基于聚类分析的气溶胶光学厚度时间变化特征研究

利用2008年3月$sim$2018年2月中分辨率成像光谱仪(Moderate resolution imaging spectroradiometer, MODIS) MOD08M3遥感反演气溶胶光学厚度(Aerosol optical depth, AOD)产品数据,结合K-means聚类分析方法,对中国中部和东部的气溶胶光学厚度时间序列进行分析。结果表明: 1)从像元尺度分析气溶胶光学厚度的时间序列变化特征,避免了规律混杂问题,得到了准确的变化规律和波动尺度。2)在年际间变化尺度上得到4个分区结果, AOD长期变化情况受人口分布的因素影响较大。3)在季节间变化尺度上得到9种变化类型区,分别是:华北平原区、长江中下游区、高原山脉区、云贵区、兰州-银川-阿拉善盟区、四川盆地区、关中陕南区、两广-湖南南部-江西南部区、东南沿海区,同时由一些变化分区的地理位置得到了该区AOD季节性波动的主导影响因素。这些结果有助于研究AOD时间序列的准确变化和东部地区的气候环境。

Temporal Characteristics of Aerosol Optical Depth Based on Cluster Analysis Method

Using the K-means cluster analysis method, the times series of aerosol optical depth (AOD) over central andeastern China from March 2008 to February 2018 were analyzed based on the data of moderate resolutionimaging spectroradiometer (MODIS) MOD08M3. The results show that: 1) The characteristics of time series ofAOD were analyzed from the pixel scale, which avoided the problem of rules mixing and leaded to the acquisitionof accurate variation rules and fluctuation scales. 2) Four regional results were obtained on the inter-annualchange scale, and the inter-annual variation of AOD was mainly influenced by the factor of population distribution.3) There are 9 types of change zones on the inter-seasonal fluctuation scale, namely, North China plain region,Middle-Lower Yangtze River region, Plateau Mountain region, Yunnan-Guizhou region, Lanzhou-Yinchuan-Alxa region,Sichuan Basin region, Guanzhong-Shaannan region, Guangdong-Guangxi-Southern Hunan-Southern Jiangxi region,and the Southeast Coastal region. And the dominant influencing factors of seasonal fluctuations of AODin the areas have been obtained based on the geographical location of some changes in the sub-regions.These results are helpful to study the precise changes of the AOD time series and the climaticenvironment in the central and eastern regions.