Hint of a truncated primordial spectrum from the CMB large-scale anomalies

TL;DR

This paper investigates whether a cutoff in the primordial power spectrum can simultaneously explain large-scale anomalies in the CMB, such as low correlation at large angles and missing low multipole power, without conflicting with standard cosmological predictions.

Contribution

It introduces a modified analysis of the primordial power spectrum with a cutoff, showing it can account for multiple large-scale CMB anomalies consistently.

Findings

The cutoff parameter k_min is consistent across different anomaly analyses.

A cutoff in P(k) explains both the low correlation and missing power at large scales.

The standard model predictions remain valid at smaller scales (ell > 30).

Abstract

Several satellite missions have uncovered a series of potential anomalies in the fluctuation spectrum of the cosmic microwave background temperature, including: (1) an unexpectedly low level of correlation at large angles, manifested via the angular correlation function, C(theta); and (2) missing power in the low multipole moments of the angular power spectrum, C_ell. Their origin is still debated, however, due to a persistent lack of clarity concerning the seeding of quantum fluctuations in the early Universe. A likely explanation for the first of these appears to be a cutoff, k_min=(3.14 +/- 0.36) x 10^{-4} Mpc^{-1}, in the primordial power spectrum, P(k). Our goal in this paper is twofold: (1) we examine whether the same k_min can also self-consistently explain the missing power at large angles, and (2) we confirm that the of this cutoff in P(k) does not adversely affect the remarkable consistency between the prediction of Planck-LCDM and the Planck measurements at ell > 30. We use the publicly available code CAMB to calculate the angular power spectrum, based on a line-of-sight approach. The code is modified slightly to include the additional parameter (i.e., k_min) characterizing the primordial power spectrum. In addition to this cutoff, the code optimizes all of the usual standard-model parameters. In fitting the angular power spectrum, we find an optimized cutoff, k_min = 2.04^{+1.4}_{-0.79} x 10^{-4} Mpc^{-1}, when using the whole range of ell's, and k_min=3.3^{+1.7}_{-1.3} x 10^{-4} Mpc^{-1}, when fitting only the range ell < 30, where the Sachs-Wolfe effect is dominant. These are fully consistent with the value inferred from C(theta), suggesting that both of these large-angle anomalies may be due to the same truncation in P(k).