Separating polarized cosmological and galactic emissions for CMB B-mode polarization experiments

In this work we study the relevance of the component separation technique based on the Independent Component Analysis (ICA) and investigate its performance in the context of a limited sky coverage observation and from the viewpoint of our ability to differentiate between cosmological models with different primordial B-mode content. We focus on the low Galactic emission sky patch, corresponding to the target of several operating and planned CMB experiments and which, in many respects, adequately represents a typical "clean" high latitude sky. We consider two fiducial observations, one operating at low (40, 90 GHz) and one at high (150, 350 GHz) frequencies and thus dominated by the synchrotron and thermal dust emission, respectively. We use a parallel version of the FASTICA code to explore a substantial parameter space including Gaussian pixel noise level, observed sky area and the amplitude of the foreground emission and employ large Monte Carlo simulations to quantify errors and biases pertinent to the reconstruction for different choices of the parameter values. We identify a large subspace of the parameter space for which the quality of the CMB reconstruction is excellent. For both the cosmological models, with and without the primordial gravitational waves, we find that FASTICA performs extremely well even in the cases when the B mode CMB signal is up to a few times weaker than the foreground contamination and the noise amplitude is comparable with the total CMB polarized emission. In addition we discuss limiting cases of the noise and foreground amplitudes, for which the ICA approach fails.