唐庆新等:Spatiotemporal fusion of multiple-satellite aerosol optical depth (AOD) products using Bayesian maximum entropy method
被阅读 933 次
2016-08-06
Spatiotemporal fusion of multiple-satellite aerosol optical depth (AOD) products using Bayesian maximum entropy method
作者:Tang, QX (Tang, Qingxin)[ 1,2,3,4,5 ] ; Bo, YC (Bo, Yanchen)[ 1,2,3,4 ] ; Zhu, YX (Zhu, Yuxin)[ 6 ]
JOURNAL OF GEOPHYSICAL RESEARCH-ATMOSPHERES
卷: 121  期: 8  页: 4034-4048
DOI: 10.1002/2015JD024571
出版年: APR 27 2016
 
摘要
Merging multisensor aerosol optical depth (AOD) products is an effective way to produce more spatiotemporally complete and accurate AOD products. A spatiotemporal statistical data fusion framework based on a Bayesian maximum entropy (BME) method was developed for merging satellite AOD products in East Asia. The advantages of the presented merging framework are that it not only utilizes the spatiotemporal autocorrelations but also explicitly incorporates the uncertainties of the AOD products being merged. The satellite AOD products used for merging are the Moderate Resolution Imaging Spectroradiometer (MODIS) Collection 5.1 Level-2 AOD products (MOD04_L2) and the Sea-viewing Wide Field-of-view Sensor (SeaWiFS) Deep Blue Level 2 AOD products (SWDB_L2). The results show that the average completeness of the merged AOD data is 95.2%, which is significantly superior to the completeness of MOD04_L2 (22.9%) and SWDB_L2 (20.2%). By comparing the merged AOD to the Aerosol Robotic Network AOD records, the results show that the correlation coefficient (0.75), root-mean-square error (0.29), and mean bias (0.068) of the merged AOD are close to those (the correlation coefficient (0.82), root-mean-square error (0.19), and mean bias (0.059)) of the MODIS AOD. In the regions where both MODIS and SeaWiFS have valid observations, the accuracy of the merged AOD is higher than those of MODIS and SeaWiFS AODs. Even in regions where both MODIS and SeaWiFS AODs are missing, the accuracy of the merged AOD is also close to the accuracy of the regions where both MODIS and SeaWiFS have valid observations.
 
通讯作者地址: Bo, YC (通讯作者)
Beijing Normal Univ, Res Ctr Remote Sensing, State Key Lab Remote Sensing Sci, Beijing 100875, Peoples R China.
通讯作者地址: Bo, YC (通讯作者)
Beijing Normal Univ, GIS, Beijing 100875, Peoples R China.
通讯作者地址: Bo, YC (通讯作者)
Beijing Normal Univ, Sch Geog, Beijing 100875, Peoples R China.
通讯作者地址: Bo, YC (通讯作者)
Beijing Normal Univ, Beijing Key Lab Remote Sensing Sci, Beijing 100875, Peoples R China.
地址:
[ 1 ] Beijing Normal Univ, Res Ctr Remote Sensing, State Key Lab Remote Sensing Sci, Beijing 100875, Peoples R China
[ 2 ] Beijing Normal Univ, GIS, Beijing 100875, Peoples R China
[ 3 ] Beijing Normal Univ, Sch Geog, Beijing 100875, Peoples R China
[ 4 ] Beijing Normal Univ, Beijing Key Lab Remote Sensing Sci, Beijing 100875, Peoples R China
[ 5 ] Liaocheng Univ, Sch Environm & Planning, Liaocheng, Shandong, Peoples R China
[ 6 ] Huaiyin Normal Univ, Sch Urban & Environm Sci, Huaian, Jiangsu, Peoples R China