Abstract
A high spectral resolution measurements of Geostationary Imaging FTS (GIFTS) simulation study is conducted to demonstrate the application of principal component analysis to measurement compression and sounding profile retrieval. This study discusses the fundamental aspects of interferogram compression scheme, noise reduction effect of compression, measurement signal degradation, and proficiency and efficiency of retrieval of temperature and water vapor. Principal Component Analysis (PCA), Principal Component Compression (PCC), and Principal Component Regression (PCR) under gaussian distribution and random noise of measurement conditions are shown to provide 1) nearly full spectral information with acceptable degradation, 2) significant measurement noise reduction, 3) measurement compression with reasonable compression ratio, and 4) tolerable loss of accuracy in temperature and water vapor retrieval. These techniques proved to be valuable tools for data compression and accurate retrieval of sounding profile parameters for coming EO-3 GIFTS and other new generation polar- orbiting satellite infrared measurements.
© 2001 Optical Society of America
PDF ArticleMore Like This
Lihang Zhou, Mitchell D. Goldberg, Walter W. Wolf, and Chris Barnet
HTuD9 Hyperspectral Imaging and Sounding of the Environment (HISE) 2005
W.L. Smith, D.T. Zhou, J.J. Tsou, and A.M. Larar
OWA1 Optical Remote Sensing (HISE) 2001
Jonathan Monsalve, Jorge Bacca, and Henry Arguello
CTh1B.4 Computational Optical Sensing and Imaging (COSI) 2017