🔭 astrophysics

Implications of the nanoHertz Gravitational-Wave Background for Galactic Feedback and Massive Black Hole Growth

This study demonstrates that PTA measurements of the nanoHertz gravitational-wave background serve as a powerful probe for feedback physics and SMBH growth, revealing that current cosmological simulations either over-suppress black hole growth (under-predicting the signal) or require inefficient feedback to match observations (failing to produce realistic galaxies), thus indicating that SMBHs likely grow more efficiently or earlier than currently modeled.

Original authors: Megan Taylor Tillman, Blakesley Burkhart, Chiara M. F. Mingarelli, J. Andrew Casey-Clyde, Lars Hernquist, Sownak Bose, Enikő Regős, César Hernández-Aguayo

Published 2026-05-27

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Megan Taylor Tillman, Blakesley Burkhart, Chiara M. F. Mingarelli, J. Andrew Casey-Clyde, Lars Hernquist, Sownak Bose, Enikő Regős, César Hernández-Aguayo

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

The Big Picture: Listening to the Universe's Hum

Imagine the universe is a giant concert hall. For a long time, we thought it was silent. But recently, scientists using "pulsar timing arrays" (which act like ultra-precise cosmic metronomes) detected a low, constant hum. This is the Gravitational Wave Background (GWB).

Think of this hum not as a single note from one instrument, but as the collective roar of thousands of distant, massive black holes orbiting each other, spinning up like figure skaters before they crash together. The paper asks: Why is this hum louder than our computer simulations predicted?

The Cast of Characters

Supermassive Black Holes (SMBHs): The "rock stars" of galaxies. They are so heavy they sit at the center of almost every galaxy.
Feedback: The "volume control" or "thermostat" of a galaxy. When a black hole eats gas, it releases massive amounts of energy (like a jet engine). This energy blows gas away, which stops the black hole from eating more and also stops new stars from forming. This is called feedback.
Simulations: Giant computer programs (like IllustrisTNG, Simba, and MillenniumTNG) that try to recreate the history of the universe from the Big Bang to today.

The Problem: The "Silent" Simulations

The scientists ran these massive computer simulations to predict how many black holes should exist and how heavy they should be. They then used those numbers to calculate what the "hum" (GWB) should sound like.

The Result: The simulations predicted a hum that was too quiet.

The real universe (measured by pulsars) is about 2 to 10 times louder than the best simulations predicted.
This means the simulations are missing something. Either there are more massive black holes than we thought, or they are growing faster than our models allow.

The Investigation: Turning the Dials

The authors treated the computer simulations like a complex sound mixing board. They wanted to see which "knobs" (feedback models) changed the volume of the hum.

The "Feedback" Knobs: In the simulations, they could turn the "AGN feedback" (the black hole's jet engine) up, down, or turn it off completely.
The Analogy: Imagine you are baking a cake (a galaxy).
- Strong Feedback: You add a lot of baking soda. The cake rises fast, but then the heat blows the batter out of the pan. The black hole stops growing, and the galaxy stays small.
- Weak/No Feedback: You don't add enough baking soda (or turn off the oven). The batter keeps rising uncontrollably. The black hole gets huge, and the galaxy becomes a monster.

What they found:

The "Realistic" Models are Too Quiet: The simulations that successfully create realistic-looking galaxies (with the right amount of stars and gas) have strong feedback. This feedback keeps the black holes small. Because the black holes are smaller, the predicted hum is too quiet compared to reality.
The "Loud" Models are Unrealistic: When they turned off the feedback (letting black holes grow wild and huge), the predicted hum became loud enough to match the real data. However, these simulations produced galaxies that looked nothing like the ones we see in the sky. They were too big, too bright, and had the wrong structure.

The Conclusion: We Need a New Recipe

The paper concludes that there is a mismatch.

If we use the recipes that make real galaxies, the black holes are too small, and the hum is too quiet.
If we use the recipes that make the loud hum, the galaxies look fake.

The Takeaway:
The universe seems to be growing supermassive black holes more efficiently or earlier than our current computer models allow. The "thermostat" (feedback) in our simulations might be set too high, cutting off black hole growth too soon.

To fix this, scientists need to rewrite the rules of how black holes are "seeded" (born) and how they grow in the early universe. The new "loudness" of the universe is a powerful new tool to help them figure out exactly how to tweak these rules so that the computer models match both the galaxies we see and the hum we hear.

Technical Summary: Implications of the nanoHertz Gravitational-Wave Background for Galactic Feedback and Massive Black Hole Growth

Problem Statement
Recent measurements by Pulsar Timing Arrays (PTAs) have provided strong evidence for a stochastic nanoHertz gravitational-wave background (GWB), likely generated by the incoherent superposition of supermassive black hole binaries (SMBHBs). The amplitude of this background is sensitive to the mass function of supermassive black holes (SMBHs). However, a tension exists: cosmological hydrodynamic simulations (such as IllustrisTNG, MillenniumTNG, and Simba) consistently under-predict the GWB amplitude inferred by PTA data. This discrepancy suggests that current models of SMBH growth, seeding, and the feedback mechanisms that regulate them (AGN and stellar feedback) may be incomplete. Specifically, it is unclear how variations in galactic feedback prescriptions alter the high-mass end of the Black Hole Mass Function (BHMF) and, consequently, the predicted GWB signal.

Methodology
The authors employ a semi-analytic framework based on Casey-Clyde et al. (2022) to link simulation-predicted BHMFs to the GWB amplitude. The methodology proceeds as follows:

Simulation Suites: The study analyzes four major simulation suites: IllustrisTNG, MillenniumTNG, Simba, and the CAMELS (Cosmology and Astrophysics with MachinE Learning Simulations) project.
Box Size and Convergence: To address the sensitivity of the GWB to the high-mass end of the BHMF, the authors conduct convergence studies using multiple box sizes (ranging from 25 Mpc/h to 500 Mpc/h). They determine that boxes smaller than ~35 Mpc/h (for TNG) and ~50 Mpc/h (for Simba) suffer from Poisson noise and insufficient sampling of massive halos ( $M_{BH} > 10^9 M_\odot$ ).
Feedback Variations: The study explores the impact of galactic feedback by:
- Comparing fiducial models across different simulation codes.
- Analyzing "1P" (one-parameter) variations in the CAMELS suites, where 23 astrophysical parameters controlling stellar and AGN feedback are varied individually.
- Examining specific Simba feedback variants where AGN jet modes are removed or altered (e.g., "SW only," "SW + Rad").
GWB Calculation: The simulation BHMFs are converted into a binary population using an Eddington Ratio Distribution Function (ERDF) to identify active quasars. This population is scaled by a normalization factor ( $N_{cal} \approx 4$ ) to account for the binary-to-quasar ratio, and secondary masses are assigned via a mass-ratio distribution. The resulting binary population is integrated to calculate the characteristic strain spectrum and the GWB amplitude at a reference frequency of $f = 1 \text{ yr}^{-1}$ .

Key Contributions

Quantifying Feedback Impact: The paper demonstrates that the choice of stellar and AGN feedback models significantly alters the high-mass end of the BHMF. Inefficient or absent AGN feedback leads to an overabundance of massive SMBHs, increasing the predicted GWB amplitude by factors of 2 (in fiducial variations) to 10 (in extreme CAMELS variations).
Convergence Analysis: The authors establish the minimum box sizes required for robust GWB predictions, identifying that $M_{BH} > 10^9 M_\odot$ statistics are not converged in the standard 25 Mpc/h CAMELS boxes.
Systematic Under-prediction: The study confirms that fiducial models from IllustrisTNG, MillenniumTNG, and Simba under-predict the PTA-inferred GWB amplitude by a factor of approximately 2. To match observations, these models would require 1.5 to 5.5 times more SMBHs in the $M_{BH} > 10^8 M_\odot$ range.

Results

Fiducial Models: All standard (fiducial) simulation models under-predict the GWB amplitude compared to NANOGrav 15-year data. This implies that current feedback prescriptions suppress SMBH growth too efficiently or that seeding occurs too late.
Feedback Variants: Simulations with reduced or absent AGN feedback (specifically Simba runs without jet modes) produce GWB amplitudes consistent with PTA data. However, these models fail to reproduce realistic galaxy populations (e.g., they fail to quench star formation effectively), rendering them physically unrealistic.
Parameter Sensitivity: In the CAMELS 1P sets, specific parameters drive the GWB amplitude:
- For CAMELS-TNG, the parameter $\beta$ (controlling the switch between low- and high-accretion AGN states) has the largest impact. A value of $\beta=0$ matches PTA data but under-predicts the low-mass stellar mass function.
- For CAMELS-Simba, the seed mass ratio parameter ( $A_{BHseed}$ ) is most critical. Lowering the seed threshold allows earlier SMBH growth, matching PTA data, but may disrupt early star formation.
Stellar Feedback: While often assumed to affect only low-mass black holes, the study finds that certain stellar feedback parameters (e.g., wind speed normalization and thermal fraction) also influence the high-mass BHMF and GWB amplitude by regulating the fuel available for black hole growth.

Significance and Claims
The paper claims that PTA measurements provide a powerful, independent probe of galactic feedback physics, complementing traditional electromagnetic constraints. The mismatch between simulation predictions and PTA data suggests that SMBH growth in the early universe is more efficient or occurs earlier than captured by current models.

The authors conclude that while modifying feedback parameters or seeding prescriptions can artificially align simulations with PTA data, such modifications often break the agreement with observed galaxy statistics (such as the stellar mass function). Therefore, the tension between the GWB amplitude and simulation predictions indicates a fundamental need to re-evaluate SMBH seeding and growth prescriptions in cosmological hydrodynamic simulations. The study emphasizes that future models must simultaneously reproduce realistic galaxy populations and the observed GWB amplitude, suggesting that current feedback models may be suppressing SMBH growth excessively. The authors note that their framework relies on simplifying assumptions (e.g., neglecting environmental hardening and merger timing details), and that a self-consistent treatment of binary evolution is required for future precision.

The Big Picture: Listening to the Universe's Hum

The Cast of Characters

The Problem: The "Silent" Simulations

The Investigation: Turning the Dials

The Conclusion: We Need a New Recipe

More like this