【作者】 刘状；

【导师】 王旭红；

【作者基本信息】 西北大学 ， 地图学与地理信息系统， 2019， 硕士

【摘要】 大气气溶胶是影响气候和大气环境的重要因子,对人类有着重要的影响。利用2001年3月至2018年2月的MODIS C6数据,结合多种气象栅格数据和地表数据对AOD的时空特征及驱动因子进行分析。利用相关分析和地理探测器方法对AOD空间分布特征驱动因子进行分析,同时研究建立了主成分驱动因子与气溶胶光学厚度空间分布的响应关系模型。基于像元尺度对AOD变化规律进行分类,从频率域和地域分异规律的角度研究了驱动因子对AOD时间变化的影响作用,并通过时间序列的几何相似性分析AOD月度变化与驱动因子月度变化的关联关系。本研究主要得到了以下研究结论:(1)在AOD空间分布方面,中国东部的AOD值在平原地区均高,在山脉高原地区均低。AOD值在平原内部的空间变化梯度较小,在平原和山脉交界地区有较大的空间变化梯度。人为源气溶胶的粒径大小有很大的不确定性,?ngstr?m常数在中国东部人口密度大的地区处于中等的水平;利用相关分析方法和地理探测器分析方法对十四种驱动因子与AOD空间分布的关系进行分析。全年通过显著性检验的四个数值量驱动因子和三个类型量驱动因子均说明中国东部AOD的空间分布主要受人为因素影响。(2)在AOD时间变化特征方面,利用ISODATA算法对AOD时间序列进行分类,聚类结果在空间上破碎度较低,整体有很好的空间自相关性,说明聚类分析有很好的效果,可以避免规律混杂问题;利用EEMD对不同典型区的AOD时间序列进行分析,偶然事件对AOD年际变化有着重要影响,其对AOD时间序列的影响大小和影响周期有不稳定的特征。AOD季节变化对AOD时间序列的影响大小和周期较稳定。RES显示七个典型区的AOD值2001年以来均先增高后降低。多个波动规律叠加在一起,构成了AOD时间序列变化的最终结果;基于灰色关联度分析不同典型区AOD月度变化的与驱动因子的几何相似性,影响AOD季节性变化的驱动因子主要是相对湿度、辐射量、植被指数和风速。(3)基于带通滤波提取AOD时间序列的季节变化特征,研究表明AOD时间序列变化与太阳直射点位置有重要关系,在一年中最接近太阳直射点的月份即为该纬度整体AOD值最高的月份。北回归线以北和南回归线以南地区的AOD变化周期为12个月,南北回归线以内地区存在一个短周期和一个长周期,赤道地区的AOD变化周期为6个月。在全球尺度下AOD季节变化与LST季节变化有一致的变化幅度。更多还原

【Abstract】 Atmospheric aerosol Optical Depth(AOD)are important factors of the climate and the atmospheric environment,and have an important impact on humans.This paper analyzed the spatio-temporal variations of Aerosol Optical Depth(AOD)and driving factors by combining various meteorological raster data and surface data,using the MODIS C6 data from March2001 to February 2018.This study established the response relationship model of the principal component driving factor to the aerosol optical thickness spatial distribution,and using the correlation analysis and geo-detector method to analyzed the AOD spatial distribution feature driving factors.The AOD variation rules are classified based on the pixel scale,the influence of driving factors on AOD time variation is studied from the perspective of frequency domain and regional differentiation law.Based on the geometric similarity of time series,the relationship between the monthly variations of AOD and the monthly change of driving factor are analyzed.This study mainly obtained the following research conclusions:(1)In terms of AOD spatial distribution,the AOD value in eastern China is high in the plain area and low in the mountain plateau.The spatial variation gradient of the AOD value inside the plain is small,and there is a large spatial variation gradient in the boundary between the plain and the mountain.The particle size of anthropogenic aerosols is highly uncertain,and the?ngstr?m constant is at a medium level in areas with high population density in eastern China;The relationship between fourteen driving factors and the spatial distribution of AOD are analyzed by correlation analysis and geographic detector analysis method.The four numerical driving factors and the three types of driving factors that passed the significance test throughout yearly indicated that the spatial distribution of AOD in eastern China is mainly affected by human factors.(2)In view of AOD temporal features,the ISODATA algorithm is used to classify AOD time series.The clustering results are less spatially fragmented and have a good overall spatial autocorrelation,indicating that clustering analysis has a good effect.which can avoid the problem of regular confounding;It analyzed the temporal variations of Aerosol Optical Depth(AOD)over different typical regions,using the Ensemble Empirical Mode Decomposition(EEMD)method.Results showed that the occasional events have an important impact on the inter-annual variation of AOD,and its impact on the AOD temporal series of the magnitude and impact period are unstable.while the effect of AOD seasonal variation on the AOD temporal series is relatively stable.RES showed that the AOD values of the seven typical areas have increased first and then decreased since 2001.The temporal variations of AOD in eastern China were resulted from multiple fluctuation rules superimposed together;Based on the grey correlation degree,the geometric similarity between the monthly change of AOD and the driving factor in different typical areas is analyzed.The driving factors affecting the seasonal change of AOD are mainly relative humidity.,radiation,vegetation index and wind speed.(3)Based on the seasonal variation characteristics of AOD temporal series extracted by band-pass filtering,the research shows that the AOD temporal series change has an important relationship with the position of the direct solar spot.The month closest to the direct sun point in the year is the highest AOD value of the latitude.The AOD change period in the area north of the Tropic of Cancer and the south of the Tropic of Cancer is 12 months.Areas in the interior the Tropic of Cancer have a short period and a long period,and the AOD period in the equatorial region is 6 months.At the global scale,the seasonal variation of AOD has a consistent change with the seasonal variation of LST.更多还原