🔭 astrophysics

First constraints on causal sources of primordial gravitational waves from BICEP/Keck, SPTpol, SPT-3G, Planck and WMAP $B$ -mode data

Using a combination of CMB $B$ -mode polarization data from BICEP/Keck, SPTpol, SPT-3G, Planck, and WMAP, this study reports the first constraints on non-inflationary, causality-limited primordial gravitational wave sources, establishing a 95% confidence upper limit of $r_{ect} < 0.0077$ that robustly restricts various early-universe models and sets a new bound on ultra-low-frequency gravitational wave energy density.

Original authors: Jessica A. Zebrowski, Aurora Ireland, Christian L. Reichardt, Kylar Greene, Gordan Krnjaic, Yuhsin Tsai, François R. Bouchet

Published 2026-01-30

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Jessica A. Zebrowski, Aurora Ireland, Christian L. Reichardt, Kylar Greene, Gordan Krnjaic, Yuhsin Tsai, François R. Bouchet

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, expanding balloon. For decades, scientists have been trying to listen to the faint "hum" of the very first moments after the Big Bang. This hum is made of gravitational waves—ripples in the fabric of space-time itself.

Usually, when we think of these waves, we imagine them coming from inflation, a theory suggesting the universe expanded faster than light in a split second. This creates a very specific, flat "hum" that scientists have been hunting for.

But this new paper asks a different question: What if the hum isn't from inflation, but from something else that happened later, like a cosmic "pop" or a collision?

Here is the breakdown of what the researchers did and found, using simple analogies:

1. The "Causality" Rule: No Instant Teleportation

The paper focuses on a fundamental rule of physics called causality. In simple terms, this means nothing can affect something else faster than the speed of light. You can't send a message to a friend across the room instantly; it takes time.

The Analogy: Imagine a crowd of people in a stadium. If one person stands up, the people next to them see it and stand up, then the next row, and so on. The "wave" of standing people moves across the stadium at a limited speed.
The Physics: If a source of gravitational waves (like a phase transition or a cosmic defect) happens in the early universe, it can only affect its immediate neighborhood. It cannot instantly ripple across the entire universe.
The Result: Because of this speed limit, any gravitational waves generated by these "non-inflationary" sources have a very specific shape: they are very quiet on large scales (the whole stadium) but get louder and louder on small scales (just a few rows). Scientists call this a "white-noise" spectrum that grows with the cube of the frequency ( $k^3$ ).

2. The Detective Work: Listening with "B-Modes"

Scientists look for these waves using the Cosmic Microwave Background (CMB), which is the afterglow of the Big Bang. It's like the static on an old TV, but it holds a map of the early universe.

The Clue: Gravitational waves leave a unique swirl pattern in the polarization of this light, called B-modes.
The Problem: There are other things that create swirls, like dust in our galaxy or the bending of light by gravity (lensing). It's like trying to hear a whisper in a room full of wind and traffic.
The Solution: The researchers combined data from five different "ears" (experiments): BICEP/Keck, SPTpol, SPT-3G, Planck, and WMAP. By looking at the data across many different angles and frequencies, they could filter out the "wind" (dust) and the "traffic" (lensing) to listen for the specific "whisper" of causal sources.

3. The Big Discovery: The "Silence"

The team created a mathematical model to see how loud these "causal" waves should be if they existed. Then, they compared that model to the actual data.

The Finding: They didn't hear the whisper. The data was too quiet.
The Limit: They calculated that if these causal sources exist, they are incredibly weak. They set a strict upper limit: the signal is less than 0.77% of the strength of the standard density fluctuations in the universe.
The Metaphor: Imagine you are looking for a specific type of bird in a forest. You have a very sensitive microphone. You don't hear the bird, so you conclude: "If this bird is here, it must be quieter than a leaf falling from a tree."

4. Why This Matters: A New Window on the Universe

This is the first time anyone has put a limit on these specific types of gravitational wave sources using CMB data.

The Reach: Traditional gravitational wave detectors (like LIGO) listen to high-frequency ripples (like the sound of a drum). Pulsar timing arrays listen to very low frequencies (like a deep bass note).
The Gap: There is a huge gap in the middle—ultra-low frequencies that no machine can currently hear.
The Paper's Contribution: By using the CMB, this paper effectively "listens" to frequencies so low ( $10^{-19}$ Hz) that they are completely inaccessible to any other technology. It's like using a telescope to see sound waves that are too low for our ears.

Summary

The paper says: "We looked for gravitational waves caused by events in the early universe that obey the speed of light (causality). We used a massive collection of telescope data to listen for their unique 'hum.' We didn't find them, but we now know exactly how quiet they must be. This rules out many theories about how the early universe might have behaved and opens a new door to studying the universe at frequencies we couldn't reach before."

In short: They didn't find the "ghost," but they proved the ghost is too shy to be heard, and they built the best "ear" we have ever had to listen for it.

Technical Summary: First Constraints on Causal Sources of Primordial Gravitational Waves

Problem and Motivation
While the imprint of primordial tensor perturbations on Cosmic Microwave Background (CMB) B-mode polarization is a primary probe for inflationary physics, non-inflationary sources of gravitational waves (GWs) in the early Universe also generate observable B-modes. Unlike inflation, which stretches sub-horizon perturbations to super-horizon scales to predict a nearly scale-invariant spectrum, post-inflationary causal sources are fundamentally limited by the impossibility of super-horizon correlations. Consequently, these sources universally yield a "white-noise" tensor power spectrum on sufficiently large scales, characterized by a dimensionless power spectrum scaling as $P_h(k) \propto k^3$ at small comoving wavenumbers $k$ .

This $k^3$ scaling is a robust, model-independent consequence of causality and finite correlation lengths for any source active before redshift $z \sim 7 \times 10^4$ . If the tensor source operates exclusively before this epoch, the turnover of the spectrum (related to the source's correlation length) is hidden by Silk damping, resulting in a universal B-mode angular power spectrum shape. Examples of such "Early Causal Tensor" (ECT) sources include first-order cosmological phase transitions, enhanced small-scale density perturbations, dark-photon perturbations, and topological defects. The challenge lies in distinguishing these signals from inflationary GWs and gravitational lensing, which can mimic ECT shapes on certain scales.

Methodology
The authors report the first constraints on the amplitude of an ECT power spectrum using a joint analysis of contemporary CMB B-mode data. The methodology involves:

Theoretical Framework: The study parametrizes the input tensor power spectrum as $P_h(k) \equiv r_{ct} A_s (k/k_{ref})^3$ , where $A_s$ is the measured scalar fluctuation amplitude, $k_{ref} = 0.01 \text{ Mpc}^{-1}$ , and $r_{ct}$ is the ratio of causal tensor power to scalar power at the reference scale. This parameterization captures the universal $k^3$ behavior required by causality for sources active before $z \sim 7 \times 10^4$ .
Data Combination: The analysis combines B-mode power spectrum measurements from:
- BICEP/Keck: Utilizing the "BK18" likelihood (through the 2018 season) covering $\ell \lesssim 300$ , incorporating cross-spectra from BICEP2, Keck Array, and BICEP3, along with external Planck and WMAP maps for foreground separation.
- SPT-3G: Covering the multipole range $32 < \ell < 502$ , focusing on the sky patch overlapping with BICEP/Keck to minimize foregrounds.
- SPTpol: Covering $52 \le \ell \le 2301$ , utilizing the 150 GHz bandpowers which offer the highest signal-to-noise.
- Planck and WMAP: Included as external maps for foreground constraints.
Likelihood Analysis: The observed B-mode spectrum is modeled as the sum of inflationary GWs ( $r$ ), ECT sources ( $r_{ct}$ ), lensing-induced B-modes (scaled by a free amplitude parameter $A_{LT}$ ), and frequency-dependent foregrounds (dust and synchrotron). The analysis marginalizes over foreground nuisance parameters and uses a Gaussian prior on the lensing amplitude based on independent measurements. Markov chain Monte Carlo (MCMC) sampling is performed using the Cobaya package with $\Lambda$ CDM cosmological parameters fixed to Planck 2018 best-fit values.

Key Results
The joint analysis yields the following constraints at 95% confidence level (CL):

Upper Limit on Causal Sources: $r_{ct} < 0.0077$ .
Upper Limit on Inflationary Tensor-to-Scalar Ratio: $r < 0.033$ .
No Significant Detection: The posterior distributions for both $r$ and $r_{ct}$ are consistent with zero; no statistically significant detection of either signal was found.

The constraint on $r_{ct}$ is driven primarily by the absence of excess power at high multipoles ( $300 < \ell < 2301$ ), where the ECT signal-to-noise is expected to peak and foregrounds are subdominant. The degeneracy between lensing and causal B-modes is broken by the external prior on the lensing amplitude.

Significance and Claims
The paper claims to provide the first-ever constraints on the amplitude of causal tensor power spectra using CMB B-mode data. The significance of these results includes:

Broad Theoretical Constraints: Because $r_{ct}$ can be directly related to the parameters of specific theories, the bound robustly constrains a wide class of well-motivated early-universe GW sources, including first-order phase transitions, scalar-induced GWs, and topological defects, without requiring model-specific spectral shapes.
Ultra-Low Frequency Probes: The authors translate the $r_{ct}$ limit into a bound on the present-day GW energy density ( $\Omega_{GW} h^2$ ) at ultra-low frequencies ( $f \sim 10^{-19}$ Hz). This frequency regime is otherwise inaccessible to traditional GW detection strategies, including pulsar timing arrays (PTAs), interferometers, and resonant cavities.
Model-Agnostic Approach: By starting with the generic $k^3$ template, the study establishes a model-agnostic upper bound applicable to any mechanism generating tensor perturbations in the early universe that respects causality.
Complementarity: The results demonstrate the unique capability of CMB observables to test new physics signals that elude traditional GW probes, extending the reach of GW astronomy to the ultra-low frequency regime.

The authors note that while the current limits are set by the absence of excess power, future work can build on this foundation to target more specific, cosmologically compelling models that may predict spectral features (peaks or turnovers) within the range of current CMB measurements.