Two-moment cosmic ray transport in RAMSES

Joki Rosdahl 1∗, Yohan Dubois 2 † {}^{2}{\dagger}, Benoît Commerçon 1, Nimatou Diallo 2, Nai Chieh Lin 1, and Alexandre Marcowith 3 · 2025

Read Two-moment cosmic ray transport in RAMSES by Joki Rosdahl 1∗, Yohan Dubois 2 † {}^{2}{\dagger}, Benoît Commerçon 1, Nimatou Diallo 2, Nai Chieh Lin 1, and Alexandre Marcowith 3 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-19447.txt.

Introduction

Cosmic rays streaming through a galaxy push on its gas and drive winds, but simulations must model their transport honestly to capture that feedback. Joki Rosdahl and five co-authors implement a two-moment cosmic ray fluid solver inside RAMSES, a widely used astrophysics code. The method tracks both the rays' energy density and its flux in three dimensions, rather than energy alone. Diffusion, streaming, and advection are all handled, and a battery of multidimensional tests, including shocks, checks the solver's robustness.

The team then runs a full simulation of an isolated galactic disc, where supernova explosions launch galaxy-wide outflows. Comparing the new solver with the code's older one-moment version shows what the extra moment buys. The implementation gives modelers a sturdier tool for the feedback physics that shapes galaxies.

Catalog record

CREATORJoki Rosdahl 1∗, Yohan Dubois 2 † {}^{2}{\dagger}, Benoît Commerçon 1, Nimatou Diallo 2, Nai Chieh Lin 1, and Alexandre Marcowith 3
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE90 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-19447

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Two-moment cosmic ray transport in RAMSES ∗ E-mail: [email protected] † E-mail: [email protected] Joki Rosdahl 1∗, Yohan Dubois 2 † {}^{2}{\dagger}, Benoît Commerçon 1, Nimatou Diallo 2, Nai Chieh Lin 1, and Alexandre Marcowith 3 Abstract Cosmic rays (CRs) are an important source of feedback in a variety of astrophysical contexts. Magneto-hydrodynamical (MHD) simulations treating CRs as a fluid have shown that how their feedback operates is strongly dependent on their transport properties such as diffusion and streaming. In this paper we introduce the numerical implementation, in the adaptive-mesh-refinement MHD code Ramses , of the grey two-moment formulation of CR fluid dynamics, which follows the energy density and its associated three-dimensional flux. This method is tested for CR diffusion, streaming, and advection in a series of multi-dimensional tests including shocks to check the robustness and stability of this numerical two-moment CRMHD solver. We finally use the new two-moment CR implementation in a complex simulation of an isolated galactic disc producing galaxy-wide outflows launched by small-scale supernova explosions, and compare it with a previously existing one-moment formulation in the same code. keywords numerical methods, cosmic rays, galaxies 1 Introduction Cosmic rays (CRs) are a population of supra-thermal particles produced in astrophysical shocks corresponding to various sources, such as supernova remnants (SNRs) and novae, magnetospheres of pulsars, superbubbles in massive star clusters and, at the highest energies, jets and shocks associated with active galactic nuclei and galaxy clusters. Among these sources, SNRs remain the most widely recognised candidates for supplying the bulk of galactic CRs, as their energetics and frequency are sufficient to explain the measured CR spectrum on Earth ( Gabici et al. 2019 ) . For sufficiently strong shocks and through the process of diffuse shock acceleration ( Bell 1978 ; Drury 1983 ; Marcowith et al. 2016 , e.g.) , the CR energy extracted from the shock kinetic energy is of the order of 10% ( Helder et al. 2012 ; Caprioli & Spitkovsky 2014 , e.g.) . Due to their close coupling with magnetic fields, CR protons exert an effective pressure on the thermal plasma, which can provide a source of feedback for the gas on galactic scales (see Ruszkowski & Pfrommer 2023 , for a recent review) . The transport of CRs through magnetized plasma is governed by their interactions with the turbulent magnetic field. Because CRs gyrate around field lines, their motion is initially guided along the ordered magnetic field direction. However, small-scale fluctuations in the field scatter their pitch angle, effectively randomizing the direction of propagation. These fluctuations may be externally driven by turbulence, or they may be generated self-consistently by CRs themselves. In particular, CRs drifting faster than the local Alfvén speed excite the so-called gyro-resonant streaming instability ( Kulsrud & Pearce 1969 ) , which amplifies magnetic perturbations that scatter the CRs. This self-regulation enforces a CR bulk drift speed of order the Alfvén velocity relative to the gas in addition to plasma (CR) heating (resp. loss), a process referred to as CR streaming. The turbulent structure of galactic magnetic fields introduces an additional layer of complexity. Magnetic field lines are tangled by turbulence, and CRs tied to these lines undergo a form of random walk, resulting in transport that appears as diffusion on large scales. The effective transport is highly anisotropic: rapid along field lines, but much slower perpendicular to them. Describing this behavior, thus, requires modeling anisotropic diffusion coefficients that can vary with environment and magnetic geometry ( Shalchi 2009 ) . These processes not only control how CRs redistribute their pressure support across galactic environments, but also determine where they ultimately deposit their energy and momentum. A consequence of this transport is that CRs are displaced away from their production sites in regions of high gas density, such as star-forming clouds and SNRs, into more diffuse environments. Since CR losses, driven by Coulomb and hadronic interactions with the background plasma, operate most efficiently in dense gas, this outward redistribution significantly reduces the overall energy loss rate of the CR population. As a result, CRs can persist for long periods and accumulate with a galactic energy density that reaches rough equipartition with the thermal, turbulent, and magnetic components, with energy densities typically around \sim 1\,\rm eV\,cm^{-3} ( Boulares & Cox 1990 ) . This ability to act as a long-lived, mobile reservoir of non-thermal energy makes CRs a potentially important driver of feedback on galactic scales. Interest in CRs as a source of astrophysical feedback has grown considerably, especially following hydrodynamical simulations that highlighted their potential to drive galactic winds and regulate star formation. Early studies employed simplified models of CR transport via isotropic diffusion or streaming ( Jubelgas et al. 2008 ; Uhlig et al. 2012 ; Booth et al. 2013 ; Salem et al. 2014 ) . More recent work incorporating anisotropic diffusion and streaming has refined our understanding of CR-driven feedback, demonstrating that the magnitude of diffusion and streaming critically influences wind launching and galactic-scale regulation ( Hanasz et al. 2013 ; Girichidis et al. 2016 ; Girichidis et al. 2018 ; Simpson et al. 2016 ; Pakmor et al. 2016 ; Ruszkowski et al. 2017 ; Chan et al. 2019 ; Dashyan & Dubois 2020 ; Hopkins et al. 2021 ; Hopkins et al. 2021 ; Farcy et al. 2022 ; Farcy et al. 2025 ; Martin-Alvarez et al. 2023 ; Rodríguez Montero et al. 2024 ) . Beyond galactic-scale feedback, CRs are believed to play a key role in several other processes: enhancing momentum deposition in SNRs ( Diesing & Caprioli 2018 ; Rodríguez Montero et al. 2022 ) , shaping the density and turbulence structure of the interstellar medium ( Commerçon et al. 2019 ; Dubois et al. 2019 ; Nuñez-Castiñeyra et al. 2022 ; Simpson et al. 2023 ; Sampson et al. 2025 ) , and potentially contributing to the reheating of the interstellar medium ( Wiener et al. 2013 ) or of galaxy cluster cores ( Sijacki et al. 2008 ; Ruszkowski et al. 2017 ; Ehlert et al. 2018 ; Yang et al. 2019 ; Su et al. 2021 ; Beckmann et al. 2022 ) . Accurately capturing these processes in numerical simulations requires robust models of CR transport. In the widely used “one-moment” formulation, the CR energy density is evolved via an advection–diffusion equation, with transport coefficients encoding diffusion and streaming. However, this approach faces significant numerical challenges. Diffusion is governed by a parabolic equation, which imposes a timestep constraint scaling as \Delta t_{\kappa}=\Delta x^{2}/(2\kappa) , where \Delta x is the volume element width and \kappa is the diffusion coefficient. This timestep is often orders of magnitude smaller than the hydrodynamical timestep, \Delta t_{\rm h}\approx\Delta x/v_{\rm max} , set by the Courant condition. Streaming makes matters worse: because the streaming velocity is discontinuous at extrema of the CR energy density, the corresponding stability condition scales as \Delta t_{\rm s}\propto\Delta x^{3} . At high resolution, these restrictions render explicit schemes impractical. To circumvent this, several strategies have been proposed, including subcycling, implicit or semi-implicit solvers ( Sharma & Hammett 2011 ; Dubois & Commerçon 2016 ; Dubois et al. 2019 ; Pakmor et al. 2016 ) , or artificial smoothing of the streaming

Read the complete text