Pith. sign in

REVIEW 3 major objections 4 minor 41 references

Future mm-VLBI of M87 can separate turbulent from reconnection electron heating and thermal from non-thermal particle distributions.

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 · grok-4.5

2026-07-14 05:35 UTC pith:ZJVPVRYA

load-bearing objection Solid multi-frequency GRMHD+GRRT diagnostics for M87 heating models; the abstract claim is a bit ahead of the idealised maps, but the work is still useful and referee-ready. the 3 major comments →

arxiv 2607.11439 v1 pith:ZJVPVRYA submitted 2026-07-13 astro-ph.HE astro-ph.GA

Probing radiation micro-physics in M 87 I. Total intensity and broad-band spectra

Christian M. Fromm , Yosuke Mizuno , Ziri Younsi , Ainara Saiz-Perez , Hector Olivares , Antonios Nathanail , Alejandro Cruz-Osorio , Matthias Kadler
show 1 more author
Karl Mannheim
This is my paper
classification astro-ph.HE astro-ph.GA
keywords M87black-hole accretionGRMHDelectron heatingkappa distributionVLBIspectral indexjets
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 asks whether next-generation millimetre VLBI arrays can distinguish how electrons are heated near M87’s black hole and whether those electrons follow a thermal or a hybrid thermal-plus-power-law distribution. It runs three-dimensional general-relativistic magnetohydrodynamic simulations of a magnetically arrested disk around a rapidly spinning Kerr black hole, applying two sub-grid heating prescriptions (turbulent cascade and magnetic reconnection) and both Maxwell–Jüttner and kappa electron distributions. Synthetic multi-frequency images and spectra from 86 GHz to 345 GHz show that total-intensity snapshots alone look similar, yet spectral-index maps, component decompositions, and optically-thin slopes differ systematically between the models. With a dynamic range of order 10^4 these differences become observationally accessible, giving a concrete route to diagnose microphysics that single-frequency images cannot resolve.

Core claim

With a dynamical range of 10^4 and simultaneous coverage from 86 GHz to 345 GHz, future VLBI arrays can observationally separate turbulent versus magnetic-reconnection electron heating and thermal versus hybrid kappa electron distributions in M87, even though the integrated spectra and single-frequency total-intensity images remain largely degenerate.

What carries the argument

Two-temperature MAD GRMHD runs with explicit electron-entropy evolution under turbulent (Kawazura-type) and reconnection (Rowan-type) heating fractions, post-processed with GRRT that includes both Maxwell–Jüttner and kappa distributions whose parameters are set by local magnetisation and plasma beta.

Load-bearing premise

The electron-heating fractions and kappa slopes taken from idealised particle-in-cell simulations remain valid when applied as sub-grid prescriptions to the global, time-dependent magnetically arrested flow.

What would settle it

A multi-frequency VLBI campaign of M87 that reaches dynamic range ~10^4 between 86 GHz and 345 GHz and measures spectral-index maps on 50–150 µas scales; if those maps show no systematic difference between the jet sheath and the highly magnetised interior, the claimed separability fails.

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

If this is right

  • Spectral-index maps between 86 GHz and 230 GHz become the primary observable for distinguishing heating mechanisms, not single-frequency total-intensity images.
  • Hybrid thermal-plus-kappa models are required to match both the radio-to-NIR spectrum and the extended jet emission at millimetre wavelengths.
  • The radial transition from steep (disk-dominated) to flat (jet-dominated) spectral index can locate the non-thermal particle injection radius.
  • Turnover frequency and turnover flux density are largely insensitive to heating model and electron distribution, while the optically thin spectral index remains diagnostic.

Where Pith is reading between the lines

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

  • If the same spectral-index contrast appears in other low-luminosity AGN, the method becomes a general diagnostic rather than an M87-specific tool.
  • Polarimetric extensions of the same multi-frequency campaign would further break remaining degeneracies between heating models once Faraday rotation and ordered-field geometry are included.
  • The requirement for dynamic range 10^4 sets a concrete performance target for ngEHT and ngVLA array design and calibration strategies.

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. This paper presents long-term 3D GRMHD simulations of MAD accretion onto a high-spin Kerr black hole (a*=0.9375) that evolve electron entropy under two sub-grid heating prescriptions (turbulent from Kawazura et al. 2019 and reconnection from Rowan et al. 2017). The snapshots are post-processed with GRRT (BHOSS) using thermal Maxwell–Jüttner and hybrid thermal–kappa electron distribution functions (with kappa and width parametrized from PIC results of Ball et al. 2018 and Meringolo et al. 2023). For M87 parameters the authors compute broadband spectra, 230 GHz images with disk/jet decompositions, 86–230 GHz spectral-index maps, optical-depth profiles, and turnover-frequency/flux maps. They conclude that, while total-intensity images and integrated spectra are largely degenerate, spectral-index structure and optically-thin slopes differ systematically between heating models and eDFs, so that future VLBI arrays with dynamical range ~10^4 over 86–345 GHz have the potential to distinguish them.

Significance. If the reported spectral-index and turnover distinctions survive realistic observing conditions, the work supplies concrete, falsifiable predictions that link sub-grid electron microphysics to multi-frequency VLBI observables of M87 on scales from the photon ring to ~1000 µas. This is timely for ngEHT and ngVLA planning. Strengths include the consistent two-temperature GRMHD implementation, PIC-calibrated heating and kappa recipes, careful component decompositions, optical-depth analysis, radial profiles, and the parameter explorations already present in the appendices. The calculations are performed with publicly documented codes (BHAC, BHOSS) and therefore in principle reproducible.

major comments (3)
  1. [Abstract and Section 6] The abstract and §6 claim that a dynamical range of 1×10^4 over 86–345 GHz is sufficient for future VLBI to distinguish turbulent versus reconnection heating and thermal versus kappa eDFs. This rests entirely on idealised, time-averaged GRRT images and maps (Figs. 6, 8–11) that are never passed through a realistic array response, thermal noise, sparse uv-coverage or imaging regularisation. The paper itself emphasises that total-intensity images are largely degenerate (§5) and that the discriminatory power resides in spectral-index structure; without a controlled recovery test it remains unproven that those differences remain detectable at the quoted dynamic range. Either synthetic ngEHT/ngVLA observations or an explicit qualification of the claim is required.
  2. [Section 3 and Appendix B] The free parameters ε = 0.5 and r_inj = 10 M are fixed to values previously chosen to fit the broadband SED and jet width, while ṁ is iterated to match 1 Jy of compact 230 GHz flux. Appendix B demonstrates that the spatial extent of the steep-to-flat spectral-index transition depends on both parameters. The main-text results and the abstract claim should quantify how the distinguishability between heating models persists (or degrades) across a plausible range of ε and r_inj rather than presenting primarily the single best-fit case.
  3. [Section 2, Eqs. (2)–(3) and (8)–(9)] The electron-heating fractions (Eqs. 2–3) and kappa parametrizations (Eqs. 8–9) are taken from idealised PIC simulations and applied as sub-grid models. The authors correctly flag the limitations (omission of compressive fluctuations, full reconnection geometries and non-local effects). Because the quantitative spectral-index differences that underpin the central claim are generated by these prescriptions, a short robustness test—e.g., modest variations of the functional forms within published uncertainties, or a direct comparison with a simple R–β model—would strengthen that the reported distinctions are not artefacts of the specific sub-grid choices.
minor comments (4)
  1. [Throughout] Several typographical errors remain (e.g., “Appdenix B”, “moti-vated”, inconsistent spacing in “eDF : thermal”). A careful proof-reading pass is needed.
  2. [Figures 2 and 4] Figure 2 and Figure 4 colour bars use non-standard symbols (æ, Øp, £e); these should be replaced by conventional σ, β_p, Θ_e for readability.
  3. [Section 3] The viewing angle is stated as ϑ = 160° in the text but the jet-axis orientation in the images is not explicitly related to the observer’s line of sight; a short clarifying sentence would help.
  4. [Section 2] The floor model and σ_cut = 3 are standard, yet the residual North–South asymmetries after time/azimuthal averaging (Fig. 2) are attributed partly to floors. A quantitative statement of how much of the jet-spine emission is discarded would be useful.

Circularity Check

2 steps flagged

Minor conventional normalizations (230 GHz flux match and ε=0.5 chosen from prior SED/jet-width fits) do not force the spectral-index distinguishability that is the paper's main claim.

specific steps
  1. fitted input called prediction [Section 3, paragraph on mass-accretion-rate normalisation]
    "For the final radiative transfer we fix the mass, distance and viewing angle (ϑ=160°) of M87 while iterating the mass accretion rate to match an average flux of 1.0 Jy at 230 GHz for the compact emission region (r∼100 µas) over 28000 M ≤ t ≤ 30000 M. The resulting accretion rates are 2.79 (2.59)×10−5 M⊙ yr−1 for magnetic reconnection heating without/with non-thermal particles and 2.72 (2.47)×10−5 M⊙ yr−1 for turbulent heating."

    The absolute flux scale of every model is forced to the observed 1 Jy compact flux at 230 GHz by construction. This is a conventional overall normalisation; it does not dictate the spectral-index or multi-frequency morphology differences that constitute the paper's distinguishability claim, so the circularity is minor.

  2. self citation load bearing [Section 3, choice of ε and r_inj]
    "The only free parameter in our GRRT setup is the fraction of magnetic energy, ε, that determines the width of the κ distribution (see Eq. 7). We study ε=0.5, as this value provides the best simultaneous fit to the broadband spectral energy distribution of M87 and the observed jet width (see Fromm et al. 2022; Cruz-Osorio et al. 2021)."

    ε=0.5 is adopted because prior papers by overlapping authors found it best matches the SED and jet width. The choice is therefore not derived from first principles inside the present work. It is not load-bearing for the heating-model contrast (the paper also shows thermal-only runs), so the circularity remains mild.

full rationale

The paper's central result is that turbulent vs reconnection heating and thermal vs kappa eDFs produce distinguishable multi-frequency signatures (especially spectral-index maps and optically-thin slopes) once idealised GRRT images are formed. Those differences arise from the distinct spatial dependence of the imported heating fractions (Eqs. 2–3) and kappa prescriptions (Eqs. 8–9) applied to the same MAD GRMHD flow; they are not tautological restatements of the inputs. The only steps that reduce to fitted or previously-chosen quantities are the overall mass-accretion-rate normalisation (iterated so that the compact 230 GHz flux equals 1 Jy) and the choice ε=0.5 (selected because earlier work by overlapping authors found it best matches the broadband SED and jet width). Both are standard practice in the field and leave the relative spectral-index contrasts free. Self-citations to Mizuno et al. (2021), Fromm et al. (2022), Cruz-Osorio et al. (2021) and Zhang et al. (2024) supply the heating implementation and the ε preference but do not constitute a uniqueness theorem that forbids alternatives; the paper itself notes the limitations of the sub-grid recipes. No self-definitional loop, no prediction that is forced by construction, and no renaming of a known empirical pattern appear. Score 2 reflects only the conventional, non-load-bearing normalisations.

Axiom & Free-Parameter Ledger

5 free parameters · 5 axioms · 0 invented entities

The central claim rests on standard GRMHD/GRRT machinery plus two literature-calibrated sub-grid heating recipes and a handful of free parameters (ε, r_inj, accretion-rate normalization, σ-cut) that are fixed by matching existing M87 data or by numerical necessity. No new physical entities are postulated; the novelty is in the multi-frequency comparison of existing models.

free parameters (5)
  • ε (magnetic-energy fraction for non-thermal particles) = 0.5
    Set to 0.5 because it previously provided the best simultaneous fit to the broadband SED and jet width of M87; controls the width of the κ distribution (Eq. 7).
  • r_inj (non-thermal injection radius) = 10 M
    Phenomenological scale set to ~10 M (stagnation surface); Appendix B shows that larger values expand the steep-spectral-index core region.
  • mass accretion rate ṁ = ≈2.5–2.8e-5 Msun/yr
    Iterated so that the compact 230 GHz flux averages 1.0 Jy; different values for each heating/eDF combination (≈2.5–2.8 × 10^{-5} M_⊙ yr^{-1}).
  • σ_cut = 3
    Regions with σ > 3 are excluded from GRRT to avoid floor-dominated jet spine; value taken from prior EHT modeling papers.
  • black-hole spin a* = 0.9375
    Fixed at 0.9375 following earlier EHT-motivated MAD models; not varied.
axioms (5)
  • domain assumption Turbulent (Kawazura et al. 2019) and reconnection (Rowan et al. 2017) heating fractions adequately capture electron energization in global MAD flows
    Adopted in §2; paper notes the prescriptions omit compressive fluctuations and full reconnection geometries.
  • domain assumption κ(σ, β_p) parametrizations from Ball et al. (2018) and Meringolo et al. (2023) correctly describe non-thermal tails
    Used in Eqs. 8–9; restricted to 3 < κ ≤ 8 for validity of synchrotron coefficients.
  • domain assumption MAD models better describe EHT observations of M87 than SANE models
    Stated in §1 citing EHT Collaboration 2019b, 2021b; justifies restricting the study to MAD.
  • domain assumption Synchrotron emission dominates from radio to NIR; Bremsstrahlung and inverse Compton can be neglected below ~10^{16} Hz
    §3; standard for mm-VLBI modeling of M87.
  • ad hoc to paper Adiabatic index γ̂ = 4/3 and floor prescriptions (ρ_fl, p_fl, electron-pressure bounds) do not qualitatively alter the observable differences outside the excluded spine
    Numerical necessities stated in §2; paper acknowledges floors affect the jet spine and therefore applies σ-cut.

pith-pipeline@v1.1.0-grok45 · 25649 in / 3195 out tokens · 45046 ms · 2026-07-14T05:35:09.771430+00:00 · methodology

0 comments
read the original abstract

Next generation Very Long Baseline Interferometers (VLBI) will provide dense sampling of the Fourier space together with high signal to noise ratios allowing to reliably observe and image faint jet structure in M 87 at mm-wavelength. The proposed next generation Event Horizon Telescope (ngEHT) and next generation Very Large Array (ngVLA) offers the unique capability to simultaneously resolve and image the accretion flow around the supermassive black hole in M 87 together with the jet launching and acceleration zone. In order to explore these capabilities and to provide theoretical expectations we perform general relativistic magnetohydrodynamic simulations of accretion on to black holes and jet launching. M 87 has been the target for multiple observations across the entire electromagnetic spectrum. Among these VLBI observations provide unique capability to resolve the jet structure down to several gravitational radii. In this work we provide possible observable signatures which will allow us to distinguish between different electron heating models and particle distributions. We use general relativistic magnetohydrodynamics and simulate the accretion of the magnetised plasma onto Kerr-black holes in 3D. The multi-frequency radiative signatures of these simulations are computed taking different electron heating and distribution functions into account. The results of our simulations show that with a dynamical range of $1\times 10^4$ and a frequency range from 86 GHz to 345 GHz observations with future VLBI arrays have the potential to tell turbulent and magnetic reconnection electron heating and the electron distribution function apart.

Figures

Figures reproduced from arXiv: 2607.11439 by Ainara Saiz-Perez, Alejandro Cruz-Osorio, Antonios Nathanail, Christian M. Fromm, Hector Olivares, Karl Mannheim, Matthias Kadler, Yosuke Mizuno, Ziri Younsi.

Figure 1
Figure 1. Figure 1: Mass accretion rates ˙m and MAD flux parameter ϕ = Φ/ √ m˙ in code units, where Φ is the magnetic flux across the horizon, for a MAD GRMHD simulation with a⋆ = 0.9375. °500 °250 0 °500 °250 0 250 500 r cos( µ) [M] r sin(µ) [M] 0 250 500 °500 °250 0 mag.recon. heating r sin(µ) [M] 0 250 500 turbulent heating -3 -2 -1 0 1 2 3 log10 æ -3 -2 -1 0 1 2 3 log10 Øp -2 -1 0 1 2 3 4 log10 £e °500 °250 0 °500 °250 0 … view at source ↗
Figure 2
Figure 2. Figure 2: Azimuthally and time-averaged (28000 M ≤ t ≤ 30000 M) dis￾tributions of the magnetisation σ (left half of panel A), plasma beta βp (right half of panel A), and electron temperature for magnetic recon￾nection heating (left half of panel B) and turbulent heating (right half of panel B). Solid black lines mark σ = 3, while dashed lines indicate the boundary between bound (−hut < 1.02) and unbound (−hut > 1.02… view at source ↗
Figure 3
Figure 3. Figure 3: Time-averaged radial profiles over 28000 M ≤ t ≤ 30000 M showing, from left to right, the rest-mass density ρ, magnetic field strength B, and dimensionless electron temperature Θe for both turbulent and magnetic reconnection heating. Profiles are further decomposed into contribu￾tions from the disk and jet, while the shaded bands indicate the corresponding 1σ variability. Article number, page 4 of 15 [PIT… view at source ↗
Figure 4
Figure 4. Figure 4: Azimuthally and time-averaged (28000 M ≤ t ≤ 30000 M) dis￾tributions of the κ (panel A) and the width of the κ-edf, w (panel B). In both panels, the left half corresponds to magnetic reconnection heating, while the right half shows the turbulent heating model. Solid black lines mark σ = 3, while dashed lines indicate the boundary between bound (−hut < 1.02) and unbound (−hut > 1.02) plasma. In panel A of … view at source ↗
Figure 5
Figure 5. Figure 5: Average broad-band spectrum (using the interval 28000 M ≤ t ≤ 30000 M) for turbulent heating (top) and magnetic reconnection heating (bottom) for thermal (left) and hybrid thermal-kappa eDF (right) the compact emission region (r ∼ 100 µas) over 28000 M ≤ t ≤ 30000 M. The resulting accretion rates are 2.79 (2.59) × 10−5 M⊙ yr−1 for magnetic reconnection heating without/with non-thermal particles and 2.72 (2… view at source ↗
Figure 6
Figure 6. Figure 6: Average 230 GHz images (left column) and their decomposition into counter jet, disk, and forward jet for turbulent (top two rows) and magnetic reconnection heating (bottom two rows). The first and third rows assume a thermal eDF, while the second and fourth include a thermal+ κ eDF. Insets show zooms of the horizon-scale structure. Images are time-averaged over 28000 M–30000 M and computed at a viewing ang… view at source ↗
Figure 7
Figure 7. Figure 7: Time-averaged opacity distribution, τν, for the turbulent heating model. Top: two-dimensional map of τν at 86 GHz. Bottom: opacity profiles along the projected jet axis (dashed line in the top panel) for 86, 230, and 345 GHz. ∼ 50–100 µas, the opacity rapidly decreases and the flow be￾comes optically thin. Consequently, opacity effects can influence the spectral index close to the ring and photon orbit, wh… view at source ↗
Figure 8
Figure 8. Figure 8: Same as [PITH_FULL_IMAGE:figures/full_fig_p009_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Flux densities at 86 GHz (left), 230 GHz (middle) and spectral index (right) along the jet axis (dashed line in Figs. 6 and 8) for different heating mechanism and eDFs. therefore directly traces the non-thermal particle population in the jet. 5. Discussion The high-frequency emission considered in this work (ν ≳ 86 GHz) is largely optically thin and therefore closely traces the underlying synchrotron emiss… view at source ↗
Figure 10
Figure 10. Figure 10: Distribution of the turnover frequency, νt (first column) and turnover flux density, S t (second column) and average optically spectral index (third and fourth column) for reconnection heating (top) and turbulent heating (bottom). The third column shows the spectral index for a thermal eDF and the fourth one for hybrid eDF consisting of thermal and kappa distribution. Notice that the turnover frequency an… view at source ↗
Figure 11
Figure 11. Figure 11: Turnover frequency (left), turnover flux density (middle) and average optically thin spectral index along the jet axis (dashed line in Figs. 10)for different heating mechanism and eDFs. 10−8 10−7 10−6 10−5 10−4 S [Jy /pix] position A position B position C position D mag. recon. thermal turb. thermal mag. recon. thermal+kappa (ε=0.5) turb. thermal+kappa (ε=0.5) 109 1010 1011 1012 ν [Hz] 10−9 10−8 10−7 10−6… view at source ↗
Figure 12
Figure 12. Figure 12: Individual spectra (top) and the decomposition (bottom) for the positions indicated in [PITH_FULL_IMAGE:figures/full_fig_p011_12.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

41 extracted references · 3 linked inside Pith

  1. [1]

    C., Balokovi´c, M., Chandra, S., et al

    Algaba, J. C., Balokovi´c, M., Chandra, S., et al. 2024, A&A, 692, A140

    2024

  2. [2]

    & Nakamura, M

    Asada, K. & Nakamura, M. 2012, ApJ, 745, L28

    2012

  3. [3]

    2018, Astrophys

    Ball, D., Sironi, L., & Özel, F. 2018, Astrophys. J., 862, 80

    2018

  4. [4]

    H., Emslie, A

    Bian, N. H., Emslie, A. G., Stackhouse, D. J., & Kontar, E. P. 2014, ApJ, 796, 142

    2014

  5. [5]

    2019, ARA&A, 57, 467

    Blandford, R., Meier, D., & Readhead, A. 2019, ARA&A, 57, 467

    2019

  6. [6]

    Blandford, R. D. & Königl, A. 1979, ApJ, 232, 34

    1979

  7. [7]

    Blandford, R. D. & Payne, D. G. 1982, Mon. Not. R. Astron. Soc., 199, 883

    1982

  8. [8]

    Blandford, R. D. & Znajek, R. L. 1977, Mon. Not. R. Astron. Soc., 179, 433

    1977

  9. [9]

    2025, MNRAS, 537, 2496

    Chael, A. 2025, MNRAS, 537, 2496

    2025

  10. [10]

    A., Johnson, M

    Chael, A. A., Johnson, M. D., Bouman, K. L., et al. 2018, ApJ, 857, 23

    2018

  11. [11]

    M., Mizuno, Y ., et al

    Cruz-Osorio, A., Fromm, C. M., Mizuno, Y ., et al. 2021, Nature Astronomy [arXiv:2111.02517]

    Pith/arXiv arXiv 2021

  12. [12]

    M., et al

    Cruz-Osorio, A., Meringolo, C., Fromm, C. M., et al. 2026, ApJ, 1001, 227

    2026

  13. [13]

    2023, Nature, 621, 711

    Cui, Y ., Hada, K., Kawashima, T., et al. 2023, Nature, 621, 711

    2023

  14. [14]

    2019, Astron

    Davelaar, J., Olivares, H., Porth, O., et al. 2019, Astron. Astrophys., 632, A2

    2019

  15. [15]

    C., & Agol, E

    Dexter, J., McKinney, J. C., & Agol, E. 2012, Mon. Not. R. Astron. Soc., 421, 1517 EHT Collaboration, Akiyama, K., Alberdi, A., et al. 2024, A&A, 681, A79 EHT Collaboration, Akiyama, K., Alberdi, A., et al. 2019a, Astrophys. J. Lett., 875, L1 EHT Collaboration, Akiyama, K., Alberdi, A., et al. 2019b, Astrophys. J. Lett., 875, L5 EHT Collaboration, Akiyama...

    2012

  16. [16]

    Fishbone, L. G. & Moncrief, V . 1976, Astrophys. J., 207, 962

    1976

  17. [17]

    Font, J. A. & Daigne, F. 2002, Astrophys. J, 581, L23

    2002

  18. [18]

    M., Cruz-Osorio, A., Mizuno, Y ., et al

    Fromm, C. M., Cruz-Osorio, A., Mizuno, Y ., et al. 2022, A&A, 660, A107

    2022

  19. [19]

    2013, Astrophys

    Hada, K., Kino, M., Doi, A., et al. 2013, Astrophys. J., 775, 70

    2013

  20. [20]

    Kawazura, Y ., Barnes, M., & Schekochihin, A. A. 2019, Proceedings of the Na- tional Academy of Science, 116, 771

    2019

  21. [21]

    Y ., Krichbaum, T

    Kim, J. Y ., Krichbaum, T. P., Lu, R. S., et al. 2018, Astron. Astrophys., 616, A188

    2018

  22. [22]

    & Fichtner, H., eds

    Lazar, M. & Fichtner, H., eds. 2021, Astrophysics and Space Science Library, V ol. 464, Kappa Distributions; From Observational Evidences via Controver- sial Predictions to a Consistent Theory of Nonequilibrium Plasmas

    2021

  23. [23]

    P., et al

    Lu, R.-S., Asada, K., Krichbaum, T. P., et al. 2023, Nature, 616, 686

    2023

  24. [24]

    2023, ApJ, 944, 122

    Meringolo, C., Cruz-Osorio, A., Rezzolla, L., & Servidio, S. 2023, ApJ, 944, 122

    2023

  25. [25]

    P., Walker, R

    Mertens, F., Lobanov, A. P., Walker, R. C., & Hardee, P. E. 2016, Astron. Astro- phys., 595, A54

    2016

  26. [26]

    M., Younsi, Z., et al

    Mizuno, Y ., Fromm, C. M., Younsi, Z., et al. 2021, MNRAS, 506, 741 4 https://bhac.science/ 5 https://achael.github.io/eht-imaging/ Article number, page 12 of 15 Christian M. Fromm et al.: Probing radiation micro-physics in M 87 Mo´scibrodzka, M. 2025, ApJ, 981, 145 Mo´scibrodzka, M., Falcke, H., & Shiokawa, H. 2016, Astron. Astrophys., 586, A38

    2021

  27. [27]

    2018, ApJ, 868, 146

    Nakamura, M., Asada, K., Hada, K., et al. 2018, ApJ, 868, 146

    2018

  28. [28]

    M., et al

    Olivares, H., Younsi, Z., Fromm, C. M., et al. 2020, Mon. Not. R. Astron. Soc., 497, 521

    2020

  29. [29]

    Pandya, A., Zhang, Z., Chandra, M., & Gammie, C. F. 2016, Astrophys. J., 822, 34

    2016

  30. [30]

    2019, arXiv e-prints, arXiv:1904.04923

    Porth, O., Chatterjee, K., Narayan, R., et al. 2019, arXiv e-prints, arXiv:1904.04923

    Pith/arXiv arXiv 2019

  31. [31]

    2017, Computational Astrophysics and Cosmology, 4, 1

    Porth, O., Olivares, H., Mizuno, Y ., et al. 2017, Computational Astrophysics and Cosmology, 4, 1

    2017

  32. [32]

    & Zanotti, O

    Rezzolla, L. & Zanotti, O. 2013, Relativistic Hydrodynamics (Oxford, UK: Ox- ford University Press)

    2013

  33. [33]

    W., et al

    Ro, H., Kino, M., Sohn, B. W., et al. 2023, A&A, 673, A159

    2023

  34. [34]

    E., Sironi, L., & Narayan, R

    Rowan, M. E., Sironi, L., & Narayan, R. 2017, ApJ, 850, 29

    2017

  35. [35]

    & Spitkovsky, A

    Sironi, L. & Spitkovsky, A. 2009, ApJ, 698, 1523

    2009

  36. [36]

    Tchekhovskoy, A., Narayan, R., & McKinney, J. C. 2011, Mon. Not. R. Astron. Soc., 418, L79

    2011

  37. [37]

    C., Hardee, P

    Walker, R. C., Hardee, P. E., Davies, F. B., Ly, C., & Junor, W. 2018, Astrophys. J., 855, 128

    2018

  38. [38]

    N., Ryan, B

    Yarza, R., Wong, G. N., Ryan, B. R., & Gammie, C. F. 2020, Astrophys. J., 898, 50

    2020

  39. [39]

    M., & Olivares, H

    Younsi, Z., Porth, O., Mizuno, Y ., Fromm, C. M., & Olivares, H. 2020, in Perseus in Sicily: From Black Hole to Cluster Outskirts, ed. K. Asada, E. de Gouveia Dal Pino, M. Giroletti, H. Nagai, & R. Nemmen, V ol. 342, 9–12

    2020

  40. [40]

    K., et al

    Zhang, M., Mizuno, Y ., Dihingia, I. K., et al. 2026, arXiv e-prints, arXiv:2605.15502

    Pith/arXiv arXiv 2026

  41. [41]

    =0.0 500µas

    Zhang, M., Mizuno, Y ., Fromm, C. M., Younsi, Z., & Cruz-Osorio, A. 2024, A&A, 687, A88 Article number, page 13 of 15 A&A proofs:manuscript no. main12page_refV1 Appendix A: Differences in turnover values across heatings and eDFs The detailed differences across heating and eDF are presented in Fig. A.1 and A.2. From the difference plots (third column in Fi...

    2024