How isotropic is dark energy?

Technical Summary: How Isotropic is Dark Energy?

Problem Statement
Recent tensions in late-time expansion data, particularly between the SH0ES $H_0$ value and DESI BAO distances, have renewed interest in models beyond the standard $\Lambda$CDM framework. While the standard model assumes a homogeneous and isotropic Friedmann-Robertson-Walker (FRW) universe, alternative frameworks allowing for anisotropic dark energy (where pressure varies with direction) offer a potential resolution. However, a significant challenge exists: direct fits of anisotropic Bianchi I models to Type Ia Supernovae (SNe) data often yield large shear parameters that violate stringent constraints from the Cosmic Microwave Background (CMB), specifically the Integrated Sachs-Wolfe (ISW) quadrupole bound. The central question addressed is whether dark energy can be sufficiently anisotropic to explain late-time observational tensions while remaining consistent with CMB limits.

Methodology
The authors investigate time-dependent anisotropic dark energy within the Bianchi I cosmological framework, characterized by non-isotropic pressure and distinct scale factors along three coordinate axes.

• Parameterization: The study employs a piecewise constant representation (top-hat bins) for the anisotropic equation of state parameter $\Delta w(a)$ over the scale factor interval $a \in [1/4, 1]$.
• CMB Constraint Mechanism: To address the ISW quadrupole bound, the authors develop a novel constrained basis function approach. They enforce a vanishing line-of-sight integral of the shear ($Q_T \approx 0$) by constructing a null-space basis via Singular Value Decomposition (SVD) with a continuation method. This ensures that while the shear $s(a)$ can be non-zero and time-varying, its integrated effect along the photon path cancels out, satisfying the CMB temperature quadrupole constraint by construction.
• Data Analysis: The framework is applied to the complete Pantheon+SH0ES SNe dataset (1701 SNe Ia) combined with DESI Year 2 BAO distance measurements. The analysis includes 13 BAO measurements spanning $z=0.295$ to $z=2.33$.
• Statistical Approach: Posterior distributions are sampled using the affine-invariant ensemble sampler emcee. The likelihood function combines SNe residuals (incorporating SALT2 corrections and Cepheid calibrators) and BAO distance ratios, marginalized over the sound horizon $r_{drag}$ using a Planck prior.

Key Results
The study compares four models: standard isotropic $w$CDM, a single-bin (constant) anisotropic model, a five-bin unconstrained anisotropic model, and a five-bin SVD-constrained model.

1. Fit Improvements:
   • The single-bin unconstrained model improves the fit over $w$CDM by $\Delta(-2 \ln L_{iso}) = 14.8$. However, it predicts a CMB quadrupole amplitude ($C_2^{aniso} \approx 10^{-3}$) that significantly exceeds observational limits, effectively ruling it out.
   • The five-bin unconstrained model yields a larger improvement of $\Delta(-2 \ln L_{iso}) = 26.6$. It captures directional structure in the Pantheon+ data and partially alleviates the tension between SH0ES and DESI by allowing $H_0 \approx 71$ km s$^{-1}$ Mpc$^{-1}$. Like the single-bin model, it violates the quadrupole constraint.
   • The five-bin SVD-constrained model achieves a fit improvement of $\Delta(-2 \ln L_{iso}) = 15.4$ while satisfying the CMB quadrupole bound ($\log_{10}(C_2^{aniso}) \approx -10$).
2. Physical Interpretation:
   • The improvement in the constrained model arises from two sources: capturing directional structure in the SNe data and partially alleviating the $H_0$ tension.
   • The best-fit constrained model exhibits sign-changing shear that integrates to zero along the line of sight. This allows for non-trivial late-time anisotropy that mimics dynamical dark energy (raising $H(z)$ at $z \lesssim 0.2$ and driving a more negative effective equation of state at $0.2 \lesssim z \lesssim 3$) without violating the ISW bound.
   • The inferred $H_0$ for the constrained model is $69.9^{+1.4}_{-0.9}$ km s$^{-1}$ Mpc$^{-1}$, lying between the standard $w$CDM value and the unconstrained anisotropic result.
3. Sensitivity: Maintaining compatibility with the quadrupole bound requires fine-tuning of the bin amplitudes $\Delta w$ at the level of $\sim 10^{-4}$.

Significance and Claims
The paper claims to provide a practical framework for testing late-time anisotropic dark energy that circumvents the "tension" between SNe-favored anisotropy and CMB constraints. By enforcing a vanishing line-of-sight shear integral, the authors demonstrate that anisotropic dark energy models can improve the fit to current expansion data (Pantheon+ and DESI) without being immediately ruled out by the CMB temperature quadrupole.

The authors note that while the constrained model offers a better fit than standard $w$CDM, the underlying equation of state $w_{de}$ remains consistent with a cosmological constant ($w \approx -1$), with the time-varying anisotropic stress mimicking dynamical dark energy. They explicitly acknowledge limitations: the current construction cancels the temperature quadrupole but does not automatically suppress the large-angle polarization quadrupole ($Q_P$), which is sourced by local shear during recombination. Furthermore, the effective equation of state generated by anisotropic stress is more constrained than general $w(a)$ parameterizations, limiting the scope for further fit improvements. The authors conclude that while anisotropic dark energy captures directional structure isotropic models miss, its viability remains contingent on fine-tuning and future tests involving CMB polarization and directional BAO analyses.