Pith. sign in

REVIEW 3 major objections 4 minor 299 references

CRESCENDO II extends a spectral cosmic-ray solver with improved energy losses and supernova-seeded spectrum injection.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-02 02:18 UTC pith:R5AL2EBU

load-bearing objection Solid spectral-CR methods paper; the solver work is credible and honest about its limitations, but the SNR seeding subgrid is validated only in an idealized regime and the paper would be stronger with code/data and quantitative error metrics. the 3 major comments →

arxiv 2607.14363 v1 pith:R5AL2EBU submitted 2026-07-15 astro-ph.HE astro-ph.IM

CRESCENDO II: Spectral cosmic rays with improved energy losses and realistic supernova seeding

Daniel Karner , Ludwig M. B\"oss , Klaus Dolag , Ildar Khabibullin This is my paper
classification astro-ph.HE astro-ph.IM
keywords cosmic raysspectral energy lossestwo-moment schemesupernova remnant injectionnon-ultra-relativistic spectranumerical methodsgalaxy formation simulationsnon-thermal emission
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper extends a spectral cosmic-ray solver used in galaxy-formation simulations so that proton and electron spectra are cooled by all major momentum-loss processes, with kinetic energy and pressure computed from the full relativistic energy-momentum relation rather than the ultra-relativistic approximation. It also replaces the usual single power-law injection with tabulated supernova-remnant spectra, coupled to the star-formation model. In idealized tests, the evolved spectra match analytical cooling solutions, the floating spectral cut-off prevents artificial curvature, and solving for the initial momentum via a root-finding procedure rather than analytical approximations keeps the slope reconstruction stable. If these methods hold in full cosmological runs, the code can deliver spectrally resolved cosmic-ray populations whose non-thermal emission can be compared with radio and gamma-ray observations.

Core claim

The central claim is that the two-moment spectral solver—which updates the number and kinetic energy of cosmic rays in each logarithmic momentum bin and then reconstructs a piecewise power-law distribution—remains accurate when the full relativistic energy-momentum relation is used and when all relevant loss processes are included. The authors show that the correct way to compute particle and energy fluxes between bins is to find the initial momentum by solving the implicit cooling equation with a numerical root-finder; analytical approximations that are not consistent with the loss rate lead to spurious slopes. They also argue that a time-dependent upper spectral cut-off is necessary to pre

What carries the argument

The central object is the piecewise power-law cosmic-ray distribution function attached to each smoothed-particle-hydrodynamics gas particle, evolved by the two-moment Fokker-Planck scheme: each bin stores the number and kinetic energy of protons or electrons per unit mass, and the slope and normalization are reconstructed after each update. The load-bearing mechanism is the implicit equation that fixes the initial momentum for each loss process; solving it numerically rather than using analytical approximations keeps number and energy fluxes consistent with the loss rate and stabilizes slope reconstruction. A floating spectral cut-off and per-bin adiabatic indices derived from the general e

Load-bearing premise

The subgrid supernova seeding model assumes that the simulation timestep is longer than the few-tens-of-thousands-of-years acceleration phase and that the smoothing length encloses where freshly accelerated cosmic rays travel in that time; no test shows these conditions hold in full cosmological runs.

What would settle it

Run the star-forming box with a deliberately reduced smoothing length so that it falls below the assumed cosmic-ray propagation distance: if the box-averaged spectrum changes appreciably with resolution, or if injected energy appears in gas particles far from the explosion site, the seeding subgrid model's volume assumption fails. A second check is to compare the electron steady-state spectral index in a run with only injection and losses against the analytical steady-state solution; a mismatch that grows with runtime would expose an error in the loss implementation.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • With non-ultra-relativistic energy and pressure integrals, Coulomb and other low-energy losses are tracked correctly, and the gas pressure response uses a per-bin adiabatic index rather than a fixed 4/3.
  • The subgrid supernova-remnant seeding model accepts external spectrum files, so improved published spectra can be plugged in without changing the numerical scheme.
  • Treating hadronic losses as continuous versus catastrophic gives measurably different spectral shapes near the pion-production threshold, and the continuous approach with exact initial-momentum solving matches the analytical cooling solution.
  • In the star-forming box test, repeated supernova injection plus cooling produces a steady-state spectrum whose shape reflects both the template spectra and the momentum-dependent cooling time.
  • In the shocktube test, the electron spectrum visibly cools downstream while the proton spectrum remains nearly unchanged, consistent with the much longer proton cooling times.

Where Pith is reading between the lines

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

  • If the floating cut-off and exact initial-momentum solve are retained, the same scheme can be extended to additional species or to energy-dependent diffusion coefficients without major restructuring.
  • The validity of the subgrid seeding model depends on resolution: in cosmological runs with large timesteps or small smoothing lengths, freshly accelerated cosmic rays may not remain inside the parent gas particle, so the injected spectra would need to be distributed to neighboring particles.
  • A testable consequence follows from the steady-state electron spectrum: the spectral peak sits at the momentum range with the longest cooling time, so the predicted radio spectral index should correlate with gas density and magnetic field strength in a way that can be checked against observed radio–gamma correlations.
  • The paper leaves spatial transport to future work; coupling these losses with explicit transport could reveal whether the spectral shapes survive advection and diffusion in realistic galactic winds.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The manuscript extends the on-the-fly spectral CR solver CRESCENDO in OpenGadget3. It introduces non-ultra-relativistic energy and pressure integrals, several momentum-loss processes for CR protons and electrons (adiabatic, Coulomb, bremsstrahlung, synchrotron/IC, hadronic), a flexible spectral cut-off, and a subgrid model that injects tabulated supernova-remnant spectra into star-forming gas. The numerical tests include single-particle cooling spectra against analytic solutions, a steady-state injection test, a periodic ISM box with SN seeding, and a shock-tube with shock acceleration. The paper also documents numerical pitfalls: fixed cut-offs cause artificial curvature (App. C), approximate p_u formulas corrupt slope reconstruction (App. D), and continuous vs. catastrophic hadronic treatments differ (App. F). The overall goal is to make a spectrally resolved CR module suitable for future cosmological production runs.

Significance. If the numerical implementation is correct, this is a useful step toward spectrally resolved CR modeling in cosmological SPH simulations: it couples multiple cooling processes, non-ultra-relativistic energy/pressure integrals, and literature SNR spectra in a single on-the-fly solver. The explicit discussion of failure modes and the use of analytic cooling solutions are strengths. The main caveat is that the 'realistic' SNR seeding is validated only in an idealized box, so its validity in production regimes is not yet demonstrated. With the technical corrections identified below, the paper would be a solid methods contribution.

major comments (3)
  1. [Sec. 2.5.1 / Sec. 3.4] The SNR-seeding model's validity conditions are stated but never demonstrated in a production-relevant regime. Sec. 2.5.1 requires the simulation timestep to exceed the CR acceleration period (few 10^4 yr) and the SPH smoothing length to enclose the CR propagation distance. The only multi-particle injection test, Sec. 3.4, is a 5 kpc box near the star-formation threshold, and no values of timestep, smoothing length, or propagation distance are reported. At dense star-forming ISM conditions (smoothing lengths of a few pc, Courant-limited timesteps possibly below 10^3 yr, GeV CRs propagating tens of pc over 10^4 yr), one or both conditions can fail, placing injected CRs in the wrong gas parcel. Since the abstract advertises 'realistic supernova seeding', this missing support is load-bearing. Please add a quantitative test or analysis at representative cosmological resolution, or explicitly
  2. [Sec. 2.4.5, Eq. (36)] The antiderivative in Eq. (36) is incorrect for the integrand 1/(sqrt(1+p_hat^2)-1). Differentiating the displayed expression p_hat*arsinh(p_hat)-1-sqrt(1+p_hat^2)/p_hat gives arsinh(p_hat)+p_hat/sqrt(1+p_hat^2)+1/(p_hat^2*sqrt(1+p_hat^2)), not 1/(sqrt(1+p_hat^2)-1). The correct primitive is arsinh(p_hat)-(sqrt(1+p_hat^2)+1)/p_hat, up to a constant. Since Eq. (36) defines the implicit p_u used for hadronic loss fluxes, the displayed formula would change the flux integrals and the results in Figs. 4, F.1, and F.2. Please correct Eq. (36) and re-verify the hadronic-loss tests against the corrected expression.
  3. [Sec. 2.3.1, Eq. (16)] Eq. (16) as displayed has a sign error in the derivation of p_u. With p(t0+Δt)=p_i and p(t0)=p_u, the correct implicit relation is Δt = ∫_{p_i}^{p_u} dp/b(p), which is positive for b>0 and p_u>p_i. The displayed expression contains minus signs before the integrals and would give a negative Δt. The variable transformation dp/ds=b(p) described in the surrounding text yields the positive form, and the final p_u formulas (e.g., Eq. 30) are consistent with the positive form. This appears to be a sign typo, but Eq. (16) is the basis for all p_u computations and should be corrected with the intermediate signs checked.
minor comments (4)
  1. [Sec. 2.3.1, Eq. (12)] The expression for the average density appears to have a sign typo: it should be (ρ(t0+Δt)+ρ(t0))/2, not (ρ(t0+Δt)−ρ(t0))/2.
  2. [Sec. 3.5, Fig. 7] It would be helpful to state explicitly which profile the red dashed analytical line in Fig. 7 refers to, since the figure shows thermal and CR pressures simultaneously.
  3. [Sec. 3.1-3.3] Agreement with analytic solutions is shown only visually. Adding quantitative residuals or convergence tests in bin count/timestep would substantially strengthen the numerical-validation claim.
  4. [General] Minor typos include 'Galacitic' (Sec. 2.5.1), 'imlicit' (App. F), and inconsistent use of 'loose' for 'lose' in a few places.

Circularity Check

0 steps flagged

No significant circularity; the new loss-process and injection tests are numerical consistency checks against the same equations, and the only self-citation supplies the base solver rather than the new results.

full rationale

The paper's central technical claims are implementations of standard CR loss processes (Coulomb, bremsstrahlung, synchrotron/IC, hadronic) with the two-moment spectral solver. The tests compare the discretized code against analytic solutions of the same Fokker–Planck evolution equation (e.g., Eq. 39 for adiabatic changes, Eq. 42 for synchrotron/IC cooling, Eq. 45 for steady-state spectra). This is code verification against the intended continuum equation, not a circular prediction derived from fitted inputs. The SNR seeding model in Sec. 2.5.1 is explicitly introduced as a subgrid prescription with two stated validity conditions (timestep longer than the CR acceleration period, and smoothing length enclosing the CR propagation distance); Sec. 3.4 demonstrates the mechanism in an idealized box but does not validate those conditions at cosmological resolution. That is an external validity / coverage gap, not a circularity: the paper does not claim to derive the validity conditions from the simulation itself. The main self-citation (Böss et al. 2023) provides the underlying CRESCENDO solver and is used for methodological background; the new loss-rate formulas and template spectra come from independent literature (e.g., Winner et al. 2019; Girichidis et al. 2020; Cristofari et al. 2021; Das et al. 2024). No load-bearing step reduces by construction to its own inputs or renamed fit. Score reflects a minor, non-load-bearing self-citation and no material circularity.

Axiom & Free-Parameter Ledger

7 free parameters · 8 axioms · 0 invented entities

The paper introduces no new physical entities, forces, or dimensions. Its central claims rest on standard Fokker-Planck/SPH modeling assumptions, several stated physical approximations for loss rates, externally computed SNR template spectra, and two explicitly stated validity conditions for subgrid injection that are not exercised in realistic simulations.

free parameters (7)
  • SN acceleration efficiency ξ_CR = 0.05
    Imported from Cristofari et al. (2021) template normalization; ratio of CR pressure to upstream ram pressure at the shock.
  • Injected proton spectral slope q = 4.3
    Used in the Cristofari et al. templates and in the artificial power-law consistency test (q=4.3 in App. G); default test slope q_ini=4.5 chosen in Sec. 3.
  • Electron-to-proton ratio K_ep = ~0.01
    Rescales electron spectra relative to protons; from Schlickeiser (2002) for q=4.2.
  • CR energy fraction of E_SN = 10%
    Standard normalization for injected CR proton energy relative to 10^51 erg supernova energy; used in tests and App. G.
  • Spectral cut-offs p_min and p_max = p_min,p ~ 1e-2, p_max,p ~ 1e6; p_min,e ~ 10, p_max,e ~ 1e6
    Chosen from Coulomb/cooling timescale arguments and the Galactic knee; not fitted to the tests.
  • Bremsstrahlung Gaunt factor bar_g = 5.7
    Fixed value evaluated at γ=200 in App. F.1.2; neglects logarithmic energy dependence.
  • Inelastic pp cross section σ_pp = 30 mb
    Constant cross section for pion production used in hadronic loss rate (Sec. 2.4.5); error grows logarithmically with energy.
axioms (8)
  • domain assumption CR distribution is isotropic and in the strong-scattering regime, so only f(x, |p|, t) is evolved.
    Invoked in Sec. 2.1 and App. B, Eq. (B.1); excludes anisotropic transport effects.
  • domain assumption The distribution function is represented as a piecewise power-law with no continuity enforced across bin boundaries.
    Eq. (1), Sec. 2.1; accuracy depends on bin resolution and the flexible cut-off.
  • domain assumption All loss rates add linearly and background quantities (n_e, u_B, u_rad, n_N, ∇·u) are constant during a simulation timestep.
    Used for separation of variables in Eq. (16) and the explicit update in Eq. (19); fails for rapidly varying environments.
  • domain assumption Synchrotron/IC losses are computed in the Thomson/ultra-relativistic limit, and proton radiative losses are negligible by (m_p/m_e)^2.
    Sec. 2.4.3 and App. F; not valid near the Klein-Nishina limit or for mildly relativistic electrons.
  • domain assumption Hadronic losses use a constant inelastic pp cross section σ_pp ≈ 30 mb and inelasticity K_p ≈ 0.5.
    Sec. 2.4.5; stated approximation with logarithmic error growth.
  • ad hoc to paper SNR injection is valid only if the timestep exceeds the CR acceleration period and the SPH smoothing length encloses the CR propagation distance.
    Sec. 2.5.1; these validity conditions are not tested in full cosmological conditions.
  • domain assumption External SNR template spectra from Cristofari et al. (2021) and Das et al. (2024) represent the true released CR spectra.
    Sec. 2.5.2; the templates are not validated in this paper and carry their own acceleration-efficiency and slope assumptions.
  • domain assumption The low-momentum transition from the CR population to the thermal pool is not modeled.
    Sec. 2.5.1; affects sub-GeV protons and low-energy electrons, which are assumed to be radiatively unimportant.

pith-pipeline@v1.3.0-alltime-deepseek · 29847 in / 17597 out tokens · 163625 ms · 2026-08-02T02:18:01.326546+00:00 · methodology

0 comments
read the original abstract

Context. Cosmological simulation codes with subgrid models for cosmic rays (CRs) help us better understand their impact on baryonic feedback and non-thermal radiation in galaxies and galaxy clusters. An accurate numerical description requires a spectrally resolved treatment of the CR population, because virtually all transport, acceleration and loss processes depend on energy. Aims. We advance the treatment of CR electrons and protons in the on-the-fly spectral CR solver CRESCENDO in OpenGadget3. Methods. We implement several new energy loss processes for both protons and electrons and improve the computation of their energies and pressures beyond the ultra-relativistic approximation. Moreover, we present a subgrid model for CR seeding by supernova remnants, in which physically motivated spectra are injected at sites of ongoing star formation. Results. We test the newly implemented loss processes and the coupling between CR injection and star formation in idealized setups. We also highlight numerical subtleties, such as the differences arising when hadronic losses are modelled as continuous or catastrophic process, and the advantages of using a flexible spectral cut-off and abandoning the ultra-relativistic approximation. Furthermore, we show that using analytical approximations to compute energy fluxes can cause the slope reconstruction to fail. Conclusions. Future applications of our spectral cosmic-ray model in large-scale, full-physics cosmological simulations will represent an important step towards building a robust and observationally verifiable link between the microphysical and macrophysical aspects of the CR component in the modern paradigm of galaxy evolution.

Figures

Figures reproduced from arXiv: 2607.14363 by Daniel Karner, Ildar Khabibullin, Klaus Dolag, Ludwig M. B\"oss.

Figure 1
Figure 1. Figure 1: Summary of the most important points discussed in the main text, namely 1) the numerical representation of the CR spectra [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Cooling of a CR proton spectrum in an adiabatically ex [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 5
Figure 5. Figure 5: Continuous injection of an electron spectrum at a rate (44) [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Injection of templated CR spectra by ongoing core-collapse supernovae in a periodic ISM box. We computed the superpo [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Sod shocktube mimicking typical ICM conditions with [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Superposition of all CR proton (left) and electron spectra (right) in the Sod shocktube shown in Fig. 7. The spectral shape of [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗

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

299 extracted references · 21 canonical work pages

  1. [1]

    , keywords =

    Rise and fall of radio haloes in simulated merging galaxy clusters. , keywords =. doi:10.1093/mnras/sts628 , archivePrefix =. 1211.3337 , primaryClass =

    Pith/arXiv arXiv

  2. [3]

    Effects on the Sunyaev-Zel'dovich effect and the X-ray emission

    Simulating cosmic rays in clusters of galaxies - I. Effects on the Sunyaev-Zel'dovich effect and the X-ray emission. , keywords =. doi:10.1111/j.1365-2966.2007.11732.x , archivePrefix =. astro-ph/0611037 , primaryClass =

    arXiv 2007

  3. [4]

    , keywords =

    Cosmic Rays Can Drive Strong Outflows from Gas-rich High-redshift Disk Galaxies. , keywords =. doi:10.1088/2041-8205/777/2/L38 , archivePrefix =. 1310.3273 , primaryClass =

    Pith/arXiv arXiv 2041

  4. [5]

    , keywords =

    Magnetic fields in spiral galaxies. , keywords =. doi:10.1007/s00159-015-0084-4 , archivePrefix =. 1509.04522 , primaryClass =

    Pith/arXiv arXiv

  5. [6]

    2013 , month = feb, number =

    Science , title =. 2013 , month = feb, number =. arXiv , doi =:1302.3307 , keywords =

    Pith/arXiv arXiv 2013

  6. [7]

    2005 , month = nov, number =

    , title =. 2005 , month = nov, number =. arXiv , doi =:astro-ph/0509673 , keywords =

    Pith/arXiv arXiv 2005

  7. [8]

    2006 , month = sep, number =

    , title =. 2006 , month = sep, number =. arXiv , doi =:astro-ph/0606592 , keywords =

    Pith/arXiv arXiv 2006

  8. [9]

    2009 , month = feb, number =

    , title =. 2009 , month = feb, number =. arXiv , doi =:0806.1223 , keywords =

    Pith/arXiv arXiv 2009

  9. [10]

    2014 , month = may, number =

    International Journal of Modern Physics D , title =. 2014 , month = may, number =. arXiv , doi =:1406.7714 , keywords =

    Pith/arXiv arXiv 2014

  10. [11]

    1977 , month = jan, volume =

    Proceedings of the 15th International Cosmic Ray Conference , keywords =. 1977 , month = jan, volume =

    1977

  11. [12]

    doi:10.1007/978-1-4614-6099-2 , url =

    2013 , address =. doi:10.1007/978-1-4614-6099-2 , url =

    doi:10.1007/978-1-4614-6099-2 2013

  12. [13]

    2013 , month = nov, number =

    , title =. 2013 , month = nov, number =. arXiv , doi =:1308.3440 , keywords =

    Pith/arXiv arXiv 2013

  13. [14]

    1978 , month = jan, pages =

    , title =. 1978 , month = jan, pages =. doi:10.1093/mnras/182.2.147 , keywords =

    doi:10.1093/mnras/182.2.147 1978

  14. [15]

    1978 , month = feb, pages =

    , title =. 1978 , month = feb, pages =. doi:10.1093/mnras/182.3.443 , keywords =

    doi:10.1093/mnras/182.3.443 1978

  15. [16]

    2004 , month = sep, pages =

    , title =. 2004 , month = sep, pages =. doi:10.1111/j.1365-2966.2004.08097.x , keywords =

    arXiv 2004

  16. [17]

    2005 , month = mar, number =

    , title =. 2005 , month = mar, number =. doi:10.1111/j.1365-2966.2005.08774.x , keywords =

    arXiv 2005

  17. [18]

    2013 , month = mar, pages =

    , title =. 2013 , month = mar, pages =. doi:10.1016/j.astropartphys.2012.05.022 , url =

    doi:10.1016/j.astropartphys.2012.05.022 2013

  18. [19]

    2013 , month =

    , title =. 2013 , month =. arXiv , doi =:1301.7264 , keywords =

    Pith/arXiv arXiv 2013

  19. [20]

    2019 , month = sep, number =

    , title =. 2019 , month = sep, number =. arXiv , doi =:1906.12240 , keywords =

    Pith/arXiv arXiv 2019

  20. [21]

    1978 , month = apr, pages =

    , title =. 1978 , month = apr, pages =. doi:10.1086/182658 , keywords =

    doi:10.1086/182658 1978

  21. [22]

    1987 , month = oct, number =

    , title =. 1987 , month = oct, number =. doi:10.1016/0370-1573(87)90134-7 , url =

    doi:10.1016/0370-1573(87)90134-7 1987

  22. [23]

    2005 , month = aug, number =

    , title =. 2005 , month = aug, number =. arXiv , doi =:astro-ph/0505351 , keywords =

    Pith/arXiv arXiv 2005

  23. [24]

    2010 , month = mar, number =

    , title =. 2010 , month = mar, number =. arXiv , doi =:0912.2053 , keywords =

    Pith/arXiv arXiv 2010

  24. [25]

    2013 , month = nov, pages =

    , title =. 2013 , month = nov, pages =. arXiv , doi =:1311.7346 , keywords =

    Pith/arXiv arXiv 2013

  25. [26]

    2019 , month = dec, number =

    Nuovo Cimento Rivista Serie , title =. 2019 , month = dec, number =. doi:10.1393/ncr/i2019-10166-0 , url =

    work page doi:10.1393/ncr/i2019-10166-0 2019

  26. [27]

    2023 , month = jul, pages =

    arXiv e-prints , title =. 2023 , month = jul, pages =. arXiv , eid =:2307.11640 , keywords =

    Pith/arXiv arXiv 2023

  27. [28]

    2023 , month = feb, number =

    , title =. 2023 , month = feb, number =. arXiv , doi =:2207.05087 , keywords =

    Pith/arXiv arXiv 2023

  28. [29]

    2022 , month = oct, number =

    , title =. 2022 , month = oct, number =. arXiv , doi =:2208.04185 , keywords =

    Pith/arXiv arXiv 2022

  29. [30]

    doi:10.1007/978-3-662-44625-6_20 , keywords =

    2015 , address =. doi:10.1007/978-3-662-44625-6_20 , keywords =

    work page doi:10.1007/978-3-662-44625-6_20 2015

  30. [31]

    2024 , month = mar, number =

    , title =. 2024 , month = mar, number =. arXiv , doi =:2308.06316 , keywords =

    Pith/arXiv arXiv 2024

  31. [32]

    2023 , month = may, number =

    , title =. 2023 , month = may, number =. arXiv , doi =:2210.14232 , keywords =

    Pith/arXiv arXiv 2023

  32. [33]

    2018 , month = feb, number =

    , title =. 2018 , month = feb, number =. arXiv , doi =:1801.08890 , keywords =

    Pith/arXiv arXiv 2018

  33. [34]

    2019 , month = feb, number =

    , title =. 2019 , month = feb, number =. arXiv , doi =:1902.00240 , keywords =

    Pith/arXiv arXiv 2019

  34. [35]

    2009 , month = may, number =

    , title =. 2009 , month = may, number =. arXiv , doi =:0807.4261 , keywords =

    Pith/arXiv arXiv 2009

  35. [36]

    2009 , month = jul, number =

    , title =. 2009 , month = jul, number =. arXiv , doi =:0807.4259 , keywords =

    Pith/arXiv arXiv 2009

  36. [37]

    2010 , month = apr, pages =

    , title =. 2010 , month = apr, pages =. arXiv , doi =:0912.2964 , primaryclass =

    Pith/arXiv arXiv 2010

  37. [38]

    2010 , month = jun, pages =

    , title =. 2010 , month = jun, pages =. arXiv , doi =:0912.2714 , primaryclass =

    Pith/arXiv arXiv 2010

  38. [39]

    2014 , month = mar, pages =

    , title =. 2014 , month = mar, pages =. arXiv , doi =:1310.2943 , keywords =

    Pith/arXiv arXiv 2014

  39. [40]

    2014 , month = oct, number =

    , title =. 2014 , month = oct, number =. arXiv , doi =:1401.7679 , keywords =

    Pith/arXiv arXiv 2014

  40. [41]

    2014 , month = oct, number =

    , title =. 2014 , month = oct, number =. arXiv , doi =:1407.2261 , keywords =

    Pith/arXiv arXiv 2014

  41. [42]

    2015 , month = oct, number =

    , title =. 2015 , month = oct, number =. arXiv , doi =:1505.06739 , keywords =

    Pith/arXiv arXiv 2015

  42. [43]

    2015 , month = jan, number =

    , title =. 2015 , month = jan, number =. arXiv , doi =:1409.8291 , keywords =

    Pith/arXiv arXiv 2015

  43. [44]

    2018 , month = jun, number =

    Journal of Plasma Physics , title =. 2018 , month = jun, number =. arXiv , doi =:1801.01510 , keywords =

    Pith/arXiv arXiv 2018

  44. [45]

    2020 , month = dec, number =

    , title =. 2020 , month = dec, number =. arXiv , doi =:2009.00007 , keywords =

    Pith/arXiv arXiv 2020

  45. [46]

    2015 , month = sep, pages =

    , title =. 2015 , month = sep, pages =. arXiv , doi =:1503.03001 , keywords =

    Pith/arXiv arXiv 2015

  46. [47]

    1982 , month = jul, pages =

    , title =. 1982 , month = jul, pages =. doi:10.1086/160132 , keywords =

    doi:10.1086/160132 1982

  47. [48]

    2019 , month = dec, number =

    , title =. 2019 , month = dec, number =. arXiv , doi =:1906.09454 , keywords =

    Pith/arXiv arXiv 2019

  48. [49]

    2022 , month = mar, adsnote =

    37th International Cosmic Ray Conference , doi =. 2022 , month = mar, adsnote =

    2022

  49. [50]

    1982 , month = jul, pages =

    , title =. 1982 , month = jul, pages =. doi:10.1086/160126 , keywords =

    doi:10.1086/160126 1982

  50. [51]

    1988 , month = nov, pages =

    , title =. 1988 , month = nov, pages =. doi:10.1086/166834 , keywords =

    doi:10.1086/166834 1988

  51. [52]

    2021 , month = apr, number =

    Reviews of Modern Physics , title =. 2021 , month = apr, number =. doi:10.1103/RevModPhys.93.025010 , eid =

    doi:10.1103/revmodphys.93.025010 2021

  52. [53]

    2013 , month = oct, number =

    , title =. 2013 , month = oct, number =. arXiv , doi =:1302.2150 , keywords =

    Pith/arXiv arXiv 2013

  53. [54]

    2021 , month = aug, number =

    Universe , title =. 2021 , month = aug, number =. arXiv , doi =:2110.07956 , keywords =

    Pith/arXiv arXiv 2021

  54. [55]

    2020 , month = dec, pages =

    , title =. 2020 , month = dec, pages =. arXiv , doi =:2007.04294 , keywords =

    Pith/arXiv arXiv 2020

  55. [56]

    2021 , month = jun, pages =

    , title =. 2021 , month = jun, pages =. arXiv , doi =:2103.02375 , keywords =

    Pith/arXiv arXiv 2021

  56. [57]

    2022 , month = may, number =

    , title =. 2022 , month = may, number =. arXiv , doi =:2203.15624 , keywords =

    arXiv 2022

  57. [58]

    doi:10.1007/978-3-662-62728-0 , url =

    2021 , address =. doi:10.1007/978-3-662-62728-0 , url =

    work page doi:10.1007/978-3-662-62728-0 2021

  58. [59]

    doi:10.1007/978-3-662-55790-7 , url =

    2017 , address =. doi:10.1007/978-3-662-55790-7 , url =

    work page doi:10.1007/978-3-662-55790-7 2017

  59. [60]

    doi:10.1007/978-3-662-49094-5 , url =

    2016 , address =. doi:10.1007/978-3-662-49094-5 , url =

    work page doi:10.1007/978-3-662-49094-5 2016

  60. [61]

    doi:10.1007/978-3-662-52884-6 , url =

    2017 , address =. doi:10.1007/978-3-662-52884-6 , url =

    work page doi:10.1007/978-3-662-52884-6 2017

  61. [62]

    2022 , issn =

    Applied Sciences , title =. 2022 , issn =. doi:10.3390/app12020705 , url =

    work page doi:10.3390/app12020705 2022

  62. [63]

    2018 , month = aug, number =

    , title =. 2018 , month = aug, number =. arXiv , doi =:1804.09731 , keywords =

    Pith/arXiv arXiv 2018

  63. [64]

    2019 , month = aug, number =

    , title =. 2019 , month = aug, number =. doi:10.1103/PhysRevLett.123.071101 , eid =

    doi:10.1103/physrevlett.123.071101 2019

  64. [65]

    2005 , month = oct, number =

    , title =. 2005 , month = oct, number =. arXiv , doi =:astro-ph/0505258 , keywords =

    Pith/arXiv arXiv 2005

  65. [66]

    2016 , month = oct, number =

    , title =. 2016 , month = oct, number =. arXiv , doi =:1603.06570 , keywords =

    Pith/arXiv arXiv 2016

  66. [67]

    1983 , month = aug, number =

    Reports on Progress in Physics , title =. 1983 , month = aug, number =. doi:10.1088/0034-4885/46/8/002 , keywords =

    doi:10.1088/0034-4885/46/8/002 1983

  67. [68]

    2011 , month = aug, pages =

    , title =. 2011 , month = aug, pages =. arXiv , doi =:1009.4799 , keywords =

    Pith/arXiv arXiv 2011

  68. [69]

    2012 , month = dec, pages =

    , title =. 2012 , month = dec, pages =. arXiv , doi =:1203.3681 , keywords =

    Pith/arXiv arXiv 2012

  69. [70]

    2018 , month = apr, pages =

    , title =. 2018 , month = apr, pages =. doi:10.1016/j.nuclphysbps.2018.07.002 , url =

    work page doi:10.1016/j.nuclphysbps.2018.07.002 2018

  70. [71]

    2020 , doi =

    Evoli, Carmelo , month = dec, title =. 2020 , doi =

    2020

  71. [72]

    2022 , month = aug, number =

    , title =. 2022 , month = aug, number =. arXiv , doi =:2208.05457 , keywords =

    Pith/arXiv arXiv 2022

  72. [73]

    1949 , month = apr, number =

    Physical Review , title =. 1949 , month = apr, number =. doi:10.1103/PhysRev.75.1169 , url =

    doi:10.1103/physrev.75.1169 1949

  73. [74]

    1954 , month = jan, pages =

    , title =. 1954 , month = jan, pages =. doi:10.1086/145789 , url =

    doi:10.1086/145789 1954

  74. [75]

    doi:10.1007/978-1-4614-5782-4 , url =

    2013 , address =. doi:10.1007/978-1-4614-5782-4 , url =

    work page doi:10.1007/978-1-4614-5782-4 2013

  75. [76]

    2016 , month = oct, pages =

    arXiv e-prints , title =. 2016 , month = oct, pages =. arXiv , doi =:1610.07638 , keywords =

    Pith/arXiv arXiv 2016

  76. [77]

    2019 , month = jan, number =

    International Journal of Modern Physics D , title =. 2019 , month = jan, number =. arXiv , doi =:1903.11584 , keywords =

    Pith/arXiv arXiv 2019

  77. [78]

    1972 , month = aug, pages =

    , title =. 1972 , month = aug, pages =. doi:10.1086/151605 , url =

    doi:10.1086/151605 1972

  78. [79]

    2018 , month = sep, number =

    , title =. 2018 , month = sep, number =. arXiv , doi =:1807.09403 , keywords =

    Pith/arXiv arXiv 2018

  79. [80]

    2020 , month = dec, number =

    , title =. 2020 , month = dec, number =. arXiv , doi =:2008.12308 , keywords =

    Pith/arXiv arXiv 2020

  80. [81]

    2021 , month = dec, number =

    Living Reviews in Computational Astrophysics , title =. 2021 , month = dec, number =. arXiv , doi =:2106.08426 , keywords =

    Pith/arXiv arXiv 2021

Showing first 80 references.