A synergic algorithm for retrieval of aerosol optical depth over land
A synergic algorithm for retrieval of aerosol optical depth over land
复制标题
DOI：
10.1007/s00376-009-7218-4
复制
发表时间：
2009-09
影响因子：
5.8
通讯作者：
Jianping Guo;Yong Xue;C. Cao;Hao Zhang;J. Guang;Xiaoye Zhang;Xiaowen Li
中科院分区：
文献类型：
--
作者：
Jianping Guo;Yong Xue;C. Cao;Hao Zhang;J. Guang;Xiaoye Zhang;Xiaowen Li
利用先进的沿着迹扫描辐射计（AATSR）的0.55、0.66和0.87 μ m双视能力，结合中分辨率成像光谱仪（MODIS）的双向反射分布函数（BRDF）模型，提出了一种新的反演陆地上空1km空间分辨率气溶胶光学厚度（AOD）的算法。通过提取AATSR中的几何信息和从MODIS BRDF/Albedo模型参数产品中提取核函数，计算出地表的BRDF特征，即该算法的先验输入参数。利用AATSR反演的北京及其周边地区2005年8月至2006年7月的大气气溶胶光学厚度（AOD）数据，与现场AERONET（Aerosol Robotic NETwork）观测数据进行了对比验证。AATSR正、底视图的AOD差值分别小于5.72%、1.9%和13.7%。与此同时，发现采用本文提出的协同算法反演AATSR比采用Lambert曲面的AATSR更有利，AATSR反演的气溶胶光学厚度与AERONET反演的气溶胶光学厚度的决定系数分别比协同算法降低了15.5%和18.5%。这进一步表明，协同算法可以潜在地用于气候变化和空气质量监测。
In this paper, a novel algorithm for aerosol optical depth(AOD) retrieval with a 1 km spatial resolution over land is presented using the Advanced Along Track Scanning Radiometer (AATSR) dual-view capability at 0.55, 0.66 and 0.87 µm, in combination with the Bi-directional Reflectance Distribution Function (BRDF) model, a product of the Moderate Resolution Imaging Spectroradiometer (MODIS). The BRDF characteristics of the land surface, i.e. prior input parameters for this algorithm, are computed by extracting the geometrical information from AATSR and reducing the kernels from the MODIS BRDF/Albedo Model Parameters Product. Finally, AOD, with a 1 km resolution at 0.55, 0.66 and 0.87 µm for the forward and nadir views of AATSR, can be simultaneously obtained.Extensive validations of AOD derived from AATSR during the period from August 2005 to July 2006 in Beijing and its surrounding area, againstin-situAErosol RObotic NETwork (AERONET) measurements, were performed. The AOD difference between the retrievals from the forward and nadir views of AATSR was less than 5.72%, 1.9% and 13.7%, respectively. Meanwhile, it was found that the AATSR retrievals using the synergic algorithm developed in this paper are more favorable than those by assuming a Lambert surface, for the coefficient of determination between AATSR derived AOD and AERONET mearured AOD, decreased by 15.5% and 18.5%, compared to those derived by the synergic algorithm. This further suggests that the synergic algorithm can be potentially used in climate change and air quality monitoring.