Gravitational waves from a curvaton model with blue spectrum
Masahiro Kawasaki, Naoya Kitajima, Shuichiro Yokoyama
TL;DR
This paper investigates gravitational waves generated at second order by blue-tilted curvature perturbations in curvaton models, focusing on quadratic and axion-like potentials. By combining inflaton and curvaton fluctuations to achieve COBE-scale normalization with a blue-tilted small-scale spectrum, it computes the induced GWB from scalar perturbations and from curvaton kinetic stress, exploring both subdominant and dominant curvaton energy density regimes. The results show that the scalar-induced GWB can reach detectable levels (up to ~10^{-10}–10^{-8} in Ω_GW) with a characteristic peak determined by curvaton decay or domination, and that future detectors like LISA, DECIGO/BBO, and SKA could probe these scenarios; the axion-like model can produce a plateau signature, while the kinetic-term contribution remains subdominant under current bounds. These findings provide a promising observational handle on curvaton dynamics, small-scale curvature perturbations, and possible PBH dark matter connections.
Abstract
We investigate the gravitational wave background induced by the first order scalar perturbations in the curvaton models. We consider the quadratic and axion-like curvaton potential which can generate the blue-tilted power spectrum of curvature perturbations on small scales and derive the maximal amount of gravitational wave background today. We find the power spectrum of the induced gravitational wave background has a characteristic peak at the frequency corresponding to the scale reentering the horizon at the curvaton decay, in the case where the curvaton does not dominate the energy density of the Universe. We also find the enhancement of the amount of the gravitational waves in the case where the curvaton dominates the energy density of the Universe. Such induced gravitational waves would be detectable by the future space-based gravitational wave detectors or pulsar timing observations.
