Pith. sign in

REVIEW 2 major objections 5 minor 30 references

Structure-Preserving Dynamical Low-Rank Approximations for Stochastic Vlasov--Poisson Equations

T0 review · 2 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read Dynamical low-rank approximations of the stochastic Vlasov–Poisson equation can preserve the model's mass, momentum, and energy balance laws if the velocity basis keeps fixed moment modes, and two augmented BUG integrators achieve this in…

desk verdict Useful stochastic extension of conservative DLR, but the discrete conservation proofs have a real truncation gap that needs patching before the claims are established. read the letter →

arxiv 2608.00397 v1 pith:6XSACRND submitted 2026-08-01 math.NA cs.NA

classification math.NAcs.NA MSC 65M9960H3565C3035Q83
keywords dynamicallow-rankapproximationstochasticVlasov–Poissonequationsstructure-preservingintegratorsbasis-updateGalerkin(BUG)Itô–Stratonovichdiscretizationmassmomentumenergyconservationtransportnoiseranktruncation
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

The paper claims that a dynamical low-rank approximation of the stochastic Vlasov–Poisson equation can inherit the model's physical balance laws—mass conserved pathwise, total momentum conserved in expectation, and a known mean-energy evolution—provided the velocity basis permanently contains the constant, linear, and quadratic moment modes. It proves this for two augmented basis-update Galerkin (BUG) integrators, one applying Euler–Maruyama to the Itô formulation and one applying Heun to the Stratonovich formulation, and it identifies the extra spatial enrichment the Heun scheme needs. If the claims hold, stochastic kinetic simulations can be compressed to low rank without systematically violating the invariants that make the model physically meaningful.

What carries the argument

The central machinery is the augmented basis-update Galerkin (BUG) integrator, a three-step projector-splitting update of the spatial basis, the moving velocity basis, and the coefficient matrix, built on a Petrov–Galerkin projection onto a low-rank manifold whose first $m \ge d+2$ velocity functions are fixed moment modes: $U_1 = 1/\|1\|_v$, $U_{i+1} = v_i/\|v_i\|_v$, and $U_{d+2} = (|v|^2-\alpha^2)/\||v|^2-\alpha^2\|_v$. Basis augmentation adds the spatial functions needed to make the projection identities (25)–(27), (39)–(40), (47), (49)–(50) exact, and conservative SVD truncation compresses only the non-fixed modes so the moment information survives rank reduction. What this machinery does is close the discrete fluxes in exactly the way the continuous balance laws close, so the low-rank approximation inherits the stochastic model's conservation structure.

What would settle it

On the two-stream test case, run the Euler–Maruyama BUG integrator and, after the conservative truncation at each step, compute the $L^2$ norm of $P_{x,n+1}(g) - g$ for $g = \tau E^n \rho^n + \sum_s \Delta\beta^n_s \sigma_s(x)\rho^n$; a nonzero residual at machine precision would mean the local momentum balance (41) is not established by the proof.

Watch

Extended reading notes

Core claim

The central claim of the paper is that fixing the first $m \ge d+2$ velocity basis modes to the functions $1$, $v_i$, and $|v|^2-\alpha^2$ makes the projected stochastic dynamics reproduce the continuous balance laws $d\rho + \nabla_x \cdot J\,dt = 0$, $dP = \sum_k (\int \rho \, \sigma_k \, dx)\,d\beta_k$, and $dH = \sum_k (\int J\cdot \sigma_k \, dx)\,d\beta_k + \frac{1}{2}\int \mathrm{Tr}(\sigma\sigma^\top)\rho\,dx\,dt$. The paper constructs two augmented BUG integrators and proves discrete analogues: local mass conservation (28) holds pathwise, local momentum balance (41) holds pathwise with expectation-preserving total momentum, and local energy balance (51) holds with an explicit electric-field residual $R^n_E$. The Euler–Maruyama variant needs only the standard conservative augmentation for the mass and momentum laws, while the Heun variant requires additional spatial enrichment so that the projection conditions (40) and (50) are exact.

Load-bearing premise

The proofs of the momentum and energy laws assume that the rank-reducing SVD step never throws away the specific spatial functions (for instance, electric field times density, or noise coefficient times density) that the conservation projections require, and the paper does not show this retention step by step.

Editorial extensions

If this is right

  • The Euler–Maruyama-based BUG integrator preserves local mass and momentum with only the standard conservative augmentation, while the Heun-based variant needs extra enrichment ($E^n\rho^n$ and $\sigma_s(x)\rho^n$) to close the momentum and energy balances.
  • Both integrators show strong order 0.5 in the $L^2$ error over independent Brownian paths, so structure preservation does not degrade the convergence rate of the underlying stochastic time integrators.
  • With the fixed moment modes ($m=3$) mass is conserved to machine precision, whereas with $m=0$ the mass drifts, confirming that the fixed velocity basis is the mechanism that enforces the invariants.
  • The discrete energy balance keeps the electric-field residual $R^n_E$ explicit, so the local energy law holds exactly without requiring a discrete Ampère-type estimate.

Reading between the lines

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

  • A testable extension is to measure the projection residual of $\tau E^n\rho^n + \sum_s \Delta\beta^n_s \sigma_s(x)\rho^n$ after conservative truncation; a nonzero residual on a concrete run would show that the proof's retention assumption fails in practice.
  • The same fixed-moment-mode construction would likely transfer to Vlasov–Maxwell or Fokker–Planck systems, where the corresponding macroscopic closures (charge, current, energy) would be inherited if the fixed velocity basis contains the relevant moments.
  • The Heun scheme's extra enrichment requirement suggests a general design principle: stochastic integrators with quadratic correction terms in their increments will need the products of the noise coefficients with the density (and with the current) added to the augmented spatial basis.
  • The explicit electric-field residual in (51) could support a post-processing step or a modified field update that makes the discrete energy law exactly conservative on average, without waiting for a convergence proof of the field.
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 manuscript proposes dynamical low-rank (DLR) approximations for the stochastic Vlasov–Poisson equation with Stratonovich transport noise. It derives continuous low-rank factor equations in both Stratonovich and Itô forms, designs two augmented basis-update Galerkin (BUG) integrators (Euler–Maruyama for the Itô formulation and Heun for the Stratonovich formulation), and claims that both schemes satisfy discrete local mass, momentum, and energy balance laws, implying pathwise mass conservation and momentum conservation in expectation. The paper also reports numerical experiments on two-stream instability and Landau damping, including Monte Carlo momentum statistics, rank adaptation, and strong convergence tests.

Significance. The paper addresses an open and worthwhile problem: constructing structure-preserving dynamical low-rank integrators for stochastic kinetic equations. The continuous projected dynamics and the identification of the moment spaces are plausible extensions of the deterministic framework in [12], and the distinction between the Itô and Stratonovich time discretizations is informative. The numerical experiments are consistent with the claimed conservation properties and include a useful comparison of the two integrators. However, the central discrete conservation results rest on an unproven assertion about the effect of the conservative SVD truncation on the post-truncation basis, so the paper's main theoretical claim is not yet fully supported. The gap appears fixable, either by proving an invariance property of the truncation or by modifying the algorithm to retain the required moment functions.

major comments (2)
  1. [Section 3.3 and Section 4] The discrete conservation theorems (4.1, 4.4, 4.7) are proved with the final spatial basis X^{n+1} obtained after conservative SVD truncation, using projection identities (25)–(27) and exact projections (39), (40), (47), (49), (50). The augmented basis eX^{n+1} is defined to contain span{X^n_i, ∇X^n_i, K^{n+1}_i}, and Lemmas 4.3 and 4.6 show that functions such as (τE^n + Σ_s Δβ^n_s σ_s)ρ^n and (τE^n + Σ_s Δβ^n_s σ_s)J^n lie in span{eX^{n+1}}. However, the conservative truncation keeps only the first m columns of K̃ = eX^{n+1} S̃^{n+1} exactly and compresses the remaining columns by SVD to r−m singular vectors; the resulting r-dimensional spatial basis X^{n+1} is a subspace of span{eX^{n+1}}, but the paper gives no argument that it contains those functions, nor that it contains X^n_i and ∇X^n_i for all i. The sentence in Section 3.3 that 'the rank reduction does not modify the components required for the discrete conservation properties' is precisely the assertion that needs proof. Consequently, the discrete local momentum balance (41) and energy balance (51) are not established for the truncated method as written; the same gap affects the use of (25)–(27) in Theorem 4.1.
  2. [Section 4.2 (condition (40)) and Section 4.3 (conditions (49)–(50))] For the Heun-based integrator, the proofs require the final spatial basis to project exactly the functions τE^nρ^n + Σ_s Δβ^n_s σ_sρ^n, Θ^n_θ, and τE^n·J^n + Σ_s Δβ^n_s σ_s·J^n. The text states that this 'can be enforced by including E^nρ^n and σ_s(x)ρ^n in the spatial augmentation before the conservative truncation,' but inclusion in the pre-truncation augmented basis does not imply retention after SVD compression to rank r. Unless the truncation is modified to keep these functions explicitly, conditions (40), (49), and (50) are additional assumptions on the post-truncation basis rather than consequences of the algorithm described in Section 3.3, and the Heun-based discrete conservation laws are not established as stated.
minor comments (5)
  1. [Section 2.1] The definitions of the inner products are inconsistent: ⟨·,·⟩_x is unweighted, ⟨·,·⟩_v is weighted by f0v, yet the text says '⟨·,·⟩_v and ⟨·,·⟩_xv denote the corresponding f0v-weighted inner products.' This makes it difficult to verify the weak formulations (6)–(10).
  2. [Section 2.2] The phrase 'and its invariant' after the choice of f0v appears to be incomplete; please rephrase.
  3. [Sections 4.1–4.3] The theorems implicitly assume that the discrete spatial operators and Poisson solver satisfy the discrete analogues of ∫∇·J dx = 0 and ∫ E ρ dx = 0, as well as the integration-by-parts identities used in Section 5.1; these assumptions should be stated in the theorem statements rather than only in the implementation section.
  4. [Equation (30) and proof of Theorem 4.1] The proof suppresses the velocity mode V^{n+1}_1 in the inner products, mixing (X^{n+1}_k, ·)_{xv} with the coefficient update in (19); rewriting with the full test functions X^{n+1}_k V^{n+1}_1 would improve readability and avoid confusion about the factors involving ∥1∥_v.
  5. [Section 5.3.1, constant-noise experiment] In the experiment with σ=0.1, the mean momentum drift of 2.81e-3 is reported with the confidence interval, but the initial discrete momentum and the pathwise standard deviation (σ_path≈2.09) appear only in the text; including them in the figure caption or a table would help the reader assess the scale.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the conservation laws are derived from the Petrov–Galerkin projection and fixed moment modes; the main caveat is an unproven truncation-retention step, which is a correctness concern rather than a circular reduction.

full rationale

The paper's derivation chain is self-contained in the relevant sense. The continuous low-rank equations (8)-(10) follow from imposing Petrov–Galerkin conditions on the residual of the stochastic Vlasov–Poisson equation, and the fixed velocity modes U1=1/||1||_v, U_{i+1}=v_i/||v_i||_v, U_{d+2}=chi^2/||chi^2||_v are chosen so that mass, momentum and energy densities become explicit linear functions of the low-rank factors. The identities (11), (12) and (13) are then obtained by direct velocity integration and the Stratonovich chain rule, not by fitting or by assuming the desired balance laws. The discrete mass law (28) is proved using the augmented-basis projection identities (25)-(26), which hold by construction of the augmented spaces, and using the fact that the stochastic increment against the constant velocity mode is zero by velocity integration by parts. The momentum law for the Euler–Maruyama scheme is supported by Lemma 4.3, which shows that (tau E^n + sum_s Delta beta_s^n sigma_s) rho^n lies in span{eX^{n+1}} because it appears in the K-step update (16); this is a real computation, not a circular definition. The energy balance (51) is likewise obtained by explicit computation of the velocity moments of the EM and Heun increments, with the Itô and Heun quadratic corrections evaluated in (56)-(60). The theorems for the Heun scheme are conditional on the projection assumptions (40), (49) and (50); these are stated hypotheses, so the proofs are valid as conditional statements. The paper's claim in Section 3.3 that the conservative truncation 'does not modify the components required for the discrete conservation properties' is asserted but not proved: the final spatial basis after SVD compression may not retain E^n rho^n, sigma_s(x) rho^n, Theta^n, E^n J^n, or sigma_s(x) J^n, which are needed for the projections in (39), (40), (47), (49) and (50). This is a genuine correctness/robustness gap, but it is not circularity: the discrete balances are not being assumed as their own conclusions; rather, an unverified invariant-retention property of the truncation is used. The paper does cite prior work by a coauthor ([13], Einkemmer, Ostermann and Scalone) for the conservative truncation framework, and it builds on [12] for the fixed-mode idea; however, these are independently published methods, and the stochastic extension, the Itô/Stratonovich comparison, the Heun correction analysis, and the numerical experiments are not a relabeling of those works.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central claims rely on standard DLR tangent-space projection and on regularity of the stochastic Vlasov-Poisson model borrowed from [6]. No physical constants are fitted; rank, fixed-mode count, and truncation tolerance are user choices. The most consequential unproved input is that the final truncated basis retains the functions required by the projection conditions in the discrete conservation theorems.

free parameters (3)
  • Fixed velocity mode count m = 3 in experiments (theory requires m >= d+2 for energy)
    The conservation analysis requires enough fixed modes to represent mass, momentum, and kinetic energy; m is a method choice, not fitted to data.
  • DLR rank r = 5, 7, or 15 in experiments
    User-selected approximation rank; the conservation claims are meant to hold for any r, but the experiments use specific values.
  • Rank-adaptation tolerance theta = 1e-4 in the adaptive experiment
    Controls when singular values are discarded in Section 3.4; an algorithmic choice, not an empirical fit.
assumptions (5)
  • domain assumption The stochastic Vlasov-Poisson model is well posed and satisfies the momentum and energy balance laws (3)-(4) under suitable coefficient assumptions from [6].
    Invoked in Section 1 to justify the continuous conservation laws that the low-rank scheme is designed to mimic.
  • domain assumption The weighted L2 manifold and tangent-space Petrov-Galerkin residual condition (6) give the correct dynamics for the low-rank factors.
    Standard DLR construction in Section 2.1, following [12]; the Stratonovich chain rule and stochastic projection are used implicitly.
  • domain assumption The velocity weight f0v is such that 1, v, and |v|^2 lie in L^2(R^d; f0v), and centered finite differences reproduce the continuous integration-by-parts identities exactly.
    Used in Sections 2.2 and 5.1; the discrete derivative of the product f0v V_l is computed directly because the product rule is not satisfied by centered differences.
  • ad hoc to paper After conservative SVD truncation, the final spatial basis still makes the projection conditions (39), (40), (47), (49), and (50) exact.
    This is assumed in Theorems 4.4 and 4.7, but no construction or proof shows the truncation preserves these spans. This is the paper's weakest load-bearing premise.
  • standard math Brownian increments are independent of F_n with E[Delta beta_s Delta beta_r | F_n] = tau delta_sr.
    Used in Theorem 4.7 to identify the conditional expectation of the Heun correction with an Itô-type drift term.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Structure-Preserving Dynamical Low-Rank Approximations for Stochastic Vlasov--Poisson Equations." pith.science (2026). https://pith.science/paper/6XSACRND

@misc{pith2026260800397,
  author       = {Pith},
  title        = {Pith review of: Structure-Preserving Dynamical Low-Rank Approximations for Stochastic Vlasov--Poisson Equations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6XSACRND}},
  note         = {Machine review of arXiv:2608.00397}
}
read the original abstract

We propose structure-preserving dynamical low-rank methods for stochastic Vlasov--Poisson equations with transport noise. We first derive the continuous low-rank evolution equations and show that, by including fixed velocity modes associated with the conserved quantities, the low-rank dynamics inherits the mass, momentum, and energy balance laws of the original stochastic model. We then develop two augmented basis-update Galerkin (BUG) integrators based on two stochastic discretizations: an Euler--Maruyama scheme applied to the equivalent It\^o formulation and a Heun scheme applied directly to the Stratonovich formulation. These two choices allow us to study how the stochastic time discretization interacts with the conservative low-rank framework. In both cases, basis augmentation and conservative rank truncation retain the relevant moment spaces and provide robustness under rank reduction. Numerical experiments demonstrate the conservation properties of the proposed methods and compare the augmentation requirements, stochastic correction terms, and momentum and energy behavior induced by the two stochastic formulations.

Figures

Figures reproduced from arXiv: 2608.00397 by the authors.

Figure 1
Figure 1. Computational results for the Heun method on the two-stream instability problem. [PITH_FULL_IMAGE:figures/full_fig_p024_1.png] view at source ↗
Figure 2
Figure 2. Computational results for the Euler–Maruyama method on the two-stream instability [PITH_FULL_IMAGE:figures/full_fig_p025_2.png] view at source ↗
Figure 3
Figure 3. Mean momentum for the linear Landau test case over [PITH_FULL_IMAGE:figures/full_fig_p025_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Euler–Maruyama method tested on the two-stream instability problem with noise [PITH_FULL_IMAGE:figures/full_fig_p026_4.png]
Figure 5
Figure 5. Figure 5: Heun method: rank adaptation over a single path (left) and mass error (right). [PITH_FULL_IMAGE:figures/full_fig_p027_5.png]
Figure 6
Figure 6. Figure 6: Mass error for the Euler–Maruyama method in the case [PITH_FULL_IMAGE:figures/full_fig_p027_6.png]
Figure 7
Figure 7. Figure 7: Strong convergence test for the DLR method applied to the two-stream instability. [PITH_FULL_IMAGE:figures/full_fig_p028_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 28 canonical work pages

  1. [12]

    Einkemmer and I

    L. Einkemmer and I. Joseph. A mass, momentum, and energy conservative dynamical low-rank scheme for the Vlasov equation.J. Comput. Phys., 443:110495, 2021

  2. [1]

    Bardos and N

    C. Bardos and N. Besse. Diffusion limit of the Vlasov equation in the weak turbulent regime.Journal of Mathematical Physics, 62(10):101505, 2021

  3. [2]

    SplittingintegratorsforlinearVlasovequationswithstochas- tic perturbations.Journal of Computational Dynamics, 11(4):494–532, 2024

    C.-E.BrehierandD.Cohen. SplittingintegratorsforlinearVlasovequationswithstochas- tic perturbations.Journal of Computational Dynamics, 11(4):494–532, 2024

  4. [3]

    Cassini and L

    F. Cassini and L. Einkemmer, Efficient 6D Vlasov simulation using the dynamical low- rank framework Ensign,Comput. Phys. Commun.,280(2022), Art. 108489

  5. [4]

    Ceruti and C

    G. Ceruti and C. Lubich. An unconventional robust integrator for dynamical low-rank approximation.BIT Numer. Math., 62:23–44, 2022

  6. [5]

    Ceruti, J

    G. Ceruti, J. Kusch, and C. Lubich. A rank-adaptive robust integrator for dynamical low-rank approximation.BIT Numer. Math., 2022

  7. [6]

    J. Cui, D. Sheng, C. Zhang, and T. Zhou. A dynamic domain semi-Lagrangian method for stochastic Vlasov equations.Journal of Computational Physics, 541:114335, 2025

  8. [7]

    Einkemmer, K

    L. Einkemmer, K. Kormann, J. Kusch, R. G. McClarren, and J.-M. Qiu. A review of low- rank methods for time-dependent kinetic simulations.Journal of Computational Physics, 538:114191, 2025

Show all 30 references
  1. [8]

    Kormann, M

    K. Kormann, M. Nazarov, and J. Wen. Positivity-preserving dynamical low-rank methods for the Vlasov equation.arXiv preprint arXiv:2606.31662, 2026

  2. [9]

    Einkemmer and C

    L. Einkemmer and C. Lubich. A quasi-conservative dynamical low-rank algorithm for the Vlasov equation.SIAM J. Sci. Comput., 41(5):B1061–B1081, 2019

  3. [10]

    Einkemmer and C

    L. Einkemmer and C. Lubich. A Low-Rank Projector-Splitting Integrator for the Vlasov– Poisson Equation.SIAM J. Sci. Comput., 40:B1330–B1360, 2018

  4. [11]

    Einkemmer, A

    L. Einkemmer, A. Ostermann, and C. Piazzola. A low-rank projector-splitting integra- tor for the Vlasov–Maxwell equations with divergence correction.J. Comput. Phys., 403:109063, 2020

  5. [13]

    Einkemmer, A

    L. Einkemmer, A. Ostermann, and C. Scalone. A robust and conservative dynamical low-rank algorithm.Journal of Computational Physics, 484:112060, 2023

  6. [14]

    Feppon and P

    F. Feppon and P. F. J. Lermusiaux. Dynamically Orthogonal Numerical Schemes for Efficient Stochastic Advection and Lagrangian Transport.SIAM Review, 60(3):595–625, 2018

  7. [15]

    Garbet, Y

    X. Garbet, Y. Idomura, L. Villard, and T. H. Watanabe. Gyrokinetic simulations of turbulent transport.Nuclear Fusion, 50:043002, 2010

  8. [16]

    Guo and J.-M

    W. Guo and J.-M. Qiu. A low rank tensor representation of linear transport and nonlinear Vlasov solutions and their associated flow maps.Journal of Computational Physics, 458:111089, 2022

  9. [17]

    Hairer, C

    E. Hairer, C. Lubich, and G. Wanner.Geometric numerical integration: structure- preserving algorithms for ordinary differential equations, volume 31. Springer Science, 2006. 29

  10. [18]

    Kieri, C

    E. Kieri, C. Lubich, and H. Walach. Discretized dynamical low-rank approximation in the presence of small singular values.SIAM J. Numer. Anal., 54(2):1020–1038, 2016

  11. [19]

    Kazashi, F

    Y. Kazashi, F. Nobile, and F. Zoccolan. Dynamical low-rank approximation for stochastic differential equations.Mathematics of Computation, 94(353):1335–1375, 2025

  12. [20]

    Kazashi, F

    Y. Kazashi, F. Nobile, and F. Zoccolan. Numerical Methods for Dynamical Low-Rank Approximations of Stochastic Differential Equations – Part I: Time Discretization.arXiv preprint arXiv:2601.21428, 2026

  13. [21]

    Koch and C

    O. Koch and C. Lubich. Dynamical Low-Rank Approximation.SIAM J. Matrix Anal. Appl., 29:434–454, 2007

  14. [22]

    Y. Li, F. Holderied, S. Possanner, and E. Sonnendrücker. Canonical variables based numerical schemes for hybrid plasma models with kinetic ions and massless electrons.J. Comput. Phys., 505, Paper No. 112916, 17, 2024

  15. [23]

    G. J. Lord, C. E. Powell, and T. Shardlow.An Introduction to Computational Stochastic Partial Differential Equations. Cambridge University Press, 2014

  16. [24]

    Musharbash, F

    E. Musharbash, F. Nobile, and T. Zhou. Error analysis of the Dynamically Orthogonal approximation of time dependent random PDEs.SIAM J. Sci. Comput., 37(2):A776– A810, 2015

  17. [25]

    Nobile and S

    F. Nobile and S. Riffaud. High-Order BUG Dynamical Low-Rank Integrators Based on Explicit Runge–Kutta Methods.Journal of Scientific Computing, 107(3):102, 2026

  18. [26]

    T. P. Sapsis and P. F. J. Lermusiaux. Dynamically Orthogonal Field Equations for Con- tinuous Stochastic Dynamical Systems.Physica D: Nonlinear Phenomena, 238(23):2347– 2360, 2009

  19. [27]

    Scalone, L

    C. Scalone, L. Einkemmer, J. Kusch, and R. G. McClarren, A multi-fidelity adaptive dynamical low-rank based optimization algorithm for fission criticality problems,J. Sci. Comput.,104(2025), Art. 27

  20. [28]

    Sonnendrücker

    E. Sonnendrücker. Numerical methods for the Vlasov equations.https://www-m16. ma.tum.de/foswiki/pub/M16/Allgemeines/NumMethVlasov/Num-Meth-Vlasov-Notes. pdf, 2013

  21. [29]

    Uschmajew and A

    A. Uschmajew and A. Zeiser. Dynamical low-rank approximation of the Vlasov–Poisson equation with piecewise linear spatial boundary.BIT Numerical Mathematics, 64(2):19, 2024

  22. [30]

    Zheng, X

    N. Zheng, X. Cai, J. Qiu and J. Qiu, A conservative semi-Lagrangian hybrid Hermite WENO scheme for linear transport equations and the nonlinear Vlasov-Poisson system SIAM J. Sci. Comput., 43(5), A3580–A3606, 2021. 30

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.