Principal Components of Variability of the Accretion Structures Near White Dwarfs

Abstract

The determination of effective colors of various contributions to the variability of accretion structures using the principal component analysis is discussed. The methods have been applied to the UBVRI observations of cataclysmic variables with various degrees of influence of the magnetic field onto accretion - polar AM Her, intermediate polar V405 Aur and nova-like system TT Ari. The ``filtering" method allows to estimate effective colors based on the assumption, that the variability consists of slow (e.g. orbital or superhump) variations, onto which are superimposed faster correlated variations with arbitrary shape and an uncorrelated noise. The wavelength-amplitude dependence and corresponding effective colors of variability may be obtained. The Principal Component Analysis (PCA) allows to split multi-color signal into orthogonal contributions, which are generally not separated in the frequency domain, and is effective with the ``running parabola" filtering. For the pseudo-multi-channel modification of the PCA, the criterion is proposed to determine the optimal length of the ``Caterpillar".