Non-gaussianity and Statistical Anisotropy in Cosmological Inflationary Models

TL;DR

This paper investigates models of cosmological inflation that produce large non-gaussianity and statistical anisotropy, analyzing scalar and vector field perturbations and their observable implications.

Contribution

It introduces detailed calculations of non-gaussianity and anisotropy levels in inflationary models with scalar and vector fields, including loop corrections and their observational relevance.

Findings

High non-gaussianity levels achievable in scalar field models.

Vector field models can produce significant anisotropy and non-gaussianity.

Some models predict non-gaussianity exceeding current observational limits.

Abstract

We study the statistical descriptors for some cosmological inflationary models that allow us to get large levels of non-gaussianity and violations of statistical isotropy. Basically, we study two different class of models: a model that include only scalar field perturbations, specifically a subclass of small-field slow-roll models of inflation with canonical kinetic terms, and models that admit both vector and scalar field perturbations. We study the former to show that it is possible to attain very high, including observable, values for the levels of non-gaussianity f_{NL} and \tao_{NL} in the bispectrum B_\zeta and trispectrum T_\zeta of the primordial curvature perturbation \zeta respectively. Such a result is obtained by taking care of loop corrections in the spectrum P_\zeta, the bispectrum B_\zeta and the trispectrum T_\zeta . Sizeable values for f_{NL} and \tao_{NL} arise even if \zeta is generated during inflation. For the latter we study the spectrum P_\zeta, bispectrum B_\zeta and trispectrum $T_\zeta of the primordial curvature perturbation when \zeta is generated by scalar and vector field perturbations. The tree-level and one-loop contributions from vector field perturbations are worked out considering the possibility that the one-loop contributions may be dominant over the tree level terms. The levels of non-gaussianity f_{NL} and \tao_{NL}, are calculated and related to the level of statistical anisotropy in the power spectrum, g_\zeta . For very small amounts of statistical anisotropy in the power spectrum, the levels of non-gaussianity may be very high, in some cases exceeding the current observational limit.