An Unbiased Estimator of the Full-sky CMB Angular Power Spectrum at Large Scales using Neural Networks

TL;DR

This paper demonstrates that neural networks can accurately and unbiasedly predict the full-sky CMB angular power spectrum at large scales from partial sky data, reducing systematic errors caused by foreground contamination.

Contribution

The study introduces a neural network-based method to estimate the full-sky CMB power spectrum from masked observations, improving accuracy and reducing covariance compared to traditional methods.

Findings

Neural networks produce unbiased full-sky power spectrum estimates.

Predictions are largely uncorrelated with observed spectra.

Covariance of neural network predictions is significantly smaller.

Abstract

Accurate estimation of the Cosmic Microwave Background (CMB) angular power spectrum is enticing due to the prospect for precision cosmology it presents. Galactic foreground emissions, however, contaminate the CMB signal and need to be subtracted reliably in order to lessen systematic errors on the CMB temperature estimates. Typically bright foregrounds in a region lead to further uncertainty in temperature estimates in the area even after some foreground removal technique is performed and hence determining the underlying full-sky angular power spectrum poses a challenge. We explore the feasibility of utilizing artificial neural networks to predict the angular power spectrum of the full sky CMB temperature maps from the observed angular power spectrum of the partial sky in which CMB temperatures in some bright foreground regions are masked. We present our analysis at large angular scales with two different masks. We produce unbiased predictions of the full-sky angular power spectrum and recover the underlying theoretical power spectrum using neural networks. Our predictions are also uncorrelated to a large extent. We further show that the multipole-space covariances of the predictions of full-sky spectra made by the ANNs are much smaller than those of the estimates obtained using the pseudo-$C_\ell$ method.