ASPCS
 
Back to Volume
Paper: Thoughts About Disentangling in Wavelength and in Fourier-Space
Volume: 318, Spectroscopically and Spatially Resolving the Components of Close Binary Stars
Page: 107
Authors: Ilijic, S.
Abstract: The spectral disentangling technique can be understood as a two-level least-squares problem. The lower level, known as spectral separation, is a linear least-squares problem where the spectra of component stars are reconstructed from a time series of spectra of a multiple stellar system, given the RVs of component stars in the observed spectra. In the upper level the parameters of orbit are optimised in order to find the setup of RVs that yields the best fit at the separation level. Spectral separation implemented in wavelength-space is computationally less efficient than in Fourier-space, but offers larger flexibility in sampling and weighting of the data. The two implementations of the separation algorithm may respond differently to systematic noise in the data such as phase correlated inconsistencies in normalisation of the observed spectra.
Back to Volume