While the standard Lambda-CDM model treats dark energy as a fixed cosmological constant with equation-of-state parameter w = -1, a new preprint (arXiv:2604.16632) from Saulo Pereira and collaborators proposes replacing the Lambda term directly in the Einstein-Hilbert action with the Kretschmann scalar K = R_{\mu\nu\rho\sigma} R^{\mu\nu\rho\sigma}, a geometric invariant that quantifies the strength of spacetime curvature independent of coordinate choice. This creates a dynamical dark energy whose density and pressure evolve naturally with the universe's expansion, without introducing new scalar fields.

The researchers derive the modified Friedmann equations and constrain two free parameters using a combined dataset of Type Ia supernovae (Pantheon+ sample, approximately 1,048 events spanning z=0 to z=2.3) and cosmic chronometer measurements (32 differential age points from passively evolving galaxies). Their Markov Chain Monte Carlo analysis shows the model fits the late-time expansion history as well as or better than Lambda-CDM at low redshifts, with the derived w(z) closely tracking phenomenological parametrizations recently favored by DESI Year-1 baryon acoustic oscillation data.

Notably, the model exhibits a phantom-crossing behavior where w evolves from above -1 at higher redshifts to below -1 today, mirroring the trend reported in the DESI collaboration's 2024 analysis (arXiv:2404.03002) that combined their BAO measurements with supernova and CMB data, yielding a 2.5-sigma preference for evolving dark energy. A related phenomenological study by Escamilla et al. (arXiv:2404.08056) on quintessence-like models also highlighted similar low-z w(z) reconstruction, which this geometric approach reproduces without fine-tuned potentials.

This work goes beyond typical phenomenological fits by grounding dynamical dark energy in spacetime geometry itself. Previous coverage and even the paper's abstract understate the theoretical pedigree: the Kretschmann scalar appears in quadratic gravity theories and is central to avoiding singularities in certain black hole solutions. The approach echoes Starobinsky's R^2 inflation model but applied to late-time acceleration, potentially linking early and late universe acceleration under geometric corrections. What the paper and most reporting miss is the potential fourth-order field equations this action generates; while the authors demonstrate viability at low z, they do not fully explore possible Ostrogradsky instabilities or additional degrees of freedom that could conflict with solar-system tests or early-universe nucleosynthesis.

Limitations are explicit: the fit excludes high-redshift CMB constraints and full DESI BAO, focusing only on z < 2 where dark energy dominates. The sample for cosmic chronometers remains small (N≈32), introducing larger uncertainties than supernova data. As a preprint, it awaits peer review and independent replication. Yet the pattern is clear: post-DESI, geometric and dynamical models are gaining traction as Lambda-CDM tensions (Hubble constant, S8) persist. If extended successfully, this Kretschmann-derived dark energy could represent a paradigm shift, suggesting acceleration emerges from curvature invariants rather than an ad-hoc constant, offering a cleaner path toward quantum gravity compatibility.

This synthesis reveals a missed connection: unlike scalar-field quintessence models that can suffer from tracking problems, a pure invariant-based approach might naturally suppress contributions at early times when curvature is high, automatically transitioning to dark energy dominance later.