利用2014年9月1日至2015年5月31日Terra/MODIS MOD 021KM数据,以京津冀地区为研究区域,采用深蓝算法和查找表法反演京津冀地区1 km分辨率的气溶胶光学厚度,并将反演的气溶胶光学厚度与NASA产品和CE-318观测的气溶胶光学厚度进行比较。结果表明:反演的气溶胶光学厚度与NASA MOD 04_L2(10 km×10 km)和MOD 04_3K(3 km×3 km)两种气溶胶产品的空间分布具有高度的一致性,且空间分辨率更高;反演的气溶胶光学厚度与石家庄站CE-318观测气溶胶光学厚度的平均绝对误差为0.07左右,二者之间的相关系数R^2=0.956。卫星过境时,1 km反演的气溶胶光学厚度与MOD 04_L2气溶胶产品的平均误差约为0.06,反演的气溶胶光学厚度与MOD 04_3K气溶胶产品的平均误差约为0.03。对反演的气溶胶光学厚度与河北省PM2.5和PM10质量浓度的空间分布进行相关性分析表明,气溶胶光学厚度AOD与PM2.5和PM10质量浓度的相关系数分别为0.745、0.663,说明1 km反演的AOD可以有效反映区域PM2.5和PM10质量浓度的空间分布。
Based on Terra/MODIS MOD021 KM data from September 1,2014 to May 31,2015,as well as deep blue algorithm and look-up-table methods,aerosol optical depths( AOD) over Jing-Jin-Ji region were retrieved with the spatial resolution of 1 km, the results were compared with NASA products and CE-318 measurements. The results show that the retrieved AODs have higher spatial resolution and relatively good agreements with the products of MOD 04_L2( 10 km × 10 km) and MOD 04_3K( 3 km × 3 km). The mean absolute error between retrieved AODs and those observed by CE-318 sun-photometer at Shijiazhuang site is about 0. 07,with a high correlation coefficient of 0. 98. When Terra passed the territory,mean errors of retrieved AOD compared with MOD 04_L2 and MOD 04_3K products were around 0. 06 and 0. 03, respectively. Further relationship analysis on the spatial contributions of retrieved AODs and the mass concentrations of PM2.5 and PM10 over Hebei province indicated that the correlation coefficients between AOD and PM2.5 and PM10 are 0. 745 and 0. 663, respectively, suggesting that the retrieved AODs with the resolution of 1 km could effectively reflect the mass concentrations of PM2.5 and PM10.