Geometric Constraints on the Pre-Recombination Expansion History from the Hubble Tension

This paper demonstrates that a model-independent reconstruction of the pre-recombination expansion history reveals a class of purely early-time solutions to the Hubble tension, characterized by a smooth $\simeq 15\%$ expansion rate enhancement around matter-radiation equality that satisfies geometric CMB constraints.

The Gist

Imagine the universe as a giant, expanding loaf of raisin bread. As the dough rises, the raisins (galaxies) move apart. For a long time, cosmologists have been arguing about exactly how fast this dough is rising right now.

On one side, we have measurements from the "local neighborhood" (nearby galaxies) that say the dough is rising fast. On the other side, we have measurements from the "ancient past" (the Cosmic Microwave Background, or CMB, which is the leftover heat from the Big Bang) that say the dough is rising slower. This disagreement is called the Hubble Tension, and it's a major headache for modern physics.

This paper by Davide Pedrotti acts like a detective trying to solve the mystery by looking at the "blueprint" of the universe's expansion before the raisins even formed.

The Problem: A Rigid Ruler

The author starts with a simple idea: To fix the speed discrepancy, we need to change the size of a "ruler" used in the ancient universe. This ruler is called the sound horizon. Think of it like a standard measuring tape used by the early universe.

If we want the universe to look like it's expanding faster today (to match the local measurements), we need to make that ancient measuring tape slightly shorter. The author calculates that we need to shrink this tape by about 7%.

However, there's a catch. The universe has three different "rulers" (angular scales) that we can measure in the ancient sky. If you try to shrink one ruler to fix the speed problem, you accidentally stretch or shrink the other two, breaking the picture of the universe we see today. It's like trying to fix a wobbly table leg by sawing off another leg; you might fix the wobble, but now the table is too short.

The Investigation: A Model-Free Reconstruction

Instead of guessing a specific new physics theory (like "invisible dark energy" or "new particles"), the author asked a different question: "What does the expansion rate have to look like mathematically to shrink that one ruler without breaking the other two?"

He used a computer to reconstruct the expansion history of the universe from the Big Bang up to the moment the first atoms formed (recombination), without assuming any specific theory. He let the math tell him the shape of the solution.

The Discovery: The "Smooth Transition"

The math revealed a very specific, rigid shape that any solution must follow. It's not a sudden explosion or a random jump. Instead, it looks like a smooth, gentle hill in the expansion rate.

Here is the analogy:
Imagine the universe's expansion rate is a car driving down a highway.

1. The Standard Model (ΛCDM): The car drives at a steady, predictable speed.
2. The Required Solution: To fix the Hubble Tension, the car must smoothly accelerate to about 15% faster than usual right before it reaches a specific checkpoint (the moment of recombination).
3. The Timing: This speed boost must happen right around the time when matter and radiation were balancing each other out (Matter-Radiation Equality). It needs to ramp up smoothly, hit that 15% peak just before the "finish line" of the early universe, and then smoothly slow back down.

The paper finds that this specific "hill" shape is the only way to shrink the sound horizon by 7% without messing up the other cosmic measurements.

The Twist: The "No-Go" Trap

The author then points out a major problem with this solution.

If the universe sped up by 15% right before the "finish line" of the early universe, and then just kept that speed forever, the universe today would be expanding way too fast. It would over-correct the problem.

To fix this, the universe would need a second transition later on (during the "Dark Ages," long after the first light but before stars formed) to slow the expansion back down.

• The Catch: While this second slowdown might look fine on paper (background level), the author suggests that if you look at the "ripples" in the universe (perturbations), this second slowdown would likely create visible "scars" or artifacts in the cosmic map that we don't see.

The Conclusion: A Blueprint for Failure?

The paper concludes that while a purely early-universe solution mathematically exists at the background level, it is incredibly fragile.

• It requires a very specific, smooth 15% speed boost.
• It likely requires a second, invisible speed adjustment later on.
• If you try to build a physical theory (like a new type of energy) to create this speed boost, it might break the delicate balance of the universe's ripples.

The author calls this a "blueprint" or a "stress test." It tells future physicists: "If you want to solve the Hubble Tension with early-universe physics, your theory must look exactly like this smooth hill. If it doesn't, it won't work."

In short, the paper suggests that the universe is very picky. It allows for a specific kind of speed-up in the past, but the rules are so strict that it might be impossible for any real physical mechanism to pull it off without breaking other parts of the cosmic puzzle.

Technical Summary

1. Problem Statement

The paper addresses the Hubble Tension, the persistent $\sim 7\sigma$ discrepancy between the local measurement of the Hubble constant ($H_0 \approx 73.5$ km/s/Mpc) and the value inferred from the Cosmic Microwave Background (CMB) assuming the standard $\Lambda$CDM model ($H_0 \approx 67.2$ km/s/Mpc).

While late-time modifications to cosmology are largely ruled out by low-redshift probes (BAO, Supernovae), early-time solutions (e.g., Early Dark Energy, extra relativistic species) face a critical geometric challenge. To resolve the tension, a model must reduce the sound horizon at recombination ($r_s^*$) by approximately 7% to match the higher $H_0$. However, this reduction must occur without altering the observed ratios of three fundamental CMB angular scales:

1. The sound horizon angle ($\theta_s^*$).
2. The photon diffusion (Silk damping) scale angle ($\theta_d^*$).
3. The matter-radiation equality scale angle ($\theta_{eq}$).

Previous models often struggle to reduce $r_s^*$ without decoupling these scales or violating the full CMB power spectrum constraints.

2. Methodology

The author employs a strictly model-agnostic, non-parametric reconstruction of the pre-recombination expansion history $H(z)$. Instead of assuming a specific physical mechanism (like a scalar field), the paper reconstructs the fractional deviation of the Hubble parameter, $\delta H/H_{\text{fid}}$, directly from geometric constraints.

Key Technical Components:

• Geometric Constraints: The reconstruction is driven by three integral conditions derived from the requirement that the angular scales remain constant:
  1. $\delta r_s^*/r_s^* \approx -7\%$
  2. $\delta r_d^*/r_d^* \approx -7\%$ (to preserve $\theta_s^*/\theta_d^*$)
  3. $\delta r_{eq}/r_{eq} \approx -7\%$ (to preserve $\theta_s^*/\theta_{eq}$)
• Kernel Formalism: The paper utilizes sensitivity kernels ($K_s(z)$ and $K_d(z)$) which describe how changes in the expansion rate $H(z)$ at a given redshift affect the sound horizon and damping scales.
  • $K_s(z)$ has a long tail extending deep into the radiation-dominated era.
  • $K_d(z)$ has narrow support, vanishing at high redshifts ($z \gtrsim 4000$).
• Reconstruction Algorithm:
  • The function $\delta H/H_{\text{fid}}(z)$ is parameterized using 20 control nodes (knots) distributed across $z \in (1090, 20000)$.
  • A cubic spline interpolates the knot amplitudes.
  • A Markov Chain Monte Carlo (MCMC) algorithm samples the likelihood space defined by Gaussian constraints on the three scales and a penalty term ($L_{\text{penalty}}$) to suppress unphysical high-frequency oscillations.
• Boundary Conditions: The reconstruction enforces that the matter-radiation equality redshift ($z_{eq}$) remains stable ($\delta z_{eq}/z_{eq} \approx 0$) and that the expansion history converges to standard $\Lambda$CDM at very high redshifts ($z > 20000$).

3. Key Contributions

• First Model-Independent Reconstruction: This is the first study to reconstruct the pre-recombination expansion history solely based on CMB geometric constraints and the Hubble tension, without assuming a specific physical Lagrangian.
• Identification of a "No-Go" Geometric Profile: The study demonstrates that purely early-time solutions are not arbitrary; they are mathematically forced into a specific shape by the interplay of the damping and sound horizon kernels.
• Quantification of the "15% Rule": The analysis reveals a precise requirement for any successful early-universe mechanism: a smooth transition around matter-radiation equality ($z_{eq}$) that peaks with a $\sim 15\%$ enhancement in the expansion rate immediately before recombination ($z \approx 1090$).

4. Results

• The Reconstructed Profile: The posterior distribution for $\delta H/H_{\text{fid}}$ shows a pronounced, smooth transition starting near $z_{eq}$ and reaching a maximum of $\approx 15\%$ just before recombination. At higher redshifts ($z > 4000$), the deviation decays smoothly to zero.
• Analytical Fit: The reconstructed mean function is well-approximated by a logistic function:
  $$\frac{\delta H}{H_{\text{fid}}}(z) \approx \frac{0.15}{1 + e^{(z-z_{eq})/1000}}$$
• Kernel Interplay:
  • The Silk damping kernel ($K_d$) forces the amplitude to be high ($\sim 15\%$) because it only "sees" the expansion history in the narrow window just before recombination. A lower amplitude would fail to reduce $r_d^*$ sufficiently.
  • The Sound horizon kernel ($K_s$) prevents the amplitude from being a simple step function. Because $K_s$ integrates contributions from high redshifts, a sharp drop immediately after the peak would over-reduce $r_s^*$. Thus, the function must decay slowly, creating the specific "tug-of-war" shape.
• Implications for Hybrid Models: The reconstruction implies that a purely pre-recombination solution is likely insufficient. If $H(z)$ is enhanced by 15% at recombination but returns to standard $\Lambda$CDM afterwards, the resulting $H_0$ would be over-enhanced. The paper suggests a second transition (a "dark ages" transition) is likely required to lower $H(z)$ back to the standard value before the late-time universe, though this introduces perturbation-level challenges (e.g., mid-ISW effects).

5. Significance and Conclusion

• Blueprint for Model Building: The paper provides a "stress test" for any proposed solution to the Hubble tension. Any viable early-universe mechanism (e.g., Early Dark Energy, varying constants) must mimic this specific smooth transition profile with a $\sim 15\%$ peak amplitude. Models that inject energy too early (high $z$) or fail to produce this specific shape are geometrically disfavored.
• Geometric Rigidity: The results highlight that the CMB geometric constraints are extremely rigid. The specific shape of the required expansion history is a mathematical necessity derived from the kernels, not an artifact of a specific model amplitude.
• Future Directions: The author concludes that while the background-level solution exists, the perturbation-level compatibility (specifically the impact on the CMB power spectrum via the Integrated Sachs-Wolfe effect) remains the primary hurdle. The paper calls for future work integrating this reconstructed profile into full Boltzmann codes (CAMB/CLASS) to test if such a drastic transition can survive the full CMB spectrum constraints.

In summary, Pedrotti demonstrates that the Hubble tension, if resolved by early-time physics, demands a very specific, high-amplitude modification to the expansion history immediately preceding recombination, effectively ruling out broad classes of "soft" early-universe modifications.