Inflationary gravitational waves in consistent $D\to 4$ Einstein-Gauss-Bonnet gravity

TL;DR

This paper explores a new inflationary scenario in consistent D→4 Einstein-Gauss-Bonnet gravity, revealing a novel attractor regime that influences primordial gravitational waves and their non-Gaussian features, with potential observational implications.

Contribution

It introduces the Gauss-Bonnet attractor regime in D→4 Einstein-Gauss-Bonnet gravity and analyzes its effects on primordial gravitational waves and non-Gaussianity.

Findings

Discovery of the Gauss-Bonnet attractor regime in inflation.

Power spectra and spectral tilts derived for curvature perturbations and GWs.

Nonlinear GW dispersion relation with observable consequences.

Abstract

We study the slow-roll single field inflation in the context of the consistent $D\to4$ Einstein-Gauss-Bonnet gravity that was recently proposed in \cite{Aoki:2020lig}. In addition to the standard attractor regime, we find a new attractor regime which we call the Gauss-Bonnet attractor as the dominant contribution comes from the Gauss-Bonnet term. Around this attractor solution, we find power spectra and spectral tilts for the curvature perturbations and gravitational waves (GWs) and also a model-independent consistency relation among observable quantities. The Gauss-Bonnet term provides a nonlinear $k^4$ term to the GWs dispersion relation which has the same order as the standard linear $k^2$ term at the time of horizon crossing around the Gauss-Bonnet attractor. The Gauss-Bonnet attractor regime thus provides a new scenario for the primordial GWs which can be tested by observations. Finally, we study non-Gaussianity of GWs in this model and estimate the nonlinear parameters $f^{s_1s_2s_3}_{\rm NL,\;sq}$ and $f^{s_1s_2s_3}_{\rm NL,\;eq}$ by fitting the computed GWs bispectra with the local-type and equilateral-type templates respectively at the squeezed limit and at the equilateral shape. For helicities $(+++)$ and $( -- )$, $f^{s_1s_2s_3}_{\rm NL,\;sq}$ is larger while $f^{s_1s_2s_3}_{\rm NL,\;eq}$ is larger for helicities $(++-)$ and $(--+)$.