Evaluating Performance of Usual ILC and Global ILC Approach in Pixel Space Over Large Angular Scales of the Sky

Abstract

During the implementation of usual Internal Linear Combination (ILC) method in pixel space, the foreground minimization is performed without taking into consideration the covariance structure of the CMB anisotropies. A new foreground model independent method, the Global ILC method, is developed where we incorporate the CMB theoretical covariance matrix in the foreground minimization algorithm. In this paper, we evaluate the performance of the usual ILC and the new Global ILC methods implemented in pixel space over large angular scales of the sky. We use the observed CMB maps provided by WMAP and Planck satellite missions. We also perform detailed Monte Carlo simulations to validate our results. We find that the cleaned CMB map and the CMB angular power spectrum obtained using the Global ILC method has significantly lower reconstruction error as compared to those obtained by usual ILC method.