CMB bispectrum constraints on DHOST inflation - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue JCAP Année : 2024

CMB bispectrum constraints on DHOST inflation

Résumé

We present the first direct constraints on a Degenerate Higher Order Scalar Tensor (DHOST) inflation model using the Planck 2018 Cosmic Microwave Background (CMB) results on non-Gaussianities. We identify that the bispectrum consists of a fixed contribution following from the power spectrum and a linear combination of terms depending on five free parameters defining the cubic perturbations to the DHOST model. The former peaks in the squeezed limit, while the latter is maximised in the equilateral limit. We directly confront the model predictions to the CMB bispectrum statistics via the public code CMB-BEST and marginalize over the free parameters. We explicitly show that there are viable DHOST inflationary models satisfying both power spectrum and bispectrum constraints from Planck. However, rather surprisingly, the constraints exclude certain models at the $6\sigma$-level even though they pass the conventional fudge factor tests. In this case and despite having a handful of free parameters, the model's large squeezed bispectrum cannot be cancelled out without introducing a large bispectrum in other limits which are strongly constrained by Planck's non-detection of primordial non-Gaussianity. We emphasize that first-order approximations such as fudge factors, albeit commonly used in the literature, may be misleading and provide weaker constraints. A proper analysis of the constraints from Planck requires a more robust approach, such as the one provided by the CMB-BEST code.

Dates et versions

hal-04328942 , version 1 (07-12-2023)

Licence

Paternité

Identifiants

Citer

Wuhyun Sohn, Andrei Lazanu, Philippe Brax, James R. Fergusson. CMB bispectrum constraints on DHOST inflation. JCAP, 2024, 04, pp.071. ⟨10.1088/1475-7516/2024/04/071⟩. ⟨hal-04328942⟩
11 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More