关键词: PM2.5, PM10, 后向轨迹, 潜在源贡献分析, 浓度权重轨迹

Abstract: Objective　This study aims to investigate the temporal and spatial variation characteristics, external transport pathways and potential pollution source areas of PM2.5 and PM10 in Yining City, Xinjiang, China, to fill the research gap in the crossregional transport of atmospheric pollutants in this area, especially the lack of studies on PM10 exogenous transport channels, and to clarify the formation mechanism of local particulate pollution, thereby providing scientific basis and data reference for atmospheric pollution prevention and control in Yining City. Methods　Taking Yining City in the Ili River Valley as the research area, this study divided 2022 into four seasons according to meteorological standards, and adopted multi-source data, including hourly PM2.5/PM10 monitoring data from 3 national-controlled monitoring stations (the Municipal Ecological Environment Bureau, the Second Water Plant, and the New Government Area) in Yining City, hourly meteorological data from National Meteorological Information Centre of China and meteorological data from Global Data Assimilation System (GDAS) of National Centers for Environmental Prediction of USA. Pearson correlation analysis was used to explore the correlation between particulate concentrations and meteorological factors. Based on GDAS data, a HYSPLIT backward trajectory model (500 m above the ground) was constructed via the TrajStat plug-in, and the Euclidean distance algorithm and TSV method were applied for airflow trajectory clustering. By introducing a weight function, the potential source contribution function (PSCF) and concentration weighted trajectory (CWT) methods were modified to obtain WPSCF and WCWT, which were used to qualitatively and quantitatively identify potential pollution sources and their contribution degrees, with the secondary concentration limits of GB 3095-2012 as the threshold and WPSCF > 0.6 defined as high-value contribution areas. Results and Discussion　It was shown that in 2022, the concentrations of PM2.5 and PM10 in Yining City were high in winter and spring, low in summer and autumn, with 68 days exceeding the standard of air quality (91% in winter), and the PM2.5/ PM10 ratio was over 0.6 in winter (dominated by anthropogenic sources) and below 0.6 in spring (affected by both anthropogenic and natural sources). The hourly concentrations of PM2.5 and PM10 in a day showed a "W" bimodal pattern in all seasons, reaching its peak during 00:00–02:00 with seasonal differences in valley values. PM2.5 and PM10 were negatively correlated with temperature, wind speed and precipitation, and positively correlated with relative humidity (with PM2.5 being more significantly affected). And the terrain-induced prevailing westerly, southwesterly and southerly winds, coupled with coal-fired heating and temperature inversion in winter, resulted in the most serious winter pollution. The airflow trajectories in four seasons were clustered into 3, 4, 3 and 4 categories, with no excessive particulate concentrations in spring, summer and autumn, while air masses from Kazakhstan, Russia's Novosibirsk Oblast and local counties in China significantly aggravating urban particulate pollution in winter. WPSCF revealed obvious seasonal differences in potential sources, with local high-value contribution areas and no long-distance large-scale sources. WCWT showed the wider distribution of particulate pollution than WPSCF, with PM2.5 high-contribution areas covering China-Kazakhstan adjacent regions and PM10 high-contribution areas concentrating in eastern Bortala Prefecture and Nilka County in Xinjiang in winter. Conclusions　The concentrations of PM2.5 and PM10 in Yining City, Xinjiang in 2022 presented significant temporal and spatial differentiation characteristics. On the annual scale, it showed high values in winter and spring and low values in summer and autumn, a sharp increase in excessive pollution days compared with the previous year, and obvious seasonal differences in the dominant pollution sources. On the hourly scale, all seasons showed a "W" shaped bimodal variation pattern, with fixed peak times and seasonal differences in valley times, and the largest concentration variation amplitude in winter due to heating. Meteorological factors had a significant regulatory effect on particulate matter concentrations in Yining City, with PM2.5 being more sensitive to meteorological changes. The unique trumpet-shaped terrain of the Ili River Valley, combined with winter adverse meteorological conditions such as temperature inversion and stable weather, was the important natural cause of severe particulate matter pollution in winter. The HYSPLIT backward trajectory model showed that the exogenous air masses had no obvious pollution contribution in spring, summer and autumn, while the transport air masses from Almaty Oblast and Jambyl Oblast of Kazakhstan, Novosibirsk Oblast of Russia, as well as Qapqal Xibe Autonomous County and Nilka County in China were the key external factors aggravating the winter particulate matter pollution in the main urban area of Yining City. The potential pollution source areas of PM2.5 and PM10 in Yining City showed clear seasonal distribution characteristics, and all high-value contribution areas were local small-scale distribution in the periphery of Ili Prefecture, with no long-distance large-scale pollution source areas. Winter was the critical period for the contribution of exogenous particulate matter concentration in Yining City. The high-contribution areas of PM2.5 covered the adjacent regions of China and Kazakhstan with a wide range, while the high-contribution areas of PM10 were relatively concentrated in eastern Bortala Mongol Autonomous Prefecture and Nilka County. This study for the first time published the original research results on the seasonal hourly concentration changes of PM2.5 and PM10, the prevailing wind direction and speed, as well as the correlation between meteorological factors and particulate matter in Yining City in 2022. It filled the research gap in the exogenous pollution transport channels of PM10 in Yining area, and deepened the understanding of the temporal and spatial variation of particulate matter pollution and the regional distribution of exogenous pollution transport in Yining City. It provided detailed data support and scientific reference for local atmospheric pollution prevention and control, especially for the formulation of cross-regional joint prevention and control measures, and also laid a foundation for subsequent research on particulate matter pollution in the Ili River Valley.

Key words: PM2.5, PM10, backward trajectory, analysis of potential source contributions, concentration weighting trajectory 676