Testing inflationary models with non-vacuum initial states against cosmological observations

A.G.SÁNCHEZ; C.G.SCÓCCOLA; S.J.LANDAU; A. GANGUI

preprint Phys Rev D.

Año: 2007

There has been recently some growing interest in cosmological models with a primordial broken scale invariant (BSI) spectrum of adiabatic perturbations. For instance, inflationary models with non--vacuum initial states for the cosmological perturbations, which can arise from generic trans--Planckian physics, lead to peculiar features in the primordial power spectrum. These features will accordingly alter the resulting Cosmic Microwave Background (CMB) temperature anisotropies and matter power spectrum, P(k), and therefore, by constraining the CMB angular power spectrum and P(k) with recent observations, one can place strong limits on the corresponding model of the early universe. In this paper a class of spatially flat models with a cosmological constant and a BSI primordial spectrum of adiabatic perturbations is confronted with the most up-to-date observational data of CMB and matter power spectrum. The theoretical model includes a parameter $n_{b}$ for the number of quanta in the non-vacuum initial state, and a privileged scale leading to the existence of a feature in the primordial power spectrum. This feature is located at comoving wavenumber $k_{b}$ and its profile is characterized by a step in $k$ with steepness $alpha$, the full set ${n_{b}, k_{b}, alpha}$ being taken as free parameters in our numerical study. By performing a detailed Markov Chain Monte Carlo analysis with {CAMB} and {CosmoMC} of the latest CMB and large scale structure measurements, including the 3-year {WMAP} and the final 2dFGRS catalogue, we derive joint constraints on 8 out of the many relevant primary parameters -both cosmological and feature- of our BSI adiabatic model. Our best-fit model yields a primordial spectrum with a step located at very large scales $log(k_{b})<-3.28$ (at 95%  c.l.). Our results for the step-like feature in the initial spectrum depart from the suggested $k_{b}= 0.05 h {Mpc}^{-1}$ or $k_{b}= 0.004 h {Mpc}^{-1}$ derived in the previous literature from various statistical tests, and place the location of the feature well beyond the combined data, meaning that the allowed range for this parameter is restricted to a small region where its effect on the CMB and large scale structure data is very small.  [NB: Este trabajo será enviado en breve a PRD. Aun no fue terminado ni enviado porque faltan correr algunas rutinas para calcular las evidencias Bayesianas de los diferentes modelos. Esto último no modificará en absoluto nuestras conclusiones del paper, pero es conveniente llevarlas a cabo para mostrar que nuestros resultados son más robustos. Lamentablemente, estas rutinas requieren mucho tiempo de CPU que, al momento, no disponemos. Esperamos poder terminar estos cálculos numéricos en un cluster de Durham, como lo vinimos haciendo ya durante el 2007.  Una versión preliminar de este trabajo fue acceptada para su publicación en el Bol. Asoc. Argent. Astron. correspondiente a la última reunión nacional.]