pith. sign in

arxiv: 2602.16508 · v2 · submitted 2026-02-18 · 🧮 math.NA · cs.NA

A Fully Discrete Nonnegativity-Preserving FEM for a Stochastic Heat Equation

Owen Hearder (1) , Claude Le Bris (2 , 3) , Ana Djurdjevac (1 , 4) ((1) Freie Universit\"at Berlin , (2) \'Ecole des Ponts , (3) INRIA , (4) University of Oxford) This is my paper

Pith reviewed 2026-05-15 21:21 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords stochastic heat equationfinite element methodnonnegativity preservationmass lumpingLie-Trotter splittingconvergence analysismultiplicative noisefully discrete scheme
0
0 comments X p. Extension

The pith

Mass-lumped finite elements with Lie-Trotter splitting preserve nonnegativity and converge for the stochastic heat equation.

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

The paper introduces a fully discrete numerical scheme for a stochastic heat equation driven by nonlinear finite-rank space-coloured multiplicative noise. The scheme uses mass-lumped finite elements in space together with a Lie-Trotter splitting in time to separate the deterministic and stochastic parts of the dynamics. Under the assumption that the continuous problem admits a unique nonnegative solution for nonnegative initial data, the discrete solutions are shown to remain nonnegative at every step and to converge to the true solution when the mesh and time step are refined. Numerical tests confirm the predicted rates and illustrate that the method continues to behave well even outside the strict hypotheses of the convergence theorem.

Core claim

The central claim is that the fully discrete scheme obtained by mass-lumped finite-element discretization in space and Lie-Trotter splitting in time preserves nonnegativity for nonnegative initial data and converges in suitable norms to the unique nonnegative mild solution of the stochastic heat equation.

What carries the argument

Mass-lumped finite-element discretization combined with Lie-Trotter operator splitting, which decouples the linear deterministic evolution from the stochastic multiplication and thereby guarantees discrete nonnegativity.

If this is right

  • Discrete solutions remain nonnegative at every time step whenever the initial vector is nonnegative.
  • The scheme converges to the continuous nonnegative solution as the spatial mesh size and time step tend to zero under the stated regularity assumptions.
  • Mass lumping is the ingredient that transfers the continuous positivity property to the finite-element level.
  • The splitting separates the deterministic and stochastic operators, simplifying both the analysis and the implementation.

Where Pith is reading between the lines

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

  • The same combination of mass lumping and splitting may be useful for other stochastic PDEs in which positivity is physically required, such as population or chemical reaction models.
  • Testing the scheme on rougher noise or on nonlinear drift terms not covered by the present theory could map the practical limits of nonnegativity preservation.
  • The method supplies a concrete template for designing positivity-preserving discretizations in related areas such as stochastic fluid models or option pricing under multiplicative uncertainty.

Load-bearing premise

The stochastic heat equation admits a unique nonnegative solution whenever the initial data is nonnegative, together with enough regularity on the solution and noise to close the convergence estimates.

What would settle it

A numerical run in which the computed solution becomes negative at some time step, despite nonnegative initial data and the given finite-rank multiplicative noise, would falsify the preservation property.

Figures

Figures reproduced from arXiv: 2602.16508 by (2) \'Ecole des Ponts, 3), (3) INRIA, 4) ((1) Freie Universit\"at Berlin, (4) University of Oxford), Ana Djurdjevac (1, Claude Le Bris (2, Owen Hearder (1).

Figure 1
Figure 1. Figure 1: Plots for the strong error, (50), when the nonlinearity is [PITH_FULL_IMAGE:figures/full_fig_p020_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Plots for the strong error squared when the nonlinearity [PITH_FULL_IMAGE:figures/full_fig_p021_2.png] view at source ↗
read the original abstract

We consider a stochastic heat equation with nonlinear finite-rank space-coloured multiplicative noise that admits a unique nonnegative solution when given nonnegative initial data. Inspired by existing results for fully discrete finite difference schemes and building on the convergence analysis of semi-discrete mass-lumped finite element approximations, a fully discrete numerical method is introduced that combines mass-lumped finite elements with a Lie-Trotter splitting strategy. This discretization preserves nonnegativity at the discrete level and is shown to be convergent under suitable regularity conditions. A rigorous convergence analysis is provided, highlighting the role of mass lumping in ensuring nonnegativity and of operator splitting in decoupling the deterministic and stochastic dynamics. Numerical experiments are presented to confirm the convergence rates and the preservation of nonnegativity. In addition, we examine several numerical examples outside the scope of the established theory, aiming to explore the range of applicability and potential limitations of the proposed method.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript introduces a fully discrete scheme for the stochastic heat equation with nonlinear finite-rank space-coloured multiplicative noise. The method combines mass-lumped finite elements with Lie-Trotter splitting to preserve nonnegativity at the discrete level. It provides a rigorous convergence analysis under suitable regularity conditions and supports the claims with numerical experiments confirming rates and nonnegativity, including examples outside the main theory.

Significance. If the convergence holds, the work supplies a practical, structure-preserving discretization for SPDEs where nonnegativity is essential. The explicit role of mass lumping for positivity and splitting for decoupling the dynamics extends prior semi-discrete results in a useful way, and the numerical tests help delineate the method's range of applicability.

major comments (1)
  1. [Section 4] Section 4 (convergence analysis): the error bounds for the splitting and FEM consistency steps rely on E[||u(t)||_{H^2}^2] < ∞ (or equivalent higher-norm control on the stochastic convolution). For mild solutions driven by finite-rank multiplicative noise the standard spatial regularity is only H^{1-ε}; the manuscript does not state the additional assumptions on the noise eigenfunctions or diffusion coefficient that would restore the required H^2 moments in the standing hypotheses.
minor comments (2)
  1. [Abstract] The abstract and introduction could more explicitly list the precise regularity hypotheses used in the convergence theorem rather than referring only to 'suitable regularity conditions'.
  2. [Section 2] Notation for the lumped mass matrix and the splitting operators should be introduced with a short display equation in Section 2 or 3 for immediate reference.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading and constructive feedback. We address the single major comment below.

read point-by-point responses

  1. Referee: [Section 4] Section 4 (convergence analysis): the error bounds for the splitting and FEM consistency steps rely on E[||u(t)||_{H^2}^2] < ∞ (or equivalent higher-norm control on the stochastic convolution). For mild solutions driven by finite-rank multiplicative noise the standard spatial regularity is only H^{1-ε}; the manuscript does not state the additional assumptions on the noise eigenfunctions or diffusion coefficient that would restore the required H^2 moments in the standing hypotheses.

    Authors: We agree that the error analysis in Section 4 requires E[||u(t)||_{H^2}^2] < ∞ (or equivalent control on the stochastic convolution), which exceeds the generic H^{1-ε} regularity of mild solutions for finite-rank multiplicative noise. The manuscript invokes “suitable regularity conditions” but does not spell out the precise extra assumptions on the noise eigenfunctions (e.g., higher Sobolev regularity) and diffusion coefficient needed to recover the H^2 moments. In the revised manuscript we will insert an explicit list of these additional hypotheses at the beginning of Section 4, making the standing assumptions complete and transparent. revision: yes

Circularity Check

0 steps flagged

No significant circularity; independent analysis on cited semi-discrete foundations

full rationale

The paper introduces a new fully discrete scheme (mass-lumped FEM + Lie-Trotter splitting) and provides a convergence proof under stated regularity conditions. It explicitly builds on prior results for semi-discrete mass-lumped FEM and finite-difference schemes rather than re-deriving them. No load-bearing step reduces by construction to a self-definition, fitted parameter renamed as prediction, or self-citation chain that collapses the central claim. The regularity hypotheses (e.g., H^2 moments) are external assumptions, not smuggled via ansatz or uniqueness theorem from the same authors. The derivation chain therefore retains independent content and receives the default low score.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on the existence of a unique nonnegative solution to the continuous SPDE and on regularity assumptions that allow the discrete analysis to go through; no free parameters or invented entities are introduced in the abstract.

axioms (2)
  • domain assumption The stochastic heat equation admits a unique nonnegative solution for nonnegative initial data.
    Stated in the abstract as the starting point for the numerical method.
  • domain assumption Suitable regularity conditions hold on the solution and the finite-rank noise.
    Required for the convergence proof; location: abstract.

pith-pipeline@v0.9.0 · 5500 in / 1270 out tokens · 17749 ms · 2026-05-15T21:21:55.273594+00:00 · methodology

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

22 extracted references · 22 canonical work pages · 1 internal anchor

  1. [1]

    Br´ ehier, D

    C.-E. Br´ ehier, D. Cohen, and J. Ulander. Analysis of a positivity- preserving splitting scheme for some semi- linear stochastic heat equations. ESAIM, Math. Model. Numer. Anal. , 58(4):1317–1346, 2024

    work page 2024

  2. [2]

    Br´ ehier and L

    C.-E. Br´ ehier and L. Gouden` ege. Weak convergence rates of splitting schemes for the stochastic Allen-Cahn equation. BIT, 60(3):543–582, 2020. 21

    work page 2020

  3. [3]

    S. C. Brenner and L. R. Scott. The mathematical theory of finite element methods , volume 15 of Texts Appl. Math. New York, NY: Springer, 3rd ed. edition, 2008

    work page 2008

  4. [4]

    J. A. Carrillo, H. Liu, and H. Yu. Positivity-preserving and energy -dissipating discontinuous Galerkin methods for nonlinear nonlocal Fokker-Planck equations. Commun. Appl. Ind. Math. , 16(1):19–40, 2025

    work page 2025

  5. [5]

    Chatzipantelidis, Z

    P. Chatzipantelidis, Z. Horv´ ath, and V. Thom´ ee. On preservation of positivity in some finite element methods for the heat equation. Comput. Methods Appl. Math. , 15(4):417–437, 2015

    work page 2015

  6. [6]

    Cresson, M

    J. Cresson, M. Efendiev, and S. Sonner. On the positivity of solu tions of systems of stochastic PDEs. ZAMM, Z. Angew. Math. Mech. , 93(6-7):414–422, 2013

    work page 2013

  7. [7]

    Da Prato and J

    G. Da Prato and J. Zabczyk. Stochastic equations in infinite dimensions , volume 152 of Encycl. Math. Appl. Cambridge: Cambridge University Press, 2nd ed. edition, 2014

    work page 2014

  8. [8]

    Debussche and J

    A. Debussche and J. Printems. Weak order for the discretizatio n of the stochastic heat equation. Math. Comput., 78(266):845–863, 2009

    work page 2009

  9. [9]

    Djurdjevac, A

    A. Djurdjevac, A. Almgren, and J. Bell. A hybrid algorithm for sys tems of noninteracting particles with an external potential. Commun. Appl. Math. Comput. Sci. , 20(1):147–174, 2025

    work page 2025

  10. [10]

    A nonnegativity-preserving finite element method for a class of parabolic SPDEs with multiplicative noise

    A. Djurdjevac, C. Le Bris, and E. S¨ uli. A nonnegativity-preserving finite element method for a class of parabolic spdes with multiplicative noise. arXiv preprint arXiv:2502.16854 , 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  11. [11]

    Djurdjevac, H

    A. Djurdjevac, H. Kremp, and N. Perkowski. Weak error analy sis for a nonlinear SPDE approximation of the Dean-Kawasaki equation. Stoch. Partial Differ. Equ., Anal. Comput. , 12(4):2330–2355, 2024

    work page 2024

  12. [12]

    Donati-Martin and E

    C. Donati-Martin and E. Pardoux. White noise driven SPDEs with r eflection. Probab. Theory Relat. Fields , 95(1):1–24, 1993

    work page 1993

  13. [13]

    I. Fried. Bounds on the extremal eigenvalues of the finite eleme nt stiffness and mass matrices and their spectral condition number. J. Sound Vib. , 22:407–418, 1972

    work page 1972

  14. [14]

    H. Fujii. Some remarks on finite element analysis of time-depende nt field problems. Theory Pract. finite Elem. struct. Anal., Proc. Semin., Tokyo 1973, 91-106 (1973)., 1973

    work page 1973

  15. [15]

    Geissert, M

    M. Geissert, M. Kov´ acs, and S. Larsson. Rate of weak conve rgence of the finite element method for the stochastic heat equation with additive noise. BIT, 49(2):343–356, 2009

    work page 2009

  16. [16]

    Kotelenez

    P. Kotelenez. Comparison methods for a class of function value d stochastic partial differential equations. Probab. Theory Relat. Fields , 93(1):1–19, 1992

    work page 1992

  17. [17]

    R. Kruse. Strong and weak approximation of semilinear stochastic evo lution equations , volume 2093 of Lect. Notes Math. Cham: Springer, 2014

    work page 2093

  18. [18]

    Liu and M

    W. Liu and M. R¨ ockner. Stochastic partial differential equations: an introductio n. Universitext. Cham: Springer, 2015

    work page 2015

  19. [19]

    Magaletti, M

    F. Magaletti, M. Gallo, S. P. Perez, J. A. Carrillo, and S. Kalliadasis . A positivity-preserving scheme for fluctuating hydrodynamics. J. Comput. Phys. , 463:19, 2022. Id/No 111248

    work page 2022

  20. [20]

    L. A. Oganesjan and L. A. Ruhovec. Variatsionno-raznostnye metody resheniya ellipticheski kh uravnenii [Variational-difference methods for solving elliptic equa tions]. Armenian Academy of Sciences, Erevan, 1979

    work page 1979

  21. [21]

    Thom´ ee

    V. Thom´ ee. On positivity preservation in some finite element met hods for the heat equation. In Numerical methods and applications. 8th international conference, N MA 2014, Borovets, Bulgaria, August 20–24, 2014. Revised selected papers, pages 13–24. Berlin: Springer, 2015

    work page 2014

  22. [22]

    X. Yang, Z. Yang, and C. Zhang. Stochastic heat equation: nu merical positivity and almost surely exponential stability. Comput. Math. Appl. , 119:312–318, 2022. 22

    work page 2022