Observational constraints of diffusive dark-fluid cosmology

TL;DR

This work studies a diffusive dark-fluid cosmology in which DM and DE exchange energy through a diffusion term, offering a unified framework to explain late-time acceleration without a cosmological constant. The authors derive the background and linear perturbation equations using the 1+3 covariant formalism and constrain the model with diverse data sets via MCMC, including DESI DR2 BAO, cosmic chronometers, multiple SNIa compilations, and RSD growth data. They find a negative interaction parameter $Q_{dm}$ across analyses, implying energy transfer from DM to DE and yielding statefinder trajectories reminiscent of Chaplygin gas, while structure growth remains DM-dominated with modest deviations from $\Lambda$CDM in $f\sigma_8$ at high redshift. Information criteria indicate the diffusive model is competitive with $\Lambda$CDM but not decisively preferred, suggesting it is a viable alternative that could help address tensions and merits further testing with upcoming CMB and large-scale structure data.

Abstract

In this manuscript, we investigate late-time cosmology and the evolution of cosmic structures using an interacting dark fluid model in which dark matter (DM) and dark energy (DE) interact through a diffusive mechanism. To provide a comprehensive understanding, we derive the background evolution and perturbation equations within this model and obtain cosmological parameters through MCMC simulations. We use recent measurements for statistical analysis and constrain the parameters $H_0$ in km/s/Mpc, $Ω_m$, $r_d$, $M$, $σ_8$, $S_8$, and the interaction term $Q_{dm}$. From the constrained values of $Q_{dm}$, we show that the diffusive model is a promising alternative DE model, capable of driving late-time cosmic acceleration due to energy exchange from DM to DE. State-finder diagnostics indicate that the model behaves like a Chaplygin gas when energy transfers from DM to DE during the Universe's expansion. We also investigate the growth of density contrast, finding $δ_m(z)\ggδ_{de}(z)$, which highlights the dominant role of DM in structure formation. Redshift space distortion and growth rate analysis show that minor deviations from $Λ$CDM at low redshifts, with larger differences at higher redshifts, indicate the impact of energy diffusion on early structure growth. Finally, we perform a detailed statistical analysis, including ${\mathcal{L}(\hatΘ|data)}$, $χ^2$, $\rm{AIC}$, and $\rm{BIC}$, which strongly supports the proposed diffusive dark-fluid model.

Figures (8)

• Figure 1: Posterior distributions of the parameters using PPS + CC + DESY5 + RSD + f for both models.
• Figure 3: Posterior distributions of the parameters using DESI DR2 BAO + CC + Union3 + RSD + f for both models.
• Figure 4: The comparison of $\Lambda$CDM and diffusive model $H_0$ values (in km/s/Mpc) with early and direct measurements (left panel), $S_8$ values between the $\Lambda$CDM and diffusive models alongside various late-time and early-time measurements (right panel).
• Figure 5: The diagrams for multiple cosmological parameters in both models: diffusive dark-fluid and $\Lambda$CDM models are highlighted in this Fig. i). With slight variations for the diffusive model at higher redshifts, the top-left panel deceleration parameter $q(z)$ depicts the change from deceleration to acceleration. ii) The effective equation of state parameter $w_{\text{eff}}(z)$ in both models is shown in the upper-right panel. iii) The bottom-left panel shows that the CC data for both models fit the Hubble parameter $H(z)$. iv) The bottom-right panel shows the distance modulus $\mu(z)$, which both models closely match the data with PPS. For illustrative purpose we use $Q_{dm} = -0.017, -0.013,-0.009$, $H_0 = 68.959$ in km/s/Mpc, $\Omega_m = 0.361$ for diffusive model and $\Omega_m = 0.310$ and $H_0 = 69.950$ in km/s/Mpc for $\Lambda$CDM model from joint analysis of PPS + CC + DESY5 + RSD + f presented in Table \ref{['table-bestfit']}.
• Figure 6: The state finder diagnostic $q$ vs. $r$ (left panel) and $s$ vs. $r$ (right panel). We use $Q_{dm} = -0.017, -0.013,-0.009$, $H_0 = 68.959$ in km/s/Mpc, $\Omega_m = 0.361$ for diffusive model and $\Omega_m = 0.310$ and $H_0 = 69.950$ in km/s/Mpc for $\Lambda$CDM model.