🔭 astrophysics

Unveiling the population of massive quenched galaxies at $z\ge2$ in the COLIBRE simulations -- I. Galaxy demographics

This paper utilizes the COLIBRE cosmological hydrodynamical simulations to demonstrate that massive quenched galaxies at $z \ge 2$ exhibit number densities and formation histories consistent with JWST observations when accounting for resolution effects, AGN feedback, and observational uncertainties, while predicting significantly lower dust and molecular gas fractions compared to their star-forming counterparts.

Original authors: Ángel Chandro-Gómez (International Centre for Radio Astronomy Research, ARC Centre for All-Sky Astrophysics in 3 Dimensions), Claudia del P. Lagos, Chris Power, William M. Baker, Alejandro Benítez-Lla

Published 2026-06-01

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Ángel Chandro-Gómez (International Centre for Radio Astronomy Research, ARC Centre for All-Sky Astrophysics in 3 Dimensions), Claudia del P. Lagos, Chris Power, William M. Baker, Alejandro Benítez-Llambay, Evgenii Chaikin, Harry G. Chittenden, Camila Correa, Carlos S. Frenk, Filip Huško, Robert J. McGibbon, Themiya Nanayakkara, Sylvia Ploeckinger, Alexander J. Richings, Matthieu Schaller, Joop Schaye, James W. Trayford

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: Finding "Cosmic Ghosts"

Imagine the universe as a giant, bustling city. For a long time, astronomers thought that in the early days of this city (when the universe was young), everything was chaotic, noisy, and full of construction crews building new stars.

However, the new James Webb Space Telescope (JWST) has acted like a super-powerful night-vision camera. It looked back in time and found something surprising: a huge number of "ghost towns." These are massive galaxies that have already stopped building stars. They are old, quiet, and dead, yet they exist when the universe was only a toddler.

This paper asks: How did these "ghost towns" get built so fast and shut down so quickly? And do our computer models of the universe actually predict them?

The Tool: The "COLIBRE" Simulator

To answer this, the authors used a new, high-tech computer simulation called COLIBRE. Think of this simulation as a massive, ultra-detailed video game engine that runs the history of the universe from the Big Bang to today.

Unlike older games that had blurry graphics or simplified physics, COLIBRE has "high-definition" settings. It specifically tracks:

Dust: Like cosmic soot.
Molecular Gas: The fuel for making stars.
Black Holes: The "engines" in the center of galaxies that can blow things apart.

The Main Findings

1. The "Ghost Towns" Are Real (and Common)

The team ran their simulation and found that these massive, dead galaxies do exist in the model.

The Tension: Early versions of the simulation didn't make enough of them, or they were too small. It was like trying to bake a cake but the oven wasn't hot enough to rise the batter.
The Fix: The authors realized that if you account for the fact that our telescopes aren't perfect (they have "fuzzy" vision), the simulation matches reality much better. When they added a layer of "real-world error" to their data, the number of ghost towns in the simulation lined up perfectly with what JWST sees.

2. The Recipe for a Ghost Town

How do these galaxies die so young? The simulation suggests a very specific, dramatic life cycle:

The Growth Spurt: These galaxies grow incredibly fast. They build up most of their stars in a short burst, like a teenager growing 6 inches in a year.
The Sudden Stop: Then, something hits the brakes hard. The simulation shows that Supermassive Black Holes in the center of these galaxies act like a cosmic firehose. They shoot out energy and jets that blow away the gas fuel.
The Result: Once the fuel is gone, the galaxy stops making stars almost instantly. It goes from a construction site to a ghost town in less than a billion years (which is a blink of an eye in cosmic time).

3. The "Dry" and "Clean" Interior

The paper also looked at what these dead galaxies are made of inside.

The Analogy: Imagine a factory that just shut down. You'd expect it to still be full of raw materials (gas) and smoke (dust).
The Discovery: In the simulation, these dead galaxies are surprisingly empty. They have 10 to 100 times less dust and gas than active, star-making galaxies.
Why it matters: This matches the few observations we have. It confirms that the "firehose" from the black hole didn't just stop the construction; it actually cleaned out the factory, leaving it dry and barren.

4. Size and Shape Don't Change Much

One might think that when a galaxy stops making stars, it shrinks or changes shape.

The Finding: The simulation shows that these dead galaxies are roughly the same size and move in the same way as their active, star-making neighbors.
The Implication: The "death" of the galaxy happens before it changes its shape. It's like a car that suddenly runs out of gas but keeps its shape and size; the transformation happens later.

Why This Matters

This paper is the first part of a series. It proves that our best computer models (COLIBRE) can successfully recreate these mysterious, early "ghost towns" if we include the right physics—specifically, how black holes blow away gas and dust.

It tells us that the universe is capable of building massive, dead galaxies very quickly, and that the "engine" responsible for killing them is likely the supermassive black hole at their center, acting as a cosmic cleanup crew that clears out the fuel supply.

In short: The universe is full of early "ghost towns," and our computer models finally have the right "ghost-busting" physics to explain how they got there.

Technical Summary: Unveiling the population of massive quenched galaxies at $z \ge 2$ in the COLIBRE simulations – I. Galaxy demographics

Problem Statement
Recent observations from the James Webb Space Telescope (JWST) have revealed a substantial population of Massive, Quenched Galaxies (MQGs; $M_\star \gtrsim 10^{10} M_\odot$ ) at high redshifts ( $z \gtrsim 2$ ). The inferred number densities and rapid formation timescales of these systems initially presented significant tensions with the hierarchical $\Lambda$ CDM model and existing galaxy formation simulations, which struggled to reproduce such massive, passive systems so early in cosmic history. While subsequent observational analyses have revised number densities downward, discrepancies remain between current models and the latest JWST data regarding the abundance, stellar mass functions, and physical properties of MQGs.

Methodology
This study utilizes the new COLIBRE cosmological hydrodynamical simulations, which introduce key innovations in sub-grid physics compared to previous large-volume simulations. The analysis focuses on the fiducial L200m6 simulation (a $200 \, \text{cMpc}^3$ volume with mean gas particle mass $m_g \sim 1.84 \times 10^6 M_\odot$ ), alongside comparative runs varying in volume, resolution, and Active Galactic Nucleus (AGN) feedback models (thermal vs. hybrid thermal+jet).

Key methodological features include:

Sub-grid Physics: Explicit tracking of the cold molecular phase and dust grains (via the HYBRID-CHIMES model), self-consistent chemical enrichment, and updated AGN feedback prescriptions (including super-Eddington accretion and jet feedback).
Sample Selection: MQGs are defined as galaxies with $M_\star > 10^{10} M_\odot$ and specific star formation rates (sSFR) $< 0.2/t_{\text{age}}$ at $z \ge 2$ .
Forward Modeling: To address observational uncertainties, the authors convolve simulated stellar masses and SFRs with a Gaussian distribution ( $\sigma = 0.3$ dex) to mimic errors inherent in Spectral Energy Distribution (SED) fitting and photometric redshifts.
Property Analysis: The study examines number densities, stellar mass functions (SMFs), star formation histories (SFHs), formation and quenching timescales ( $t_{50}$ and $t_q$ ), dust and molecular hydrogen ( $H_2$ ) fractions, and kinematics ( $v_\star/\sigma_\star$ ).

Key Contributions and Results

Number Densities and Resolution Dependence:
- The number density of MQGs in COLIBRE depends on both mass resolution and AGN feedback implementation. Higher-resolution simulations (e.g., L200m7) produce fewer quenched galaxies because higher resolution leads to denser gas around black holes, requiring stronger feedback to quench, and because AGN feedback becomes less bursty.
- Hybrid AGN feedback (thermal + jets) produces significantly fewer quenched galaxies at $z \gtrsim 2$ compared to the fiducial thermal model, as black holes grow more slowly in the early universe in the hybrid scenario.
- When observational uncertainties (0.3 dex) are forward-modeled, the number densities predicted by the fiducial L200m6 simulation are in broad agreement with the latest JWST spectroscopic and photometric constraints, significantly reducing the tension seen in previous studies.
Stellar Mass Functions (SMF):
- The SMF of MQGs in COLIBRE peaks around $M_\star \approx 10^{10.5} M_\odot$ , consistent with observational data from Baker et al. (2025c) and Stevenson et al. (2026).
- Unlike other models (e.g., EAGLE, IllustrisTNG, Simba) that struggle to match both the low- and high-mass ends simultaneously, COLIBRE reproduces the SMF across the full mass range when observational errors are accounted for. This is attributed to the explicit modeling of the cold gas phase, which allows for higher gas densities, super-Eddington black hole accretion, and consequently more efficient early quenching.
Star Formation Histories (SFHs):
- Formation Timescales ( $t_{50}$ ): MQGs in COLIBRE form rapidly, with median formation times of $t_{50} \approx 0.5 - 1.5$ Gyr. This indicates extended formation followed by rapid quenching.
- Quenching Timescales ( $t_q$ ): The simulations predict rapid quenching with median timescales $t_q \lesssim 0.6$ Gyr, often driven by strong starburst episodes.
- The SFHs frequently exhibit multiple starburst episodes before final quenching, a feature that complicates observational recovery but is qualitatively consistent with inferences from JWST data.
Dust and Molecular Gas Content:
- The model predicts that MQGs possess dust and $H_2$ fractions more than 1 dex lower than their massive star-forming counterparts.
- These low fractions are consistent with the scarce observational estimates and upper limits available, supporting the view that feedback processes efficiently deplete gas reservoirs and suppress dust growth in these systems.
Sizes and Kinematics:
- MQGs and massive star-forming systems in COLIBRE show broadly similar sizes and kinematics ( $v_\star/\sigma_\star$ ). This suggests that significant size or morphological transformations in these galaxies occur after quenching, rather than being a prerequisite for the quenched state at $z \ge 2$ .

Significance
This paper establishes that the COLIBRE simulations, leveraging state-of-the-art sub-grid physics (particularly in dust, cold gas, and AGN feedback), can robustly reproduce the demographics of massive quenched galaxies at high redshift. The study demonstrates that apparent tensions between simulations and JWST observations are largely alleviated when realistic observational uncertainties are forward-modeled. By providing a consistent framework for MQG abundance, mass assembly, and quenching timescales, this work offers a viable pathway to understanding the origins of these early passive systems and sets the stage for Paper II, which will investigate the specific physical mechanisms driving their rapid formation and quenching.