Title: Hidden in the background: a local approach to CMB anomalies

Abstract: We investigate a framework aiming to provide a common origin for the large-angle anomalies detected in the Cosmic Microwave Background (CMB), which are hypothesized as the result of the statistical inhomogeneity developed by different isocurvature fields of mass $m\sim H$ present during inflation. The inhomogeneity arises as the combined effect of $(i)$ the initial conditions for isocurvature fields (obtained after a fast-roll stage finishing many $e$-foldings before cosmological scales exit the horizon), $(ii)$ their inflationary fluctuations and $(iii)$ their coupling to other degrees of freedom. Our case of interest is when these fields (interpreted as the precursors of large-angle anomalies) leave an observable imprint only in isolated patches of the Universe. When the latter intersect the last scattering surface, such imprints arise in the CMB. Nevertheless, due to their statistically inhomogeneous nature, these imprints are difficult to detect, for they become hidden in the background similarly to the Cold Spot. We then compute the probability that a single isocurvature field becomes inhomogeneous at the end of inflation and find that, if the appropriate conditions are given (which depend exclusively on the preexisting fast-roll stage), this probability is at the percent level. Finally, we discuss several mechanisms (including the curvaton and the inhomogeneous reheating) to investigate whether an initial statistically inhomogeneous isocurvature field fluctuation might give rise to some of the observed anomalies. In particular, we focus on the Cold Spot, the power deficit at low multipoles and the breaking of statistical isotropy.