Advanced statistical analysis of linear and nonlinear Regge trajectories for light non-strange mesons

Abstract A rigorous statistical analysis of Regge-type trajectories for light non-strange meson mass spectra is carried out, which explicitly accounts for experimental uncertainties. We test three scenarios: a linear model with distinct radial ( n ) and orbital ( l ) slopes (Model 1), a linear model with a universal slope (Model 2), and a nonlinear model featuring a universal slope and a Dirac-Coulomb-type term $$\\propto (l+1)^{-1}$$ ∝ ( l + 1 ) - 1 (Model 3). Optimization is performed using a nonlinear $$\\chi ^2$$ χ 2 minimization framework, where the intrinsic theoretical model uncertainty is determined self-consistently by enforcing $$\\chi ^2/\\text {dof} = 1$$ χ 2 / dof = 1 . To ensure robust model selection, we employ the Akaike and Bayesian information criteria alongside complementary statistical tests. Our analysis demonstrates that moderate deviations from linearity in the Regge spectrum are predominantly localized within the S -wave resonance sector. These distortions are successfully accommodated by the nonlinear correction in Model 3, from which an approximate Coulomb-type degeneracy, $$m^2(n,l) \\propto n+l$$ m 2 ( n , l ) ∝ n + l , emerges as a statistically robust feature. Furthermore, we show that the l -dependent correction to the principal quantum number in light non-strange mesons is consistent with the leading-order relativistic correction to the Coulomb problem. The statistical inferences gained in our analysis can provide an efficient tool for identifying non- $$q\\bar{q}$$ q q ¯ candidates in the light meson spectrum. For example, the masses of three main candidates for the lightest scalar glueball – the scalar mesons $$f_0(1370)$$ f 0 ( 1370 ) , $$f_0(1500)$$ f 0 ( 1500 ) and $$f_0(1710)$$ f 0 ( 1710 ) – lie on the corresponding radial trajectories of the ordinary $$q\\bar{q}$$ q q ¯ -states.