The hemispherical asymmetry from a scale-dependent inflationary bispectrum

TL;DR

This paper develops a method to analyze how scale-dependent primordial bispectra can produce CMB hemispherical asymmetry, showing it is possible to do so without conflicting with observational bispectrum constraints.

Contribution

It introduces a general response function approach to evaluate the impact of large-scale modes on the two-point function in local-like bispectrum models.

Findings

A new method to compute the response of the two-point function in any local-like bispectrum model.

It is possible to generate observable hemispherical asymmetry without violating bispectrum constraints.

A concrete model demonstrating compatibility with Planck data constraints.

Abstract

If the primordial bispectrum is sufficiently large then the CMB hemispherical asymmetry may be explained by a large-scale mode of exceptional amplitude which perturbs the zeta two-point function. We extend previous calculations, which were restricted to one- or two-source scenarios, by providing a method to compute the response of the two-point function in any model yielding a 'local-like' bispectrum. In general, this shows that it is not the reduced bispectrum fNL which sources the amplitude and scale-dependence of the mode coupling but rather a combination of 'response functions'. We discuss why it is difficult to construct successful scenarios and enumerate the fine-tunings which seem to be required. Finally, we exhibit a concrete model which can be contrived to match the observational constraints and show that to a Planck-like experiment it would appear to have |fNL-local| ~ |fNL-equi| ~ |fNL-ortho| ~ 1. Therefore, contrary to previous analyses, we conclude that it is possible to generate the asymmetry while respecting observational constraints on the bispectrum and low-ell multipoles even without tuning our location on the long-wavelength mode.