🔭 astrophysics

Updated Constraints on Infrared Cutoff Models and Implications for Large-Scale CMB Anomalies

This study revisits infrared cutoff models to address large-scale CMB anomalies and potential reionization optical depth tensions, finding that while these models suppress low-multipole power, they offer no statistically significant improvement over the standard power-law spectrum and fail to resolve current cosmological tensions.

Original authors: Ujjwal Upadhyay, Yashi Tiwari, Tarun Souradeep

Published 2026-02-19

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ujjwal Upadhyay, Yashi Tiwari, Tarun Souradeep

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the universe as a giant, cosmic drum. When it was born, it didn't just sit there; it vibrated. These vibrations created the seeds for everything we see today: stars, galaxies, and the cosmic microwave background (CMB), which is essentially the "afterglow" or the echo of the Big Bang.

Scientists have a standard recipe for how these vibrations should sound. They call it the Power-Law Model. Think of this like a perfectly tuned piano string: it produces a smooth, consistent sound across all notes, from the deepest bass to the highest treble. For decades, this recipe has worked incredibly well to explain the universe.

However, when scientists listen to the very lowest, deepest notes of the cosmic drum (the largest scales of the universe), something feels a bit off. The sound is quieter than the recipe predicts. It's like the bass note on the piano is slightly muffled. This is known as the "low quadrupole anomaly."

This paper is like a group of detectives (the authors) revisiting the case to see if we need a new recipe to explain why the bass is so quiet.

The New Recipes (IR Cutoff Models)

The authors tested several "new recipes" called Infrared (IR) Cutoff Models.

The Analogy: Imagine the standard recipe says the drum should vibrate at every size. The new recipes suggest that maybe the drum was too small to vibrate at the very largest sizes. It's like saying, "The drum just doesn't have the physical space to make that super-low sound."
The Models: They tried different mathematical ways to describe this "missing bass," such as an exponential fade-out (the sound just drops off) or a sudden break (like a drumhead with a tear).

The Investigation

The authors took the latest, most precise recordings of the universe (from satellites like Planck, and telescopes like ACT and SPT) and tried to fit these new recipes against the data. They also checked if these new recipes could solve a different mystery: a recent disagreement between two different ways of measuring the universe's expansion (the "CMB-BAO tension").

Here is what they found, broken down simply:

1. The "Quiet Bass" Mystery

The Hope: Maybe the new recipes explain why the bass is quiet.
The Reality: The new recipes do predict a quieter bass. In fact, they fit the low notes slightly better than the standard recipe.
The Catch: To get this better fit, the new recipes require adding extra knobs and dials (extra parameters) to the theory. In science, if you have to add more complexity to get a slightly better result, it's often not worth it.

The Verdict: The improvement is so small that it doesn't justify the extra complexity. The standard "Power-Law" recipe is still the winner. The universe's bass is quiet, but the standard recipe is still the best explanation we have, even with that quiet note.

2. The "Expansion Tension" Mystery

The Hope: Recently, two different ways of measuring how fast the universe is expanding gave conflicting answers. Some scientists thought maybe the "missing bass" (the IR cutoff) could trick our measurements and fix this disagreement.
The Reality: The authors checked this carefully. They found that the "missing bass" is so tightly constrained by the rest of the data (the higher notes) that it can't wiggle enough to fix the expansion problem.

The Verdict: These new recipes cannot save the standard model from the expansion tension. The disagreement remains, and these specific models don't help solve it.

3. The "Parity" Mystery

The Hope: There's another weird thing: the universe seems to have a preference for "odd" notes over "even" notes in the low range.
The Reality: The new recipes actually make this odd/even imbalance worse, not better. While they fix the volume of the bass, they mess up the rhythm.

The Verdict: The standard recipe is still the most balanced choice.

The Big Picture: Why is this hard?

The authors explain that studying the "largest scales" of the universe is like trying to hear a whisper in a hurricane. Because there is only one universe, we only have one sample of these giant waves. This creates a lot of "cosmic noise" (called cosmic variance). It's hard to tell if the bass is quiet because of a specific rule of physics, or just because we got unlucky with the one drum we have.

The Conclusion

The paper concludes that while it's fun to imagine the universe had a "cutoff" where the biggest waves couldn't exist, the current data doesn't give us enough evidence to say that's true. The standard model, with its simple, smooth power law, still holds up best.

In a nutshell:
The universe's "bass note" is quieter than expected. The authors tried several fancy new theories to explain this, but the data says: "Thanks, but no thanks. The old theory is still good enough, and these new theories don't fix the other problems we have."

The authors suggest that future, more sensitive telescopes (like LiteBIRD) might one day hear these deep notes clearly enough to finally tell us if the universe really does have a "cutoff" or if it's just a cosmic coincidence.

1. Problem Statement

The standard $\Lambda$ CDM cosmological model assumes a nearly scale-invariant, power-law primordial power spectrum (PPS) for curvature perturbations. While this fits most Cosmic Microwave Background (CMB) data, the largest angular scales (low multipoles, $\ell < 30$ ) remain weakly constrained due to cosmic variance. This has led to persistent "large-scale anomalies," most notably the low quadrupole power and parity asymmetry.

These anomalies have motivated phenomenological models proposing an Infrared (IR) cutoff in the PPS, where power is suppressed on the largest scales. Additionally, recent tensions between CMB data and Baryon Acoustic Oscillation (BAO) measurements (specifically from DESI DR2) within the $\Lambda$ CDM framework have suggested that a higher reionization optical depth ( $\tau_{\text{reio}}$ ) might alleviate the discrepancy. Since IR cutoffs suppress large-scale power, they could theoretically bias the inference of $\tau_{\text{reio}}$ (which is constrained by large-scale polarization), potentially resolving the CMB-BAO tension.

The paper aims to:

Re-evaluate the viability of various IR cutoff models using the latest high-precision CMB data (Planck, ACT, SPT).
Determine if these models can resolve the CMB-BAO tension by allowing for a higher $\tau_{\text{reio}}$ .
Assess the statistical significance of IR cutoffs in explaining CMB anomalies using a Bayesian framework.

2. Methodology

Models Considered

The authors analyze five classes of primordial power spectrum models, parametrized by functional forms:

Power Law (PL): The standard baseline ( $P_R(k) \propto k^{n_s-1}$ ).
Exponential Cutoff (EC): $P_R(k) = P_0(k)[1 - e^{-(k/k_c)^\alpha}]$ .
Starobinsky Break (SB): Features a smooth suppression followed by damped oscillations due to a sharp change in the inflaton potential derivative.
Exponential–Starobinsky (ESB): Combines an exponential cutoff with the Starobinsky break.
Pre-inflationary Radiation Domination (RD): Assumes a radiation-dominated phase before inflation, leading to suppression and oscillatory features.

Datasets

The analysis utilizes a comprehensive combination of datasets to constrain parameters:

CMB: Planck PR4 (low- $\ell$ TT/EE, high- $\ell$ Plik), Planck PR4 lensing, ACT DR6, and SPT-3G D1.
BAO: DESI DR2 measurements.
Supernovae: Pantheon+ (Type Ia SNIa).

Statistical Framework

Parameter Estimation: Markov Chain Monte Carlo (MCMC) sampling using Cobaya and the CLASS Boltzmann solver.
Model Comparison: The Akaike Information Criterion (AIC) is used to penalize models for increased complexity (extra parameters like $k_c$ , $\alpha$ , $R_*$ ).
Anomaly Analysis: A unique Bayesian approach is adopted to analyze anomaly statistics ( $C_2$ for quadrupole and $P$ for parity asymmetry). Instead of just comparing model predictions to observed values, the authors derive the posterior distribution of the "true theory" (given the data and assumptions of statistical isotropy) and compare it against the model-predicted posteriors.

3. Key Contributions

Updated Constraints with High-Resolution Data: Unlike previous studies relying heavily on Planck alone, this work incorporates high- $\ell$ data from ACT and SPT. This is crucial because many IR cutoff models predict oscillatory features at smaller scales that could be detected by these experiments.
DESI DR2 Context: The paper is one of the first to rigorously test IR cutoff models against the specific CMB-BAO tension highlighted by the recent DESI DR2 release.
Bayesian Anomaly Statistics: The authors move beyond frequentist "p-values" for anomalies, using a Bayesian framework to compare the posterior distributions of anomaly statistics derived from models against the "true theory" posterior derived directly from observations.

4. Key Results

A. Viability of IR Cutoff Models

Marginal Improvement: While IR cutoff models (specifically EC, SB, and ESB) show a modest improvement in the fit to low- $\ell$ temperature data ( $\chi^2_{\text{low-}\ell}$ ), this does not translate to a significant reduction in the total $\chi^2$ .
AIC Penalty: When accounting for model complexity via the Akaike Information Criterion, no IR cutoff model is statistically preferred over the standard power-law $\Lambda$ CDM model. The SB and RD models are statistically comparable to the baseline, while EC and ESB are mildly disfavoured.
Tight Constraints on Cutoff Scale ( $k_c$ ): High-multipole data (ACT/SPT) tightly constrain the cutoff scale $k_c$ to be very small ( $k_c \lesssim k_h$ ). This forces the suppression to occur at scales so large that they barely affect the observable CMB spectrum, effectively rendering the models indistinguishable from the standard power law in most regimes.

B. CMB-BAO Tension and $\tau_{\text{reio}}$

No Resolution of Tension: The authors find that IR cutoff models cannot resolve the CMB-BAO tension.
Mechanism Failure: The hypothesis that large-scale suppression would bias $\tau_{\text{reio}}$ to higher values (alleviating tension) fails because the cutoff features are too tightly constrained by CMB temperature anisotropies. The suppression is limited to $\ell \sim 2$ , leaving the large-scale polarization spectrum (which constrains $\tau_{\text{reio}}$ ) largely unaffected.
Exception (RD Model): The Pre-inflationary Radiation Domination (RD) model shows a slightly better overlap with DESI BAO constraints in the $\Omega_m$ – $H_0$ plane. However, the authors attribute this not to the IR cutoff itself, but to asymmetric error bars in the cosmological parameters caused by high-frequency oscillations in the RD power spectrum, which distort the posterior distribution.

C. CMB Anomalies (Quadrupole and Parity)

Quadrupole ( $C_2$ ): IR cutoff models predict a lower quadrupole amplitude, shifting the mean closer to the observed Planck value. However, within the Bayesian "true theory" framework (which accounts for cosmic variance), the standard power-law model is already consistent with the observation. The improvement offered by IR models is not statistically significant.
Parity Asymmetry ( $P$ ): Interestingly, while IR models suppress the quadrupole, they tend to predict $P > 1$ (excess even power), moving away from the observed Planck value ( $P < 1$ ). The oscillatory features in models like SB exacerbate this deviation.
Conclusion on Anomalies: The presence of large-scale anomalies does not constitute a significant discrepancy for the standard model when cosmic variance is properly accounted for. IR cutoff models do not provide a statistically robust solution.

5. Significance and Conclusion

This paper provides a definitive update on the status of IR cutoff models in the era of high-precision CMB data (Planck, ACT, SPT) and modern BAO surveys (DESI).

Robustness of $\Lambda$ CDM: The standard power-law spectrum remains the most robust description of the primordial universe. The "anomalies" at large scales are consistent with statistical fluctuations (cosmic variance) rather than evidence for new physics like IR cutoffs.
Limits of Phenomenology: The study demonstrates that adding phenomenological parameters to fix large-scale anomalies is penalized by model selection criteria (AIC) and constrained by small-scale data.
Future Outlook: The authors conclude that current data is insufficient to distinguish IR cutoff signatures from the standard model due to cosmic variance limits. Future experiments with improved large-scale polarization sensitivity (e.g., LiteBIRD, CMB-Bharat) will be required to further probe these scales, though cosmic variance will likely remain the fundamental limit for temperature anisotropies.

In summary, the paper argues that IR cutoff models are not favored by current data, do not resolve the CMB-BAO tension, and do not offer a statistically significant explanation for large-scale CMB anomalies over the standard $\Lambda$ CDM paradigm.