[Paper Review] The impact of cosmic rays on the interstellar medium and galactic outflows of Milky Way analogues

Francisco Rodríguez Montero, Sergio Martin-Alvarez+4 more|arXiv (Cornell University)|Jul 25, 2023

Galaxies: Formation, Evolution, PhenomenaPhysics and Astronomy3 citations

TL;DR

This study investigates cosmic ray (CR) feedback in a cosmological zoom-in simulation of a Milky Way-like galaxy, showing that CRs reduce stellar mass by factors of ~10 at high redshift and ~4 at cosmic noon by suppressing star formation and enhancing supernova efficiency. CR pressure gradients drive multi-phase galactic outflows dominated by warm ionized gas, with significant cold and neutral phases at high redshift, offering testable predictions for JWST observations.

ABSTRACT

During the last decade, cosmological simulations have managed to reproduce realistic and morphologically diverse galaxies, spanning the Hubble sequence. Central to this success was a phenomenological calibration of the few included feedback processes, whilst glossing over higher complexity baryonic physics. This approach diminishes the predictive power of such simulations, preventing to further our understanding of galaxy formation. To tackle this fundamental issue, we investigate the impact of cosmic rays (CRs) and magnetic fields on the interstellar medium (ISM) and the launching of outflows in a cosmological zoom-in simulation of a Milky Way-like galaxy. We find that including CRs decreases the stellar mass of the galaxy by a factor of 10 at high redshift and $\sim 4$ at cosmic noon, leading to a stellar mass to halo mass ratio in good agreement with abundance matching models. Such decrease is caused by two effects: i) a reduction of cold, high-density, star-forming gas, and ii) a larger fraction of SN events exploding at lower densities, where they have a higher impact. SN-injected CRs produce enhanced, multi-phase galactic outflows, which are accelerated by CR pressure gradients in the circumgalactic medium of the galaxy. While the mass budget of these outflows is dominated by the warm ionised gas, warm neutral and cold gas phases contribute significantly at high redshifts. Importantly, our work shows that future JWST observations of galaxies and their multi-phase outflows across cosmic time have the ability to constrain the role of CRs in regulating star formation.

Motivation & Objective

• To investigate the role of cosmic rays (CRs) and magnetic fields in regulating star formation and driving galactic outflows in a cosmological Milky Way-like galaxy simulation.
• To assess how CR feedback alters the interstellar medium (ISM) structure and thermodynamics compared to standard hydrodynamic (HD) and magnetohydrodynamic (MHD) simulations.
• To determine whether CR-driven feedback can naturally reproduce observed stellar mass–halo mass relations without ad-hoc calibration of feedback processes.
• To explore the multi-phase nature of galactic outflows and the contribution of CR pressure gradients in accelerating outflows in the circumgalactic medium (CGM).

Proposed method

• Conducted a cosmological zoom-in simulation of a Milky Way-like galaxy using the RAMSES-Adaptive Mesh Refinement (AMR) code with self-consistent CR transport and energy deposition.
• Implemented a CR transport model that includes diffusion, streaming, and energy loss processes, coupled with thermal and magnetic pressure dynamics.
• Compared three simulation variants: hydrodynamics (HD), magnetohydrodynamics (MHD), and CR-MHD (CRMHD), to isolate CR feedback effects.
• Tracked the evolution of gas phases (cold, warm neutral, warm ionized, hot) in the ISM and outflows, analyzing their mass loading factors and thermodynamic properties.
• Used spatially resolved analysis of SN explosion sites to quantify changes in ambient density and cooling efficiency due to CR feedback.
• Evaluated outflow morphology and acceleration mechanisms, distinguishing contributions from CR pressure gradients versus thermal pressure.

Figure 1: Overview of the CRMHD model of the nut simulation at $z=1.5$ . Background: Colour composite of the zoom-in region centred on the nut galaxy, with gas density (silver blue), gas temperature (orange) and CR energy density (yellow), showing 600 kpc along the x-axis. Three cold filaments are f

Experimental results

Research questions

• RQ1 How do cosmic rays affect the stellar mass assembly of Milky Way-like galaxies across cosmic time?
• RQ2 What is the impact of CR feedback on the multi-phase structure and thermodynamics of the interstellar medium?
• RQ3 How do CR pressure gradients contribute to the acceleration and morphology of galactic outflows in the circumgalactic medium?
• RQ4 To what extent do CRs alter the efficiency of supernova feedback by modifying explosion environments and cooling losses?
• RQ5 Can CR-driven outflows reproduce observed multi-phase outflow properties detectable by the James Webb Space Telescope (JWST)?

Key findings

• Including cosmic rays reduces the stellar mass of the simulated Milky Way-like galaxy by a factor of ~10 at high redshift and ~4 at cosmic noon, aligning with abundance matching models.
• CR feedback suppresses cold, high-density, star-forming gas and increases the fraction of supernovae exploding in lower-density environments, enhancing feedback efficiency.
• Galactic outflows in the CRMHD simulation are dominated in mass by the warm ionized phase (9000–10⁵ K), with significant contributions from cold and warm neutral phases at high redshift.
• Outflows in the CRMHD simulation exhibit higher mass loading factors than HD and MHD runs, with biconical hot outflows and filamentary, clumpy cold phases embedded within them.
• CR pressure gradients provide the primary acceleration mechanism for outflows in the warm and cold phases, while thermal pressure support remains negligible.
• The CRMHD simulation predicts that CR-driven outflows will display a multi-phase structure with clumpy, filamentary cold components embedded in biconical hot outflows—testable with upcoming JWST observations.

Figure 2: Projections of the nut galaxy for the three main models explored in this work: HD (top 2 rows), MHD (middle 2 rows), and CRMHD (bottom 2 rows), generated at $z=1.5$ . For each model, the plot includes density-weighted projections of gas density (first column), gas temperature (second colum