DocumentCode
446119
Title
Robustness of the NN Approach to emulating atmospheric long wave radiation in complex climate models
Author
Krasnopolsky, Vladimir M. ; Fox-Rabinovitz, Michael S. ; Chou, Ming-Dah
Volume
4
fYear
2005
fDate
July 31 2005-Aug. 4 2005
Firstpage
2661
Abstract
In this paper we present comparisons of NN emulations developed for two different complex climate models: NCAR CAM model and NASA NSIPP. These models have different dynamics, different horizontal and vertical resolutions and different physics (including different long wave radiation schemes). Comparison of two (NCAR and NASA) NN emulations shows their profound similarity in terms of the accuracy of emulation vs. the original parameterizations and complexity of emulating NNs, i.e. the methodological robustness and portability of our NN emulation approach.
Keywords
atmospheric radiation; atmospheric techniques; geophysics computing; neural nets; National Center for Atmospheric Research; atmospheric long wave radiation; community atmospheric model; complex climate models; neural networks; Atmospheric modeling; Atmospheric waves; CADCAM; Computer aided manufacturing; Emulation; Geoscience; NASA; Neural networks; Physics; Robustness;
fLanguage
English
Publisher
ieee
Conference_Titel
Neural Networks, 2005. IJCNN '05. Proceedings. 2005 IEEE International Joint Conference on
Conference_Location
Montreal, Que.
Print_ISBN
0-7803-9048-2
Type
conf
DOI
10.1109/IJCNN.2005.1556323
Filename
1556323
Link To Document