Blind component separation in wavelet space. Application to CMB analysis

Abstract

It is a recurrent issue in astronomical data analysis that observations are unevenly sampled or incomplete maps with missing patches or intentionaly masked parts. In addition, many astrophysical emissions are non stationary processes over the sky. Hence spectral estimation using standard Fourier transforms is no longer reliable. Spectral matching ICA (SMICA) is a source separation method based on covariance matching in Fourier space which is successfully used for the separation of diffuse astrophysical emissions in Cosmic Microwave Background observations. We show here that wavelets, which are standard tools in processing non stationary data, can profitably be used to extend SMICA. Among possible applications, it is shown that gaps in data are dealt with more conveniently and with better results using this extension, wSMICA, in place of the original SMICA. The performances of these two methods are compared on simulated CMB data sets, demonstrating the advantageous use of wavelets.