乌鲁木齐市气溶胶光学厚度时空变化及影响因素研究

气溶胶光学厚度（AOD）在一定程度上可以反映城市上空大气污染状况，本文以乌鲁木齐市为研究区，利用2015～2019年MODIS L1B 1 km数据，在IDL平台下调用6S辐射传输模型构建查找表（LUT），结合地面暗像元法反演得到近5年乌鲁木齐AOD，采用标准差椭圆、空间自相关分析和趋势分析等方法分析研究区AOD的时空分布特征，并利用统计法分析气象因子对AOD的影响。结果表明：（1）2015～2019年，乌鲁木齐市AOD先增加后减小，整体没有明显变化，趋势近于0，年均AOD在0～1.95之间。在空间上，AOD呈现西南-东北方向浓度逐渐增加的趋势。AOD高值区（1.67～1.95）主要分布在达坂城区、头屯河区及沙依巴克区。（2）研究区冬季和秋季AOD值相对春季和夏季偏高，冬季AOD平均值可以达到0.63。（3）2015～2019年研究区AOD呈正的空间自相关。AOD的Moran I值呈先增加后减小的趋势，最大值和最小值分别为0.991和0.979。近5年研究区AOD的空间分布中心主要集中在达坂城区和乌鲁木齐县，AOD在空间上呈现先扩散后逐渐集中的趋势，标准偏差椭圆先增大后减小。（4）研究区年总降水量、风速分别与AOD负相关，R²分别为0.04和0.26。相对湿度与AOD呈正相关，R²为0.27。本研究可以为西北干旱区城市气溶胶监测提供参考。

Temporal and Spatial Variation of Aerosol and Its Influencing Factors in Urumqi

Aerosol Optical Depth (AOD) reflects the air pollution over the city to a certain extent through the reduction of light. This paper uses MODIS L1B 1km-day data in Urumqi during 2015 to 2019 to generate the search Table (LUT) by using IDL to call the 6S radiation transmission model, with the combination of the ground dark pixel method, standard deviation ellipse and spatial auto-correlation analysis and other methods, the AOD over Urumqi for the past five years from 2015 to 2019 was obtained. The results showed that:(1) AOD concentration in Urumqi showed a trend of gradual increase from southwest to northeast in recent five years. The annual average value of AOD is between 0 and 1.95, and the AOD high value area (1.67-1.95) is mainly distributed in Dabancheng District, Toutunhe District and Shaybuck District. (2) AOD values in winter and autumn in the study area were higher than those in spring and summer, and the average AOD value in winter could reach 0.63. (3) AOD in the study area presented positive spatial auto-correlation during 2015-2019. The Moran I value of AOD showed a trend of first increasing and then decreasing, and the maximum and minimum values were 0.991 and 0.979. In recent 5 years, the spatial distribution center of AOD in the study area was mainly concentrated in Daban City District and Urumqi County. From 2015 to 2019, AOD first diffused and then gradually concentrated in space, and the standard deviation ellipse first increased and then decreased. (4) AOD at monthly and annual scales in the study area is closely related to precipitation, relative humidity and wind speed. Precipitation is negatively correlated with AOD (R²=0.04). Under the action of precipitation and dust removal, AOD will decrease. The relative humidity was positively correlated with AOD (R²=0.27). The increase of relative humidity affected the increase of the scattering rate of hygroscopic aerosol particles, leading to the increase of AOD. Wind speed is negatively correlated with AOD (R²=0.26), and wind speed has an effect on AOD diffusion speed. This study can provide reference for urban aerosol monitoring in arid areas of Northwest China.