Effects of Varied Cosmic Ray Feedback from AGN on Massive Galaxy Properties

This study demonstrates that while varied cosmic ray feedback mechanisms from active galactic nuclei successfully quench star formation in massive galaxies to match observed scaling relations, they produce significantly different circumgalactic medium properties, suggesting that multi-wavelength observations of halo gas could constrain the specific physical nature of AGN feedback.

The Gist

The Big Problem: Why Are Giant Galaxies "Dead"?

Imagine the universe as a giant garden. In this garden, there are massive, ancient trees (galaxies) that used to grow new leaves (stars) all the time. But today, astronomers see that these giant trees have stopped growing new leaves. They are "red and dead."

For a long time, computer simulations of the universe had a problem: they predicted these giant trees should still be growing furiously, filling up with new stars. But in reality, they are quiet. Something must be turning off the water and fertilizer (gas) that feeds the stars. This is known as the "Quenching Problem."

Scientists know that the supermassive black holes at the center of these galaxies act like a sprinkler system, blasting energy out to stop the gas from cooling down and forming stars. But how exactly does this sprinkler work? Is it a high-pressure water jet? A heat lamp? Or something else?

The New Idea: The "Cosmic Ray" Sprinkler

This paper investigates a specific type of energy called Cosmic Rays. Think of cosmic rays not as water, but as a swarm of invisible, super-fast bees buzzing around the galaxy.

• The Theory: When the black hole eats gas, it shoots out these "bees" (cosmic rays). These bees carry energy and pressure. If they buzz around the gas cloud enough, they heat it up or push it away, preventing it from collapsing into new stars.
• The Mystery: We don't know exactly how many bees the black hole shoots out, or how fast they fly through the gas. Do they get stuck in a cage? Do they fly straight out? Do they bounce around randomly?

The Experiment: A Virtual Galaxy Lab

The researchers used a supercomputer to run a "virtual reality" simulation of four massive galaxies (called the FIRE-3 project). They didn't just run one simulation; they ran many, changing the rules of the game to see what happened.

They tested two main variables:

1. How many bees? (The injection efficiency): Did the black hole shoot out a few bees or a massive swarm?
2. How do the bees move? (The transport model):
   • Model A (Constant Diffusion): The bees fly through the gas like they are in a foggy room, bouncing off everything at a steady, predictable rate.
   • Model B (Variable Diffusion): The bees move differently depending on where they are. In some areas, the gas is thick and they get stuck; in others, they zoom through like they are on a highway.

The Results: The "Dead" Galaxies Look the Same, But the "Air" is Different

Here is the surprising part of the study:

1. The Outcome is the Same (The "Dead" Tree)
No matter which rules they used for the bees (whether they shot out a few or a million, or whether they flew fast or slow), all the simulations successfully stopped the galaxies from forming new stars.

• Analogy: It's like trying to stop a fire. Whether you use a garden hose, a fire extinguisher, or a bucket of sand, the fire goes out. The final result (a dead galaxy) looks the same to the naked eye. The black hole's "bees" successfully turned off the star formation in all scenarios.

2. The Inside Story is Totally Different (The "Air" in the Room)
While the galaxies looked the same on the outside, the environment surrounding them (the Circumgalactic Medium, or CGM) was completely different depending on the rules used.

• Analogy: Imagine two identical houses. In House A, the air inside is still and calm. In House B, the air is swirling violently with tornadoes. To an outsider looking at the roof, they look the same. But if you could fly inside, the experience would be worlds apart.
• The "Variable Diffusion" models created a very different distribution of gas and pressure compared to the "Constant" models. Some models trapped the cosmic rays close to the center, while others let them spread out far into the galaxy's halo.

Why Does This Matter?

The paper concludes that while we can't tell which model is correct just by looking at how many stars a galaxy has (because they all work), we can tell the difference by looking at the gas surrounding the galaxy.

• The Future Detective Work: Astronomers need to build better "cameras" to look at the invisible gas and the pressure waves around these galaxies. By comparing real telescope data with these different computer models, we might finally figure out exactly how the black hole's "bees" work.
• The Takeaway: The universe has a few different ways to turn off a galaxy's star factory. The black hole is very good at the job, but the specific "mechanics" of how it does it leave a unique fingerprint in the gas surrounding the galaxy.

Summary in One Sentence

The paper shows that while different theories about how cosmic rays from black holes stop star formation all produce "dead" galaxies, they create very different invisible environments around them, giving astronomers a new way to solve the mystery of how galaxies die.

Technical Summary

1. Problem Statement

The "quenching problem" remains a central discrepancy in galaxy formation theory: observations show that massive galaxies ($M_{\text{halo}} \gtrsim 10^{12.5} M_\odot$) are "red and dead" with suppressed star formation, yet standard simulations without specific feedback mechanisms predict excessive gas cooling and star formation (the "overcooling problem"). While Active Galactic Nuclei (AGN) feedback is widely accepted as the necessary mechanism to suppress this cooling, the specific physical channels (radiative, mechanical, or cosmic ray) and their coupling to the interstellar/circumgalactic medium (ISM/CGM) are not fully understood.

Specifically, the role of Cosmic Rays (CRs) as an AGN feedback channel is uncertain. While CRs can provide non-thermal pressure and heat gas via ionization, the theoretical parameter space is vast due to order-of-magnitude uncertainties in:

1. Injection Efficiency: The fraction of AGN accretion energy converted into CRs ($\epsilon_{\text{CR}}^{\text{BH}}$).
2. Transport Physics: How CRs scatter and diffuse through the magnetized plasma of the CGM, which depends on unresolved magnetic field structures.

This study aims to determine how varying these CR parameters affects the bulk properties of massive galaxies and the state of their circumgalactic medium.

2. Methodology

The authors utilized high-resolution, cosmological "zoom-in" simulations from the Feedback In Realistic Environments (FIRE-3) project, employing the GIZMO hydrodynamics solver in meshless-finite-mass mode.

• Simulation Suite: Four distinct dark matter halos ($M_{\text{halo}} \sim 10^{13} M_\odot$) were evolved (h206, h113, h029, h236).
• Physics Included:
  • Standard FIRE-3 physics: Star formation, stellar feedback (radiation, SNe, winds), and multi-band radiative cooling.
  • AGN Feedback: Explicit modeling of radiative, mechanical (jets/winds), and CR feedback from central supermassive black holes (SMBHs).
  • CR-MHD: Full spectrum propagation (MeV–TeV) of CRs along magnetic field lines, including scattering, Lorentz forces, and energy losses.
• Variable Parameters:
  1. Injection Efficiency ($\epsilon_{\text{CR}}^{\text{BH}}$): Varied by $\sim 1.5$ dex (from $10^{-4}$ to $3 \times 10^{-3}$).
  2. Transport Models:
     • CD (Constant Diffusion): A temporally/spatially constant power-law scattering rate ($\kappa_{\text{eff}} \propto R^{0.6}$).
     • VD (Variable Diffusion): A diffusion coefficient dependent on local ISM properties, calibrated to reproduce Milky Way observations (Voyager/AMS-02) and including an "external driving" turbulent term.
• Analysis: The study compared bulk galaxy scaling relations (Stellar Mass-Halo Mass, $M_{\text{BH}}-\sigma$), CR injection histories, and radial profiles of CR energy density, gas density, and effective diffusivity ($\kappa_{\text{eff}}$) at $z=0$.

3. Key Contributions

• Systematic Parameter Exploration: This is one of the first studies to systematically vary both AGN CR injection efficiency and transport physics within a multi-channel feedback framework (including radiative and mechanical modes) in high-resolution cosmological simulations.
• Decoupling Bulk vs. Halo Properties: The study demonstrates that while bulk galaxy properties (star formation rates, stellar mass) are robust across a wide range of CR parameters, the Circumgalactic Medium (CGM) properties are highly sensitive to these micro-physical assumptions.
• Non-Linear Feedback Response: The authors identify a non-linear relationship between CR injection efficiency and AGN growth, highlighting a transition between "feedback-regulated" (quasar-mode) and "fueling-regulated" (radio/maintenance-mode) phases driven by the responsiveness of the CR transport model.

4. Key Results

A. Bulk Galaxy Properties (Robustness)

• Quenching: All parameterizations successfully produced quenched galaxies with suppressed star formation rates (SFRs), generally falling 1–2 dex below the main sequence.
• Scaling Relations: The simulations reproduced observed empirical scaling relations:
  • Stellar Mass–Halo Mass (SMHM): All models fell within the observed scatter, indicating self-regulation.
  • $M_{\text{BH}}-\sigma$ Relation: Black hole masses and stellar velocity dispersions aligned with observational constraints (Greene et al. 2020).
• Insensitivity: The bulk properties were relatively insensitive to the specific CR transport model or injection efficiency, provided the feedback was active. Higher CR efficiency generally led to stronger suppression of star formation and lower final black hole masses.

B. CR Injection and AGN Responsiveness

• Transport Efficiency: The VD (Variable Diffusion) models were more efficient at suppressing black hole growth than the CD (Constant Diffusion) models, even when the VD models had lower injection efficiencies. This suggests that variable transport (confining CRs more effectively in the inner CGM) enhances feedback coupling.
• Timing vs. Aggregate Energy: The regulation of galaxy growth was driven more by the timing and responsiveness of the injection (chaotic bursts linked to BH fueling) rather than the total cumulative energy injected.
• Stellar vs. AGN CRs: In VD models, AGN CR energy was comparable to or slightly higher than stellar CR energy. In CD models, AGN CR energy could be up to 3.3 times higher than stellar CR energy, despite lower injection efficiencies, due to less efficient BH suppression in the CD runs.

C. Circumgalactic Medium (CGM) Properties (High Sensitivity)

• Radial Profiles: While bulk properties were similar, the CGM showed orders-of-magnitude differences:
  • CR Energy Density ($e_{\text{CR}}$): CD models exhibited power-law-like profiles. VD models showed flatter profiles, particularly at late times.
  • Non-Linear Ordering: In VD models, the $e_{\text{CR}}$ profile did not scale linearly with injection efficiency. The intermediate efficiency model (VDMidCR) often produced the highest CR energy density at large radii ($r \gtrsim 15$ kpc), while the highest efficiency model (VDHiCR) showed the most temporal variation and lowest density at $z=0$.
  • Gas Density: Gas density profiles flattened over time, correlating with CR pressure gradients. The VDHiCR model showed the strongest modulation, suggesting CRs drive outflows that dilute the CGM.
  • Diffusivity Troughs: The VD models exhibited "troughs" in effective diffusivity ($\kappa_{\text{eff}}$) that corresponded to local CR over-densities, indicating regions where CRs are trapped, creating local pressure gradients.

5. Significance and Future Directions

• Observational Constraints: The study concludes that while bulk galaxy properties cannot distinguish between different CR feedback models, multi-wavelength synthetic observations of the CGM can.
  • Proposed probes include Inverse Compton emission (tracing CR leptons), X-ray surface brightness, and the thermal Sunyaev–Zel'dovich (tSZ) effect.
  • Differences in CGM gas density and CR pressure profiles between models suggest these observables could constrain the "micro-physical" parameters of CR scattering and injection.
• Implications for Quenching: The results suggest that AGN feedback naturally produces the necessary "maintenance mode" to quench massive galaxies, but the specific physical mechanism (CR transport) leaves a distinct fingerprint on the halo gas.
• Future Work: The authors plan to generate synthetic observations (synchrotron, IR, UV, X-ray, SZ) from larger simulation samples to compare directly with stacked observational data (e.g., eROSITA) and to explore alternative CR injection sites (e.g., jet termination shocks) which may further alter quenching behavior.

In summary, the paper establishes that while CR feedback is a viable mechanism for quenching massive galaxies regardless of specific transport assumptions, the state of the circumgalactic medium is a critical diagnostic for understanding the physical nature of AGN feedback.