Constraining Spatial Curvature with Priors from Swampland Conjectures

TL;DR

This paper investigates how string theory-inspired priors, especially from swampland conjectures, influence the inferred spatial curvature in dark energy models using cosmological data.

Contribution

It introduces a framework incorporating swampland-motivated priors into dark energy models and analyzes their impact on cosmological parameter estimation.

Findings

Swampland priors mildly shift the inferred values of spatial curvature.

The dynamical system analysis reveals fixed points supporting acceleration in open universes.

Imposing swampland constraints narrows the parameter space consistent with observational data.

Abstract

We study a string-motivated theoretical prior on the quintessential dark energy model with exponential potential, \( V(\phi) = V_0 e^{-\lambda \phi} \), allowing for non-zero spatial curvature. First, we formulate the corresponding dynamical system and investigate its cosmological evolution numerically, illustrating the phase-space behaviour and the influence of curvature on the background dynamics. In open universes (\( \Omega_k > 0 \)), it has been suggested that a curvature-related fixed point may support accelerated expansion even for relatively steep potentials compatible with swampland considerations. Next, we explicitly impose swampland-motivated priors on the slope parameter $\lambda$, restricting it to values consistent with the de Sitter conjecture that excludes the (curved) $\Lambda$CDM limit. Furthermore, we restrict our considerations to the range of field excursion that is consistent with the swampland distance conjecture. Our primary interest is the possibility that such theoretically-motivated priors may shift values of cosmological parameters inferred by observational data, compared with the standard analysis based on theory-agnostic priors such as a sufficiently wide flat prior. We examine this possibility using a combination of Planck CMB data, DESI BAO measurements, and recent Type Ia supernova samples, performing a Bayesian inference of the model parameters. Our analysis indicates that the swampland-motivated prior mildly shifts the values of $\Omega_k$.