Properties of Galactic Outflows Driven by Starburst at Cosmic Noon: Insights from Hydrodynamical Simulations

arXiv:2602.07592 · astro-ph.GA, astro-ph.SR · Submitted 2026-02-07 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: Today's paper: "Properties of Galactic Outflows Driven by Starburst at Cosmic Noon".

Jocelyn: Galactic outflows driven by starbursts in low-mass galaxies at cosmic noon are investigated using high-resolution 3D hydrodynamical simulations to provide insights into their multiphase structure and scaling relations.

Vera: First, who's behind it and why it matters.

Title and authors: Vera: So, to get into the summary of "Properties of Galactic Outflows Driven by Starburst at Cosmic Noon: Insights from Hydrodynamical Simulations," they detail how their simulations successfully reproduce the complex, multiphase structure seen in M82's outflows.

Jocelyn: They explain that this multiphase nature means they are tracking how gas gets heated into different temperature states, like cool, warm, and hot gas as it moves out of the galaxy <ref:2602.07592#pg1>.

Subrahmanyan: This multiphase structure is quite important because it tells us that stellar feedback doesn't just push a uniform cloud; instead, it creates a complex medium with various thermal states within the outflow <ref:2602.07592#pg1>.

Vera: They specifically model these starbursts to last for about twenty to thirty million years, and they find peak star formation rates can range from two solar masses per year up to sixty-eight solar masses per year depending on the initial setup <ref:2602.07592#pg0>.

Jocelyn: And when they report on the resulting outflows themselves, they document velocities spanning from several hundred to a couple of thousand kilometers per second and mass outflow rates between zero point three and twenty solar masses per year <ref:2602.07592#pg0>.

Subrahmanyan: The most interesting part is how the mass loading factors vary, as they can range from as low as zero point two four up to six point two six, showing the wide variety in how efficiently these galaxies eject their gas into the surrounding medium <ref:2602.07592#pg0>.

The paper's summary: Vera: Now, when we talk about what the authors suggest for next steps or ways to improve this work, they point out several areas where things could get even more detailed and thorough. They aren't just stopping at the current results from their "Properties of Galactic Outflows Driven by Starburst at Cosmic Noon: Insights from Hydrodynamical Simulations."

Jocelyn: They suggest that while their twenty to thirty million year snapshot is useful, running the simulations over a full evolutionary timescale would provide a better picture of how these outflows actually change as time goes on <ref:2602.07592#pg1>.

Subrahmanyan: Theoretically, they admit that there are still major uncertainties concerning the exact fraction of core-collapse supernovae per massive star, which could significantly impact how much energy is actually injected into the interstellar medium <ref:2602.07592#pg1>.

Vera: They also mention that because these simulations rely on idealized initial conditions, like testing gas fractions ranging from thirty percent to eighty-five percent, the real world likely has more complex variations in galaxy structure than the current models can perfectly capture <ref:2602.07592#pg0>.

Jocelyn: The authors also mention that they refined how they handle supernova feedback, specifically by using the RM ratio to ensure momentum and energy conservation stay consistent in their results <ref:2602.07592#pg1>.

The paper's improvements: Vera: So, wrapping up our discussion on "Properties of Galactic Outflows Driven by Starburst at Cosmic Noon: Insights from Hydrodynamical Simulations," the paper confirms that these simulations align quite well with many of the real galactic outflows we observe across cosmic time.

Jocelyn: They confirm that the cool phase gas is indeed what dominates the outflow, and they have established scaling relations between those outflow properties and characteristics of the host galaxy, even if some of those correlations aren't as straightforward as we might hope <ref:2602.07592#pg1>.

Subrahmanyan: The implications for galaxy evolution are that we need to be cautious when comparing simulation results with observations because the paper points out differences in how those measurements are made, which is a crucial distinction for theoretical modeling <ref:2602.07592#pg2>.

Vera: It really gives us a solid set of constraints on the physics governing starburst winds, and it opens up new avenues for testing different feedback models that we've been exploring <ref:2602.07592#pg1>.

Jocelyn: We’ve got a good foundation here for understanding how these outflows shape the environment around low-mass galaxies specifically at cosmic noon <ref:2602.07592#pg1>.

Subrahmanyan: I think this work helps us zero in on the specific physics governing these winds during that critical epoch when galaxy growth was peaking, providing a better benchmark for theoretical models <ref:2602.07592#pg1>.

Vera: That's all we have time for today on "Properties of Galactic Outflows Driven by Starburst at Cosmic Noon: Insights from Hydrodynamical Simulations." We’ll be back with more papers soon.

Conclusion: Vera: So, we've been deep in "Properties of Galactic Outflows Driven by Starburst at Cosmic Noon: Insights from Hydrodynamical Simulations," and what we found is that these models give us a pretty solid handle on how gas gets expelled during those intense starburst phases.

Jocelyn: Exactly, Vera. The multiphase structure they reproduced in M82's outflows really shows us the complexity of these winds, which is fascinating to see confirmed by the data we collect from surveys like ours <ref:2602.07592#pg1>.

Subrahmanyan: From a theoretical viewpoint, this helps us connect the energy injection from stellar feedback directly to measurable galactic outflows at a specific cosmic epoch, which is vital for our larger models of galaxy assembly.

Vera: And that variety in mass loading factors they found, from zero point two four up to six point two six, really tells us there's a lot of physical flexibility in how these galaxies are responding to their star formation <ref:2602.07592#pg0>.

Jocelyn: I think that range is significant because it means we can’t just assume a single outcome for every low-mass galaxy; there's real diversity there that we need to account for in our observational interpretations.

Subrahmanyan: That diversity reinforces the idea that environment and internal processes play a huge role in determining how much gas is driven out, which feeds back into how we model structure formation across cosmic time.

Vera: So, to recap, the paper provides detailed insights into the multiphase structure and scaling relations of starburst-driven galactic winds in low-mass galaxies at cosmic noon <ref:2602.07592#pg0>.

Jocelyn: And it sets up a clear path forward by showing us where we need to focus our observational efforts when looking for these types of structures across the sky <ref:2602.07592#pg1>.

Subrahmanyan: Ultimately, this work offers a stronger benchmark for theoretical models trying to describe the transition from gas-rich disk growth to more evolved galaxy states.

Vera: It really gives us a strong set of constraints on the physics involved in these processes and opens up new avenues for testing how different feedback mechanisms behave in these environments.

Jocelyn: I think understanding this better will help us better interpret the spectral signatures we see from high-redshift galaxies that might be experiencing similar starbursts.

Subrahmanyan: We should keep looking at how these scaling relations evolve as we move further back in time, because that’s where the real story of galaxy growth lies.

School of Physics and Astronomy, Sun Yat-sen University

astro-ph.GA, astro-ph.SR

Submitted: 2026-02-07

Updated: 2026-02-07

Comments: 33 pages, 12 figures, accepted for publication in ApJ

Journal ref: Huan Chen et al 2026 ApJ 1000 158

DOI: 10.3847/1538-4357/ae45a8

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 77/100

The gist: Galactic outflows driven by starbursts in low-mass galaxies at cosmic noon are investigated using high-resolution 3D hydrodynamical simulations to provide insights into their multiphase structure and

Key concepts

Multiphase Outflows
These are galactic winds that contain different types of gas at various temperatures, such as hot, warm, and cool phases. The simulation tracks these distinct components to understand the complex structure of the outflow driven by star formation.
Stellar Feedback
This refers to the physical processes where stars and their remnants influence the surrounding interstellar medium. This includes radiation pressure from massive stars, stellar winds from these stars, and explosions from core-collapse supernovae that inject energy back into the gas.
Mass Loading Factor (ηM)
This factor measures how much mass is ejected by a galaxy's outflow relative to the mass of the galaxy itself. A higher value indicates a more powerful outflow, showing how efficiently starburst activity drives material out of the system.
Cosmic Noon
This refers to the epoch in cosmic history (redshift z~1-2) when star formation rates peaked in low-mass galaxies. The study focuses on this specific time and environment to understand galaxy evolution during that crucial period.

Terminology

Summary

Galactic outflows driven by starbursts in low-mass galaxies at cosmic noon are investigated using high-resolution 3D hydrodynamical simulations to provide insights into their multiphase structure and scaling relations.

How it works

The study employs a framework capable of reproducing the multiphase outflows observed in M82, focusing on starbursts lasting 20–30 Myr with peak star formation rates of 2–68 M⊙ yr−1. The simulations are performed on 14 idealized isolate low-mass disk galaxies at redshifts between z ∼ 1 and z ∼ 2, varying gas fractions from 30% to 85%. These models incorporate physical modules for radiative cooling, star formation, and stellar feedback.

Key physical processes included in the simulations are:

: Radiative cooling and heating are computed using the Grackle library with metallicity-dependent functions. The initial gas metallicity is set to 0.02 Z⊙, tracked via passive scalar fields. The star formation is modeled using a sink particle module, where stellar mass conversion occurs when a sink particle exceeds a mass threshold (Mcnv). Stellar feedback includes radiation pressure and heating, stellar winds based on the metallicity-dependent model of J. Vink & A. Sander (2021), and core-collapse supernovae (CCSNe). The treatment for supernova feedback is determined by the ratio RM = MSNR/Msf, where injection methods are chosen based on this ratio to ensure conservation of momentum and energy.

Starburst Characteristics

The simulations produce starbursts with a peak SFR of 2–68 M⊙ yr−1 within the first 1–8 Myr, followed by a decline due to stellar feedback and gas depletion. The peak star formation rate is dependent on the initial gas fraction, total gas disk mass, and disk compactness. Higher initial gas fractions produce larger SFR peaks, while more compact disks drive more intense starbursts due to higher gas column densities and rapid gravitational collapse. The framework can reproduce more intense starbursts when the gas disk is more compact, as seen in the Z2 C simulation which peaks at 68.0 M⊙ yr−1 around t ∼ 1 Myr.

Outflow Properties

The simulations produce galactic-scale multiphase outflows with velocities ranging from 50–1000 km s−1, mass outflow rates of 0.3–20 M⊙ yr−1, and mass loading factors ηM of 0.24–6.26. The cool phase (800 < T ≤ 2×10 4 K) dominates the outflow, and properties are broadly consistent with observations, with typical cool phase velocities around 100–400 km s−1 and mass loading factors ranging from 0.25 to 4 for the total outflow.

Scaling Relations and Comparison to Observations

The study explores correlations between outflow properties and host galaxy characteristics. The results show that outflow velocity exhibits no clear correlation with stellar mass or SFR, whereas the mass outflow rate and mass loading factor generally decrease with increasing stellar mass and increase with SFR. The total mass loading factor exhibits a negative trend with stellar mass, both in unweighted fitting and fitting weighted by stellar mass and gas fraction, which is broadly consistent with trends reported in previous simulations.

The total ηM values in the simulated range of 109–1010 M⊙ are lower than those in EAGLE (by 0.50 dex) and TNG50 (by 0.86 dex), but slightly higher than FIRE-2 by 0.06 dex, although accounting for methodological differences reduces these gaps to 0.2–0.4 dex, suggesting discrepancies may arise from inconsistent definitions and measurement methods rather than fundamental physics differences alone.

Temporal and Spatial Evolution

Outflow properties evolve substantially in time and space; the simulations are halted at 30 Myr, which is relatively short compared to the full evolutionary timescale of galactic outflows. The mass outflow rate at a radius of 3 kpc increases with time, likely due to the growing volume and density of the outflow. The inclusion of gas return processes accelerates the increase in ηM between t = 18 and t = 30 Myr, although it results in only a modest change in the final ηM. The radial profiles show that while properties can vary significantly with radius within a single simulation, the mass outflow rate and mass loading factor of the cool and warm phases change moderately between r = 1.5 and 3.0 kpc.

Caveats

The study notes that due to computational constraints, simulations are limited to a narrow stellar mass range of ∼ 109–1010 M⊙, a short duration of 30 Myr, and idealized initial conditions.

Improvements for AI systems

As a fastidious and diligent researcher, I have analyzed this paper on Properties of Galactic Outflows Driven by Starburst at Cosmic Noon: Insights from Hydrodynamical Simulations. The core scientific contributions lie in providing high-resolution, multiphase hydrodynamical simulations that bridge the gap between theoretical models and observational constraints for starburst-driven galactic winds.

Here are specific improvements to AI systems, categorized by the capabilities they can gain:

)

)

AI System Improvements: Specific Capabilities Enabled by This Research

  1. Enhanced Simulation Fidelity and Parameter Space Exploration (for Hydrodynamical Models):

  2. Improved Observational Data Interpretation and Systematic Error Mitigation (for Machine Learning/Statistical Analysis):

  3. Advanced Scaling Relation Prediction and Uncertainty Quantification (for Galaxy Evolution Modeling):

  4. Enhanced Simulation Fidelity and Parameter Space Exploration (for Hydrodynamical Models):

The paper details a complex, multiphase simulation framework involving stellar feedback mechanisms (radiation pressure, stellar winds, CCSNe), radiative cooling/heating modules (Grackle library), and sophisticated sink-particle models for star formation.

AI Systems can be improved by:

  • Developing AI agents capable of rapidly exploring the vast parameter space defined in Table 1 (e.g., varying initial gas fractions from 30% to 85%, different bulge/disk mass ratios, and redshift-dependent halo concentrations).

  • Creating generative models that predict outflow morphology (Figure 4) and phase distribution (Figure 2) based on input galaxy parameters, allowing researchers to quickly visualize the impact of specific physical settings (like gas return processes or supernova clustering) without running full simulations.

  • Implementing AI to rapidly test What-if scenarios regarding subgrid physics, such as varying the supernova energy injection scheme (e.g., testing RM ratios for different feedback regimes) and measuring their effect on critical outputs like mass loading factors or outflow velocity profiles (Figure 10).

  1. Improved Observational Data Interpretation and Systematic Error Mitigation (for Machine Learning/Statistical Analysis):

The paper extensively discusses the apple-to-apple comparison challenges, noting that observational inferences are highly sensitive to assumptions about geometry, temperature, and measurement radius (e.g., density/temperature assumptions in Section 4.1).

AI Systems can be improved by:

  • Building Machine Learning classifiers trained on the systematic errors identified: An AI system could ingest observational spectral data (like Hα or OIII line ratios) and predict the likely systematic biases (e.g., overestimation of velocity due to beam smearing or temperature assumptions) based on known physical dependencies discussed in Section 4.1 and 4.2.

  • Developing sophisticated uncertainty quantification models that integrate multiple sources of error (methodological, sample size, temporal snapshot) to provide a probabilistic range for derived quantities like the mass loading factor (e.g., predicting the likely scatter in the total ηM–M∗ relation).

  • Creating AI tools that perform systematic comparisons between simulation results and observations by automatically adjusting measurement parameters (like opening angle or radius) until a statistically optimal agreement is found, helping to quantify exactly how much discrepancy is due to methodology versus fundamental physics.

  1. Advanced Scaling Relation Prediction and Uncertainty Quantification (for Galaxy Evolution Modeling):

The paper highlights the ambiguity in scaling relations (e.g., velocity vs. M/SFR) due to limited sample sizes and short simulation durations (30 Myr).

AI Systems can be improved by:

  • Developing Bayesian Inference models that use the simulation results as a prior, allowing the AI to infer the most probable functional form of scaling relations (e.g., ηM–M) while explicitly accounting for the systematic biases introduced by different measurement radii or definitions (as discussed in Section 4.2).

  • Creating predictive tools that forecast how scaling relations might change if key physical processes are modified (e.g., predicting the effect of including cosmic ray feedback or Type Ia SNe, which are currently omitted) based on the established trends in the current simulation suite.

  • Implementing AI to perform weighted linear fits (as done in Section 3.5) by dynamically adjusting weighting factors based on galaxy properties (like gas fraction scatter) to derive more robust scaling relations than those derived from unweighted fits alone.

Sources

Related papers