w0w_aCDM Dark Energy Model

• The w0w_aCDM model is a flexible parameterization that describes a time-varying dark energy equation of state via the parameters w0 and wa.
• It is widely used in cosmology to test dynamical dark energy against observations from CMB, BAO, Type Ia supernovae, and growth measurements.
• Recent analyses indicate a moderate (∼2σ) preference for evolving dark energy over ΛCDM, despite internal tensions among different data sets.

The spatially flat $w_0w_a$CDM model is a leading phenomenological parameterization designed to capture possible time variation in the dark energy equation of state, beyond the standard $\Lambda$CDM scenario. It is defined by an equation of state parameter $w(z) = w_0 + w_a z/(1+z)$ or, equivalently, $w(a) = w_0 + w_a(1-a)$ in terms of the scale factor $a = 1/(1+z)$. This two-parameter extension (Chevallier–Polarski–Linder, CPL) offers a flexible yet tractable framework—capable of modeling "thawing", "freezing", and crossing ("quintom") behaviors of dark energy—while enabling global analyses against cosmological data sets. The $w_0w_a$CDM model is widely employed in large-scale structure, CMB, and supernova cosmology for testing the origin, evolution, and possible dynamical nature of cosmic acceleration.

1. Formal Definition and Dynamical Evolution

The $w_0w_a$CDM model assumes a dark energy fluid characterized by an equation of state evolving linearly with scale factor: $w(a) = w_0 + w_a (1 - a)$ or, equivalently, as a function of redshift: $w(z) = w_0 + w_a \frac{z}{1+z}$ Here $w_0$ is the present-day value and $w_a$ quantifies its rate of change. The conservation equation for dark energy,

$$\frac{d\rho_{de}}{dz} = \frac{3[1+w(z)]}{1+z} \rho_{de}(z)$$

yields the analytic solution: $$\rho_{de}(z) = \rho_{de,0} (1+z)^{3(1+w_0+w_a)} \exp\left(-\frac{3w_a z}{1+z}\right)$$ The Friedmann equation in a spatially flat universe reads: $$H^2(z) = H_0^2 \left[ \Omega_{m}(1+z)^3 + \Omega_{r}(1+z)^4 + (1-\Omega_{m}-\Omega_{r}) (1+z)^{3(1+w_0+w_a)} \exp\left(-3w_a \frac{z}{1+z}\right) \right]$$ This framework encompasses the cosmological constant ($w_0=-1, w_a=0$), thawing and freezing quintessence, and can describe a wide array of dynamical histories (Park et al., 2024).

2. Data Sets and Statistical Methodology

Current constraints on $w_0w_a$CDM leverage joint analyses of cosmological probes:

• CMB: Planck 2018 TT, TE, EE spectra and lensing (“P18+lensing”).
• BAO: Pre-DESI data (BOSS, eBOSS, DES Y3, 6dFGS, SDSS MGS, Ly$\alpha$), or DESI DR1/DR2.
• Type Ia Supernovae: Pantheon+, Union3, DES-Y5.
• Cosmic Chronometers: $H(z)$ measurements.
• Growth measurements: $f\sigma_8(z)$.

Parameter inference is performed using MCMC methods (e.g., CosmoMC, MontePython, Cobaya+PolyChord), with convergence criteria $R-1 < 0.01$ or stricter. Model selection employs information criteria—AIC, DIC—and Bayesian evidence. Dataset consistency is quantified by tension metrics (tension $\sigma$, suspiciousness log $S$, dimensionality $d_G$) (Park et al., 2024, Ong et al., 13 Nov 2025). Notably, the correlation coefficient between $w_0$ and $w_a$ is typically highly negative ($\sim -0.85$), reflecting an extended degeneracy direction.

3. Current Constraints and Model Comparison

Recent constraints—excluding DESI BAO—on the flat $w_0w_a$CDM parameterization from Planck 2018 CMB, lensing, Pantheon+ SNe, pre-DESI BAO, $H(z)$, and $f\sigma_8(z)$ data are (Park et al., 2024): $$w_0 = -0.850 \pm 0.059 \qquad w_a = -0.59^{+0.26}_{-0.22}$$ with derived parameters

$$H_0 = 67.80 \pm 0.64~{\rm km\,s^{-1}Mpc^{-1}} \qquad \Omega_m = 0.3094 \pm 0.0063 \qquad \sigma_8 = 0.8108 \pm 0.0091$$

Compared to $\Lambda$CDM, there is a likelihood-ratio preference at the level $\Delta\chi^2_{min} = -6.25$ (AIC and DIC differences $-2.25$ and $-2.45$, i.e., "positive" evidence per Jeffreys’ scale for $-6<\Delta{\rm DIC}<-2$). Analyzing the DESI 2024 BAO plus CMB and PantheonPlus, the DESI collaboration reports similar, slightly looser bounds: $w_0^{({\rm DESI})} = -0.827\pm0.063$, $w_a^{({\rm DESI})} = -0.75^{+0.29}_{-0.25}$ (Park et al., 2024). The newer non-DESI compilation yields $1\sigma$ error bars $\sim 6\%$ smaller on $w_0$, reflecting the inclusion of $H(z)$ and $f\sigma_8$ data.

These results consistently show $w_0 > -1, w_a < 0$ is favored over $(-1,0)$. The displacement with respect to DESI 2024 is $\Delta w_0 = -0.27\sigma$, $\Delta w_a = +0.44\sigma$, statistically insignificant.

4. Statistical Significance, Tensions, and Interpretation

The combined non-DESI cosmological data favor evolving dark energy in $w_0w_a$CDM at approximately $2\sigma$ relative to the cosmological constant (Park et al., 2024). Tension between CMB and low-$z$ data in $w_0w_a$CDM is quantified at $\sigma=2.7$, exceeding the DESI-DR1 BAO vs. CMB $\Lambda$CDM tension of $1.9\sigma$. Excluding Pantheon+ SNe from the non-CMB compilation preserves tension at $2.4$–$2.5\sigma$, and still disfavors $\Lambda$CDM at $\sim 2\sigma$.

Model comparison via $\Delta$AIC, $\Delta$DIC, and likelihood-ratio approaches consistently indicate "positive" (but not "strong") evidence for $w_0w_a$CDM over $\Lambda$CDM. The best-fit values of $w_0$ and $w_a$ lie within $1\sigma$ of the $\Lambda$CDM values, but the $\sim2\sigma$ deviation is robust to compendium datasets.

5. Comparison with DESI 2024 and Impact of Dataset Choices

The comparison between non-DESI constraints and DESI 2024 results shows close agreement. Inclusion of BAO data from DESI modestly shifts and slightly loosens the contours in the $w_0$–$w_a$ plane, but the best-fit values remain statistically consistent at $<0.5\sigma$ shifts. Additional non-CMB probes ($H(z)$, $f\sigma_8(z)$) tighten the orthogonal direction of the likelihood ellipsoid.

A residual $\sim 2.7\sigma$ tension between CMB and low-$z$ constraints persists even after the exclusion of SNe data. When the Pantheon+ dataset is omitted, evidence against $\Lambda$CDM remains at $\sim2\sigma$, indicating the preference is not dominated by the supernovae.

6. Caveats: Parameterization, Model Systematics, and Future Prospects

The $w_0w_a$CDM model is a phenomenological parameterization, not a microphysical scalar field or modified gravity model. Its linear-in-scale-factor construction makes it a practical benchmark, but it does not correspond to a physically self-consistent scalar-field potential in all cases. The significance of dynamical dark energy is sensitive to systematic effects in the CMB (notably the lensing anomaly $A_L$), internal tensions among probes, and prior choices.

Residual internal tensions among datasets and the moderate evidence for $w_0w_a$CDM underscore the need for caution. High-precision, next-generation data from DESI, Euclid, the Vera Rubin Observatory, and CMB-S4 will be necessary to robustly distinguish between evolving and constant dark energy. Physically motivated models—dynamical scalar fields, modified gravity, and effective field theory treatments—will be required to interpret any true detection of $w(z) \ne -1$.

7. Summary Table: Key Constraints from Non-DESI Data (Park et al., 2024)

	$w_0$	$w_a$	$H_0$ (${\rm km\,s^{-1}Mpc^{-1}}$)	$\Omega_m$	$\sigma_8$
P18+lensing+non-CMB	$-0.850 \pm 0.059$	$-0.59^{+0.26}_{-0.22}$	$67.80 \pm 0.64$	$0.3094 \pm 0.0063$	$0.8108 \pm 0.0091$
DESI 2024	$-0.827 \pm 0.063$	$-0.75^{+0.29}_{-0.25}$	$68.03 \pm 0.72$	$0.3085 \pm 0.0068$

The $w_0w_a$CDM model remains a data-driven standard for probing dynamical dark energy signatures, with current global evidence amounting to a persistent but moderate ($\sim2\sigma$) preference for evolution relative to $\Lambda$CDM. The phenomenological approach is robust across non-DESI and DESI datasets, but ultimate resolution will require both new data and theoretical developments in the modeling of cosmic acceleration (Park et al., 2024).