Comparison of double-inflationary models with observations.

Abstract

We consider double-inflationary models with two noninteracting scalar fields: a ``light'' scalar field ${\mathrm{\ensuremath{\varphi}}}_{\mathit{l}}$ with the potential 1/2${\mathit{m}}_{\mathit{l}}^{2}$${\mathrm{\ensuremath{\varphi}}}_{\mathit{l}}^{2}$ and a ``heavy'' scalar field ${\mathrm{\ensuremath{\varphi}}}_{\mathit{h}}$ with the potential \ensuremath{\lambda}/n${\mathrm{\ensuremath{\varphi}}}_{\mathit{h}}^{\mathit{n}}$, n=2,4, whose effective mass is larger than ${\mathit{m}}_{\mathit{l}}$ during inflation with ${\mathrm{\ensuremath{\varphi}}}_{\mathit{h}}$>${\mathit{M}}_{\mathit{P}}$. Cold dark matter (CDM) with the initial spectrum of adiabatic perturbations produced in these models is compared with observations. These models contain two more free parameters than the standard CDM model with an initial scale-invariant spectrum. We normalize our spectra to the COBE DMR and compare the predictions with observations on the biasing factor, large-scale peculiar velocities, quasar and galaxy formation, and the Stromlo-APM counts-in-cells analysis. The model with n=4 is excluded by the data while for the n=2 model, taking cosmic variance into account, a small window of parameters compatible with observations is found.