基于Prophet-LSTM+P-Bshade融合模型的中国陆地MAIAC AOD补值研究  

MAIAC AOD complement study of Chinese terrene based on Prophet-LSTM+P-Bshade fusion model

作　　者：熊英杰 杜宁[1] 王莉[1] 张显云[1] 王耀 张洪飞 XIONG Yingjie;DU Ning;WANG Li;ZHANG Xianyun;WANG Yao;ZHANG Hongfei(Mining College,Guizhou University,Guiyang 550025,China)

机构地区：[1]贵州大学矿业学院,贵阳550025

出　　处：《中国科技论文》2024年第8期877-883,共7页China Sciencepaper

基　　金：贵州省省级科技计划项目(黔科合支撑[2022]一般204)。

摘　　要：因受云层等气象条件干扰和卫星传感器性能影响,卫星获取的气溶胶光学厚度(aerosol optical depth,AOD)存在数据缺失的情况,为了提高AOD数据的覆盖度,利用AOD的时空相关性,提出Prophet-LSTM+P-Bshade融合模型,对2016年3月—2021年2月我国1 km×1 km分辨率的多角度大气校正算法(multiangle implementation of atmospheric correct, MAIAC) AOD数据进行时空补值,并与地面监测站点AERONET AOD进行对比验证。结果表明:提出的融合模型补值成果精度明显优于其他对比模型,验证精度指标R、MAE和RMSE分别为0.905、0.237和0.375。2016年3月—2021年2月中国陆地AOD值分布时间上季度性变化较明显,整体呈现冬季>春季>夏季>秋季的季节分布特征,空间上总体呈现东高西低和塔里木盆地局部较高的特点。Data missing occurs commonly in the aerosol optical thickness(AOD)acquired by satellite due to atmospheric conditions such as cloud cover and satellite sensor performance.In order to improve the coverage of AOD data,Prophet-LSTM+P-Bshade model was proposed using the spatial and temporal correlation of AOD to make up the spatial and temporal MAIAC(multiangle implementation of atmospheric correct)AOD data with a resolution of 1 km×1 km from March 2016 to February 2021 and validate the results with the AERONET AOD at ground monitoring stations.The results show that the accuracy of the proposed fusion model supplementation results is obviously better than that of the other comparative models,and the accuracy indicators of valida-tion,such as R,MAE,and RMSE,are 0.905,0.237 and 0.375,respectively.The distribution of AOD values in China from March 2016 to February 2021 is characterized by quarterly variations in time.Specifically,the seasonal distribution is winter>spring>summer>autumn,and spatially,the distribution of AOD values in China is generally higher in the east,lower in the west,and locally higher in the Tarim basin.

关 键 词：MAIAC AOD 时空相关性 Prophet-LSTM P-Bshade 时空补值 

分 类 号：P237[天文地球—摄影测量与遥感]