Mathematical algorithms for processing and analysis of near-infrared data from a satellite-borne Fourier transform spectrometer

摘要

Mathematical algorithms for processing and analysis of near-infrared data from a satellite-borne Fourier transform spectrometer have been designed. The algorithms involve non-parametric approaches based on neural networks, random forest, and empirical orthogonal functions. The results obtained from processing of model data and real measurements performed by means of the GOSAT-borne Fourier transform spectrometer are presented.