Tracking the multifield dynamics with cosmological data: a Monte Carlo approach

Abstract

We introduce a numerical method specifically designed for investigating generic multifield models of inflation where a number of scalar fields ϕ K are minimally coupled to gravity and live in a field space with a non-trivial metric 𝒢> IJ (ϕ K ). Our algorithm consists of three main parts. Firstly, we solve the field equations through the entire inflationary period, deriving predictions for observable quantities such as the spectrum of scalar perturbations, primordial gravitational waves, and isocurvature modes. We also incorporate the transfer matrix formalism to track the behavior of adiabatic and isocurvature modes on super-horizon scales and the transfer of entropy to scalar modes after the horizon crossing. Secondly, we interface our algorithm with Boltzmann integrator codes to compute the subsequent full cosmology, including the cosmic microwave background anisotropies and polarization angular power spectra. Finally, we develop a novel sampling algorithm able to efficiently explore a large volume of the parameter space and identify a sub-region where theoretical predictions agree with observations. In this way, sampling over the initial conditions of the fields and the free parameters of the models, we enable Monte Carlo analysis of multifield scenarios. We test all the features of our approach by analyzing a specific model and deriving constraints on its free parameters. Our methodology provides a robust framework for studying multifield inflation, opening new avenues for future research in the field.