⚛️ general relativity

Eccentric or circular? A reanalysis of binary black hole gravitational wave events for orbital eccentricity signatures

This paper reanalyzes 17 gravitational-wave events using a new eccentric waveform model (IMRPhenomTEHM) to demonstrate that two specific events show evidence for orbital eccentricity, while others exhibit potential eccentric features, thereby highlighting the necessity of incorporating eccentric models to avoid biases in parameter estimation and better understand binary black hole formation channels.

Original authors: Maria de Lluc Planas, Antoni Ramos-Buades, Cecilio García-Quirós, Héctor Estellés, Sascha Husa, Maria Haney

Published 2026-01-15

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Maria de Lluc Planas, Antoni Ramos-Buades, Cecilio García-Quirós, Héctor Estellés, Sascha Husa, Maria Haney

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, dark dance floor. For years, astronomers have been listening to the music of this dance floor: the gravitational waves created when two black holes spin around each other and crash together.

For a long time, scientists assumed these black holes danced in perfect circles, like a couple waltzing smoothly. This assumption made the math easier, but it might have been wrong. In reality, some black holes might have been thrown together by chaotic collisions in crowded star clusters, causing them to dance in wild, stretched-out ovals (ellipses) before they crash.

This paper is like a group of detectives re-examining 17 old crime scene recordings (gravitational wave events) to see if they missed any "oval dance" clues. They used a brand-new, super-fast computer tool called IMRPhenomTEHM to listen to the signals again.

Here is what they found, explained simply:

1. The New Tool: A Faster, Sharper Ear

Previous attempts to find these oval orbits were slow and clunky, like trying to run a marathon while carrying a heavy backpack. The new model (IMRPhenomTEHM) is like a lightweight, high-tech running shoe. It allows scientists to check for "eccentricity" (the oval shape of the orbit) much faster and more accurately than before.

2. The Big Discovery: Two Clear "Oval" Dancers

Out of the 17 events they checked, they found strong evidence that two of them were indeed dancing in an oval, not a circle:

GW200129: This event is the strongest candidate. The signal suggests the black holes were swinging in a stretched-out path. Even when the scientists tried to clean up "static" (glitches) from the recording, the oval shape remained the best explanation.
GW200208_22: This event also showed signs of an oval orbit, though the evidence was a bit fuzzier, like a song played through a slightly broken speaker.

3. The "Maybe" Dancers: Two High-Mass Mysteries

They also looked at two very heavy black hole pairs (GW190701 and GW190929). These signals were short and faint, making them hard to analyze.

The data hinted they might have been oval dancers, but because the signals were so short (like hearing only the last few seconds of a song), it's hard to be 100% sure.
The scientists caution that with these heavy pairs, the current tools assume the dance becomes a circle right before the crash. If the crash happened while they were still in an oval, our current tools might miss it.

4. The "Glitch" Problem: Cleaning the Record

One of the events (GW200129) had a known "pop" or "crackle" in the recording (a glitch) that confused the computers. The team tried different ways to scrub this noise out:

Method A (gw_subtract): Like using a noise-canceling headphone to remove a specific hum. This kept the "oval" evidence strong.
Method B (BayesWave): Like using a smart filter that guesses what the noise sounds like and removes it. This made the "oval" evidence weaker, but it was still there.
Conclusion: No matter how they cleaned the noise, the "oval" explanation still fit the data better than the "perfect circle" explanation.

5. Why This Matters: The Origin Story

Why do we care if the dance was a circle or an oval?

Perfect Circles usually mean the black holes were born together as a pair and evolved slowly over billions of years (like a couple growing old together).
Oval Orbits are the "smoking gun" of a dynamical formation. This means the black holes met by accident in a crowded place, like a dense star cluster or the center of a galaxy, and were thrown into a wild spin before crashing.

The Bottom Line

This paper proves that we need to stop assuming all black hole dances are perfect circles. By using their new, fast tool, the team found that at least two (and possibly four) of the black holes they studied were likely "wild dancers" with oval orbits. This suggests that the universe has more chaotic, collision-based black hole families than we previously thought.

They also noted that ignoring these oval shapes can trick scientists into getting the wrong numbers about how heavy the black holes are or how fast they were spinning. So, to get the true story of the universe's history, we must listen for the oval rhythms, not just the circular ones.

Technical Summary: Eccentric or circular? A reanalysis of binary black hole gravitational wave events for orbital eccentricity signatures

Problem Statement
The formation channels of binary black holes (BBHs) remain a central open question in gravitational-wave (GW) astrophysics. While isolated binary evolution is expected to produce quasi-circular (QC) orbits by the time they enter the sensitive frequency band of ground-based detectors, dynamical formation scenarios (e.g., in globular clusters or galactic nuclei) can produce mergers with significant orbital eccentricity. Detecting eccentricity is a "smoking gun" for dynamical formation. However, previous attempts to identify eccentric signatures in GW events have been constrained by computational limitations, often relying on reweighting techniques, machine learning, or restricting analyses to non-eccentric parameters. Furthermore, neglecting eccentricity in parameter estimation (PE) can introduce biases in inferred source properties, such as chirp mass and effective spin.

Methodology
This work presents a reanalysis of 17 BBH events detected during the first three observing runs (O1–O3) of the Advanced LIGO and Advanced Virgo detectors. The analysis utilizes the IMRPhenomTEHM model, a new phenomenological time-domain multipolar waveform model for aligned-spin BBHs in elliptical orbits. This model extends the QC aligned-spin model IMRPhenomTHM by incorporating two eccentric parameters: eccentricity ( $e$ ) and mean anomaly ( $l$ ), including full 3PN orbit-averaged dynamics and eccentric corrections up to $O(e^6)$ .

Key methodological components include:

Model Comparison: The authors compare the eccentric IMRPhenomTEHM model against the underlying QC model (IMRPhenomTHM) and, for specific events, precessing models (IMRPhenomTPHM and NRSur7dq4) to distinguish between eccentricity and spin precession.
Priors: Two distinct priors for eccentricity at the reference frequency of 10 Hz are tested: a uniform prior ( $e \in [0, 0.5]$ ) and a log-uniform prior ( $e \in [10^{-4}, 0.5]$ ).
Bayesian Inference: Using the bilby framework and the dynesty nested sampler, the authors compute posterior distributions and Bayes factors ( $B_{E/QC}$ ) to quantify the evidence for eccentricity versus the QC hypothesis.
Systematics and Glitch Mitigation: For the event GW200129, which is known to contain a glitch, the authors perform analyses using raw GWOSC data, data processed with gw_subtract, and data processed with BayesWave (using three distinct glitch draws) to assess the impact of data treatment on eccentricity recovery.
High-Mass Events: For high-mass events where only the merger-ringdown is visible, the authors investigate the impact of sampler settings (increasing live points to $n_{live}=2000$ ) and waveform systematics.

Key Results
The reanalysis identifies four events with potential evidence for eccentricity, though the strength of the evidence varies:

GW200129: This event shows the strongest evidence for eccentricity.
- Using non-deglitched GWOSC data, the log-10 Bayes factor favors the eccentric hypothesis significantly ( $\log_{10} B_{E/QC} \approx 5.14$ for uniform prior, $4.19$ for log-uniform).
- Analyses using gw_subtract data also yield strong support ( $\log_{10} B_{E/QC} \approx 4.00$ and $3.35$).
- Even with BayesWave mitigation (specifically the BayesA draw, which shows the least support for dynamical features), the eccentric hypothesis remains favored over both aligned-spin and precessing QC scenarios ( $\log_{10} B_{E/QC} \approx 1.30$ ).
- The recovered eccentricity is consistent across priors ( $e_{GW}^{10Hz} \approx 0.26$ for gw_subtract data).
- The eccentric model posteriors align better with the precessing surrogate NRSur7dq4 than with the aligned-spin QC model, suggesting a preference for dynamical effects over simple spin precession.
GW200208_22: This event shows moderate support for eccentricity, heavily dependent on the prior choice.
- With a uniform prior, the log-10 Bayes factor is $\approx 1.14$ , favoring the eccentric hypothesis.
- With a log-uniform prior, the support diminishes ( $\log_{10} B_{E/QC} \approx 0.49$ ), though the eccentric hypothesis remains preferred over QC precessing-spin models.
- The inclusion of eccentricity significantly tightens constraints on binary parameters, reducing the inferred total mass from $\sim 174 M_\odot$ (QC model) to $\sim 65 M_\odot$ .
GW190701 and GW190929: These high-mass events show only minor or inconclusive support for eccentricity.
- While uniform priors suggest a preference for eccentricity (e.g., $\log_{10} B_{E/QC} \approx 0.56$ for GW190929), log-uniform priors suppress this evidence.
- The authors note that for these high-mass events, the signal duration is short, and current eccentric models assume circularization at merger. Consequently, distinguishing eccentricity from waveform systematics near the merger is challenging, and strong claims are not supported.

Significance and Claims
The paper claims that the high computational efficiency of IMRPhenomTEHM enables the routine application of eccentric waveform models to large datasets, a capability previously limited by computational costs. The study highlights several critical findings:

Bias Mitigation: Ignoring eccentricity can lead to systematic biases in parameter estimation, particularly underestimating total mass and misinterpreting effective spin.
Dynamical Formation: The robust detection of eccentricity in GW200129 (and to a lesser extent GW200208_22) provides evidence for dynamical formation channels, such as those in dense stellar environments.
Degeneracy: There is a notable degeneracy between orbital eccentricity and spin precession; the authors note that distinguishing between them requires generic spin-precessing eccentric models, which are not yet fully implemented in their current framework.
Future Outlook: The authors emphasize that as detector sensitivity improves and the diversity of the BBH population increases, incorporating eccentric (and eventually precessing-eccentric) models will be essential for accurate population studies and formation channel identification. They explicitly state that their results support the presence of orbital eccentricity in GW200129, reinforcing interpretations of a dynamical origin, while acknowledging that current models assume circularization at merger, limiting strong claims for high-mass, short-duration signals.