A Partial-Sky Gibbs ILC Approach for the Estimation of CMB Posterior over Large Angular Scales of the Sky

TL;DR

This paper introduces a novel partial-sky Gibbs ILC method to estimate the CMB posterior and power spectrum from masked sky maps, enabling accurate analysis without explicit foreground modeling.

Contribution

The paper develops a formalism to incorporate partial-sky maps into the Gibbs ILC algorithm, allowing for full-sky CMB analysis from masked observations.

Findings

Successfully estimates full-sky CMB power spectrum from partial-sky data.

Validates the method with realistic simulations including foregrounds and noise.

Produces partial-sky cleaned CMB maps suitable for cosmological analysis.

Abstract

In this article we present a formalism to incorporate the partial-sky maps to the Gibbs ILC algorithm to estimate the joint posterior density of the Cosmic Microwave Background (CMB) signal and the theoretical CMB angular power spectrum given the observed CMB maps. In order to generate the partial-sky maps we mask all the observed CMB maps provided by WMAP and Planck satellite mission using a Gaussian smoothed mask formed on the basis of thermal dust emissions in Planck 353 GHz map. The central galactic region from all the input maps is removed after the application of the smoothed mask. While implementing the Gibbs ILC method on the partial-sky maps, we convert the partial-sky cleaned angular power spectrum to the full-sky angular power spectrum using the mode-mode coupling matrix estimated from the smoothed mask. The main products of our analysis are partial-sky cleaned best-fit CMB map and an estimate of the underlying full-sky theoretical CMB angular power spectrum along with their error estimates. We validate the methodology by performing detailed Monte Carlo simulations after using realistic models of foregrounds and detector noise consistent with the WMAP and Planck frequency channels used in our analysis. We can estimate the posterior density and full-sky theoretical CMB angular power spectrum, without any need to explicitly model the foreground components, from partial-sky maps using our method. Another important feature of this method is that the power spectrum results along with the error estimates can be directly used for cosmological parameter estimations.