摘要
开展气溶胶光学厚度反演研究对监测和评估大气空气质量具有重要意义。针对现有气溶胶光学厚度反演精度低的问题,将AERONET站点实测的气溶胶光学厚度数据与预处理后的Himawari-8影像数据进行时空匹配,建立样本数据集,从而构建气溶胶光学厚度反演模型;通过深度神经网络模型与深度置信网络模型对比分析,验证了深度置信网络反演模型的精度;以AERONET站点实测的气溶胶光学厚度数据和MCD19A2气溶胶光学厚度数据为验证数据,验证了深度置信网络反演气溶胶光学厚度的可行性。结果表明,基于深度置信网络反演的气溶胶光学厚度,精度较高、拟合效果好且误差小。
The study of aerosol optical depth retrieval is of great significance for monitoring and assessing the quality of the atmospheric environment.Aiming at the low accuracy of the existing aerosol optical depth retrieval,the aerosol optical depth data measured at the AERONET site were temporally and spatially matched with the preprocessed Himawari-8 image data to establish a sample data set so as to build an aerosol optical depth retrieval model.The accuracy of the deep belief network retrieval model was verified by comparing and analyzing the deep neural network model with the deep belief network model,and the feasibility of the deep belief network retrieval of aerosol optical depth was verified by using the measured aerosol optical depth data from the AERONET site and the MCD19 A2 aerosol optical depth data as the validation data.The results show that the aerosol optical depth based on deep belief network retrieval has high accuracy,good fitting,and a small error.
作者
祖维涛
陈优良
王兆茹
刘星根
黄孝斌
ZU Weitao;CHEN Youliang;WANG Zhaoru;LIU Xinggen;HUANG Xiaobin(School of Civil and Surveying&Mapping Engineering,Jiangxi University of Science and Technology,Ganzhou 341000,China;School of Resources and Environmental Engineering,Jiangxi University of Science and Technology,Ganzhou 341000,China;The Engineering&Technical College,Chengdu University of Technology,Leshan 614000,China)
出处
《江西冶金》
2024年第6期472-480,共9页
Jiangxi Metallurgy
基金
国家自然科学基金项目(42261072)。
