Unveiling the Pitfalls of CPL Parametrization at High Redshifts: A Critical Assessment of the $ω_{0}$$ω_a$CDM Model with DESI DR2 BAO Data

Abstract: We critically assess the w0waCDM model, employing the Chevallier-Polarski-Linder (CPL) parametrization, w(z) = w0 + wa * z / (1 + z), using Baryon Acoustic Oscillation (BAO) measurements from DESI Data Release 2 (DR2). While DESI BAO data broadly align with the standard LambdaCDM model, giving Omega_m = 0.3, our Markov Chain Monte Carlo fits show notable variation in best-fit values and degeneracy directions for w0waCDM across combinations of BAO observables (DV/rd, DM/rd, DH/rd, and DM/DH). When constrained solely by DESI BAO data, the w0waCDM model shows internal inconsistencies. Unlike LambdaCDM and wCDM, which fit all observables consistently, w0waCDM predicts lower values for rd-dependent observables and higher values for DM/DH. This conflict, including unphysical parameter compensation at high redshift, reflects limitations of the CPL form, especially its behavior at high redshift. The model compensates for poor individual fits by inflating Omega_m and driving wa to large negative values. Our findings suggest that DESI DR2 BAO data alone do not robustly constrain w0waCDM due to degeneracies and CPL limitations. These issues affect both rd-based and ratio observables like DM/DH. We recommend applying dynamical dark energy parametrizations directly to DV, DM, and DH measurements to better constrain wa across redshifts.