pith. machine review for the scientific record. sign in

arxiv: 2604.11898 · v1 · submitted 2026-04-13 · 🌌 astro-ph.HE

Recognition: unknown

Beyond the Diffusion Coefficient: Propagators and Memory in Cosmic Ray Transport

Naixin Liang , S. Peng Oh
Authors on Pith no claims yet

Pith reviewed 2026-05-10 15:49 UTC · model grok-4.3

classification 🌌 astro-ph.HE
keywords cosmic ray transportpropagatormemory kernelMontroll-Weiss formalismdiffusion coefficientmultiphase mediumparticle tracingsuperdiffusion
0
0 comments X

The pith

Cosmic ray transport requires the full position probability propagator to capture memory effects ignored by a single diffusion coefficient.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Standard models use only a diffusion coefficient to describe how cosmic rays spread, but this tracks only the variance of positions and misses the complete probability distribution of where particles end up. Different mechanisms can produce the same diffusion rate yet yield different distributions and escape behaviors, especially in structured, multiphase, and time-varying astrophysical environments such as supernova remnants or molecular clouds. The paper develops a framework built around the propagator P(x,t), the probability a particle is at position x after time t, or its Fourier-Laplace transform P(k,s); this object is statistically complete and directly reveals memory kernels arising when trapping or phase changes are averaged over. Applying the Montroll-Weiss formalism to trajectories shows how to extract these kernels, represent them efficiently, and simulate transport in media where slow regions control escape and evolving traps prevent reaching a static diffusive limit.

Core claim

The transport process is fully characterized by the propagator P(x,t) or its Fourier-Laplace transform P(k,s), which encodes all statistical information and exposes non-local memory effects in the flux; using the Montroll-Weiss formalism, memory kernels are recovered from trajectories, slow regions in multiphase media are shown to regulate escape without dominating residence time, and an accelerated Monte Carlo method demonstrates that dynamically evolving trapping structures prevent the system from always reaching the static long-time diffusive limit.

What carries the argument

The propagator P(k,s), the Fourier-Laplace transform of the particle position probability distribution P(x,t), which encodes the complete transport process and allows extraction of memory kernels via the Montroll-Weiss formalism.

If this is right

  • Slow regions in multiphase media regulate cosmic ray escape even when they do not dominate the total residence-time budget.
  • Memory kernels can be measured directly from trajectories and represented compactly with a Prony expansion.
  • The static long-time diffusion limit need not be reached when trapping structures evolve while particles are still sampling them.
  • An accelerated Monte Carlo method enables efficient coarse-grained transport simulations in such media.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The framework could supply improved time-dependent closures for cosmic-ray MHD simulations by incorporating explicit memory kernels.
  • Particle-tracing measurements in simulations of supernova remnants or pulsar wind nebulae could directly test whether observed transport variations match predicted propagators.
  • The approach may generalize to other transport problems with unresolved phases, such as energetic particle motion in turbulent plasmas.

Load-bearing premise

The Montroll-Weiss continuous-time random walk formalism can be applied without modification to cosmic-ray trajectories in realistic multiphase, time-dependent astrophysical media.

What would settle it

Extracting a memory kernel from particle trajectories in a controlled multiphase simulation with time-evolving traps and then finding that the resulting propagator fails to reproduce the directly measured position distributions or escape times in the same simulation.

Figures

Figures reproduced from arXiv: 2604.11898 by Naixin Liang, S. Peng Oh.

Figure 1
Figure 1. Figure 1: Green’s functions 𝑃(𝑥, 𝑡) for three benchmark CTRWs: standard diffusion (top), subdiffusion (middle), and superdiffusion (bottom). Solid curves show the inverse Fourier–Laplace transform of the Montroll–Weiss propagator constructed from the prescribed jump and waiting-time statistics; dashed curves show the known analytic Green’s functions; points show direct Monte Carlo realizations of the same CTRWs. The… view at source ↗
Figure 2
Figure 2. Figure 2: The MW propagator against wavenumber k, measured from Monte Carlo simulations (dot) and constructed by analytics (solid line, equations 15–17). Top: standard diffusion (𝛼 = 2, 𝛽 = 1); Middle: superdiffusion (𝛼 = 1.5, 𝛽 = 1); Bottom: subdiffusion (𝛼 = 2, 𝛽 = 0.7). All panels show agreement at three different Laplace frequencies 𝑠. memory of previously sampled environments. To make this concrete, we now cons… view at source ↗
Figure 3
Figure 3. Figure 3: Statistics from the Monte Carlo (MC) data and the inverse Laplace transform of the MW propagator measured by Eq.20. Top: Probability den￾sity function 𝑃(𝑥, 𝑡) for standard diffusion at various diffusivities. The MC data (stepped histogram) show agreement with the measured MW inversion (solid) and theory (dotted). The Gaussian approximation (dashed) is provided for comparison, highlighting the difference in… view at source ↗
Figure 4
Figure 4. Figure 4: The Monte-Carlo results across different correlation times 𝑡corr. Top: MSD ⟨𝑥 2 (𝑡) ⟩ as a function of time 𝑡. The simulated data for varying 𝑡corr values (solid) overlap entirely, exhibiting the linear growth characteristic of standard diffusion, in perfect agreement with the theoretical prediction 2⟨𝐷⟩𝑡 (dashed). Bottom: Convergence to Gaussianity measured by the excess kurtosis 𝛼2 (𝑡). For finite correl… view at source ↗
Figure 5
Figure 5. Figure 5: Memory structure in the finite correlation time simulation (§2.3), which has a linear MSD but non-trivial memory effect. Top: The propagator in Fourier-Laplace space, 𝑃(𝑘, 𝑠), at small wavenumber 𝑘 = 0.1, 0.01. Bottom: Memory Kernel 𝐾ˆ (𝑠), which acts as a frequency-dependent diffusion coefficient. This frequency dependence confirms the non-Markovian nature of the transport, which persists even when MSD is… view at source ↗
Figure 6
Figure 6. Figure 6: The memory kernel 𝐾˜ (𝑠) in the Laplace domain for different subdiffusion stability parameters 𝛼 = 0.3, 0.5, and 0.7, showing the power￾law scalings characterized by 1 − 𝛼 that matches theoretical predictions in Eq. 42. The upper limit in s is roughly 1/10𝜏0, where particles have jumped multiple times. The lower limit in s is roughly 1/𝑇max, determined by the simulation time. future evolution is independen… view at source ↗
Figure 7
Figure 7. Figure 7 [PITH_FULL_IMAGE:figures/full_fig_p009_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Prony reconstruction of the time-domain memory kernel 𝐾 (𝑡) for subdiffusive CTRWs. The measured Laplace-space kernel 𝐾˜ (𝑠) is fit with Eq. (59), yielding a finite sum of exponential modes (Eq. 58) which ap￾proximates 𝐾 (𝑡). The reconstructed kernels accurately recover the expected power-law scaling 𝐾 (𝑡) ∝ 𝑡 − (2−𝛼) for 𝛼 = 0.3, 0.5, and 0.7 over several decades in time with only 𝑁 = 10 modes, demonstrat… view at source ↗
Figure 10
Figure 10. Figure 10: Comparison of the running diffusion coefficient 𝜅 (𝑡) obtained from coarse-grained accelerated MC and full MC simulations, with constant patch sizes and 𝜅ℎ/𝜅𝑙 = 10. 𝜅 (𝑡) starts from the arithmetic mean 𝜅𝐴 at early times and gradually decreases to the harmonic mean 𝜅𝐻 at long times (dotted lines). Vertical gray dashed lines mark the characteristic diffusion timescales 𝑙 2 /𝜅ℎ and 𝑙 2 /𝜅𝑙 . two-cell dynami… view at source ↗
Figure 9
Figure 9. Figure 9: Validation of the coarse-grained two-cell first-passage model. Top: Laplace-transformed exit-time distributions 𝑈𝑙 (𝑠) and 𝑈ℎ (𝑠) from Eq. (66) compared with Monte Carlo measurements for several diffusivity contrasts 𝜅ℎ/𝜅𝑙 at fixed cell sizes. Bottom: Monte Carlo measurements of the slow-side exit probability 𝛼𝑙 compared with the analytic prediction in Eq. (65). Here 𝑅𝑙 = 𝐿𝑙/(2𝜅𝑙 ) and 𝑅ℎ = 𝐿ℎ/(2𝜅ℎ ) are t… view at source ↗
Figure 11
Figure 11. Figure 11: The MC results compared to theoretical prediction in the Laplace domain (𝑠). Top: The MW propagator 𝑃(𝑘 = 0.1, 𝑠) shows agreement. Mid￾dle: The waiting time distribution 𝜓(𝑠) matches at small 𝑠, though it diverges at large 𝑠 (short times). Bottom: The Laplace space kernel 𝐾 (𝑠) matches at small 𝑠 near 𝑘𝐻, but diverges at large 𝑠 as the MC data approaches 𝑘𝐴. From equation 76 and associated discussion, thi… view at source ↗
Figure 12
Figure 12. Figure 12: Normalized running diffusion coefficient for variable-cell media with patch lengths drawn from 𝑃(ℓ ) ∝ ℓ −𝛼. Solid curves show Monte Carlo results and dotted curves show the scaling estimate above. For a uniform initial distribution, transport interpolates from the short-time arithmetic mean 𝜅𝐴 to the long-time harmonic mean 𝜅𝐻. Initial conditions restricted to fast (dashed) or slow (dash-dotted) cells in… view at source ↗
Figure 14
Figure 14. Figure 14: Running diffusion coefficient for quenched and annealed environ￾ments. Top: both cases interpolate from the short-time arithmetic mean 𝜅𝐴 to the long-time harmonic mean 𝜅𝐻. In the parameter range shown, the annealed curve departs only modestly from the quenched one, entering the crossover slightly earlier. Bottom: the ratio 𝜅ann (𝑡)/𝜅quench (𝑡) isolates this difference. Both cases share the same short-tim… view at source ↗
Figure 15
Figure 15. Figure 15: Recovery of coarse-grained transport statistics in the variable-cell medium. Top: propagator 𝑃(𝑘, 𝑠) at fixed 𝑠 = 0.1. The quenched Monte Carlo result agrees with the quenched theory, while the annealed propaga￾tor remains close at small 𝑘 and deviates only mildly at larger 𝑘. Middle (quenched only): joint transform Φ(𝑘, 𝑠) at fixed 𝑠 = 0.1. The directly measured kernel agrees with the reconstruction obta… view at source ↗
Figure 17
Figure 17. Figure 17: The slow fractions and resistance dominance across different parameters. Top: Grouped bar chart shows the residence fraction ( 𝑓 (𝑠) 𝜏 ), resistance fraction ( 𝑓 (𝑠) 𝑅 ), and total slow-phase volume fraction ( 𝑓slow) across sweeps of the parameters 𝑞𝑅, 𝑞𝜏, 𝛼, and Δ. Bottom: The dominance 𝐷 = log10 ( 𝑓𝑅/ 𝑓𝜏 ) as a function of the swept parameter values. All tests lead to positive values (𝐷 > 0), indicating… view at source ↗
Figure 16
Figure 16. Figure 16: Diagnosing separability with the MW kernel 𝑀(𝑘, 𝑠) and the joint transform Φ(𝑘, 𝑠). Top: in the constant-cell case, normalizing 𝑀(𝑘, 𝑠) by a reference wavenumber collapses the curves across different 𝑠, consistent with the separable form 𝑀(𝑘, 𝑠) = 𝐴(𝑠)𝐵(𝑘). The inset shows that both 𝑀(𝑘, 𝑠) and Φ(𝑘, 𝑠) are nearly rank one. Middle: variable cells spoil this collapse, and higher singular modes remain finite… view at source ↗
Figure 18
Figure 18. Figure 18: Running diffusion coefficient 𝐾 (𝑡) for a particle in a 1D alternat￾ing diffusivity landscape (𝑘𝑆 = 0.1, 𝑘𝐹 = 10.0) with constant cell size. With￾out resetting (circles), 𝐾 (𝑡) converges to the harmonic mean (dotted). With fast resetting (𝜏renew ∼ 0.02𝜏𝑠, triangles), the environment renews rapidly, leading to breakdown of the convergence. Slow resetting (𝜏renew ∼ 0.5𝜏𝑠, squares) shows a crossover behavior… view at source ↗
read the original abstract

Cosmic ray (CR) transport is usually modeled with a single diffusion coefficient, but this description captures only the growth of the variance and not the full transport process. Distinct transport mechanisms can share the same effective diffusion coefficient while producing different particle distributions and approaches to the diffusive limit. This limitation is especially relevant in realistic multiphase, structured, and time-dependent media, and is also reflected in observed environmental variations in CR transport near pulsar wind nebulae, supernova remnants, and molecular clouds. Particle-tracing studies also show clear departures from standard diffusion, including both superdiffusion and subdiffusion. We therefore develop a propagator-based framework centered on $P(x,t)$, the probability distribution of particle positions, or equivalently its Fourier-Laplace transform $P(k,s)$. This object is compact and statistically complete, and naturally exposes memory: the CR flux can depend on earlier gradients when unresolved trapping or phase changes are coarse-grained away. Using the Montroll-Weiss formalism, we show how to measure $P(k,s)$ directly from trajectories, how to recover the associated memory kernel, and how to represent broad kernels efficiently with a Prony expansion. Applied to a multiphase medium, the framework shows that slow regions can regulate escape without dominating the total residence-time budget. We also introduce an accelerated Monte Carlo method for coarse-grained transport, and show that if trapping structures evolve while particles are still sampling them, the static long-time limit need not be reached. This paper provides the foundation for future observational applications, particle-tracing measurements, and CR-MHD closures.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript argues that a single diffusion coefficient is insufficient to describe cosmic-ray transport in complex media because it only captures variance growth and not the full statistical process. It develops a propagator-based framework centered on the position probability distribution P(x,t) (or its Fourier-Laplace transform P(k,s)), employs the Montroll-Weiss formalism to extract memory kernels directly from trajectories, demonstrates that slow regions in multiphase media can regulate escape without dominating residence times, introduces a Prony expansion for efficient kernel representation, and presents an accelerated Monte Carlo method for coarse-grained transport while noting that time-evolving trapping structures may prevent reaching the static long-time diffusive limit.

Significance. If the framework is shown to be robust, it supplies a statistically complete description of CR transport that naturally incorporates memory effects and non-diffusive regimes observed in particle-tracing simulations and near astrophysical sources. This could improve CR-MHD closures, enable direct extraction of transport properties from trajectories, and support more accurate modeling of environmental variations around PWNe, SNRs, and molecular clouds. The emphasis on measurement protocols and efficient numerical representations constitutes a concrete methodological advance.

major comments (1)
  1. [Montroll-Weiss formalism and memory-kernel section] § on Montroll-Weiss application and memory-kernel recovery: the central claim that the standard Montroll-Weiss relation supplies the memory kernel for the flux in realistic multiphase, time-dependent media rests on the assumption of a stationary waiting-time distribution. The abstract acknowledges that evolving trapping structures can prevent the static long-time limit, yet the derivation appears to omit any convective correction term that would arise from explicit time dependence in the medium. A concrete derivation or numerical test showing that the claimed integro-differential equation for the flux remains unmodified under non-stationary conditions is required; otherwise the applicability to the stated target media is not yet established.
minor comments (2)
  1. The Prony-expansion representation of broad kernels is introduced without a short self-contained definition or reference; adding one sentence or a one-line formula would improve accessibility.
  2. [Introduction or formalism section] Notation for the Fourier-Laplace transform P(k,s) is introduced in the abstract but the precise conventions (signs, normalization) should be restated at first use in the main text.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading and constructive feedback on our manuscript. We address the single major comment below and will revise the paper accordingly to strengthen the discussion of applicability.

read point-by-point responses

  1. Referee: [Montroll-Weiss formalism and memory-kernel section] § on Montroll-Weiss application and memory-kernel recovery: the central claim that the standard Montroll-Weiss relation supplies the memory kernel for the flux in realistic multiphase, time-dependent media rests on the assumption of a stationary waiting-time distribution. The abstract acknowledges that evolving trapping structures can prevent the static long-time limit, yet the derivation appears to omit any convective correction term that would arise from explicit time dependence in the medium. A concrete derivation or numerical test showing that the claimed integro-differential equation for the flux remains unmodified under non-stationary conditions is required; otherwise the applicability to the stated target media is not yet established.

    Authors: We agree that the standard Montroll-Weiss formalism assumes a stationary waiting-time distribution and that our abstract explicitly flags the possibility that time-evolving trapping structures can prevent the static long-time diffusive limit. The propagator framework and kernel extraction in the manuscript are presented for quasi-stationary intervals, which are the relevant regime for most cosmic-ray transport applications. To directly address the request, we will add a short derivation in the revised manuscript (new appendix) showing that, when the medium evolves on timescales much longer than individual trapping events, the leading-order integro-differential equation for the flux remains unmodified; convective correction terms appear only at higher order in the ratio of evolution to trapping timescales. We will also include a numerical test in a slowly time-dependent multiphase medium to verify that the unmodified equation reproduces the measured flux to within a few percent. This revision will clarify the domain of applicability without altering the core claims. revision: yes

Circularity Check

0 steps flagged

No circularity: Montroll-Weiss applied as external formalism to propagator extraction

full rationale

The paper's central chain adopts the pre-existing Montroll-Weiss CTRW relation to convert measured P(k,s) into a memory kernel and Prony expansion, then applies the result to multiphase media. No equation reduces a claimed prediction to a fitted parameter defined by the same data, no self-citation supplies a uniqueness theorem or ansatz, and no renaming of known results is presented as new derivation. The framework remains an independent measurement procedure whose validity rests on the standard stationary-waiting-time assumption rather than on internal redefinition.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Only the abstract is available, so the ledger is necessarily incomplete and limited to elements explicitly named.

axioms (1)
  • domain assumption Montroll-Weiss formalism applies directly to cosmic-ray trajectories in multiphase media
    Invoked to extract P(k,s) and the memory kernel from particle paths.

pith-pipeline@v0.9.0 · 5580 in / 1317 out tokens · 32183 ms · 2026-05-10T15:49:41.293856+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

52 extracted references · 48 canonical work pages · 1 internal anchor

  1. [2]

    U., Albert , A., Alfaro , R., et al

    Abeysekara A. U., et al., 2017b, @doi [Science] 10.1126/science.aan4880 , 358, 911–914

    work page doi:10.1126/science.aan4880

  2. [3]

    Ackermann M., et al., 2011, @doi [ ] 10.1088/0004-637X/726/2/81 , http://adsabs.harvard.edu/abs/2011ApJ...726...81A 726, 81

    work page doi:10.1088/0004-637x/726/2/81 2011

  3. [4]

    Aharonian F., Yang R., de O \ n a Wilhelmi E., 2019, @doi [Nature Astronomy] 10.1038/s41550-019-0724-0 , https://ui.adsabs.harvard.edu/abs/2019NatAs...3..561A 3, 561

    work page doi:10.1038/s41550-019-0724-0 2019

  4. [5]

    Almada Monter S., Gronke M., 2024, @doi [ ] 10.1093/mnrasl/slae074 , https://ui.adsabs.harvard.edu/abs/2024MNRAS.534L...7A 534, L7

    work page doi:10.1093/mnrasl/slae074 2024

  5. [6]

    Beylkin G., Monz \'o n L., 2010, Applied and Computational Harmonic Analysis, 28, 131

    2010

  6. [7]

    Bloemen J. B. G. M., Dogiel V. A., Dorman V. L., Ptuskin V. S., 1993, , https://ui.adsabs.harvard.edu/abs/1993A&A...267..372B 267, 372

    1993

  7. [8]

    R., 2005, @doi [ ] 10.1086/428919 , https://ui.adsabs.harvard.edu/abs/2005ApJ...624..213B 624, 213

    Boldyrev S., Gwinn C. R., 2005, @doi [ ] 10.1086/428919 , https://ui.adsabs.harvard.edu/abs/2005ApJ...624..213B 624, 213

    work page doi:10.1086/428919 2005

  8. [9]

    Bouchaud J.-P., Georges A., 1990, @doi [Physics Reports] https://doi.org/10.1016/0370-1573(90)90099-N , 195, 127

    work page doi:10.1016/0370-1573(90)90099-n 1990

  9. [10]

    S., Hopkins P

    Butsky I. S., Hopkins P. F., Kempski P., Ponnada S. B., Quataert E., Squire J., 2024, @doi [ ] 10.1093/mnras/stae276 , https://ui.adsabs.harvard.edu/abs/2024MNRAS.528.4245B 528, 4245

    work page doi:10.1093/mnras/stae276 2024

  10. [11]

    V., Seno F., Metzler R., Sokolov I

    Chechkin A. V., Seno F., Metzler R., Sokolov I. M., 2017, @doi [Physical Review X] 10.1103/physrevx.7.021002 , 7

    work page doi:10.1103/physrevx.7.021002 2017

  11. [12]

    C., Viswanathan G

    Cressoni J. C., Viswanathan G. M., Ferreira A. S., da Silva M. A. A., 2012, @doi [Phys. Rev. E] 10.1103/PhysRevE.86.022103 , 86, 022103

    work page doi:10.1103/physreve.86.022103 2012

  12. [13]

    Dijkstra M., 2019, @doi [Saas-Fee Advanced Course] 10.1007/978-3-662-59623-4_1 , https://ui.adsabs.harvard.edu/abs/2019SAAS...46....1D 46, 1

    work page doi:10.1007/978-3-662-59623-4_1 2019

  13. [14]

    Effenberger F., et al., 2025, @doi [Space Science Reviews] 10.1007/s11214-025-01203-4 , 221, 75

    work page doi:10.1007/s11214-025-01203-4 2025

  14. [15]

    Eugene Engelbrecht , F

    Engelbrecht N. E., et al., 2022, @doi [ ] 10.1007/s11214-022-00896-1 , https://ui.adsabs.harvard.edu/abs/2022SSRv..218...33E 218, 33

    work page doi:10.1007/s11214-022-00896-1 2022

  15. [16]

    Evoli C., Dupletsa U., 2023, Phenomenological models of Cosmic Ray transport in Galaxies , @doi 10.48550/arXiv.2309.00298 , https://ui.adsabs.harvard.edu/abs/2023arXiv230900298E

    work page doi:10.48550/arxiv.2309.00298 2023

  16. [17]

    Cosmic-ray transport in inhomogeneous media.MNRAS2026,545, staf2108, [arXiv:astro-ph.HE/2507.19044]

    Ewart R. J., et al., 2025, @doi [arXiv e-prints] 10.48550/arXiv.2507.19044 , https://ui.adsabs.harvard.edu/abs/2025arXiv250719044E p. arXiv:2507.19044

    work page doi:10.48550/arxiv.2507.19044 2025

  17. [18]

    Gabici S., Evoli C., Gaggero D., Lipari P., Mertsch P., Orlando E., Strong A., Vittino A., 2019, @doi [International Journal of Modern Physics D] 10.1142/S0218271819300222 , https://ui.adsabs.harvard.edu/abs/2019IJMPD..2830022G 28, 1930022

    work page doi:10.1142/s0218271819300222 2019

  18. [19]

    R., Mazur J

    Giacalone J., Jokipii J. R., Mazur J. E., 2000, @doi [ ] 10.1086/312564 , https://ui.adsabs.harvard.edu/abs/2000ApJ...532L..75G 532, L75

    work page doi:10.1086/312564 2000

  19. [20]

    Cosmic Rays on Galaxy Scales: Progress and Pitfalls for CR-MHD Dynamical Models

    Hopkins P. F., 2025, @doi [arXiv e-prints] 10.48550/arXiv.2509.07104 , https://ui.adsabs.harvard.edu/abs/2025arXiv250907104H p. arXiv:2509.07104

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2509.07104 2025

  20. [21]

    F., Squire, J., Butsky, I

    Hopkins P. F., Squire J., Butsky I. S., Ji S., 2022, @doi [ ] 10.1093/mnras/stac2909 , https://ui.adsabs.harvard.edu/abs/2022MNRAS.517.5413H 517, 5413

    work page doi:10.1093/mnras/stac2909 2022

  21. [22]

    Jiang S., Zhang J., Zhang Q., Zhang Z., 2017, @doi [Communications in Computational Physics] 10.4208/cicp.OA-2016-0136 , 21, 650

    work page doi:10.4208/cicp.oa-2016-0136 2017

  22. [23]

    A., Moskalenko I

    J \'o hannesson G., Porter T. A., Moskalenko I. V., 2019, @doi [ ] 10.3847/1538-4357/ab258e , https://ui.adsabs.harvard.edu/abs/2019ApJ...879...91J 879, 91

    work page doi:10.3847/1538-4357/ab258e 2019

  23. [24]

    Kempski P., Quataert E., 2022, @doi [ ] 10.1093/mnras/stac1240 , https://ui.adsabs.harvard.edu/abs/2022MNRAS.514..657K 514, 657

    work page doi:10.1093/mnras/stac1240 2022

  24. [25]

    Cosmic ray transport in large-amplitude turbulence with small-scale field reversals.MNRAS2023,525, 4985–4998, [arXiv:astro-ph.HE/2304.12335]

    Kempski P., Fielding D. B., Quataert E., Galishnikova A. K., Kunz M. W., Philippov A. A., Ripperda B., 2023, @doi [ ] 10.1093/mnras/stad2609 , https://ui.adsabs.harvard.edu/abs/2023MNRAS.525.4985K 525, 4985

    work page doi:10.1093/mnras/stad2609 2023

  25. [26]

    B., Quataert E., Ewart R

    Kempski P., Fielding D. B., Quataert E., Ewart R. J., Grete P., Kunz M. W., Philippov A. A., Stone J., 2025, @doi [arXiv e-prints] 10.48550/arXiv.2507.10651 , https://ui.adsabs.harvard.edu/abs/2025arXiv250710651K p. arXiv:2507.10651

    work page doi:10.48550/arxiv.2507.10651 2025

  26. [27]

    Laitinen , A

    Laitinen T., Kopp A., Effenberger F., Dalla S., Marsh M. S., 2016, @doi [ ] 10.1051/0004-6361/201527801 , https://ui.adsabs.harvard.edu/abs/2016A&A...591A..18L 591, A18

    work page doi:10.1051/0004-6361/201527801 2016

  27. [28]

    1999 Reconnection in a Weakly Stochastic Field.Astrophys

    Lazarian A., Vishniac E. T., 1999, @doi [ ] 10.1086/307233 , https://ui.adsabs.harvard.edu/abs/1999ApJ...517..700L 517, 700

    work page doi:10.1086/307233 1999

  28. [29]

    Lazarian A., Xu S., 2021, @doi [ ] 10.3847/1538-4357/ac2de9 , https://ui.adsabs.harvard.edu/abs/2021ApJ...923...53L 923, 53

    work page doi:10.3847/1538-4357/ac2de9 2021

  29. [30]

    Lemoine M., 2023, @doi [Journal of Plasma Physics] 10.1017/S0022377823000946 , https://ui.adsabs.harvard.edu/abs/2023JPlPh..89e1701L 89, 175890501

    work page doi:10.1017/s0022377823000946 2023

  30. [31]

    A., 2020, @doi [ ] 10.1093/mnras/staa3131 , https://ui.adsabs.harvard.edu/abs/2020MNRAS.499.4972L 499, 4972

    Lemoine M., Malkov M. A., 2020, @doi [ ] 10.1093/mnras/staa3131 , https://ui.adsabs.harvard.edu/abs/2020MNRAS.499.4972L 499, 4972

    work page doi:10.1093/mnras/staa3131 2020

  31. [32]

    P., 2025, @doi [ ] 10.1093/mnras/staf1474 , https://ui.adsabs.harvard.edu/abs/2025MNRAS.543.1911L 543, 1911

    Liang N., Oh S. P., 2025, @doi [ ] 10.1093/mnras/staf1474 , https://ui.adsabs.harvard.edu/abs/2025MNRAS.543.1911L 543, 1911

    work page doi:10.1093/mnras/staf1474 2025

  32. [33]

    arXiv:2505.18155

    L \"u bke J., Reichherzer P., Aerdker S., Effenberger F., Wilbert M., Fichtner H., Grauer R., 2025a, @doi [arXiv e-prints] 10.48550/arXiv.2505.18155 , https://ui.adsabs.harvard.edu/abs/2025arXiv250518155L p. arXiv:2505.18155

    work page doi:10.48550/arxiv.2505.18155

  33. [35]

    arXiv:2509.15320

    L \"u bke J., Effenberger F., Wilbert M., Fichtner H., Grauer R., 2025c, @doi [arXiv e-prints] 10.48550/arXiv.2509.15320 , https://ui.adsabs.harvard.edu/abs/2025arXiv250915320L p. arXiv:2509.15320

    work page doi:10.48550/arxiv.2509.15320

  34. [36]

    E., et al., 2023, @doi [Physics of Plasmas] 10.1063/5.0147683 , https://ui.adsabs.harvard.edu/abs/2023PhPl...30e0501M 30, 050501

    Malandraki O. E., et al., 2023, @doi [Physics of Plasmas] 10.1063/5.0147683 , https://ui.adsabs.harvard.edu/abs/2023PhPl...30e0501M 30, 050501

    work page doi:10.1063/5.0147683 2023

  35. [37]

    E., Mason G

    Mazur J. E., Mason G. M., Dwyer J. R., Giacalone J., Jokipii J. R., Stone E. C., 2000, @doi [ ] 10.1086/312561 , https://ui.adsabs.harvard.edu/abs/2000ApJ...532L..79M 532, L79

    work page doi:10.1086/312561 2000

  36. [38]

    Metzler R., Klafter J., 2000, Physics reports, 339, 1

    2000

  37. [39]

    W., Weiss G

    Montroll E. W., Weiss G. H., 1965, Journal of Mathematical Physics, 6, 167

    1965

  38. [40]

    Mori H., 1965, @doi [Progress of Theoretical Physics] 10.1143/PTP.33.423 , 33, 423

    work page doi:10.1143/ptp.33.423 1965

  39. [41]

    Pecora F., et al., 2021, @doi [ ] 10.1093/mnras/stab2659 , https://ui.adsabs.harvard.edu/abs/2021MNRAS.508.2114P 508, 2114

    work page doi:10.1093/mnras/stab2659 2021

  40. [42]

    R., Kirk J

    Ragot B. R., Kirk J. G., 1997, @doi [ ] 10.48550/arXiv.astro-ph/9708041 , https://ui.adsabs.harvard.edu/abs/1997A&A...327..432R 327, 432

    work page doi:10.48550/arxiv.astro-ph/9708041 1997

  41. [43]

    Ruszkowski M., Pfrommer C., 2023, @doi [ ] 10.1007/s00159-023-00149-2 , https://ui.adsabs.harvard.edu/abs/2023A&ARv..31....4R 31, 4

    work page doi:10.1007/s00159-023-00149-2 2023

  42. [44]

    L., Beattie J

    Sampson M. L., Beattie J. R., Krumholz M. R., Crocker R. M., Federrath C., Seta A., 2022, arXiv e-prints, https://ui.adsabs.harvard.edu/abs/2022arXiv220508174S p. arXiv:2205.08174

    work page arXiv 2022

  43. [45]

    K., Kim C., Talkner P., Lee E

    Shin H. K., Kim C., Talkner P., Lee E. K., 2010, @doi [Chemical Physics] 10.1016/j.chemphys.2010.05.019 , 375, 316–326

    work page doi:10.1016/j.chemphys.2010.05.019 2010

  44. [46]

    M., Matthews J

    Taylor A. M., Matthews J. H., Bell A. R., 2023, @doi [ ] 10.1093/mnras/stad1716 , https://ui.adsabs.harvard.edu/abs/2023MNRAS.524..631T 524, 631

    work page doi:10.1093/mnras/stad1716 2023

  45. [47]

    Whitman K., et al., 2023, @doi [Advances in Space Research] 10.1016/j.asr.2022.08.006 , https://ui.adsabs.harvard.edu/abs/2023AdSpR..72.5161W 72, 5161

    work page doi:10.1016/j.asr.2022.08.006 2023

  46. [48]

    Xu S., Yan H., 2013, @doi [ ] 10.1088/0004-637X/779/2/140 , https://ui.adsabs.harvard.edu/abs/2013ApJ...779..140X 779, 140

    work page doi:10.1088/0004-637x/779/2/140 2013

  47. [49]

    Yan H., Lazarian A., 2008, @doi [ ] 10.1086/524771 , http://adsabs.harvard.edu/abs/2008ApJ...673..942Y 673, 942

    work page doi:10.1086/524771 2008

  48. [50]

    Yan K., Yan H., Pavaskar P., Hou C., Liu R.-Y., 2026, Non-Markovian Cosmic-Ray Pitch-Angle Transport from Mirror Interactions ( @eprint arXiv 2603.19037 ), https://arxiv.org/abs/2603.19037

    work page arXiv 2026

  49. [52]

    Yang R.-z., Li G.-X., Wilhelmi E. d. O., Cui Y.-D., Liu B., Aharonian F., 2023b, @doi [Nature Astronomy] 10.1038/s41550-022-01868-9 , https://ui.adsabs.harvard.edu/abs/2023NatAs...7..351Y 7, 351

    work page doi:10.1038/s41550-022-01868-9

  50. [53]

    Zhang C., Xu S., 2023, @doi [ ] 10.3847/2041-8213/ad0fe5 , https://ui.adsabs.harvard.edu/abs/2023ApJ...959L...8Z 959, L8

    work page doi:10.3847/2041-8213/ad0fe5 2023

  51. [54]

    Zimbardo G., Perri S., 2013, @doi [ ] 10.1088/0004-637X/778/1/35 , https://ui.adsabs.harvard.edu/abs/2013ApJ...778...35Z 778, 35

    work page doi:10.1088/0004-637x/778/1/35 2013

  52. [55]

    2017 , bdsk-url-1 =

    Zweibel E. G., 2017, @doi [Physics of Plasmas] 10.1063/1.4984017 , 24, 055402

    work page doi:10.1063/1.4984017 2017