Non-gaussianity and cosmic uncertainty in curvaton-type models

TL;DR

This paper extends curvaton-type models by incorporating scale dependence in the Gaussian perturbation δσ and its quadratic contribution to the curvature perturbation ζ, while addressing cosmic uncertainty and anthropic considerations. It develops a box-averaging framework and uses the δN formalism to derive the full spectra, bispectrum, and trispectrum, including tilt effects and the resulting f_NL and τ_NL. A master curvaton formula is presented, with detailed analysis of special cases (e.g., Ω_σ = 1, induced isocurvature) and BD versus general distributions, highlighting how box size, tilt, and post-inflationary evolution influence observational predictions. The work discusses the status of curvaton-type models in light of WMAP year three results, showing that negative tilt can be accommodated and that anthropic arguments can shape the allowed parameter space, while offering concrete criteria to distinguish curvaton scenarios from inflaton-driven perturbations.

Abstract

In curvaton-type models, observable non-gaussianity of the curvature perturbation would come from a contribution of the form $(δσ)^2$, where $δσ$ is gaussian. I analyse this situation allowing $δσ$ to be scale-dependent. The actual curvaton model is considered in more detail than before, including its cosmic uncertainty and anthropic status. The status of curvaton-type models after WMAP year three data is considered.