Meson Spectroscopy from Bayesian MCMC: Probing Confinement and State Mixing
arXiv:2604.04846
Abstract
We present a Bayesian study of the meson spectrum using three non-relativistic confining potentials, namely the Cornell potential, a logarithmically modified extension, and a screened Cornell form. Parameters for each potential are sampled using Markov chain Monte Carlo (MCMC), retaining correlations among the fitted parameters. The resulting parameter distributions are used to calculate the spectrum through the multiplet, including masses, spin-dependent splittings, mixing angles, and wave-function observables. The predicted masses are further examined through Regge trajectories. The low-lying spectrum is relatively stable across the three potentials, while their predictions become increasingly separated with excitation as the states probe larger distances, accompanied by growing parameter-induced uncertainties. The Regge trajectories show stronger non-linearity for low-lying states and progressively approach linear behavior with excitation. These results quantify the sensitivity of the predicted excited spectrum to the chosen confining interaction and provide uncertainty estimates for states that remain experimentally less explored.
73 pages, 13 figures, 18 tables, Revised version