Statistical consistency of sign-switching vacuum energy with cosmological observations

Using exact non-Gaussian consistency diagnostics on CMB, BAO, and supernova data, this study finds that while Gaussian metrics may overstate tensions, both the standard $\Lambda$CDM and its sign-switching extension $\Lambda_{\rm s}$CDM show excellent consistency between observations, with the latter offering modest geometric improvements at intermediate redshifts.

The Gist

Here is an explanation of the paper, translated into everyday language with some creative analogies.

The Big Picture: A Cosmic Dispute

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

The problem is that we have two very different ways of measuring this speed, and they don't agree:

1. The "Baby Photo" Method (Early Universe): We look at the Cosmic Microwave Background (CMB), which is like a baby photo of the universe taken 13.8 billion years ago. Based on this photo, the universe should be expanding at a certain speed (let's call it Speed A).
2. The "Adult Photo" Method (Late Universe): We look at supernovas (exploding stars) and galaxies nearby, which are like photos of the universe today. These measurements suggest the universe is expanding faster than the baby photo predicts (let's call it Speed B).

This disagreement is called the "Hubble Tension." It's like looking at a child's height chart and predicting they will be 5 feet tall as an adult, but when they grow up, they are actually 6 feet tall. Something is wrong with our prediction model.

The Proposed Fix: The "Sign-Switching" Vacuum

The standard model of cosmology (called $\Lambda$CDM) assumes that "Dark Energy" (the force pushing the universe apart) is constant, like a steady wind blowing on the balloon.

The authors of this paper are testing a new idea called $\Lambda_s$CDM. Imagine the "wind" of Dark Energy isn't constant. Instead, imagine it was blowing in the opposite direction (pulling the balloon in) in the early universe, and then suddenly flipped to blowing outward (pushing the balloon out) at a specific moment in time.

This is the "Sign-Switching" idea. The authors wanted to see if this sudden flip could fix the disagreement between the "Baby Photo" and the "Adult Photo."

The Investigation: How They Checked the Math

The authors didn't just look at the numbers; they looked at how they were looking at the numbers. This is the most important part of their paper.

The Analogy: The Ruler vs. The Map
Imagine you are trying to measure the distance between two cities.

• Method 1 (Gaussian Metrics): You use a standard ruler. It works great if the road is straight. But if the road is winding, bumpy, or shaped like a weird loop, a straight ruler gives you a misleading answer. It might tell you the cities are "very far apart" just because the road is curvy, even if they are actually close.
• Method 2 (Exact Non-Gaussian Metrics): You use a GPS map that accounts for every curve, hill, and detour. It gives you the true distance, even if the road is weird.

The authors found that most previous studies used the "Standard Ruler" (Gaussian statistics). When they combined very precise data (like the CMB) with data that is a bit "fuzzy" or spread out (like the supernova data), the Standard Ruler screamed, "Huge Error! These datasets don't match!"

However, when the authors used the "GPS Map" (Exact Non-Gaussian statistics), they found that the "Baby Photo" (CMB) and the "Intermediate Photo" (Galaxy surveys) actually match perfectly. The "Ruler" was just being fooled by the shape of the data.

The Results: Did the New Model Work?

1. The Good News:
The new "Sign-Switching" model ($\Lambda_s$CDM) does help a little bit. It makes the geometry of the universe fit the data slightly better at intermediate times. It's like adjusting the wind speed on the balloon so the growth curve looks a bit more realistic.

2. The Bad News:
Even with this new model, the "Adult Photo" (local supernova measurements) still doesn't fully match the "Baby Photo."

• The new model shifts the predicted speed up a little bit, getting closer to the observed speed.
• But, the observed speed is still an "outlier." It's like the model predicts the adult is 5'8", and the measurement says 6'0". It's closer than before, but they still don't agree.

3. The "Posterior Predictive" Test (The Crystal Ball):
The authors also ran a "Crystal Ball" test. They asked: "If our model is true, what kind of universe should we see today?"

• Standard Model: The crystal ball says, "You should see a slow expansion." But we see a fast one. The model fails.
• Sign-Switching Model: The crystal ball says, "You should see a slightly faster expansion." This is closer to reality, but the fast expansion we actually see is still too extreme for the model to comfortably explain.

The Conclusion: What Does It All Mean?

The paper teaches us two main lessons:

1. Don't trust the "Ruler" blindly: When dealing with complex, weird-shaped data in cosmology, simple statistical tools can make things look much more broken than they really are. We need more sophisticated tools (the "GPS maps") to tell the truth.
2. The Mystery Remains: The "Sign-Switching" vacuum energy is a clever idea that improves the situation slightly, but it does not solve the Hubble Tension. The universe is still expanding faster than our best theories predict, even with this new twist.

In a nutshell: The authors fixed the measuring tape so we can see the problem more clearly. They found that the new "Sign-Switching" theory helps a little, but the universe is still acting stranger than our current theories can fully explain. We need a bigger breakthrough to solve the mystery of the expanding balloon.

Technical Summary

Here is a detailed technical summary of the paper "Statistical consistency of sign-switching vacuum energy with cosmological observations" by Khandelwal, Capistrano, and Kumar.

1. Problem Statement

The paper addresses the persistent Hubble tension, a $\sim 5-7\sigma$ discrepancy between early-universe estimates of the Hubble constant ($H_0$) derived from the Cosmic Microwave Background (CMB) and late-universe measurements from Type Ia supernovae (SNe Ia) and Baryon Acoustic Oscillations (BAO).

While various theoretical extensions to the standard $\Lambda$CDM model have been proposed to resolve this, including the $\Lambda_s$CDM model (where the cosmological constant undergoes an abrupt sign switch from negative to positive at a transition redshift $z^\dagger$), previous assessments of these models have relied heavily on Gaussian tension metrics. The authors argue that these standard metrics can be misleading when comparing datasets with vastly different constraining powers and non-Gaussian posterior distributions, potentially overstating inconsistencies or mischaracterizing the nature of the tension.

2. Methodology

The authors employ a rigorous, multi-faceted statistical framework to assess the internal consistency of the standard $\Lambda$CDM model and the sign-switching $\Lambda_s$CDM model against current cosmological data.

Datasets Used:

• CMB: Planck 2018 (TT, TE, EE, lensing), ACT DR6, and SPT-3G.
• BAO: DESI Data Release 2 (DR2) covering $0.295 \le z \le 2.330$.
• Low-z Supernovae: PantheonPlus+SH0ES (PPS), including Cepheid calibration.

Models:

• $\Lambda$CDM: Standard six-parameter model with a constant cosmological constant ($w=-1$).
• $\Lambda_s$CDM: An extension introducing a single parameter $z^\dagger$ where the cosmological constant $\Lambda$ switches sign from negative (AdS-like) to positive (dS-like).

Statistical Diagnostics:
To avoid biases inherent in Gaussian approximations, the study utilizes a suite of diagnostics:

1. Gaussian Metrics: Parameter Difference (DM), Updated Difference in Mean (UDM), and Maximum A Posteriori Difference (DMAP). These rely on mean shifts and covariance matrices.
2. Exact Non-Gaussian Parameter Shift: A metric that calculates the probability density of the parameter shift ($\Delta\theta$) without assuming Gaussianity, accounting for the full shape of the posterior distributions.
3. Posterior Predictive Consistency (PPC): A Bayesian model-checking technique. The authors generate replicated datasets ($y_{rep}$) from the posterior predictive distribution (conditioned on CMB+DESI) and compare them to observed low-redshift data ($H_0$ and distance modulus $\mu$). This tests if the model can predict the observed late-time expansion.

Implementation:

• MCMC simulations were performed using MontePython interfaced with the CLASS Boltzmann solver (modified to implement the abrupt sign-switching behavior).
• Convergence was verified using the Gelman-Rubin criterion ($R-1 < 0.01$).
• The Tensiometer framework was used to compute tension metrics directly from MCMC chains.

3. Key Contributions

• Critique of Gaussian Metrics: The paper demonstrates that standard Gaussian tension metrics (like DM and UDM) can significantly overstate inconsistencies when combining tightly constrained datasets (CMB/DESI) with broad, non-Gaussian datasets (PPS).
• Validation of Exact Non-Gaussian Shifts: The authors show that the exact non-Gaussian parameter shift provides a more robust assessment, revealing that CMB and BAO data are actually highly consistent in both models, contrary to what some Gaussian metrics might suggest when PPS is involved.
• Predictive Adequacy Testing: By applying PPC tests specifically to $H_0$ and $\mu$, the study distinguishes between "parameter space incompatibility" and "predictive failure," showing that $\Lambda_s$CDM improves marginal predictions but fails to resolve joint discrepancies.

4. Key Results

A. CMB vs. DESI DR2 (Intermediate Redshifts):

• $\Lambda$CDM: Shows moderate tension ($\sim 2.3\sigma$) using Gaussian metrics.
• $\Lambda_s$CDM: Shows excellent consistency ($\sim 0.06\sigma$). The sign-switching extension improves geometric compatibility at intermediate redshifts, allowing the models to better align CMB and BAO constraints without altering early-universe physics.
• Non-Gaussian Insight: The exact non-Gaussian shift confirms substantial posterior overlap between CMB and DESI in both models, validating their mutual consistency.

B. CMB+DESI vs. PantheonPlus+SH0ES (Low Redshifts):

• Gaussian Metrics: Indicate extreme tension ($\sim 5.5\sigma$ for $\Lambda$CDM and $\sim 7.7\sigma$ for $\Lambda_s$CDM). The authors attribute this largely to the asymmetry caused by the broad, non-Gaussian posterior of the PPS dataset compressing the covariance of the combined dataset.
• Exact Non-Gaussian Shift: Confirms that the posteriors for CMB+DESI and PPS occupy disjoint regions of parameter space. The tension is genuine, not an artifact of Gaussian approximation.
• PPC Results:
  • $\Lambda$CDM: The observed local $H_0$ ($73.04 \pm 1.04$km/s/Mpc) lies deep in the tail of the predictive distribution derived from CMB+DESI ($p_{ppc} \approx 5 \times 10^{-5}$).
  • $\Lambda_s$CDM: The model shifts the predictive distribution toward higher $H_0$ values, increasing the $p_{ppc}$ to $\approx 4 \times 10^{-4}$. This represents a partial alleviation of the tension.
  • Joint Discrepancy: Crucially, while the marginal prediction for $H_0$ improves, the joint discrepancy (considering $H_0$ and distance modulus $\mu$ together) remains statistically significant. The model does not fully reconcile the local expansion rate with the CMB/BAO calibrated distance ladder.

5. Significance and Conclusion

The paper concludes that while the $\Lambda_s$CDM model offers a statistically improved description of the data compared to standard $\Lambda$CDM (particularly in reducing geometric tension between CMB and BAO), it does not fully resolve the Hubble tension.

• Methodological Impact: The study establishes that robust cosmological model assessment requires exact, non-Gaussian, and predictive diagnostics. Relying solely on Gaussian parameter shifts can lead to erroneous conclusions about dataset consistency, especially when dealing with the complex, non-Gaussian posteriors typical of modern cosmological data.
• Physical Implication: The sign-switching mechanism successfully modifies the late-time expansion history to shift $H_0$ predictions upward, but it is insufficient to decouple the CMB/BAO calibration from the local distance ladder measurements to the degree required to eliminate the tension.
• Future Work: The authors suggest that future extensions should focus on dynamical dark energy parameterizations and further explore the interplay between posterior geometry and dataset dominance to achieve full consistency.

In summary, the paper provides a rigorous statistical benchmark, demonstrating that $\Lambda_s$CDM softens but does not solve the Hubble tension, and emphasizes the necessity of moving beyond Gaussian approximations in cosmological consistency checks.