Primordial power spectrum: a complete analysis with the WMAP nine-year data

TL;DR

This paper introduces an improved deconvolution algorithm to reconstruct the primordial power spectrum from WMAP 9-year data, achieving higher accuracy and efficiency, and revealing potential features beyond the standard power-law model.

Contribution

The authors develop a more straightforward, error-sensitive Richardson-Lucy deconvolution method applicable directly to unbinned CMB data, enabling more precise primordial spectrum reconstruction.

Findings

Achieved over 300 improvement in  fit compared to power-law

Reconstructed spectrum consistent with power-law within error margins

Method capable of detecting features beyond standard models

Abstract

We have improved further the error sensitive Richardson-Lucy deconvolution algorithm making it applicable directly on the un-binned measured angular power spectrum of Cosmic Microwave Background observations to reconstruct the form of the primordial power spectrum. This improvement makes the application of the method significantly more straight forward by removing some intermediate stages of analysis allowing a reconstruction of the primordial spectrum with higher efficiency and precision and with lower computational expenses. Applying the modified algorithm we fit the WMAP 9 year data using the optimized reconstructed form of the primordial spectrum with more than 300 improvement in \chi^2 with respect to the best fit power-law. This is clearly beyond the reach of other alternative approaches and reflects the efficiency of the proposed method in the reconstruction process and allow us to look for any possible feature in the primordial spectrum projected in the CMB data. Though the proposed method allow us to look at various possibilities for the form of the primordial spectrum, all having good fit to the data, proper error-analysis is needed to test for consistency of theoretical models since, along with possible physical artefacts, most of the features in the reconstructed spectrum might be arising from fitting noises in the CMB data. Reconstructed error-band for the form of the primordial spectrum using many realizations of the data, all bootstrapped and based on WMAP 9 year data, shows proper consistency of power-law form of the primordial spectrum with the WMAP 9 data at all wave numbers. Including WMAP polarization data in to the analysis have not improved much our results due to its low quality but we expect Planck data will allow us to make a full analysis on CMB observations on both temperature and polarization separately and in combination.