maximum frequency is between 200 m and 300 m (Fig. 3). The mean sea level pressure of ERA5 reanalysis data shows that the observation area is mainly affected by two types of weather background situations during the study period, which is one reason why the LLJ is concentrated in the southwest and northeast (Fig. 6). Correlation analysis indicates that strong horizontal pressure gradient provides favorable conditions for LLJ formation in the boundary layer. In winter, the wind direction distribution of LLJ is different from that of the background wind field (Fig. 5). Because the land temperature is lower than the ocean temperature at most time in winter (Fig. 7), the temperature gradient can enhance the horizontal pressure gradient difference between the northwest and southeast, which is conducive to the formation of the northeast jet and plays a role in hindering the formation of the southwest jet (Table 2). Doppler wind lidar could catch the process of the atmospheric boundary layer LLJ from formation to extinction. Through the standard deviation of vertical wind speed provided by the lidar and PM_2.5 mass concentration data of Huludao City, LLJ has accelerated the reduction of PM_2.5 mass concentration by enhancing the vertical diffusion ability of the near-surface atmosphere (Fig. 8). The relative change of PM_2.5 mass concentration at the time of LLJ is shown in Fig. 9, and the size of the circle in the figure is determined by the standard deviation of vertical wind speed.

Huludao City is located in the mid-latitude coastal area, with the coastline trending southwest-northeast. The atmospheric boundary layer LLJ in this region has obvious seasonal characteristics. Its wind direction is mainly affected by the seasonal characteristics of the large-scale weather situations and the trend of the coastline, with obvious regional characteristics. For coastal areas, differences between the thermal properties of land and sea will provide favorable or unfavorable conditions for LLJ formation in different background wind fields, thus affecting the LLJ distribution under different wind directions. When the LLJ occurs at night, the decrease in PM_2.5 mass concentration can be accelerated or its growth rate can be weakened by enhancing the vertical diffusion ability of the near-surface atmosphere. The conclusions are of certain reference significance for subsequent LLJ observation in the boundary layer and the study on air pollution in similar areas. The coastline of China is extensive with complex topography of coastal areas, and the current single-station observation is far from covering LLJ's characteristics. Additionally, the relationship between the horizontal distribution characteristics of the LLJ in the boundary layer and the coastal tomography needs further exploration.