Machine Learning for Retrieving Cloud Optical Thickness from Observed Reflectance: 3D Effects CyberTraining: Big Data + High-Performance Computing + Atmospheric Sciences
Links to Fileshttp://hpcf-files.umbc.edu/research/papers/CT2020Team5.pdf
MetadataShow full item record
Type of Work11 pages
Citation of Original PublicationAngelof, Kallista; Bergstrom, Kirana; Le, Tianhao; Xu, Chengtao; Rajapakshe, Chamara; Zheng, Jianyu; Zhang, Zhibo; Machine Learning for Retrieving Cloud Optical Thickness from Observed Reflectance: 3D Effects CyberTraining: Big Data + High-Performance Computing + Atmospheric Sciences (2020); http://hpcf-files.umbc.edu/research/papers/CT2020Team5.pdf
RightsThis item is likely protected under Title 17 of the U.S. Copyright Law. Unless on a Creative Commons license, for uses protected by Copyright Law, contact the copyright holder or the author.
Clouds are inherently three dimensional (3D), and simulating radiative transfer (RT) properties accurately requires models that take their 3D effects into account. Because 3D models are complex and computationally expensive, RT models often use simplified 1D models to retrieve cloud properties, which suffer from retrieval uncertainty and sometimes significant biases due to 3D effects. Recent advancements in machine learning may lead to a retrieval algorithm that is capable of taking these effects into account. We will develop a machine-learning based cloud property retrieval algorithm that is able to reconstruct the 3D structure of clouds based on observed cloud radiative signatures.