本研究以成都市2015年监测数据为研究对象，系统地分析了成都城区大气污染物特征、气象特征、PM2.5化学组分特征、PM2.5来源解析、气溶胶光学性质等。分析了冬季典型灰霾案例，梳理了灰霾的形成机制。主要结论如下： （1）PM10、PM2.5冬季超标率高达48%、69%；O3春夏两季平均浓度较高，超标率达17%和30%；NO2冬季超标率较高，达15%。成都市整体风速较小（V=1.7m/s）、湿度较高（RH=62.5%）。SO42-、NO3-、NH4+浓度平均值为10.4μg/m3、9.1μg/m3、6.1μg/m3，冬季各离子浓度远高于其他季节。SOR、NOR全年均值为0.34、0.15，气粒二次转化冬季最为强烈。OC、EC、OC/EC的全年平均值分别为10.2μg/m3、3μg/m3、3.8。POM、POA和SOA的年平均质量浓度分别为16.44μg/m3，7.37μg/m3和9.07μg/m3。 （2）2015年全年，SO42-、NO3-和NH4+（SNA），有机物，EC分别占PM2.5质量浓度的34.6%，20.4%，4.1%。PMF模型解析的PM2.5源解析结果表明，二次无机盐、机动车排放、燃煤、生物质燃烧、扬尘和工业对2015年全年PM2.5质量浓度的贡献率分别为33.4%、27.9%、10.5%、9.7%、8.6%和6.8%。秋冬两季的二次无机盐污染最为严重。春季扬尘污染、机动车排放污染问题突出。冬季生物质燃烧的贡献比例比其他季节突出。PSCF结果显示，成都春夏秋季PM2.5的潜在源区为成都南部，而冬季PM2.5主要来源于城市的局地污染，也有东南方向的远距离跨区域输送。 （3）消光系数σext的全年、春、夏、秋、冬季平均值分别为825.8Mm-1、665.9Mm-1、505.0Mm-1、795.8Mm-1、1336.0Mm-1。σsp、σap、σag 和σext与PM2.5浓度变化趋势基本一致。能见度与颗粒物浓度、湿度呈负相关关系。实际观测到的消光系数与IMPROVE方法计算得到的消光系数相关性较好（R2=0.73），硫酸铵、硝酸铵、有机物、元素碳和粗粒子对消光系数平均贡献率为30.8%、23.2%、26%、11.7%和8.3%。少量的SNA和EC质量浓度的增加会对消光产生更大的影响。因此，着重控制SNA和EC的排放，对灰霾现象的减轻有重要作用。 （4）灰霾案例（2015年1月6日-1月16日）中，污染过程的大气边界层（PBL）高度明显下降，空气污染物垂直扩散受到抑制。在清洁天、污染过程第1阶段（EP.1）和污染过程第2阶段（EP.2）中，[NO3-]/[SO42-]比值分别为0.61、0.76和0.88，表明移动源对成都空气污染贡献越来越重要。气溶胶光学吸湿性增长在灰霾形成过程中起到重要作用。PSCF和大气后向轨迹结果表明，污染主要来自偏南方。PM2.5源解析结果显示，二次无机气溶胶、机动车排放、燃煤、生物质燃烧、工业和扬尘在EP.1中的贡献率分别为34.1%、24.1%、12.7%、12.3%、7.6%和7.2%、在EP.2中的贡献率分别为28.9%、23.1%、9.4%、9.5%、20.3%和7.5%。

Based on monitoring data from Chengdu in 2015, this study systematically analysed the air pollutant characteristics, meteorological characteristics, chemical compositions of PM2.5, optical properties of aerosols and source apportionment. And a typical haze episode was analyzed to clarify the haze formation mechanism in Chengdu. The conclusions of this study are summarized below: (1) There are 48%, 69% days exceeded the PM10, PM2.5 standard (GB 3095-2012) in winter; the average concentration of O3 is high in spring and summer；15% days exceeded the NO2 standard in winter. The overall wind speed in Chengdu was low (V=1.7m/s) and the humidity was high (RH=62.5%). The average concentrations of SO42-, NO3-, and NH4+ were 10.4μg/m3, 9.1μg/m3, and 6.1μg/m3, respectively. The concentrations of ions in winter were much higher than those in other seasons. The annual averages of SOR and NOR were 0.34 and 0.15, and the secondary transformation of gas-particle was worse in winter. The annual averages of OC, EC and OC/EC were 10.2μg/m3, 3μg/m3, and 3.8, respectively. The annual average mass concentrations of POM, POA, and SOA were 16.44 μg/m3, 7.37μg/m3, and 9.07μg/m3, respectively. (2) The contribution ratio of SNA (SO42-, NO3-, and NH4+), organic matter, and EC to PM2.5 were 34.6%, 20.4%, and 4.1%, respectively. 5. Six major source factors were identified to have largely contributed to PM2.5. The secondary inorganic aerosols, vehicle emissions, coal combustion, biomass burning, dust and industry contributed 33.4%、27.9%、10.5%、9.7%、8.6% and 6.8% to PM2.5 masses,respectively. The secondary inorganic salt pollution was the most serious in autumn and winter. Dust and vehicle emissions in the spring were prominent. The results of PSCF indicated that， the potential sources of PM2.5 were distributed near the south in spring, summer and autumn，while in the winter, PM2.5 mainly comes from the local as well as long-distance and cross-regional transportation in the southeast direction. (3) The average extinction coefficient (σext) during whole year, spring, summer, autumn, and winter were 665.9Mm-1，505.0 Mm-1，795.8 Mm-1 and 1336.0 Mm-1 respectively. σsp, σap, σag, and σext showed similar increases in patterns with PM2.5. Visibility was negatively correlated with particulate matter concentration and humidity. The observed extinction coefficient correlated with the extinction coefficient calculated by the IMPROVE method (R2 = 0.73). The average contribution rates of ammonium sulfate, ammonium nitrate, organic matter, elemental carbon and coarse particles to the extinction coefficient were 30.8%, 23.2%, 26%, 11.7% and 8.3%. SNA and elemental carbon have great effect on light extinction. Therefore, focusing on the control of SNA and elemental carbon emissions plays an important role in reduce the visibility. (4) The planetary boundary layer (PBL) height decreased distinctly during the haze episodes and restrained air pollutant vertical dispersion. The [NO3?] / [SO42?] ratio was 0.61, 0.76 and 0.88 during a non-haze period, episode 1 and episode 2, respectively, indicating that the mobile source of the air pollution is increasingly predominant in Chengdu. The hygroscopic growth of aerosols played a vital role in the formation of haze. The PSCF and backward trajectories of the air masses indicated that the pollution mainly came from the south. The secondary inorganic aerosols, vehicle emissions, coal combustion, biomass burning, industry, and dust contributed 34.1%, 24.1%, 12.7%, 12.3%, 7.6%, and 7.2% to PM2.5 masses in episode 1 and 28.9%, 23.1%, 9.4%, 9.5%, 20.3% and 7.5% in episode 2.