An Application of Global ILC Algorithm over Large Angular Scales to Estimate CMB Posterior Using Gibbs Sampling

Abstract: In this work, we formalize a new technique to investigate joint posterior density of Cosmic Microwave Background (CMB) signal and its theoretical angular power spectrum given the observed data, using the global internal-linear-combination (ILC) method first proposed by Sudevan & Saha (2017). We implement the method on low resolution CMB maps observed by WMAP and Planck satellite missions, using Gibbs sampling, assuming that the detector noise is negligible on large angular scales of the sky. The main products of our analysis are best fit CMB cleaned map and its theoretical angular power spectrum along with their error estimates. We validate the methodology by performing Monte Carlo simulations that includes realistic foreground models and noise levels consistent with WMAP and Planck observations. Our method has an unique advantage that the posterior density is obtained without any need to explicitly model foreground components. Secondly, the power spectrum results with the error estimates can be directly used for cosmological parameter estimations.