Pith. sign in

REVIEW 3 major objections 4 minor 44 references

Data-driven reduced modeling of streamer discharges in air

T0 review · 3 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read A single length scale ahead of a streamer tip predicts its velocity, radius, and channel conductivity, so reduced models can skip picosecond steps and run 3D branching discharges on a desktop.

desk verdict The reduced-model framework and public dataset are a real contribution; the 3D branching claim is an acknowledged extrapolation, not a validation. read the letter →

arxiv 2501.06093 v3 pith:KNK6EUYC submitted 2025-01-10 physics.plasm-ph

classification physics.plasm-ph
keywords electricdischargestreamerreducedmodeldata-drivenconductingcylindersbranchingPoissonequationpositivestreamers
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 tries to show that the expensive microscopic physics of positive streamers in air can be coarse-grained into a single measurable quantity: the size $L_E$ of the high-field zone just ahead of a streamer tip. From 1000 axisymmetric fluid simulations, the paper derives simple fit expressions for streamer velocity, radius, and line conductivity (conductance per unit length) as functions of $L_E$, and builds a reduced model in which each streamer is a chain of conducting cylindrical segments on a numerical mesh. The reduced model reproduces the axisymmetric simulations' velocity, radius, conductivity, and on-axis field profiles while using time steps up to about 1 ns instead of about 2 ps and grid spacings up to hundreds of micrometers instead of a few micrometers. Because the electron dynamics no longer have to be resolved, 3D simulations with 20+ branching streamers run in 4-8 minutes on a desktop computer and produce morphologies that resemble experimental discharges. If the approach holds, it opens a path to simulating large multi-streamer systems, leaders, and sprites that are currently out of reach of direct fluid models.

What carries the argument

The load-bearing object is the length scale $L_E$: the distance from the streamer head to the point where the on-axis electric field falls below 50 kV/cm. All three fitted quantities - velocity, radius, and line conductivity - are functions of $L_E$ alone, making it the single state variable that carries the reduced model. The computational carrier is the framework in which each channel is grown as cylindrical segments with a semi-spherical cap; the segment conductivity is mapped onto a tree-structured adaptive mesh, and the next potential comes from solving $\nabla\cdot[(\varepsilon_0+\Delta t\,\sigma)\nabla\phi]=-\rho/\varepsilon_0$ with geometric multigrid. Branching is modeled as a memoryless Poisson process with mean time $\bar{\tau}_{\mathrm{branch}} = c_b (R_\sigma/v)(1+L_b^2/R_\sigma^2)$, whose two parameters control branch frequency and the suppression of thin channels.

What would settle it

Run the reduced model in a regime the training set did not cover - for example, a positive streamer propagating in a background field below about 10 kV/cm or a gap over 30 mm - and compare its predicted head velocity, radius, and line conductivity against a full 3D drift-diffusion simulation or an experiment. If the $v=1.78\times10^9 L_E$ relation or the radius and conductivity fits systematically miss outside the trained range, the transferability assumption is falsified.

Watch

Extended reading notes

Core claim

The central claim is that the dynamics of a positive streamer head in air are determined, to a good approximation, by the extent $L_E$ of the region in front of the head where the electric field exceeds 50 kV/cm. Using this single feature, the paper fits closed-form expressions (Eqs. 16-18) for the streamer's head velocity $v$, its electrodynamic radius $R_E$ (the radius at which the radial electric field peaks), and the line conductivity $\sigma_h$, and then embeds these fits in a mesh-based model of conducting cylinders with semi-spherical caps. The resulting reduced model agrees well with the drift-diffusion simulations it was trained against, and it is numerically stable for time steps up to about 1 ns and grid cells hundreds of micrometers wide, because the potential update solves an implicit variable-coefficient Poisson equation rather than tracking electron density. This speedup is what makes branching 3D simulations of 20+ channels practical on a desktop computer. The paper also shows that a simple grid-spacing correction removes most of the dependence of the measured $L_E$ on resolution, and that a two-parameter Poisson branch model produces experimentally plausible discharge trees.

Load-bearing premise

The fitting formulas come from axisymmetric simulations in one geometry and are applied, without recalibration, to 3D branching discharges in a different geometry; everything rests on the assumption that the size of the high-field zone ahead of a streamer tip is the only information needed to predict how fast it moves, how thick it grows, and how conductive it becomes.

Editorial extensions

If this is right

  • Time steps in the reduced model can be up to about 1 ns, roughly three orders of magnitude larger than the ~2 ps steps of the fluid simulations, so multi-streamer discharges can be evolved over much longer physical times.
  • Grid spacings of hundreds of micrometers suffice, compared with a few micrometers for fluid models, and a one-parameter correction for grid resolution makes the predicted streamer velocity nearly independent of $\Delta x$.
  • 3D simulations with 20+ branching streamers in a 4 cm gap reproduce qualitative experimental features - stagnation of overtaken branches, near-horizontal propagation near the electrode, fastest vertical propagation around 1.1 mm/ns - and complete in 4-8 minutes on a desktop computer.
  • Because the framework only needs a rule for advancing position, radius, and line conductivity, it can accept other growth models, such as physics-based reduced models or machine-learned surrogates, without changing the field solver.
  • The implicit potential update removes the dielectric-relaxation time restriction, so the model is stable for time steps much larger than $\tau_{\mathrm{drt}}=\varepsilon_0/\sigma$.

Reading between the lines

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

  • If $L_E$ is truly a sufficient predictor, the same fitting pipeline should transfer to negative streamers, other gas mixtures, or sprite discharges, provided the training dataset is regenerated; the paper lists these as future work, but the transferability is a direct consequence of the single-feature assumption.
  • A sharper, testable consequence is that the reduced model should fail precisely where the $L_E$-scaling breaks down: in low background fields where streamers become thin and stagnate, a regime the dataset excludes; comparing predicted radius and velocity against full 3D fluid simulations in such fields would probe the boundary of the method.
  • The branching parameters $c_b$ and $L_b$ are only qualitatively calibrated; a quantitative check would compare the model's branch-angle and branch-spacing distributions against high-speed imaging statistics from experiments.
  • Since the conductivity field lives on the mesh, the model is naturally positioned to be coupled to gas dynamics for ohmic heating, making a streamer-to-leader transition simulation a plausible near-term extension rather than a separate framework.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper presents a reduced modeling framework (cocydimo) for filamentary discharges, in which streamer channels are represented as conducting cylindrical segments moving on a numerical mesh. A 1000-run axisymmetric fluid simulation dataset is used to fit simple expressions for streamer head radius, velocity, and line conductivity as functions of the high-field length scale L_E (Eqs. 16-18). These expressions are then embedded in a reduced model with a stochastic branching rule, and the model is tested against the axisymmetric dataset, studied for time-step and grid-spacing sensitivity, and demonstrated in 3D simulations with branching in a 4 cm gap.

Significance. If the closure and the 3D transfer hold, this would be a substantial practical advance: 3D simulations with 20+ interacting streamers in minutes on a desktop computer, with public code and dataset, would enable parameter studies and large-scale discharge morphologies that full fluid simulations cannot reach. The time-step insensitivity (Fig. 10) and the grid-spacing correction (Fig. 11) are convincing for the tested case, and the open-source availability of both the code and the dataset is a clear strength. However, the current evidence does not yet establish the predictive accuracy of the model in the regimes where it is most useful: branching, streamer interaction, low background fields, and small-radius branches are all outside the dataset used to fit the closure, and the 3D experimental comparison is qualitative with a ~30% velocity discrepancy and undetermined branching parameters.

major comments (3)
  1. [§3.3, §5.1] The validation in §5.1 is not an out-of-sample test of Eqs. (16)-(18). Section 3.3 states that the data were split into 70% training and 30% test sets, but it also says there was 'essentially no overfitting', and §5.1 does not state whether the eight displayed runs in Figs. 7-9 belong to the training or test portion. Since the same dataset was used both to fit and to assess the closure, the R² values in Fig. 5 and the qualitative agreement in Figs. 7-9 overstate predictive skill. Given that the R² values are only 0.80 for σ_h and 0.79 for R_E, this distinction matters. Please report test-set metrics separately, identify the displayed runs, and ideally validate on parameter ranges or geometries excluded from the training set.
  2. [§4.2, §5.3, §3.4] The central 3D claim rests on applying Eqs. (16)-(18) outside their training regime. The dataset contains only isolated axisymmetric positive streamers in a 30 mm gap, and Section 3.4 explicitly lists the missing regimes: no branching, no streamer interactions, and no low-background-field or stagnating cases with small radii. The 3D simulations of Section 5.3 operate in exactly these regimes: branches can approach the stagnation radius R_E,min = 0.15 mm, and each head sees fields from neighboring channels. This is a load-bearing extrapolation that the paper acknowledges but does not quantitatively close. A concrete test would be to extract L_E, v, R_E, and σ_h from a full 3D fluid simulation of a branched discharge (or from a two-head interaction setup) and compare them with Eqs. (16)-(18); alternatively, run the reduced model and a full fluid reference on the same small 3D case and quantify errors in velocity, radius, and conductivity.
  3. [§5.3] The experimental comparison does not currently constrain the predictive accuracy of the 3D model. The fastest simulated streamer velocity is about 1.1 mm/ns versus 0.8 ± 0.2 mm/ns in the experiments, a roughly 30% difference that is attributed to the voltage rise time, but no simulation with a finite rise time is presented to test that explanation. The branching parameters c_b and L_b are varied and judged only qualitatively, and Section 5.3 concludes that their values 'could not accurately be determined'. Please provide quantitative morphology metrics (branch counts, branch angles, velocity distributions, channel radii) and a calibration or sensitivity statement for c_b and L_b, or explicitly present Section 5.3 as a feasibility demonstration rather than a validation.
minor comments (4)
  1. [Fig. 11 caption] The caption reads 'a) c1 = 0x' for the uncorrected case; this appears to be a typo for 'c1 = 0' or 'c1 = 0.0'.
  2. [Fig. 5 and Eq. (16)] The axes of Fig. 5 give σ_h in units of A m/MV, while the text describing Eq. (16) lists units of A m/V; please unify the notation so that the fit coefficients and the plotted quantities are immediately comparable.
  3. [§4.2] The text states that L_E,min = 0.1 mm corresponds to R_E,min = 0.15 mm, but Eq. (17) with L_E = 0.1 mm gives R_E ≈ 0.16 mm; please check the rounding or clarify how the threshold was obtained.
  4. [§3.3] The sentence introducing Eqs. (16)-(18) says the quantities are 'made dimensionless using the following units'; this wording is confusing because the expressions remain dimensionful. Consider saying that the fits are written in SI units.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the fitted closure is openly fitted, and the reduced model's validation retains independent self-consistent content.

full rationale

The derivation chain is self-contained and transparent. The reduced model's per-step velocity, radius, and head conductivity indeed come from the explicit fit expressions Eqs. (16)-(18), but the paper never presents these fits as first-principles predictions; Section 3.3 states the data were split 70/30 into training and test sets and that there was essentially no overfitting. The Section 5.1 comparison against the dataset is therefore a consistency check of a data-driven closure, not a circular derivation: the reduced model computes L_E self-consistently from its own electrostatic solution and then evolves channel geometry and mesh conductivity, so the resulting conductivity and field profiles are not mere restatements of the fitted targets. The 3D simulations rest on an acknowledged extrapolation of the axisymmetric fits into branching and low-field regimes, with limitations explicitly listed in Section 3.4; extrapolation risk is not circularity. No load-bearing uniqueness theorem or self-citation chain is invoked. The central computational claims (large time steps, grid spacing, 3D runtime) follow from the framework's implicit Poisson solve and AMR implementation, independent of the fitted closure.

Assumptions & free parameters 13 free parameters · 7 assumptions · 0 invented entities

The central model rests on fitted closure coefficients, several hand-chosen constants, and assumptions that the axisymmetric training data transfer to 3D branching cases. No new physical entities are introduced.

free parameters (13)
  • L_E threshold E_threshold = 50 kV/cm
    Chosen threshold to define the high-field region size L_E; authors state it is somewhat arbitrary and tried 45-60 kV/cm (Section 3.2).
  • Grid spacing correction coefficient c1 = 0.75
    Introduced in Eq. (26) to correct L_E for grid spacing dependence; c1=1.0 over-corrects, so 0.75 is chosen by hand.
  • Field sampling offset c_ahead = 0.5
    Distance ahead of streamer head at which E direction is sampled, Eq. (19); authors state it is somewhat arbitrary.
  • L_E smoothing coefficient beta = 0.5
    Exponential smoothing factor in Eq. (21); default value, varied in time-step study (beta=1).
  • Conductivity update delay tau_delay = 1 ns
    Delay before channel cells are updated with Eq. (24), chosen for examples.
  • Branching scale c_b = 10-20
    Branching time parameter in Eq. (25); ranges tested, not determined from experiments.
  • Branching radius scale L_b = 0.2-0.8 mm
    Branching time parameter in Eq. (25); ranges tested, not determined.
  • Initial radius fraction R_sigma,0/R_sigma = 0.5
    Initial streamer radius at electrode tip set to half the fitted radius (Sections 5.1 and 5.3).
  • Stagnation threshold L_E,min = 0.1 mm
    Streamers halt when L_E drops below this value, corresponding to R_E,min=0.15 mm.
  • Radius conversion factor R_sigma = 1.2 R_E = 1.2
    Approximate conversion from electrodynamic radius to conductivity radius from Appendix A; authors call it rather rough.
  • sigma_h fit coefficients (Eq. 16) = 1e-8, 1.40, -1.41e-6, 2.80e-3
    Fitted to the 1000-simulation dataset with R2=0.80.
  • R_E fit coefficients (Eq. 17) = 2.90e-5, 1.30, 6.31e-4, 0.627
    Fitted to the dataset with R2=0.79.
  • v fit coefficient (Eq. 18) = 1.78e9
    Fitted to the dataset with R2=0.93.
assumptions (7)
  • domain assumption The drift-diffusion fluid model with local field approximation accurately describes positive streamer dynamics in air.
    Used to generate the 1000-simulation dataset (Section 3.1); all fits inherit this assumption.
  • domain assumption Electron transport and reaction data from Phelps cross sections via BOLSIG+ are accurate for air.
    Underpins the fluid model and Figure 1 (Section 3.1).
  • domain assumption Photoionization is described by the Zheleznyak model with Helmholtz approximation.
    Affects streamer properties in the dataset (Section 3.1).
  • domain assumption Channels propagate parallel to the electric field ahead of them and have a rounded head with fixed radial conductivity profile f_r(x)=max(0,2(1-x^2)).
    Core geometric assumption of the framework (Section 2 and Eq. (5)).
  • ad hoc to paper The streamer head state is fully determined by L_E, the size of the high-field region ahead of the streamer.
    The reduced model uses only L_E as input to predict v, R, and sigma_h (Section 3.3); this single-feature sufficiency is assumed, not derived.
  • domain assumption Conductivity in channels evolves as d sigma/dt = sigma * S(E), with S = alpha_bar * mu_e * E.
    Used to update channel conductivity in Eq. (23); valid only where electron impact ionization and attachment dominate.
  • ad hoc to paper Branching can be modeled as a memoryless Poisson process with mean time given by Eq. (25).
    Stochastic branching model (Section 4.5); parameters c_b and L_b are not determined from experiments.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Data-driven reduced modeling of streamer discharges in air." pith.science (2026). https://pith.science/paper/KNK6EUYC

@misc{pith2026250106093,
  author       = {Pith},
  title        = {Pith review of: Data-driven reduced modeling of streamer discharges in air},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KNK6EUYC}},
  note         = {Machine review of arXiv:2501.06093}
}
read the original abstract

We present a computational framework for simulating filamentary electric discharges, in which channels are represented as conducting cylindrical segments. The framework requires a model that predicts the position, radius, and line conductivity of channels at a next time step. Using this information, the electric conductivity on a numerical mesh is updated, and the new electric potential is computed by solving a variable-coefficient Poisson equation. A parallel field solver with support for adaptive mesh refinement is used, and the framework provides a Python interface for easy experimentation. We demonstrate how the framework can be used to simulate positive streamer discharges in air. First, a dataset of 1000 axisymmetric positive streamer simulations is generated, in which the applied voltage and the electrode geometry are varied. Fit expressions for the streamer radius, velocity, and line conductivity are derived from this dataset, taking as input the size of the high-field region ahead of the streamers. We then construct a reduced model for positive streamers in air, which includes a stochastic branching model. The reduced model compares well with the axisymmetric simulations from the dataset, while allowing spatial and temporal step sizes that are several orders of magnitude larger. 3D simulations with the reduced model resemble experimentally observed discharge morphologies. The model runs efficiently, with 3D simulations with 20+ streamers taking 4-8 minutes on a desktop computer.

Figures

Figures reproduced from arXiv: 2501.06093 by the authors.

Figure 1
Figure 1. The effective ionization coefficient ¯α and electron mobility µe for electrons in air (80% N2, 20% O2) at 1 bar and 300 K, computed from Phelps’s cross sections [12, 13]. inverse of the effective electron impact ionization rate, but it can also be thought of as a resulting from a CFL-like condition ∆t ≲ ∆x/vd, with ∆x given by equation (1). The above time step restriction also applies to models that use implicit tim… view at source ↗
Figure 2
Figure 2. Illustration of how cylindrical segments with a semi-spherical cap can [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Example of the electric field distribution in one of the axisymmetric [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Pairwise relationships between parameters extracted from 1000 axisymmetric positive streamer simulations in air at 1 bar and 300 K. On the diagonal the [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: A comparison of the simple model given by equations (16)–(18) [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Schematic illustration of the approach used to determine the propa [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 8
Figure 8. Figure 8: Comparison of line conductivity between fluid simulations and the [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 10
Figure 10. Figure 10: The dependence of the reduced model on the time step. Shown are [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]
Figure 9
Figure 9. Figure 9: Comparison of on-axis electric field between fluid simulations and [PITH_FULL_IMAGE:figures/full_fig_p011_9.png]
Figure 11
Figure 11. Figure 11: The dependence of the reduced model on the grid spacing. For a) and [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 12
Figure 12. Figure 12: 3D simulations of branching positive streamers in air with the re [PITH_FULL_IMAGE:figures/full_fig_p013_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 44 canonical work pages

  1. [1]

    The physics of streamer discharge phenomena

    Sander Nijdam, Jannis Teunissen, and Ute Ebert. The physics of streamer discharge phenomena. Plasma Sources Sci. Technol., 29(10):103001, November 2020

  2. [2]

    Influence of the pre- ionization background and simulation of the optical emission of a streamer discharge in preheated air at atmospheric pressure between two point electrodes

    A Bourdon, Z Bonaventura, and S Celestin. Influence of the pre- ionization background and simulation of the optical emission of a streamer discharge in preheated air at atmospheric pressure between two point electrodes. Plasma Sources Sci. Technol., 19(3):034012, May 2010

  3. [3]

    M. M. Becker and D. Loffhagen. Enhanced reliability of drift-diffusion approximation for electrons in fluid models for nonthermal plasmas. AIP Advances, 3(1):012108, January 2013

  4. [4]

    Luque and U

    A. Luque and U. Ebert. Density models for streamer discharges: Beyond cylindrical symmetry and homogeneous media.Journal of Computational Physics, 231(3):904–918, February 2012

  5. [5]

    Simulation of pulsed positive streamer discharges in air at high temperatures

    Atsushi Komuro, Shuto Matsuyuki, and Akira Ando. Simulation of pulsed positive streamer discharges in air at high temperatures. Plasma Sources Sci. Technol., 27(10):105001, October 2018. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 r (mm) 0.0 0.2 0.4 0.6 0.8(r) (A V/m) simulation 1 c1 × max[0, 1 (r/R1)2] simulation 2 c2 × max[0, 1 (r/R2)2] Figure A.14: Two examples of rad...

  6. [6]

    An adaptive Cartesian embedded boundary approach for fluid simulations of two- and three-dimensional low temperature plasma filaments in complex geometries

    Robert Marskar. An adaptive Cartesian embedded boundary approach for fluid simulations of two- and three-dimensional low temperature plasma filaments in complex geometries. Journal of Computational Physics, 388:624–654, July 2019

  7. [7]

    3D fluid modeling of positive streamer discharges in air with stochastic photoionization

    Robert Marskar. 3D fluid modeling of positive streamer discharges in air with stochastic photoionization. Plasma Sources Sci. Technol., 29(5):055007, May 2020

  8. [8]

    Generation of the single-filament pulsed positive streamer discharge in atmospheric-pressure air and its com- parison with two-dimensional simulation

    Ryo Ono and Atsushi Komuro. Generation of the single-filament pulsed positive streamer discharge in atmospheric-pressure air and its com- parison with two-dimensional simulation. J. Phys. D: Appl. Phys., 53(3):035202, January 2020

Show all 44 references
  1. [9]

    3D simulations of positive streamers in air in a strong external magnetic field

    Zhen Wang, Anbang Sun, Sasa Dujko, Ute Ebert, and Jannis Teunissen. 3D simulations of positive streamers in air in a strong external magnetic field. Plasma Sources Sci. Technol., January 2024

  2. [10]

    Jiang, Y

    M. Jiang, Y . Li, H. Wang, and C. Liu. 3D PIC-MCC simulations of pos- itive streamers in air gaps. Physics of Plasmas, 24(10):102112, October 2017

  3. [11]

    Comparison of six simulation codes for positive streamers in air

    B Bagheri, J Teunissen, U Ebert, M M Becker, S Chen, O Ducasse, O Eichwald, D Loffhagen, A Luque, D Mihailova, J M Plewa, J van Dijk, and M Yousfi. Comparison of six simulation codes for positive streamers in air. Plasma Sources Sci. Technol., 27(9):095002, September 2018

  4. [12]

    Phelps database, www.lxcat.net, retrieved on august 19, 2021

  5. [13]

    L. C. Pitchford and A. V . Phelps. Comparative calculations of electron- swarm properties in N 2 at moderate E N values.Phys. Rev.A, 25(1):540– 554, January 1982

  6. [14]

    Montijn, W

    C. Montijn, W. Hundsdorfer, and U. Ebert. An adaptive grid refine- ment strategy for the simulation of negative streamers. Journal of Computational Physics, 219(2):801–835, December 2006

  7. [15]

    Pancheshnyi, P

    S. Pancheshnyi, P. Ségur, J. Capeillère, and A. Bourdon. Numerical sim- ulation of filamentary discharges with parallel adaptive mesh refinement. Journal of Computational Physics, 227(13):6574–6590, June 2008

  8. [16]

    Kolobov and R.R

    V .I. Kolobov and R.R. Arslanbekov. Towards adaptive kinetic-fluid sim- ulations of weakly ionized plasmas. Journal of Computational Physics, 231(3):839–869, February 2012

  9. [17]

    Simulating streamer discharges in 3D with the parallel adaptive Afivo framework.Journal of Physics D: Applied Physics, 50(47):474001, October 2017

    Jannis Teunissen and Ute Ebert. Simulating streamer discharges in 3D with the parallel adaptive Afivo framework.Journal of Physics D: Applied Physics, 50(47):474001, October 2017

  10. [18]

    Niemeyer, L

    L. Niemeyer, L. Pietronero, and H. Wiesmann. Fractal Dimension of Di- electric Breakdown. Physical Review Letters, 52(12):1033–1036, March 1984

  11. [19]

    Pasko, Umran S

    Victor P. Pasko, Umran S. Inan, and Timothy F. Bell. Fractal structure of sprites. Geophysical Research Letters, 27(4):497–500, February 2000

  12. [20]

    Measurement and simulation of electrical tree growth and partial discharge activity in epoxy resin

    M D Noskov, M Sack, A S Malinovski, and A J Schwab. Measurement and simulation of electrical tree growth and partial discharge activity in epoxy resin. J. Phys. D: Appl. Phys., 34(9):1389–1398, May 2001

  13. [21]

    3D simulations of streamer branching in air

    Mose Akyuz, Anders Larsson, Vernon Cooray, and Gustav Strandberg. 3D simulations of streamer branching in air. Journal of Electrostatics, 59(2):115–141, September 2003. 15

  14. [22]

    Growing discharge trees with self- consistent charge transport: The collective dynamics of streamers

    Alejandro Luque and Ute Ebert. Growing discharge trees with self- consistent charge transport: The collective dynamics of streamers. New Journal of Physics, 16(1):013039, January 2014

  15. [23]

    Streamer discharges as advancing imperfect conductors: Inhomogeneities in long ionized chan- nels

    A Luque, M González, and F J Gordillo-Vázquez. Streamer discharges as advancing imperfect conductors: Inhomogeneities in long ionized chan- nels. Plasma Sources Sci. Technol., 26(12):125006, November 2017

  16. [24]

    Afivo: A framework for quadtree/octree AMR with shared-memory parallelization and geometric multigrid meth- ods

    Jannis Teunissen and Ute Ebert. Afivo: A framework for quadtree/octree AMR with shared-memory parallelization and geometric multigrid meth- ods. Computer Physics Communications, 233:156–166, December 2018

  17. [25]

    Luque, H

    A. Luque, H. C. Stenbaek-Nielsen, M. G. McHarg, and R. K. Haaland. Sprite beads and glows arising from the attachment instability in streamer channels: DYNAMICS OF SPRITE CHANNELS. J. Geophys. Res. Space Physics, 121(3):2431–2449, March 2016

  18. [26]

    Peter L. G. Ventzek. Two-dimensional modeling of high plasma density inductively coupled sources for materials processing.J. Vac.Sci. Technol. B, 12(1):461, January 1994

  19. [27]

    Hagelaar and G.M.W

    G.J.M. Hagelaar and G.M.W. Kroesen. Speeding Up Fluid Models for Gas Discharges by Implicit Treatment of the Electron Energy Source Term. Journal of Computational Physics, 159(1):1–12, March 2000

  20. [28]

    Geometric multigrid method for solving Poisson’s equation on octree grids with irregular boundaries

    Jannis Teunissen and Francesca Schiavello. Geometric multigrid method for solving Poisson’s equation on octree grids with irregular boundaries. Computer Physics Communications, 286:108665, May 2023

  21. [29]

    Weber, Hari Krishnan, Thomas Fogal, Allen Sander- son, Christoph Garth, E

    Hank Childs, Eric Brugger, Brad Whitlock, Jeremy Meredith, Sean Ahern, David Pugmire, Kathleen Biagas, Mark Miller, Cyrus Harrison, Gunther H. Weber, Hari Krishnan, Thomas Fogal, Allen Sander- son, Christoph Garth, E. Wes Bethel, David Camp, Oliver Rübel, Marc Durant, Jean M. ...

  22. [30]

    Comparing simulations and experiments of positive streamers in air: Steps toward model validation

    Xiaoran Li, Siebe Dijcks, Sander Nijdam, Anbang Sun, Ute Ebert, and Jannis Teunissen. Comparing simulations and experiments of positive streamers in air: Steps toward model validation. Plasma Sources Sci. Technol., 30(9):095002, September 2021

  23. [31]

    Solving the Boltzmann equation to obtain electron transport coefficients and rate coefficients for fluid mod- els

    G J M Hagelaar and L C Pitchford. Solving the Boltzmann equation to obtain electron transport coefficients and rate coefficients for fluid mod- els. Plasma Sources Science and Technology, 14(4):722–733, November 2005

  24. [32]

    M. B. Zheleznyak, A. K. Mnatsakanian, and S. V . Sizykh. Photoioniza- tion of nitrogen and oxygen mixtures by radiation from a gas discharge. Teplofizika Vysokikh Temperatur, 20:423–428, November 1982

  25. [33]

    Stagnation dynamics of a cathode-directed streamer discharge in air

    S V Pancheshnyi and A Yu Starikovskii. Stagnation dynamics of a cathode-directed streamer discharge in air. Plasma Sources Science and Technology, 13(3):B1–B5, August 2004

  26. [34]

    Underlying mech- anism of the stagnation of positive streamers

    M Niknezhad, O Chanrion, J Holbøll, and T Neubert. Underlying mech- anism of the stagnation of positive streamers. Plasma Sources Sci. Technol., 30(11):115014, November 2021

  27. [35]

    A computational study of steady and stagnating positive streamers in N2– O2 mixtures

    Xiaoran Li, Baohong Guo, Anbang Sun, Ute Ebert, and Jannis Teunissen. A computational study of steady and stagnating positive streamers in N2– O2 mixtures. Plasma Sources Sci. Technol., page 15, 2022

  28. [36]

    Probing photo-ionization: Experiments on positive stream- ers in pure gases and mixtures

    S Nijdam, F M J H van de Wetering, R Blanc, E M van Veldhuizen, and U Ebert. Probing photo-ionization: Experiments on positive stream- ers in pure gases and mixtures. Journal of Physics D: Applied Physics, 43(14):145204, April 2010

  29. [37]

    Positive and negative streamers in ambient air: Measuring diameter, ve- locity and dissipated energy

    T M P Briels, J Kos, G J J Winands, E M van Veldhuizen, and U Ebert. Positive and negative streamers in ambient air: Measuring diameter, ve- locity and dissipated energy. J. Phys. D: Appl. Phys., 41(23):234004, November 2008

  30. [38]

    Quantitative modeling of streamer discharge branching in air

    Zhen Wang, Siebe Dijcks, Yihao Guo, Martijn Van Der Leegte, An- bang Sun, Ute Ebert, Sander Nijdam, and Jannis Teunissen. Quantitative modeling of streamer discharge branching in air. Plasma Sources Sci. Technol., 32(8):085007, August 2023

  31. [39]

    Statistical analysis on branching char- acteristics of positive streamer discharges in N 2 –O 2 mixtures

    Yihao Guo and Sander Nijdam. Statistical analysis on branching char- acteristics of positive streamer discharges in N 2 –O 2 mixtures. Plasma Sources Sci. Technol., 33(4):045006, April 2024

  32. [40]

    Positive streamers in air and nitrogen of varying density: Experiments on similarity laws

    T M P Briels, E M van Veldhuizen, and U Ebert. Positive streamers in air and nitrogen of varying density: Experiments on similarity laws. Journal of Physics D: Applied Physics, 41(23):234008, December 2008

  33. [41]

    Nanosec- ond repetitively pulsed discharges in N 2 –O 2 mixtures: Inception cloud and streamer emergence

    She Chen, L C J Heijmans, Rong Zeng, S Nijdam, and U Ebert. Nanosec- ond repetitively pulsed discharges in N 2 –O 2 mixtures: Inception cloud and streamer emergence. Journal of Physics D: Applied Physics, 48(17):175201, May 2015

  34. [42]

    Macroscopic param- eterization of positive streamer heads in air, November 2024

    Dennis Bouwman, Jannis Teunissen, and Ute Ebert. Macroscopic param- eterization of positive streamer heads in air, November 2024

  35. [43]

    Pancheshnyi, M

    S. Pancheshnyi, M. Nudnova, and A. Starikovskii. Development of a cathode-directed streamer discharge in air at different pressures: Experi- ment and comparison with direct numerical simulation. Physical Review E, 71(1), January 2005

  36. [44]

    A computa- tional study of accelerating, steady and fading negative streamers in am- bient air

    Baohong Guo, Xiaoran Li, Ute Ebert, and Jannis Teunissen. A computa- tional study of accelerating, steady and fading negative streamers in am- bient air. Plasma Sources Sci. Technol., 31(9):095011, September 2022. 16

Pith tools

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