Viscous f(T, L_m) Model Yields H0 of 70.5 km/s/Mpc Consistent with Pantheon+SH0ES and DESI DR II
A viscous extension of f(T, L_m) gravity fits current cosmological datasets and returns observationally viable H0 and w values. The modeling strategy parallels data-constrained machine-learned potentials in biochemistry. Future growth and lensing measurements will decide if the model survives beyond background expansion.
The study solves modified Friedmann equations for a barotropic viscous fluid and fits four free parameters via Markov Chain Monte Carlo sampling against four datasets. Best-fit values place the effective equation of state in the quintessence regime at late times while keeping viscous pressure negative throughout cosmic history. The deceleration parameter transitions from positive to negative at z approximately 0.6, matching the observed acceleration. This analytic Hubble function reproduces the background expansion without invoking a separate dark-energy component. The approach mirrors machine-learned potentials now used in enzyme catalysis, where high-dimensional parameter spaces are constrained against experimental observables to achieve quantum-level accuracy; both replace ad-hoc assumptions with data-driven functional forms that generalize across scales. Similar MCMC pipelines in drug-design workflows have already accelerated lead optimization by reducing false positives in binding free-energy predictions. Extending the cosmology pipeline to include growth-rate data from upcoming surveys would test whether the viscous term remains stable under perturbation theory.
Behera et al.: Next DESI data release will shift best-fit viscous parameter outside 1-sigma if effective w remains above -0.95 at z=0.3.
Sources (3)
- [1]Primary Source(https://arxiv.org/abs/2609.09169)
- [2]Supporting Source(https://arxiv.org/abs/2609.09293)
- [3]Supporting Source(https://arxiv.org/abs/2305.16647)