⚛️ general relativity

Environmental effects vs. modified gravity in the LISA massive black hole binary population

This paper demonstrates that while environmental effects like accretion and modified gravity can formally degenerate in individual LISA massive black hole binary signals, hierarchical population analysis of realistic astrophysical scenarios shows these effects are unlikely to bias tests of general relativity or mimic deviations from it.

Original authors: Lorenzo Copparoni, Enrico Barausse

Published 2026-07-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Lorenzo Copparoni, Enrico Barausse

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 concert hall. For the last decade, we've been listening to the music of colliding black holes using ground-based detectors like LIGO. But soon, a new, space-based instrument called LISA will tune into a lower, deeper frequency. This new instrument will listen to the "heavyweights" of the black hole world: massive black hole binaries (MBHBs) that are millions of times heavier than our sun.

The scientists in this paper, Lorenzo Copparoni and Enrico Barausse, asked a crucial question: Can the "noise" of the universe mess up our ability to hear the true music of gravity?

Here is the breakdown of their investigation using simple analogies:

1. The Two Suspects: "Dirty" vs. "Broken"

When two black holes spiral toward each other, they create ripples in space-time called gravitational waves. According to Einstein's General Relativity, these waves should follow a very specific, clean pattern (like a perfect violin note).

However, the universe isn't always empty. There are two things that could distort this pattern:

The "Dirty" Environment (Astrophysics): Imagine the black holes are swimming in a thick soup of gas and dust (a circumbinary disk). As they swim, they might eat some of the soup (accretion) or get pushed by the current (viscous migration). This changes how they spiral, slightly altering the "note" they play.
The "Broken" Physics (Modified Gravity): Imagine that the fundamental laws of the universe are slightly different than Einstein predicted. For example, maybe the strength of gravity (Newton's constant, $G$ ) is slowly changing over time. This would also alter the "note" in a very specific way.

The Problem: Both the "soup" and the "broken physics" change the sound in exactly the same way. If you listen to just one black hole collision, you cannot tell if the sound is distorted because the black hole was eating gas or because the laws of physics are wrong. They are "formally degenerate," meaning they look identical in a single event.

2. The Detective Work: Listening to the Whole Choir

Since you can't solve the mystery with one note, the authors decided to listen to the entire choir of black hole collisions that LISA will detect over 10 years.

The Logic: If the distortion is caused by gas, it should only happen to black holes that are actually swimming in gas. Some will be in thick soup, some in thin soup, and some in a vacuum. The "distortion" will vary from song to song.
The Logic: If the distortion is caused by changing gravity, it should happen to every single black hole in the universe, equally. It's a universal rule, not a local condition.

By analyzing the statistical pattern of hundreds of events, the scientists hoped to separate the "local gas noise" from the "universal law change."

3. The Investigation Results

The researchers ran complex simulations using different models of how black holes form and evolve. They asked: If LISA listens for 10 years, will it be fooled?

The "Gas" Effect: They found that even if they assumed an extreme scenario where 50% of black holes are actively eating gas (and some are eating at super-fast rates), the effect is still too weak to be clearly heard. Most black holes are too far away or too massive for the gas to make a noticeable difference in the signal.
The "Bias" Check: They checked if the gas could trick them into thinking the laws of physics were broken. The answer is no. Even with extreme assumptions, the data still looks perfectly consistent with Einstein's vacuum theory. The "gas noise" is too quiet to drown out the true signal.
The "Universal" Limit: They also calculated how well LISA could measure a changing gravitational constant. They found LISA would be much better than current ground detectors, but still not as precise as our measurements within our own Solar System.

4. The Verdict

The paper concludes with a reassuring message for physicists:

Don't worry about the "soup."

Even under the most optimistic (and extreme) assumptions about how much gas surrounds these black holes, the environmental effects are unlikely to fake a signal of "new physics." LISA's tests of General Relativity using massive black holes will remain clean and reliable.

The Caveat: The authors add one small warning: Their analysis assumed the black holes were moving in perfect circles. If the black holes are moving in very stretched, oval-shaped (eccentric) orbits, the gas might have a louder effect. But for now, in the standard scenarios, the universe's "background noise" won't ruin the music.

Summary Analogy

Imagine you are trying to detect if a piano is out of tune (Modified Gravity) by listening to a choir of pianists.

The Fear: You worry that some pianists are playing in a windy room (Gas), which might make their notes sound flat, tricking you into thinking the piano itself is broken.
The Study: The authors listened to a thousand pianists. They found that even if half of them were in windy rooms, the wind wasn't strong enough to make the whole choir sound out of tune.
The Conclusion: You can trust your ears. If the choir sounds out of tune, it's likely the piano, not the wind. And if the piano is fine, the wind is just too quiet to matter.

Technical Summary: Environmental Effects vs. Modified Gravity in the LISA Massive Black Hole Binary Population

Problem Statement
Gravitational-wave (GW) signals from massive black hole binaries (MBHBs) observed by the Laser Interferometer Space Antenna (LISA) carry imprints of both their astrophysical environments and potential deviations from General Relativity (GR). A critical challenge arises because specific environmental effects and certain modified-gravity (MG) theories enter the GW waveform at the same negative post-Newtonian (PN) order. Specifically, gas accretion, viscous migration in circumbinary disks, and a secular time-variation of the gravitational constant ( $\dot{G}$ ) all induce a dephasing at the $-4$ PN order. Consequently, these effects are formally degenerate at the single-event level; a single MBHB signal cannot distinguish between an environmental interaction and a fundamental modification of gravity. The paper investigates whether these environmental effects leave a detectable imprint on the LISA MBHB population and whether they could mimic $\dot{G}$ signals, thereby biasing tests of GR.

Methodology
The authors employ a multi-stage approach combining analytical forecasting with hierarchical Bayesian inference:

Waveform Modeling: The study utilizes the parameterized post-Einsteinian (ppE) framework to capture the dominant phase deformation at $-4$ PN order. Three specific channels are modeled:
- Modified Gravity: A secular drift of Newton's constant, $G(t) \approx G_N[1 + \dot{G}(t-t_0)]$ .
- Environmental (Accretion): Mass growth of binary components due to gas accretion, conserving angular momentum.
- Environmental (Viscous Migration): Torques exerted by a circumbinary disk on the binary.
  All three are described by a common phase-deformation parameter, making them indistinguishable for individual sources.
Population Models: The analysis relies on five semi-analytic population models spanning different formation scenarios:
- Heavy-Seed (HS): Seeds from proto-galactic disk collapse ( $\sim 10^5 M_\odot$ ).
- Light-Seed (LS): Seeds from Population III star remnants.
  These models vary in assumptions regarding delay times between galaxy and MBH mergers and the impact of supernova (SN) feedback on accretion.
Single-Event Analysis: Fisher-matrix forecasts are computed across the chirp-mass and redshift plane to estimate parameter estimation errors ( $\sigma_{\dot{G}}$ , $\sigma_{f_{\text{Edd}}}$ ) and systematic biases introduced by unmodeled environmental effects. Synthetic catalogs are generated for a 10-year LISA mission.
Hierarchical Bayesian Inference: To break the single-event degeneracy, the authors perform a population-level analysis. They compare three recovery hypotheses using nested sampling and Bayes factors:
- $H_V$ : Vacuum GR (no environmental or MG effects).
- $H_E$ : Environmental effects (a fraction $\lambda$ of binaries are active with an Eddington-ratio distribution slope $\alpha$ ).
- $H_{\dot{G}}$ : Universal $\dot{G}$ drift.
  The analysis marginalizes over single-event parameters to constrain the population hyperparameters ( $\lambda, \alpha$ for $H_E$ ; $\dot{G}$ for $H_{\dot{G}}$ ).

Key Contributions and Results

Single-Event Degeneracy and Sensitivity:
- Environmental effects and $\dot{G}$ are indistinguishable for single sources due to identical frequency dependence.
- Sensitivity to $-4$ PN deformations is concentrated in low-mass ( $M_c \lesssim 10^5 M_\odot$ ), low-redshift ( $z \lesssim 1$ ) binaries.
- Accretion: Even under extreme assumptions (active fraction of 50%, super-Eddington accretion), individual events show negligible bias. At realistic Eddington ratios ( $f_{\text{Edd}} \approx 1$ ), no event exhibits a chirp-mass shift exceeding 10% of the statistical uncertainty.
- Viscous Migration: This channel is more promising due to its stronger scaling with mass ratio ( $\eta^{-2}$ vs. $\eta^{-1}$ for accretion). In Heavy-Seed models, up to $\sim 8$ events per year (with optimistic disk coupling $\xi \sim 10$ ) could show chirp-mass biases comparable to statistical errors. However, these detections remain relatively noisy.
Population-Level Constraints:
- $\dot{G}$ Limits: The hierarchical analysis yields a population-level statistical error of $\sigma_{\dot{G}} \sim 10^{-7} \text{ yr}^{-1}$ . This is significantly tighter than current ground-based GW bounds but remains orders of magnitude weaker than Solar-System limits and constraints from other LISA sources (e.g., EMRIs).
- Environmental Detection: When injecting environmental signals (including a high active fraction $\lambda=0.5$ and super-Eddington tails), the recovered population posteriors remain fully compatible with the vacuum hypothesis ( $H_V$ ). The data do not provide sufficient information to distinguish the environmental model from vacuum, though they yield non-trivial upper limits on the active fraction $\lambda$ and a mild lower bound on the Eddington-ratio slope $\alpha$ .
- Model Comparison: Bayes factors consistently favor the vacuum hypothesis ( $H_V$ ) over the environmental hypothesis ( $H_E$ ) for realistic injections. A clear preference for the modified-gravity hypothesis ( $H_{\dot{G}}$ ) only emerges for injected $\dot{G}$ values several orders of magnitude above Solar-System limits (e.g., $G_1 \sim 10^{-6} \text{ yr}^{-1}$ ), which lack strong theoretical motivation.

Significance and Conclusions
The paper concludes that, under astrophysically realistic assumptions, environmental contamination is unlikely to bias LISA's tests of General Relativity using massive black hole binaries. While environmental effects and modified gravity are formally degenerate at the single-event level, the population-level analysis—leveraging the universality of $\dot{G}$ versus the stochastic nature of environmental interactions—fails to detect a significant environmental signal or a false positive for modified gravity.

The authors emphasize that while a population-wide detection of environmental imprints is unlikely in the scenarios considered, this does not rule out environmental effects entirely. A key caveat noted is that the analysis is restricted to circular inspirals; recent work suggests environmental dephasing could be substantially amplified in eccentric orbits, a factor that could alter both detection prospects and bias risks in future studies. Ultimately, the study provides reassurance that LISA's precision tests of GR with MBHBs will not be compromised by standard astrophysical environmental effects.