Accelerating Universe without dark energy: matter creation after DESI DR2

TL;DR

This paper explores an alternative cosmological model where matter creation drives the universe's acceleration, matching observations and potentially replacing dark energy, with two specific models showing promising results against standard cosmology.

Contribution

It introduces and tests two matter creation models that can replicate the universe's accelerated expansion without dark energy, validated against recent observational data.

Findings

Both models mimic ΛCDM behavior.

Matter creation is statistically significant in both models.

Model I is statistically equivalent to ΛCDM; Model II is favored with DESI data.

Abstract

We investigate the cosmological matter creation scenario, an alternative approach to both dark energy and modified gravity theories, after the recent DESI DR2-BAO release. We consider that the total matter sector consists of three independently evolving components, namely, radiation, baryons, and dark matter, with the latter being governed by an adiabatic matter creation process that leads to a modified continuity equation. Due to the violation of the standard conservation law, a creation pressure appears, and under a proper choice of dark-matter particle creation rate one can obtain the present accelerating phase as well as the past thermal history of the Universe. We study two specific matter creation rates. By applying the dynamical-system analysis we show that both Model I and Model II can mimic a $\Lambda$CDM-like behavior. Furthermore, we perform a detailed observational confrontation using a series of latest observational datasets including Cosmic Chronometers (CC), Supernovae Type Ia (SNIa) (Pantheon+, DESY5 and Union3 samples) and DESI Baryon Acoustic Oscillations (BAO) (DR1 and DR2 samples). In both Model I and Model II we find evidence of matter creation at many standard deviations. Finally, applying the AIC and BIC information criteria we find that Model I is statistically equivalent with $\Lambda$CDM scenario, while Model II shows a mixed picture, namely for most datasets $\Lambda$CDM scenario is favoured, however when DESI data are included matter creation Model II is favoured over $\Lambda$CDM paradigm.