Information-Geometric Perspective on the Hubble Tension: Eigenmode Rotation and Curvature Suppression in wCDM

This paper analyzes the Hubble tension within the wCDM model using an information-geometric framework, revealing that the apparent relief in tension arises not from new physical agreement but from a geometric reconfiguration where varying the dark energy equation of state suppresses Planck's constraint curvature while DESI DR2 data introduces a stiff geometric wall that limits tension relief.

The Gist

The Big Problem: The "Hubble Tension"

Imagine two groups of people trying to measure the speed of a car driving away from them.

• Group A (The Early Universe): They look at the car's blueprint and the smoke from its exhaust (the Cosmic Microwave Background, or CMB) to guess how fast it must be going based on physics. They calculate a speed of 67 km/s.
• Group B (The Late Universe): They stand on the side of the road with a radar gun and measure the car right now. They get a speed of 73 km/s.

The numbers don't match. This disagreement is called the Hubble Tension. For years, scientists have wondered: Is the blueprint wrong? Is the radar gun broken? Or is there a new force of nature we don't understand?

The Paper's New Idea: It's Not Just the Numbers, It's the "Stiffness"

Most scientists have been focusing on the difference between the numbers (67 vs. 73). This paper argues that the difference is only half the story. The other half is how confident each group is in their measurement.

The author uses a concept called Information Geometry. Think of the universe's parameters (like the speed of the car) as a landscape.

• The "Shift": How far apart the two groups' best guesses are.
• The "Stiffness" (Curvature): How "rigid" or "tight" the measurement is.

The Analogy of the Rubber Band:
Imagine the measurement is a rubber band stretched between two points.

• If the rubber band is loose and floppy (low stiffness), the two groups can be far apart, and it's not a big deal. They might just be slightly off.
• If the rubber band is tight and stiff (high stiffness), even a tiny distance between the two groups creates a huge amount of tension.

This paper says the Hubble Tension is so high not just because the numbers are different, but because the "rubber bands" holding the data are incredibly stiff.

The Main Discovery: Changing the Rules Doesn't Help

Scientists tried to fix the tension by changing the rules of the game. They added a new variable called $w$ (which describes how Dark Energy behaves). They hoped that by making the rules more flexible, the rubber band would get looser, and the tension would disappear.

What the paper found:

1. The "Loose" Illusion: When they added this new variable ($w$), the Planck data (Group A) did look looser. The rubber band seemed to stretch out. The tension number went down.
2. The Trap: But this wasn't because the universe changed. It was because the "rubber band" got so floppy that it stopped holding the measurement tight. It's like loosening a screw so much that the whole machine falls apart. The data didn't actually agree better; it just became less certain.
3. The "Geometric Wall": Then, a new dataset called DESI (a massive survey of galaxies) arrived. DESI is like a super-precise laser measurement.
   • In the old, loose scenario, DESI slammed into the data like a wall.
   • Because DESI is so precise, it "stiffened" the rubber band again.
   • Result: The tension didn't go away. In fact, the new data reinforced the original "67" number, pushing the "73" measurement even further away.

The "Sound Horizon" Twist

The paper also looks at how we measure the "ruler" used by these groups (called the Sound Horizon, $r_d$).

• Scenario 1 (The Ruler is Floating): If we let the ruler's size float freely, the new data (DESI) can't really help measure the speed. It's like trying to measure a car's speed with a ruler that keeps changing length. In this case, the tension is just a battle between the Blueprint (Planck) and the Radar (SH0ES).
• Scenario 2 (The Ruler is Fixed): If we lock the ruler's size to a known value, the new data (DESI) becomes incredibly powerful. It acts like a third, super-strong arm pulling the measurement.
  • The Surprise: When the ruler is fixed, DESI provides 90% of the "stiffness." It becomes the dominant force.
  • The Result: This "Geometric Wall" created by DESI is so strong that it crushes any hope of the "phantom" solutions (where the universe behaves strangely) working. It forces the universe back to the standard model, making the tension with the local measurements (SH0ES) even more obvious.

The Conclusion: Why We Can't Just "Fix" It

The author concludes that the Hubble Tension isn't just a math error or a missing variable. It is a geometric collision.

• The Blueprint (Planck) is rigid and holds the universe at a specific speed.
• The Radar (SH0ES) is rigid and holds it at a different speed.
• The New Data (DESI) is so precise that it acts as a massive anchor, locking the Blueprint in place.

Adding more variables (like changing Dark Energy) doesn't solve the problem because it just makes the rubber band floppy without actually moving the two groups closer together. To truly solve the tension, we either need to:

1. Move the groups closer (find a physical reason why the speeds should match).
2. Or, find a way to make the "rubber bands" softer (which the new data is actively preventing).

In short: The universe is telling us that the standard model is very rigid. The new, high-precision data is acting like a concrete wall, preventing us from sneaking around the tension by just tweaking the rules. The tension is real, and it's likely a sign of something fundamental we still don't understand, rather than a simple calculation error.

Technical Summary

1. Problem Statement

The Hubble tension—the persistent discrepancy between early-Universe measurements of the Hubble constant ($H_0$) from the Cosmic Microwave Background (CMB, e.g., Planck) and late-Universe measurements from the local distance ladder (e.g., SH0ES)—is typically quantified by the statistical significance of the shift in best-fit parameters. However, standard scalar metrics often conflate two distinct physical ingredients:

1. Parameter Displacement: The magnitude of the shift between best-fit values.
2. Directional Stiffness (Curvature): The "rigidity" of the likelihood constraints along the direction of that shift.

The paper argues that recent model extensions (such as allowing the dark energy equation of state $w$ to vary, i.e., $w$CDM) appear to alleviate the tension not by resolving the physical discrepancy, but by geometrically altering the constraint structure. Specifically, the paper investigates how extending the parameter space reshapes the Fisher information geometry, potentially diluting the statistical penalty (curvature) associated with the $H_0$ discrepancy without introducing new independent information.

2. Methodology

The author employs an Information-Geometric framework based on the Fisher Matrix formalism within a local Gaussian approximation.

• Quadratic Tension Decomposition: The tension $T^2$ along a unit direction $\hat{u}$ is factorized into the squared parameter shift ($\Delta_{\parallel}^2$) and the directional curvature ($\kappa$):
  $$T^2(\hat{u}) = \kappa_{\text{joint}}(\hat{u}) \cdot \Delta_{\parallel}^2$$
  where $\kappa_{\text{joint}} = \sum \kappa_i$ is the sum of curvatures from independent datasets (Planck, DESI DR2, SH0ES).
• Eigenmode Analysis: The Fisher matrix is diagonalized to identify principal eigenmodes (stiff directions) and eigenvalues (curvature magnitudes). The study tracks how these eigenmodes rotate and how eigenvalues change when moving from $\Lambda$CDM to $w$CDM.
• Probe Comparison: The analysis compares three configurations:
  1. Planck alone: Analyzing the acoustic-scale constraint.
  2. Planck + DESI DR2: Examining how BAO data injects curvature.
  3. Planck + DESI DR2 + SH0ES: Analyzing the full three-probe tension.
• Sound Horizon Treatments: The study contrasts two scenarios for the sound horizon ($r_d$):
  • $r_d$FREE: $r_d$ is marginalized (free parameter), creating a degeneracy between $h$ and $r_d$.
  • $r_d$FIX: $r_d$ is fixed to a theoretical value, breaking the degeneracy.

3. Key Contributions and Results

A. Curvature Suppression and Eigenmode Rotation in $w$CDM

• Suppression: Extending $\Lambda$CDM to $w$CDM suppresses the leading Fisher eigenvalue of the Planck constraint by a factor of $\sim 37.5$ (in the $(\Omega_{m0}, \ln H_0, w)$ space). This corresponds to a broadening of the posterior uncertainty by a factor of $\sim 6$.
• Rotation: The dominant eigenmode (associated with the acoustic scale $\theta_*$) rotates by $\sim 87^\circ$ in the $(\Omega_{m0}, \ln(h r_d))$ plane.
• Physical Interpretation: This is not a new physical direction but a geometric dilution. The acoustic constraint is spread over a larger parameter volume. The apparent reduction in tension is an "Ockham penalty" effect: the model gains flexibility (lower $\chi^2$) but loses geometric stiffness (lower curvature), making the parameter shift statistically less significant.

B. The Role of DESI DR2 as "Curvature Injection"

• In $\Lambda$CDM: DESI DR2 reinforces the existing stiff acoustic constraint, leading to only a mild shift in parameters.
• In $w$CDM: Because the Planck constraint has been "softened" (curvature suppressed) by the $w$ extension, the high-precision DESI data acts as a curvature injection. It re-establishes a "geometric wall" by adding significant stiffness along the expansion-rate direction, effectively blocking the "phantom escape" ($w < -1$) that $w$CDM alone might suggest.

C. Three-Probe Geometry and the $r_d$ Treatment

The paper reveals a critical dependency on how the sound horizon is treated:

• $r_d$FREE (Degenerate): When $r_d$ is free, DESI contributes zero curvature along the pure $H_0$ axis because it only constrains the product $h r_d$. The system reduces to a binary balance between Planck (acoustic rigidity) and SH0ES (local stiffness).
• $r_d$FIX (Fixed): When $r_d$ is fixed, the degeneracy is broken. DESI acquires a massive projection onto the $H_0$ axis, contributing ~90% of the total directional curvature.
• Result: In the $r_d$FIX scenario, the tension becomes a "geometric collision" between the Planck degeneracy (pulling toward phantom $w$) and the extremely stiff DESI constraint (pulling toward $\Lambda$CDM). The joint solution is stabilized near $w \approx -1$ not because of new physics, but because the combined curvature is too rigid to allow large shifts.

D. Low-Dimensional Curvature Structure

The analysis confirms that the Hubble tension is dominated by a single, highly rigid Fisher eigenmode. The cumulative curvature fraction ($C_k$) saturates after the first or second mode. This low-dimensional structure explains why Bayesian evidence often disfavors complex extensions: the tension is a localized conflict along a specific stiff axis, not a broad multi-dimensional discrepancy.

4. Significance and Implications

• Reframing the Tension: The paper shifts the perspective from "discrepancy in values" to "discrepancy in geometric constraints." It demonstrates that the Hubble tension is a geometric collision between early-Universe acoustic rigidity and late-Universe distance-ladder stiffness.
• Explaining Model Failures: It provides a physical explanation for why many $w$CDM or Early Dark Energy (EDE) models fail to resolve the tension in global fits: they often merely redistribute or suppress curvature (diluting information) rather than generating a genuine physical shift in the preferred $H_0$ values that aligns with late-time data.
• Diagnostic Tool: The proposed "Shift-Curvature Decomposition" offers a calculation-free diagnostic. By analyzing the partition of Fisher curvature ($f_X$) among probes, one can predict the stability of a joint solution without exhaustive MCMC sampling.
• Future Directions: The paper suggests that resolving the tension requires either:
  1. Shift Reduction: Physically moving the preferred $H_0$ values closer (e.g., via Early Universe physics modifications).
  2. Curvature Softening: Genuinely weakening the dominant acoustic eigenmode (difficult without violating other constraints).
  3. New Probes: Introducing independent curvature (e.g., from standard sirens or time-delay cosmography) that competes with the existing hierarchy, rather than just reinforcing it.

Conclusion

The paper concludes that the persistence of the Hubble tension in extended models is a predictable consequence of geometric collision. The high precision of modern data (like DESI DR2) creates a "geometric wall" that prevents parameter shifts from escaping the $\Lambda$CDM solution. Apparent alleviations of tension in $w$CDM are largely artifacts of curvature suppression (dilution of information) rather than genuine physical convergence. The framework provides a transparent, geometric basis for interpreting why model extensions often fail to resolve the tension despite lowering $\chi^2$.