Description
The nature of dark energy has been a growing point of debate in recent years, particularly after the DESI measurements of the Baryon Acoustic Oscillations. While frequentist metrics appear to indicate a growing preference for a dynamical dark energy, some bayesian approaches indicate otherwise. Beyond this, there also lies the question of whether there exists a physical motivation behind any phenomenological parametrisation of dynamical dark energy. The model described in this talk is motivated from computations of the non-perturbative QCD vacuum, which we parametrise into a dynamic, non-local contribution to the energy density, described by two additional cosmological parameters. I will describe how this model compares against both CPL and ΛCDM on the latest cosmological datasets, employing both frequentist χ² statistics and Bayesian model comparison. For the latter, the evidence is estimated from MCMC chains via the learnt harmonic mean estimator method which bypasses the need for the computationally expensive, nested sampling to obtain estimates of the evidence. This marks one of the first use cases of this method in cosmological bayesian model comparison.