Paper
20 January 2005 Neural network retrieval of deuterium to hydrogen ratio in atmosphere from IMG/ADEOS spectra
Author Affiliations +
Proceedings Volume 5655, Multispectral and Hyperspectral Remote Sensing Instruments and Applications II; (2005) https://doi.org/10.1117/12.579496
Event: Fourth International Asia-Pacific Environmental Remote Sensing Symposium 2004: Remote Sensing of the Atmosphere, Ocean, Environment, and Space, 2004, Honolulu, Hawai'i, United States
Abstract
A feedforward neural network has been developed for retrieval of the Deuterium to Hydrogen ratio (D/H) in atmospheric water vapour from high resolution atmospheric radiances observed from space. The learning and test sets for the neural network training were created by forward simulation of atmospheric emission spectra using FIRE - ARMS for a large set of given temperature, humidity and D/H vertical profiles. The D/H profiles were generated using output from an atmospheric GCM including isotope tracers. The developed neural network was applied for retrieval of total atmospheric column D/H from IMG/ADEOS data over the ocean. A latitudinal distribution of D/H was obtained. The results are in agreement with latitudinal distribution of D/H in the atmosphere obtained from the IMG/ADEOS data earlier by using conventional retrieval methodology. However, the neural network has better accuracy. The stability of the neural network retrieval scheme with di®erent noise levels of the sensor is investigated, and we discuss the possibility of applying the neural network technique to the retrieval of D/H vertical profiles from TES/AURA spectra.
© (2005) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Konstantin G. Gribanov, Ryoichi Imasu, Gavin A. Schmidt, Alexander Yu. Toptygin, and Vyacheslav I. Zakharov "Neural network retrieval of deuterium to hydrogen ratio in atmosphere from IMG/ADEOS spectra", Proc. SPIE 5655, Multispectral and Hyperspectral Remote Sensing Instruments and Applications II, (20 January 2005); https://doi.org/10.1117/12.579496
Lens.org Logo
CITATIONS
Cited by 1 scholarly publication.
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Neural networks

Neurons

Hydrogen

Atmospheric sensing

Humidity

Atmospheric modeling

Databases

Back to Top