Non-Gaussian Signatures of a Thermal Big Bang

TL;DR

This paper explores the non-Gaussian signatures of a thermal Big Bang model within a bimetric theory, providing precise predictions for non-Gaussianity parameters and discovering a new shape of flattened non-Gaussianity.

Contribution

It adapts non-Gaussianity calculations to a thermal bimetric model, predicting specific non-Gaussianity shapes and amplitudes, including a novel flattened shape due to thermal initial conditions.

Findings

Predicted $f^{\rm local}_{\rm NL} = -3/2$

Predicted $f^{\rm equil}_{\rm NL} \simeq 0.4$

Discovered a new flattened non-Gaussianity shape

Abstract

What if Big Bang was hot from its very inception? This is possible in a bimetric theory where the source of fluctuations is thermal, requiring the model to live on a critical boundary in the space of parameters and can be realized when an anti-DBI brane moves within an $EAdS_2 \times E_3$ geometry. This setup renders the model unique, with sharp predictions for the scalar spectral index and its running. We investigate the non-Gaussian signatures of this thermal bimetric model, or "bi-thermal" for short. We adapt the standard calculation of non-Gaussianities for $P(X,\phi)$ models to the thermal nature of the model, emphasising how the bi-thermal peculiarities affect the calculation and alter results. This leads to precise predictions for the shape and amplitude of the three-point function of the bi-thermal model (at tree-level): $f^{\rm local} _{\rm NL} = -3/2$ and $f^{\rm equil} _{\rm NL} = -2 + 4 \sqrt{3}\pi/9 \simeq 0.4$. We also discover a new shape of flattened non-gaussianity $\propto (k_1+k_2-k_3)^{-3/2} +$ permutations, which is expected due to the excited thermal initial conditions. These results, along with our earlier predictions for the scalar power spectrum, provide sharp targets for the future generation of cosmological surveys.