Pith. sign in

REVIEW 2 major objections 5 minor 1 cited by

The Role of Electric Dominance for Particle Injection in Relativistic Reconnection

T0 review · 2 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper claims that in relativistic reconnection with zero guide field, regions where the electric field exceeds the magnetic field ($E>B$) supply more than 80% of the energy that lifts particles past the injection threshold, and that…

desk verdict Careful 2D PIC study with a real measurement, but the headline 80% result is tied to the fiducial box size and the abstract overstates its universality. read the letter →

arxiv 2501.00979 v1 pith:L35TKTVV submitted 2025-01-01 astro-ph.HE hep-ph

Sanya Gupta , Navin Sridhar , Lorenzo Sironi This is my paper · ORCID
classification astro-ph.HEhep-ph
keywords relativisticreconnectionparticleinjectionelectricdominanceE>Bregionsnonthermalaccelerationpair-plasmaparticle-in-cellsimulationmagnetization
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper asks which electromagnetic fields perform the early-stage 'injection' of particles to relativistic energies in magnetic reconnection with no guide field. Using two-dimensional particle-in-cell simulations, the authors measure, for every particle, what fraction of its energy was gained inside regions of electric dominance ($E>B$) before it crossed the injection threshold $\epsilon^\ast=\sigma/4$. They find that at magnetizations $\sigma\gtrsim50$, particles that will end up with $\epsilon_T\gtrsim8\sigma$ obtain more than 80% of their pre-threshold energy inside $E>B$ regions. This matters because it redirects the search for the reconnection accelerator from the ideal electric fields of outflows and plasmoids to the thin, non-ideal, electrically dominated dissipation regions.

What carries the argument

The load-bearing device is the fractional-contribution function $\zeta(\epsilon^\ast,\epsilon_T)$: for particles binned by their energy $\epsilon_T$ at time $T$, it records, at the instant each particle crosses the injection threshold $\epsilon^\ast$, the mean fraction of its total energy that was accumulated in regions where $\chi=(E^2-B^2)/(E^2+B^2)>0$. Evaluating this function requires backtracking each particle's energy record and comparing it with the local field invariant $\chi$, which identifies electric dominance locally without reference to a fluid frame. The companion ingredient is the test-particle experiment, in which the energy of tracer particles is held fixed while they are inside $E>B$ regions; the contrast between their spectra and the self-consistent spectra exposes how much of the nonthermal tail is produced by electric-dominance energization.

What would settle it

Run a 3D particle-in-cell simulation of zero-guide-field relativistic pair-plasma reconnection with outflow boundaries and measure $\zeta(\epsilon^\ast=\sigma/4,\epsilon_T)$ for $\sigma\gtrsim50$; if the fractional contribution for $\epsilon_T/\sigma\gtrsim8$ falls well below 80%, the central claim is falsified.

Watch

Extended reading notes

Core claim

The central claim is that in two-dimensional relativistic pair-plasma reconnection with vanishing guide field, the dominant agent of particle injection is not the ideal-field Fermi mechanism but the non-ideal electric fields found in $E>B$ regions. This is quantified by $\zeta(\epsilon^\ast,\epsilon_T)\equiv\langle \epsilon_\chi/\epsilon_{\rm tot}\rangle$ evaluated at the moment a particle's energy first reaches $\epsilon^\ast=\sigma/4$, where $\epsilon_\chi$ is the energy acquired while the local fields satisfy $\chi=(E^2-B^2)/(E^2+B^2)>0$. At high magnetization ($\sigma\gtrsim50$), $\zeta$ rises with the particle's final energy, reaching $\gtrsim80\%$ for $\epsilon_T/\sigma\gtrsim8$, and it is independent of box size when $\epsilon_T$ is normalized to the maximum particle energy, which scales as $\epsilon_{\max}\propto L_x^{1/2}$ in 2D. The distribution of the individual $E>B$ energy gains follows $dN/d\epsilon_\chi\propto \epsilon_\chi^{-0.35}\exp[-(\epsilon_\chi/0.06\,\sigma)^{0.5}]$, and test particles whose energization is artificially frozen inside $E>B$ regions develop much steeper nonthermal spectra. The authors conclude that electric dominance provides a large—possibly dominant—fraction of the non-ideal-field work throughout the injection stage.

Load-bearing premise

The claim rests on two-dimensional simulations; if three-dimensional reconnection reduces the frequency or energy yield of $E>B$ regions, the observed 80% fractional contribution could be a 2D artifact.

Editorial extensions

If this is right

  • For $\sigma\gtrsim50$, the early energization that lifts particles past $\epsilon^\ast=\sigma/4$ is dominated by $E>B$ regions rather than by ideal fields, so the injection stage cannot be modeled as a purely ideal Fermi process.
  • The fractional contribution $\zeta$ increases with both magnetization and final particle energy, so higher-$\sigma$ reconnection and the most energetic particles are the ones most dependent on electric dominance.
  • The box-size independence of $\zeta$ at fixed $\epsilon_T/\epsilon_{\max}$ means the injection physics is local and robust, and that $\epsilon_{\max}\propto L_x^{1/2}$ is the correct normalizing scale for comparing simulations.
  • The analytic fit $dN/d\epsilon_\chi\propto\epsilon_\chi^{-0.35}\exp[-(\epsilon_\chi/0.06\,\sigma)^{0.5}]$ gives a concrete prediction for the statistics of $E>B$ energy gains that can be checked in other simulations.
  • Suppressing energization in $E>B$ regions substantially steepens the particle spectrum, so electric dominance is required to produce the hardest nonthermal component in zero-guide-field reconnection.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the 2D result carries to 3D, global models of black-hole and neutron-star magnetospheres should locate particle injection in electrically dominated dissipation layers near X-points, not in the ideal outflow regions that dominate the highest-energy acceleration.
  • The measured energy-gain distribution suggests a physical picture the paper does not fully spell out: individual $E>B$ sites act as localized voltage drops, each contributing energy of order $0.06\sigma$, with the exponential cutoff encoding the maximum voltage a site can sustain; this could be tested by correlating $\epsilon_\chi$ with the local reconnection electric field at the moment of crossi
  • A natural extension would be to measure $\zeta$ in electron-ion reconnection; since ions decouple from the magnetic field at lower energies, the fractional contribution of $E>B$ regions may be species-dependent, providing a testable baseline for comparing pair-plasma and proton-electron simulations.
  • The box-size independence of $\zeta$ implies that sub-grid injection recipes in large-scale reconnection simulations could set the injected particle distribution using the local fraction $\zeta$ without resolving the layer, provided the $E>B$ region statistics are parametrized.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 5 minor

Summary. The paper uses 2D particle-in-cell simulations of relativistic pair-plasma reconnection with zero guide field and outflow boundaries to quantify the fractional contribution ζ of electric-dominance (E>B) regions to particle energization up to an injection threshold ε* = σ/4. The main measurement, Eq. (2), is the mean over particles in a final-energy bin εT of the ratio εχ/εtot evaluated at the moment the particle first reaches ε*. The authors find that ζ increases with magnetization and with εT/σ, reaching ≳80% for σ≳50 and εT/σ≳8 in the fiducial box (Lx = 768 c/ωp). They argue that the box-size dependence is removed when εT is normalized to the maximum particle energy, which scales as εmax ∝ Lx^{1/2} in 2D, as shown in Fig. 6. They also fit the distribution of energy gains εχ acquired in E>B regions with a power law times a stretched exponential (Eq. 3), and use a test-particle control to argue that E>B energization shapes the high-energy end of the particle spectrum. The paper concludes that electric dominance, rather than ideal-field Fermi processes alone, is the dominant injection agent in this regime, with caveats about 3D generalization and electron-ion plasmas acknowledged in Section 4.

Significance. If correct, this is a valuable quantitative step in a long-standing debate about particle injection in relativistic reconnection. The study is methodologically transparent: it presents convergence tests in Appendix A, time-averaging of ζ, a direct outflow-versus-periodic boundary comparison in Fig. 5, and a controlled test-particle experiment in Fig. 8 that isolates the role of E>B energization in shaping the high-energy spectrum. The fitting formula (Eq. 3) is a falsifiable prediction that can be checked in other codes and geometries. The explicit 2D caveat and the discussion of behavior toward the non-relativistic limit are appropriate. These strengths make the paper a solid contribution to the reconnection-acceleration literature, provided the quantitative headline is stated with the correct box-size qualifications.

major comments (2)
  1. [Abstract; Section 3.2; Section 4] The headline number 'for σ≳50 and εT/σ≳8, ≳80% of the energy gain occurs in E>B regions' is presented without the box size attached, but Fig. 6 shows that at fixed εT/σ the value of ζ systematically decreases with increasing Lx, and the curves for different box sizes overlap only when the abscissa is rescaled to εT/(σ Lx^{1/2}). In the fiducial box (Lx=768, √Lx≈27.7) the threshold εT/σ≳8 corresponds to εT/(σ√Lx)≳0.29. To make the claim universal, either the threshold should be expressed in the rescaled variable or the fiducial box size should be explicitly attached. As written, the abstract and the first paragraph of Section 4 overstate the universality of the 80% result, and the statement 'ζ is independent of simulation box size' in the abstract is misleading without immediately specifying the normalization condition. This is a load-bearing caveat because the paper's central quantitative conclusion is otherwise quoted as a single number independent of the simulation domain.
  2. [Section 3.2, Eq. (2)] The algorithm for accumulating εχ, the energy acquired in E>B (χ>0) regions, is not fully specified. The text defines εχ as 'the amount of kinetic energy acquired in regions of electric dominance' and 'equivalently, as the work done by E>B electric fields,' but it does not state whether this is computed by integrating q E·v over all timesteps in which the particle resides in χ>0 cells, or by summing the changes in Lorentz factor over such intervals, or by some hybrid procedure. Also ambiguous is how the spatial boundary of a χ>0 region is assigned (e.g., cell-by-cell using the local χ value, or requiring a minimum residence time). Because Eq. (2) is the central observable of the paper, a precise algorithmic definition is needed for reproducibility and for interpreting the quantitative values of ζ.
minor comments (5)
  1. [Table 1] The row 'outflow 50 1536 64' appears twice in the table; one of the duplicates should be removed.
  2. [Appendix D] The figure in Appendix D is referred to as 'Fig. C2' in the text, but the appendix is labeled D and the previous figure is C1. Rename it 'Fig. D1' for consistency.
  3. [Section 3.3, Eq. (3)] The best-fit parameters B=-0.35, D=0.5, and A=0.06σ are reported without uncertainties or a measure of goodness of fit. Since the fit is based on only two magnetizations and the stretched-exponential form has degenerate parameters, please provide at least a qualitative statement of the fitting range and sensitivity.
  4. [Section 4] In the sentence 'this scaling is appropriate only in 2D, see; e.g., Zhang et al. (2021, 2023)', the punctuation 'see; e.g.' should be 'see, e.g.,'.
  5. [Abstract and Section 1] The abstract states 'We find that ζ is independent of simulation box size Lx' but the full qualification is that this holds only after normalizing εT to εmax. Consider rewording to 'ζ depends only on εT/εmax, not on Lx separately' to avoid an unqualified independence claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the central ζ measurement is a direct simulation diagnostic, and the εmax scaling used for normalization is an independent, tested input rather than a derivation from the target claim.

full rationale

The paper's central claim—that ≳80% of the energy gain before reaching ε*=σ/4 occurs in E>B regions for σ≳50 and εT/σ≳8—is a direct measurement from PIC simulations, not a quantity derived from an assumed result. The diagnostic ζ(ε*,εT) is defined in Eq. 2 and computed by tracking individual particles' energy gains in χ>0 regions; this is self-consistent but not circular. Equation 3 is explicitly labeled as a best-fit model to the distribution of energy gains, and its later use to infer a mean energy gain is a derived consistency check, not a prediction from an input. The only prior scaling invoked, εmax∝Lx^{1/2}, is used to renormalize the horizontal axis in Fig. 6 and is attributed to previous theoretical and simulation work; the resulting curve collapse is presented as an empirical test rather than an input. The paper explicitly notes that without this normalization ζ decreases with box size, so the box-size dependence is disclosed as a scope condition, not hidden. While some citations are to prior work by co-author Sironi, none serve as a load-bearing uniqueness or derivation step; the 2D/3D agreement is mentioned only as a caveat, and the absence of a 3D E>B study is acknowledged. The skeptic's box-size concern is a correct caveat about the universality of the headline threshold but is a correctness/scope issue, not circularity.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim rests on domain assumptions about the PIC method, the quasi-steady state average, the 2D-to-3D transferability of injection physics, and the epsilon_max scaling used for box-size normalization. No new physical entities are introduced. The only fitted constants are the parameters of the empirical distribution in Eq. 3 and the adopted injection threshold epsilon* = sigma/4.

free parameters (4)
  • epsilon* (injection threshold) = sigma/4 (adopted, not fitted)
    The quoted 80 percent figure is defined relative to this threshold. The paper tests epsilon* = sigma/8, sigma/2, sigma, 2 sigma; the general trend persists but the numerical value depends on the choice.
  • B (power-law index in Eq. 3) = -0.35
    Best-fit index for dN/d epsilon_chi, obtained from sigma = 50 and 200 simulation data; no uncertainty is reported.
  • D (stretched-exponential index in Eq. 3) = 0.5
    Best-fit stretched-exponential index for the E>B energy-gain distribution; no uncertainty is reported.
  • A(sigma) (characteristic energy scale in Eq. 3) = 0.06 sigma
    Best-fit characteristic energy scale for the E>B energy-gain distribution; no uncertainty is reported.
assumptions (5)
  • domain assumption The PIC code TRISTAN-MP with a Vay pusher correctly evolves relativistic collisionless pair-plasma reconnection on the simulated scales.
    Invoked in Section 2; the entire study is a numerical experiment whose physical validity depends on this.
  • domain assumption After T vA/Lx > 1.5, the system is quasi-steady and time-averaging over windows of about 0.6 Lx/vA yields converged values of zeta.
    Used in Section 3.2 and Appendix A to define fiducial measurements; relies on the reconnection rate being stationary.
  • domain assumption Two-dimensional simulations with outflow boundaries are representative of 3D injection physics for zero-guide-field reconnection.
    Explicitly a caveat in Section 4; the authors cite prior 2D/3D agreement for injection but note that no dedicated 3D E>B study exists.
  • domain assumption The maximum particle energy in 2D scales as epsilon_max proportional to Lx^{1/2}, due to plasmoid trapping and compression.
    Used in Section 3.2 and Fig. 6 to normalize the horizontal axis; the box-size independence claim depends on this scaling.
  • domain assumption The injection threshold epsilon* = sigma/4 separates the injection stage from later acceleration.
    Adopted in Section 3.2 following Totorica et al. (2023); the mean E>B gain < epsilon_chi > about 0.18 sigma is used as supporting consistency.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Role of Electric Dominance for Particle Injection in Relativistic Reconnection." pith.science (2026). https://pith.science/paper/L35TKTVV

@misc{pith2026250100979,
  author       = {Pith},
  title        = {Pith review of: The Role of Electric Dominance for Particle Injection in Relativistic Reconnection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L35TKTVV}},
  note         = {Machine review of arXiv:2501.00979}
}
abstract

Magnetic reconnection in relativistic plasmas -- where the magnetization $\sigma\gg1$ -- is regarded as an efficient particle accelerator, capable of explaining the most dramatic astrophysical flares. We employ two-dimensional (2D) particle-in-cell simulations of relativistic pair-plasma reconnection with vanishing guide field and outflow boundaries to quantify the impact of the energy gain occurring in regions of electric dominance ($E>B$) for the early stages of particle acceleration (i.e., the ``injection'' stage). We calculate the mean fractional contribution $\zeta(\epsilon^\ast,\epsilon_{\rm T}$) by $E>B$ fields to particle energization up to the injection threshold energy, $\epsilon^\ast=\sigma/4$; here, $\epsilon_{\rm T}$ is the particle energy at time $T$. We find that $\zeta$ monotonically increases with $\sigma$ and $\epsilon_{\rm T}$; for $\sigma\gtrsim 50$ and $\epsilon_{\rm T}/\sigma\gtrsim 8$, we find that $\gtrsim 80\%$ of the energy gain obtained before reaching $\epsilon^\ast=\sigma/4$ occurs in $E>B$ regions. We find that $\zeta$ is independent of simulation box size $L_x$, as long as $\epsilon_{\rm T}$ is normalized to the maximum particle energy, which scales as $\epsilon_{\rm max}\propto L_{\rm x}^{1/2}$ in 2D. The distribution of energy gains $\epsilon_{\chi}$ acquired in $E>B$ regions can be modeled as $dN/d\epsilon_{\chi}\propto\epsilon_{\chi}^{-0.35}\exp[-(\epsilon_{\chi}/0.06\,\sigma)^{0.5}]$. Our results help assess the role of electric dominance in relativistic reconnection with vanishing guide fields, which may be realized in the magnetospheres of black holes and neutron stars.

Figures

Figures reproduced from arXiv: 2501.00979 by the authors.

Figure 1
Figure 1. Reconnection rate (i.e., inflow velocity) in units of the Alfvén speed, 𝑣in/𝑣A, as a function of time (in units of 𝐿x/𝑣A). Colors represent different magnetizations: 𝜎 = 12.5 (blue), 50 (orange), 200 (green). We show cases with outflow boundary conditions (denoted ‘o’) using solid curves and cases with periodic boundaries (denoted ‘p’) using dashed translucent curves. we employ two moving injectors—that constantly i… view at source ↗
Figure 2
Figure 2. Structure of the reconnection layer at time 𝑇 𝑣A/𝐿x ∼ 2.8 (after the system has settled into a quasi-steady state) for our fiducial outflow run with magnetization 𝜎 = 50 and box size 𝐿x = 768 𝑐/𝜔p. We show a subset of the domain: the range in 𝑥 is −0.8 < 𝑥/𝐿x < 0.8, while along 𝑦 we focus on −0.25 < 𝑦/𝐿x < 0.25 for panels [a],[b],[c] and −0.15 < 𝑦/𝐿x < 0.15 for panels [d],[e]. We present: [a] the particle number den… view at source ↗
Figure 3
Figure 3. Average fractional contribution of 𝐸 > 𝐵 regions to particle energization, for our fiducial simulation (𝜎 = 50, 𝐿x = 768 𝑐/𝜔p, and outflow boundaries). We plot 𝜁 ( 𝜖 ∗ , 𝜖T) ≡ ( 𝜖𝜒/𝜖tot) | 𝜖tot=𝜖 ∗ on the vertical axis and 𝜖T/𝜎 on the horizontal axis. Each panel refers to a different injection threshold, 𝜖 ∗ = 𝜎/4, 𝜎/2, 𝜎 from left to right. In each panel, different curves refer to different time intervals over whic… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Same as [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Same as [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Same as [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: Left panel: the distribution of 𝜖𝜒 = 𝛾𝜒 − 1 > 0, i.e., of the energy gains obtained in regions of electric dominance before time 𝑇 = 2.8 𝐿x/𝑣A. Right panel: distribution of particle energies for all the particles (𝜖T; solid lines) and for the particles contributing to …
Figure 8
Figure 8. Figure 8: Dashed lines: self-consistent particle spectra from our simulations—time-averaged from 𝑇 𝑣A/𝐿x ≃ 2.0 to 2.8—as a function of magnetization (see legend). Solid lines: energy spectra of test-particles ini￾tialized and evolved as regular particles, but such that their ene…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Particle Injection Problem in Magnetic Reconnection and Turbulence

    physics.plasm-ph 2025-06 conditional novelty 3.0 of 10

    A review of the particle injection problem in magnetic reconnection and turbulence, arguing that injection is set by direct acceleration, Fermi kicks, and pickup processes, not by E>B diffusion regions.

Reference graph

Works this paper leans on

25 extracted references · 4 canonical work pages · cited by 1 Pith paper

  1. [1]

    Chernoglazov A., Hakobyan H., Philippov A., 2023, @doi [ ] 10.3847/1538-4357/acffc6 , https://ui.adsabs.harvard.edu/abs/2023ApJ...959..122C 959, 122

    doi:10.3847/1538-4357/acffc6 2023

  2. [2]

    A., 2023, @doi [The Astrophysical Journal] 10.3847/1538-4357/acb7dd , 948, 19

    French O., Guo F., Zhang Q., Uzdensky D. A., 2023, @doi [The Astrophysical Journal] 10.3847/1538-4357/acb7dd , 948, 19

    doi:10.3847/1538-4357/acb7dd 2023

  3. [3]

    Guo F., Li H., Daughton W., Liu Y.-H., 2014, @doi [Phys. Rev. Lett.] 10.1103/PhysRevLett.113.155005 , 113, 155005

    doi:10.1103/physrevlett.113.155005 2014

  4. [4]

    Guo F., Liu Y.-H., Daughton W., Li H., 2015, @doi [ ] 10.1088/0004-637X/806/2/167 , https://ui.adsabs.harvard.edu/abs/2015ApJ...806..167G 806, 167

    doi:10.1088/0004-637x/806/2/167 2015

  5. [5]

    Guo F., Li X., Daughton W., Kilian P., Li H., Liu Y.-H., Yan W., Ma D., 2019, @doi [ ] 10.3847/2041-8213/ab2a15 , https://ui.adsabs.harvard.edu/abs/2019ApJ...879L..23G 879, L23

    doi:10.3847/2041-8213/ab2a15 2019

  6. [6]

    Guo F., Liu Y.-H., Li X., Li H., Daughton W., Kilian P., 2020, @doi [Physics of Plasmas] 10.1063/5.0012094 , https://ui.adsabs.harvard.edu/abs/2020PhPl...27h0501G 27, 080501

    doi:10.1063/5.0012094 2020

  7. [7]

    Guo F., et al., 2023, @doi [ ] 10.1103/PhysRevLett.130.189501 , https://ui.adsabs.harvard.edu/abs/2023PhRvL.130r9501G 130, 189501

    work page doi:10.1103/physrevlett.130.189501 2023

  8. [8]

    Guo F., Liu Y.-H., Zenitani S., Hoshino M., 2024, @doi [ ] 10.1007/s11214-024-01073-2 , https://ui.adsabs.harvard.edu/abs/2024SSRv..220...43G 220, 43

    doi:10.1007/s11214-024-01073-2 2024

Show all 25 references
  1. [9]

    Hakobyan H., Petropoulou M., Spitkovsky A., Sironi L., 2021, @doi [ ] 10.3847/1538-4357/abedac , https://ui.adsabs.harvard.edu/abs/2021ApJ...912...48H 912, 48

    2021 doi

  2. [10]

    Hoshino M., Lyubarsky Y., 2012, @doi [SSRv] 10.1007/s11214-012-9931-z , http://adsabs.harvard.edu/abs/2012SSRv..173..521H 173, 521

    2012 doi

  3. [11]

    Kagan D., Sironi L., Cerutti B., Giannios D., 2015, @doi [Space Science Reviews] 10.1007/s11214-014-0132-9 , https://ui.adsabs.harvard.edu/abs/2015SSRv..191..545K 191, 545

    2015 doi

  4. [12]

    A., Lovelace R

    Larrabee D. A., Lovelace R. V. E., Romanova M. M., 2003, @doi [ ] 10.1086/367640 , https://ui.adsabs.harvard.edu/abs/2003ApJ...586...72L 586, 72

    2003 doi

  5. [13]

    Lyubarsky Y., Liverts M., 2008, @doi [ ] 10.1086/589640 , https://ui.adsabs.harvard.edu/abs/2008ApJ...682.1436L 682, 1436

    2008 doi

  6. [14]

    A., Cerutti B., Werner G

    Nalewajko K., Uzdensky D. A., Cerutti B., Werner G. R., Begelman M. C., 2015, @doi [The Astrophysical Journal] 10.1088/0004-637X/815/2/101 , 815, 101

    2015 doi

  7. [15]

    Petropoulou M., Sironi L., 2018, @doi [ ] 10.1093/mnras/sty2702 , https://ui.adsabs.harvard.edu/abs/2018MNRAS.481.5687P 481, 5687

    2018 doi

  8. [16]

    Sironi L., 2022, @doi [ ] 10.1103/PhysRevLett.128.145102 , https://ui.adsabs.harvard.edu/abs/2022PhRvL.128n5102S 128, 145102

    2022 doi

  9. [17]

    Sironi L., Giannios D., Petropoulou M., 2016, @doi [ ] 10.1093/mnras/stw1620 , https://ui.adsabs.harvard.edu/abs/2016MNRAS.462...48S 462, 48

    2016 doi

  10. [18]

    Spitkovsky A., 2005, @doi [AIP Conference Proceedings] 10.1063/1.2141897 , 801, 345

    2005 doi

  11. [19]

    R., Zenitani S., Matsukiyo S., Machida M., Sekiguchi K., Bhattacharjee A., 2023, @doi [The Astrophysical Journal Letters] 10.3847/2041-8213/acdb60 , 952, L1

    Totorica S. R., Zenitani S., Matsukiyo S., Machida M., Sekiguchi K., Bhattacharjee A., 2023, @doi [The Astrophysical Journal Letters] 10.3847/2041-8213/acdb60 , 952, L1

    2023 doi

  12. [20]

    A., 2022, @doi [Journal of Plasma Physics] 10.1017/S0022377822000046 , https://ui.adsabs.harvard.edu/abs/2022JPlPh..88a9014U 88, 905880114

    Uzdensky D. A., 2022, @doi [Journal of Plasma Physics] 10.1017/S0022377822000046 , https://ui.adsabs.harvard.edu/abs/2022JPlPh..88a9014U 88, 905880114

    2022 doi

  13. [21]

    Vay J.-L., 2008, @doi [Physics of Plasmas] 10.1063/1.2837054 , 15

    2008 doi

  14. [22]

    R., Uzdensky D

    Werner G. R., Uzdensky D. A., Cerutti B., Nalewajko K., Begelman M. C., 2016, @doi [ ] 10.3847/2041-8205/816/1/L8 , https://ui.adsabs.harvard.edu/abs/2016ApJ...816L...8W 816, L8

    2016 doi

  15. [23]

    Zenitani S., Hoshino M., 2001, @doi [ ] 10.1086/337972 , https://ui.adsabs.harvard.edu/abs/2001ApJ...562L..63Z 562, L63

    2001 doi

  16. [24]

    Zhang H., Sironi L., Giannios D., 2021, @doi [ ] 10.3847/1538-4357/ac2e08 , https://ui.adsabs.harvard.edu/abs/2021ApJ...922..261Z 922, 261

    2021 doi

  17. [25]

    Zhang H., Sironi L., Giannios D., Petropoulou M., 2023, @doi [ ] 10.3847/2041-8213/acfe7c , https://ui.adsabs.harvard.edu/abs/2023ApJ...956L..36Z 956, L36

    2023 doi

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.