SPT-3G D1: Axion Early Dark Energy with CMB experiments and DESI

This paper presents updated constraints on axion early dark energy (AEDE) using CMB data from SPT, ACT, and Planck alongside DESI BAO measurements, revealing that while CMB data alone shows no significant evidence for AEDE, the inclusion of DESI data induces a mild preference for the model and significantly reduces the Hubble tension, a shift attributed to existing discrepancies between DESI and CMB datasets in the standard $\Lambda$CDM model.

The Gist

Imagine the universe as a giant, expanding balloon. For decades, scientists have been trying to measure exactly how fast this balloon is inflating today. This speed is called the Hubble Constant ($H_0$).

The problem is that we have two different ways of measuring this speed, and they disagree.

1. The "Local" Method: Astronomers look at nearby stars and supernovas (like checking the speedometer on a car right next to you). This method says the universe is expanding fast: about 73 units.
2. The "Ancient" Method: Physicists look at the Cosmic Microwave Background (CMB), which is the "baby photo" of the universe taken 13.8 billion years ago. By analyzing this ancient light, they calculate how fast the universe should be expanding today. This method says the speed is slower: about 67 units.

This disagreement is known as the Hubble Tension. It's like if your car's speedometer said 70 mph, but your GPS (based on the road map) said 60 mph, and you couldn't figure out who was wrong.

The Proposed Fix: Axion Early Dark Energy (AEDE)

To fix this, scientists proposed a new theory called Axion Early Dark Energy (AEDE).

Think of the early universe as a race car. In the standard model ($\Lambda$CDM), the car runs on a steady fuel mix. But the AEDE theory suggests that for a very brief moment just before the "baby photo" was taken, the car had a nitrous oxide boost.

• This "boost" (the axion field) made the universe expand slightly faster in its early days.
• This extra speed changes the "baby photo" in a way that allows the ancient calculation to match the faster, modern speed of 73.
• The "nitrous" then faded away, leaving the universe looking mostly normal today, but with a higher final speed.

What This Paper Did

The authors of this paper acted like detectives testing this "nitrous oxide" theory. They gathered the latest, most precise data from three major cosmic observatories:

• SPT-3G: A telescope at the South Pole.
• ACT: A telescope in the Atacama Desert in Chile.
• Planck: A space telescope that took the original "baby photo."
• DESI: A project mapping the positions of millions of galaxies to measure the universe's structure.

They asked: "Does adding this 'nitrous oxide' (AEDE) to our model actually fix the speedometer disagreement?"

The Findings

1. Looking at the "Baby Photo" Alone (CMB Data Only)

When the team looked only at the ancient light (the CMB data from SPT, ACT, and Planck), the answer was no.

• The data didn't show a strong need for the "nitrous oxide."
• The "speedometer" (Hubble constant) calculated from the ancient light only moved from 67 to about 68.
• This is still far from the modern measurement of 73. The tension between the two methods dropped slightly (from a 6.4-sigma disagreement to a 3.6-sigma disagreement), but it's still a significant gap.
• Verdict: The ancient data alone doesn't prove the "nitrous oxide" exists.

2. Adding the "Galaxy Map" (DESI Data)

Then, the team added data from DESI, which maps the current structure of the universe (like a detailed map of the road the car is driving on).

• The Shift: When they combined the ancient light with the galaxy map, the "nitrous oxide" theory suddenly looked a bit more promising. The data started to slightly prefer the idea that the boost happened.
• The Result: The calculated speed of the universe moved up to about 69.8.
• The Tension: The disagreement between the ancient and modern methods dropped significantly, from a 6.4-sigma gap down to 2.6 sigma. This is much better, but it's not a perfect match yet.

The Catch: Is it Real?

Even though the numbers improved when they added the galaxy map, the paper concludes that we still don't have enough proof to say AEDE is the solution.

• The improvement in the fit wasn't "statistically significant" enough to rule out the standard model (the car without nitrous).
• The authors point out a crucial twist: The reason the numbers shifted when they added the DESI data might not be because of "nitrous oxide" at all. It might just be because the ancient light data and the galaxy map data don't perfectly agree with each other in the standard model.
• Think of it this way: If you try to fix a speedometer by changing the road map, and the speedometer finally matches the car, it might mean the road map was wrong, not that the engine has nitrous.

The Bottom Line

This paper is a rigorous check-up on a popular theory.

• Did it solve the Hubble Tension? Not completely. The gap is smaller, but still there.
• Is AEDE the winner? Not yet. The data is "mildly" in favor of it when combining all sources, but not strongly enough to declare it the new standard.
• What's next? The authors suggest that as we get even better data from future telescopes, we will finally know if this "nitrous oxide" is real physics or just a glitch in our measurements.

In short: The "nitrous oxide" theory is a good candidate, but the evidence is currently just a whisper, not a shout.

Technical Summary

Technical Summary: SPT-3G D1: Axion Early Dark Energy with CMB experiments and DESI

Problem Statement
The paper addresses the persistent "Hubble tension," a significant discrepancy between the Hubble constant ($H_0$) inferred from the Cosmic Microwave Background (CMB) under the standard $\Lambda$CDM model and the value measured via the classical distance ladder (SH0ES collaboration). Within $\Lambda$CDM, the combination of SPT-3G D1, ACT-DR6, and Planck data yields $H_0 = 67.19 \pm 0.38$ km/s/Mpc, which is in $6.4\sigma$ tension with the SH0ES result of $73.17 \pm 0.86$ km/s/Mpc. The authors investigate the Axion Early Dark Energy (AEDE) model as a potential solution. In this model, an axion field contributes a fraction of the energy budget prior to recombination, increasing the Hubble parameter at that epoch and reducing the sound horizon, thereby allowing for a higher inferred $H_0$. The study aims to update constraints on AEDE using the latest datasets: SPT-3G D1, ACT-DR6, and DESI-DR2, assessing whether the model can statistically resolve the tension.

Methodology
The authors employ a Bayesian and frequentist statistical framework to constrain the AEDE model parameters using Markov Chain Monte Carlo (MCMC) analysis.

• Model: The AEDE potential is defined as $V(\phi) = m^2 f^2 [1 - \cos(\theta)]^n$ with $n=3$. The model introduces three additional parameters relative to $\Lambda$CDM: the axion decay constant fraction $f_{\text{EDE}}$, the critical redshift $z_c$ (where the field becomes dynamical), and the initial field value $\theta_i$.
• Datasets: The analysis utilizes various combinations of:
  • CMB: SPT-3G D1 (2019–2020 main field), ACT-DR6, and Planck 2018 (PR3/PR4). Lensing information is included where available.
  • BAO: Dark Energy Spectroscopic Instrument (DESI) Data Release 2 (DR2).
  • Priors: Uniform priors are applied to model parameters. The reionization optical depth ($\tau_{\text{reio}}$) is constrained by a Gaussian prior from [70] rather than low-$\ell$ Planck EE data to avoid degeneracies.
• Assessment Metrics: To evaluate the viability of AEDE as a solution to the Hubble tension, three specific metrics are used:
  1. Marginalized Posterior Compatibility Level (QMPCL): Quantifies the deviation of the inferred $H_0$ from the SH0ES value. A passing score is $Q_{\text{MPCL}} \leq 3\sigma$.
  2. Difference of the Maximum A Posteriori (QDMAP): A frequentist measure of the improvement in $\chi^2$ when SH0ES data is added. A passing score is $Q_{\text{DMAP}} \leq 3\sigma$.
  3. Akaike Information Criterion ($\Delta$AIC): Compares the goodness of fit of AEDE against $\Lambda$CDM, penalizing for extra parameters ($N=3$). A passing score requires $\Delta$AIC $\leq -6.91$ (indicating more than a "weak preference").

Key Results

• CMB Data Alone:

  • Using CMB data only (SPT-3G D1, SPT+ACT, or the combined CMB-SPA baseline), the authors find no statistically significant evidence for AEDE.
  • The upper limits on the fractional energy density are $f_{\text{EDE}} < 0.12$ (95% CL) for SPT-3G D1 alone and $f_{\text{EDE}} < 0.070$ for the combined CMB-SPA dataset.
  • The Hubble tension is reduced from $6.4\sigma$ in $\Lambda$CDM to $3.6\sigma$ in AEDE with CMB-SPA, but this reduction is insufficient to resolve the tension.
  • The model fails all three assessment metrics: $Q_{\text{MPCL}} = 3.6\sigma$ (fails), $Q_{\text{DMAP}} = 3.0\sigma$ (fails), and $\Delta$AIC $= 0.91$ (fails, indicating no preference over $\Lambda$CDM).

• CMB + DESI BAO Data:

  • The inclusion of DESI-DR2 BAO data shifts the parameter constraints. Unlike the CMB-only case, there is a mild preference for non-zero $f_{\text{EDE}}$.
  • For the CMB-SPA + DESI combination, the constraint is $f_{\text{EDE}} = 0.055^{+0.024}_{-0.047}$ (68% CL), representing a $1.2\sigma$ deviation from zero.
  • The inferred $H_0$ increases to $69.81^{+0.95}_{-1.2}$ km/s/Mpc, reducing the tension with SH0ES to $2.6\sigma$.
  • Metric Performance:
    • $Q_{\text{MPCL}}$ improves to $2.6\sigma$ (passes the $3\sigma$ threshold).
    • $Q_{\text{DMAP}}$ improves to $2.4\sigma$ (passes the $3\sigma$ threshold).
    • $\Delta$AIC is $-2.5$. While this indicates a better fit than $\Lambda$CDM, it does not meet the strict threshold of $-6.91$ required for a "weak preference" on the Jeffreys scale.

Significance and Claims
The paper concludes that while AEDE is not a definitive solution to the Hubble tension based on CMB data alone, the addition of DESI BAO data creates a scenario where the model is statistically favored over $\Lambda$CDM in terms of fit quality ($Q_{\text{DMAP}}$) and tension reduction ($Q_{\text{MPCL}}$). However, the authors emphasize that this preference is weak according to information criteria ($\Delta$AIC).

Crucially, the authors interpret the shift in parameters when adding DESI data not necessarily as proof of new physics, but as a manifestation of the existing discrepancy between DESI and CMB experiments within the concordance $\Lambda$CDM model. The AEDE model effectively bridges this gap by allowing non-zero early dark energy, but the authors caution that future data (e.g., from further SPT releases, the Simons Observatory, or updated DESI data) is required to determine if this discrepancy points to new physics or systematic errors. The work serves as an update to previous constraints, highlighting that current data does not yet provide strong, independent evidence for AEDE over the standard model.