Pith. sign in

REVIEW 4 major objections 4 minor 27 references

A Directional-ODE Framework for Discretization of Advection-Diffusion Equations

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Discretizing advection-diffusion equations as directional ODEs yields closed-form update formulas with unconditional stability of the discrete scheme.

desk verdict Plausible exponential-integrator extension with real new formulas, but the stability and accuracy claims are not yet proven. read the letter →

arxiv 2506.06543 v1 pith:T42VU7B4 submitted 2025-06-06 math.AP cs.NAmath-phmath.MPmath.NA

Amin Jafarimoghaddam , Manuel Soler , Irene Ortiz This is my paper · ORCID
classification math.APcs.NAmath-phmath.MPmath.NA MSC 35K5565M0665M1265M20
keywords advection-diffusionequationsdirectional-ODEdiscretizationfinitedifferenceschemesrepresentativeODEsmethodofcharacteristicsoperatorsplittingSegmentedAdomianDecompositionstochasticdiffusioncoefficient
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 tries to establish that discretizing advection-diffusion equations need not choose between explicit and implicit schemes: the discrete equations can be rewritten as ODEs in one direction, either time or space, and then solved exactly. The resulting update formulas carry the continuous exponential relaxation of diffusion inside the discrete step, which the paper argues improves stability, efficiency, and spatiotemporal resolution. The framework generalizes earlier linear, single-direction constructions to nonlinear, non-autonomous, multidimensional, and stochastic problems, and it claims unconditional stability of the discrete scheme for the diffusion operator. A sympathetic reader should care because the method offers closed-form updates for a class of problems where numerical schemes ordinarily supply only algebraic update rules.

What carries the argument

Representative ODE: an ordinary differential equation obtained by compressing a discretized ADE along one direction, freezing space to get a temporal ODE or freezing time to get a spatial ODE. The temporal-ODE machinery assumes that the sum of neighboring node values $U(\tau) = u_{i-1}(\tau) + u_{i+1}(\tau)$ can be represented as a degree-$P$ polynomial; the polynomial coefficients are recovered by a predictor-corrector sampling procedure, and the closed-form solution becomes the update. The spatial-ODE machinery solves a second-order boundary value problem between neighboring grid points, producing an implicit exponential scheme. These ODEs carry the exponential factors $\mathrm{e}^{-A\tau}$ or $\mathrm{e}^{\pm\sqrt{A}x}$ that standard finite-difference updates lack, and this factor is what the paper identifies as the structural property of the diffusion operator.

What would settle it

Take the linear heat equation with an initial condition whose exact solution is known; compute $U(\tau) = u_{i-1}(\tau) + u_{i+1}(\tau)$ exactly, fit its best degree-$P$ polynomial, and compare the predictor-corrector coefficients to that fit. If the error between the two does not shrink at the claimed order as $P$ grows, or if the resulting update does not converge to the exact solution at the claimed rate on a stiff, high-CFL test, the arbitrary-order claim is refuted.

Watch

Extended reading notes

Core claim

The paper's central claim is that advection-diffusion equations can be discretized by compressing the discrete equations into directional ODEs. In the temporal-ODE scheme, spatial derivatives are expanded and neighboring node values are treated as known or as a polynomial in time, giving a first-order ODE $\mathrm{d}u_i/\mathrm{d}\tau = -A u_i + B$ whose exact solution is the update formula (8); in the spatial-ODE scheme, time is frozen and the resulting second-order ODE $\mathrm{d}^2u/\mathrm{d}x^2 - A u = B$ is solved exactly on an interval between neighboring grid points to give an implicit update (12). The paper claims these exponential update formulas embed the continuous relaxation of the diffusion operator into the discrete step, making the discrete scheme unconditionally stable in the sense that as $\Delta t \to \infty$ the update converges to a Gauss-Seidel or fully implicit iteration. The same compression is applied to nonlinear diffusion coefficients, solved via the Segmented Adomian Decomposition Method when no closed form exists, and to uncertain coefficients, where the expected update is derived in closed form.

Load-bearing premise

The claim of arbitrary-order accuracy depends on the assumption that the evolving sum of neighboring node values $U(\tau) = u_{i-1}(\tau) + u_{i+1}(\tau)$ can be captured by a degree-$P$ polynomial on each time step, with its coefficients recovered from a predictor-corrector sampling; the paper provides no error bound or convergence proof for that polynomial representation.

Editorial extensions

If this is right

  • The temporal-ODE update (8) replaces the classical explicit Euler diffusion step with an exponential relaxation; because the factor $\mathrm{e}^{-A\Delta t}$ decays for any positive time step, the scheme does not suffer the CFL-driven blow-up that limits standard explicit methods.
  • The spatial-ODE update (12) yields an implicit system that can be rearranged into tridiagonal form; as $\Delta t \to \infty$, the discrete solution approaches equilibrium rather than oscillating, inheriting the steady-state behavior of the continuous operator.
  • In the limit $\Delta t \to \infty$, the $P=0$ temporal-ODE update reduces to the explicit Gauss-Seidel iteration and the $P \neq 0$ updates reduce to the fully implicit iteration, which is the paper's mechanism for unconditional stability of the discrete scheme.
  • For problems with a random diffusion coefficient, the expectation of the update is available in closed form (Eq. 29), so the mean field can be propagated without Monte Carlo sampling of the coefficient.
  • Applied to the streamfunction-vorticity Navier-Stokes equations (Example 3), the zeroth-order temporal-ODE scheme reaches steady state in far fewer iterations than the classical ADI method under identical settings.

Reading between the lines

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

  • The exponential update structure points toward a comparison with exponential integrators for stiff reaction-diffusion systems; a natural extension is to test whether the representative-ODE updates match the accuracy of standard exponential time-differencing schemes on problems with strongly nonlinear sources.
  • The predictor-corrector loop that refines the polynomial coefficients is structurally similar to deferred correction; one could make the arbitrary-order claim precise by analyzing how the iteration count and polynomial degree control the local truncation error.
  • The probabilistic formula (29) computes only the expectation; a straightforward extension would derive variance or other moments of the solution under a random diffusion coefficient, directly quantifying predictive uncertainty.
  • The framework's claim that it integrates continuous temporal and spatial history between grid points could be tested on problems with known oscillatory or front solutions, where the exponential structure should visibly outperform classical schemes in phase accuracy.
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

4 major / 4 minor

Summary. The paper proposes a 'directional-ODE' discretization framework for advection-diffusion equations, in which the spatially or temporally discretized equations are reinterpreted as ordinary differential equations along one direction. A temporal-ODE scheme solves a scalar ODE for each node, modeling the neighboring sum as a polynomial in time and using a predictor-corrector procedure to determine the polynomial coefficients; a spatial-ODE scheme solves a boundary-value ODE across each cell. The paper claims closed-form analytical update formulas, arbitrary order of accuracy, unconditional stability, and extension to stochastic diffusion coefficients. The algebraic derivations appear internally consistent under the stated polynomial and local-uniformity assumptions, but the analysis does not establish the advertised accuracy and stability properties, and the numerical validations are partly circular.

Significance. If the central claims were rigorously established, the framework would be a useful contribution: it offers explicit exponential-type update formulas, a unified reinterpretation of explicit and implicit schemes, and a novel treatment of uncertain diffusion coefficients. The closed-form updates for P=0,1,2 are checkable and the stochastic extension in Eq. (29) is original. However, the two main theoretical claims — arbitrary order of accuracy and unconditional stability — are currently unsupported, and the examples do not provide independent validation. The approach is therefore promising but not yet ready for publication in its present form.

major comments (4)
  1. [§3.2, Eqs. (18)–(23)] The central claim of 'arbitrary order of accuracy' is not supported. Equation (20) is the exact solution of the ODE (19) only under the ansatz that U(τ)=u_{i−1}(τ)+u_{i+1}(τ) is exactly a degree-P polynomial on [t_n,t_{n+1}]; in the actual scheme, U is replaced by the interpolating polynomial through samples produced by the P=0 predictor (21)–(22). No truncation-error estimate, no bound on the interpolation error in U, and no proof of convergence of the iterative corrector loop are provided. Consequently, P is a polynomial degree, not an established order of accuracy in Δt.
  2. [§3.4 and Supplementary §8] The 'unconditional stability' claim is not established. Supplementary Section 8 only proves the Δt→∞ limits in Eqs. (85)–(87), showing that the updates reduce to Gauss–Seidel or fully implicit steady-state iterations. This asymptotic property does not bound the amplification factor of the update over finite time steps; a von Neumann or energy stability analysis of the finite-Δt operator is absent. The statement in §3.4 that t→∞ is equivalent to Δt→∞ conflates the asymptotic stability of the continuous ODE with the finite-step stability of the discrete update.
  3. [§4.2 and §4.3] The numerical validation is partly circular. In Example 2, the reference solution is itself a second-order temporal-ODE scheme with loop correction — a variant of the method being tested — so Fig. 6 measures consistency between two members of the same family rather than accuracy against the PDE or an independent reference. Example 3 compares only steady-state iteration counts with ADI and does not report solution error or convergence rates; hence it does not support the claimed superiority in accuracy or spatiotemporal resolution.
  4. [§2.1.2, Eqs. (9)–(12)] The spatial-ODE scheme rests on the assumption that u^n(x) and f(x,t_n,u^n(x)) are locally uniform on each cell [x_i−Δx, x_i+Δx]; this approximation is asserted without an error estimate. The closed-form solution (12) is exact for the constant-coefficient ODE (10), but the reduction from the PDE to that ODE is not analyzed, so the spatial accuracy of the scheme is not established.
minor comments (4)
  1. [Data and Code Availability] The section states that 'the computational codes are publicly available in link' and similar placeholders appear in the Supplementary Information; actual repository URLs are needed.
  2. [§4.3, Fig. 7] The table header 'wall speed rate 1' should be 'wall speed 1' or similar, and the table would be clearer with explicit units and column alignment.
  3. [Eqs. (29) and (90)] The notation ̅a is reused with different meanings: in Eq. (29) it denotes 1/(Δx)², while in Eq. (90) it denotes the sum of reciprocal squared spacings in all dimensions; please align the notation to avoid confusion.
  4. [Supplementary §8, title] The section is titled 'Proof of the Convergence for the Predictor-Corrector Algorithm' but its content proves only a Δt→∞ limit; either retitle the section or add a genuine convergence proof for the corrector iteration.

Circularity Check

2 steps flagged · score 6.0 of 10

The update formulas are derived analytically, but the paper's central 'unconditionally stable' claim is defined into existence via the Δt→∞ limit, and the main examples validate against other variants of the same method family.

  1. self definitional [Main paper Section 3.4 (Stability Analysis of the Diffusion Operator) and Supplementary Information Section 8, Eqs. (77)-(87)]
    "the condition t → ∞ is equivalent to Δt → ∞, corresponding to the steady-state solution. In other words, in a discrete format, we check whether the discrete scheme can converge to a fixed value for arbitrary time steps. Therefore, the stability of a discrete format refers to its asymptotic behavior as Δt → ∞. ... In Supplementary Information, Section 3, we present an exclusive stability analysis of the multi-stage temporal-ODE scheme, demonstrating that the discrete approach is unconditionally stable."

    Stability is explicitly redefined as convergence to a fixed value as Δt → ∞, and the Supplementary proof establishes exactly that limit: lim_{Δt→∞} u(Δt, P=0) = Gauss-Seidel update and lim_{Δt→∞} u(Δt, P≠0) = fully implicit update. Calling this 'unconditionally stable' is therefore a definitional equivalence: the proved asymptotic limit is the paper's own definition of stability. No finite-Δt amplification or von-Neumann-type bound is given, so the advertised unconditional stability claim reduces to the paper's redefined meaning of stability rather than to an independent finite-step stability result.

  2. other [Section 4.2, Example 2, reference-solution sentence after Eq. (34), and Fig. 6 caption]
    "The results are then compared against those obtained using the zeroth-order temporal-ODE scheme. For this comparison, a second-order temporal-ODE scheme with loop correction and a highly refined temporal grid is used as the reference solution."

    The 'reference solution' used to compute errors is itself produced by another member of the same directional-ODE temporal-ODE family being proposed. The comparison therefore measures internal consistency among variants of the method (zeroth-order vs second-order vs third-order SADM), not agreement with an independent exact or externally computed solution. The reported error is a difference within the method class by construction of the test design, making the validation self-referential rather than an external check of the framework's accuracy.

full rationale

The core derivation of the update formulas is not circular: Eq. (8) solves the linear representative ODE exactly, and Eq. (20) is the exact solution of Eq. (19) for a polynomial neighbor sum U(τ), with coefficients computed from an interpolation system Eq. (23). No fitted target data enters this derivation chain. The dominant circularity is the stability claim: stability is redefined as steady-state convergence as Δt → ∞, the Supplementary proof establishes precisely that limit (Eqs. 85-87), and the paper then labels the scheme 'unconditionally stable.' That conclusion follows from the paper's own definition of stability, while the finite-Δt behavior normally denoted by unconditional stability is not analyzed. Secondary self-referential validation appears in Example 2, where the reference solution is another temporal-ODE variant, and in Example 1's stochastic test, where direct sampling reproduces the analytic expectation of the same approximate update formula, confirming only sampling consistency. The 'arbitrary order of accuracy' claim is unsupported because no truncation-error or convergence analysis for the polynomial representation of U(τ) is provided; this is a correctness gap rather than a circularity, since the polynomial degree P is not shown to control time-stepping error.

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

The paper introduces no fitted constants; the parameters in the examples are problem-specific. The key assumptions are: polynomial representation of neighbor sums, convergence of Adomian series, validity of operator splitting, and local uniformity in the spatial-ODE derivation. None are independently verified; the first is central to the claimed order of accuracy.

assumptions (4)
  • ad hoc to paper The sum of neighboring node values U(τ) can be represented by a P-th degree polynomial over each time interval.
    Section 3.2, Eqs. (18)-(23). The multi-stage temporal-ODE scheme's closed-form solution assumes this representation; no error estimate is provided for non-polynomial U.
  • domain assumption The Adomian series in SADM converges to the solution of the nonlinear representative ODE.
    Section 3.3.1; used in Example 2 for nonlinear diffusion, no convergence proof is given in the paper.
  • domain assumption The operator splitting error is negligible or controlled in the test cases.
    Section 2; all examples use first-order or Strang splitting, but the reported errors are not decomposed into splitting error and discretization error.
  • ad hoc to paper In the spatial-ODE scheme, u^n(x) and f(x,t_n,u^n(x)) are locally uniform on each cell [x_i-Δx, x_i+Δx].
    Section 2.1.2; this assumption is introduced to derive the constant-coefficient ODE (10); it does not hold in general.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Directional-ODE Framework for Discretization of Advection-Diffusion Equations." pith.science (2026). https://pith.science/paper/T42VU7B4

@misc{pith2026250606543,
  author       = {Pith},
  title        = {Pith review of: A Directional-ODE Framework for Discretization of Advection-Diffusion Equations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/T42VU7B4}},
  note         = {Machine review of arXiv:2506.06543}
}
read the original abstract

We present a novel approach that redefines the traditional interpretation of explicit and implicit discretization methods for solving a general class of advection-diffusion equations (ADEs) featuring nonlinear advection, diffusion operators, and potential source terms. By reformulating the discrete ADEs as directional ordinary differential equations (ODEs) along temporal or spatial dimensions, we derive analytical solutions that lead to novel update formulas. In essence, the information of discrete ADEs is compressed into these directional ODEs, which we refer to as representative ODEs. The analytical update formulas derived from the representative ODEs significantly enhance stability, computational efficiency, and spatiotemporal resolution. Furthermore, we extend the framework to systems with uncertain parameters and coefficients, showcasing its versatility in addressing complex ADEs encountered in modeling and simulation across diverse scientific and engineering disciplines.

Figures

Figures reproduced from arXiv: 2506.06543 by the authors.

Figure 1
Figure 1. Comparison of temporal-ODE and spatial-ODE schemes [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Schematic diagram of the proposed method for solving advection-diffusion equations [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Example 1 — Performance and probabilistic solution comparisons for the 1D Burgers’ equation. 17 [PITH_FULL_IMAGE:figures/full_fig_p017_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Example 2 — The evolution of particles’ concentration with β = 10 and smooth velocity field 20 [PITH_FULL_IMAGE:figures/full_fig_p020_4.png]
Figure 5
Figure 5. Figure 5: Example 2 — The evolution of particles’ concentration with β = 10 and noisy velocity field [PITH_FULL_IMAGE:figures/full_fig_p021_5.png]
Figure 6
Figure 6. Figure 6: Example 2 — Comparison of errors between the zeroth-order linear closed-form dis￾cretization and the third-order nonlinear SADM discretization for β = 750 (left) and β = 1000 (right). 21 [PITH_FULL_IMAGE:figures/full_fig_p021_6.png]
Figure 7
Figure 7. Figure 7: Example 3 — Flow configuration schematic (top) and test case parameters (bottom): u1 is wall speed rate 1, u2 is wall speed rate 2, h1 is step height 1, and h2 is step height 2 23 [PITH_FULL_IMAGE:figures/full_fig_p023_7.png]
Figure 8
Figure 8. Figure 8: Example 3 — Cross-section velocities for the considered configurations. 24 [PITH_FULL_IMAGE:figures/full_fig_p024_8.png]
Figure 9
Figure 9. Figure 9: Example 3 — Stream functions for the considered configurations. 25 [PITH_FULL_IMAGE:figures/full_fig_p025_9.png]
Figure 10
Figure 10. Figure 10: Example 3− Comparison between the zeroth-order temporal-ODE scheme and the classic ADI method− (a) to (e) represent first to fifth configurations considered respectively. 5 Conclusion In this work, the directional-ODE discretization approach was introduced as a novel …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 27 canonical work pages

  1. [1]

    Basdevant, M

    C. Basdevant, M. Deville, P. Haldenwang, J. M. Lacroix, J. Ouazzani, R. Peyret, P. Orlandi, and A. T. Patera. Spectral and finite difference solutions of the burgers equation.Computers and Fluids, 14(1):23–41, 1986

    work page 1986

  2. [2]

    Afastdirectsolverfortheadvection-diffusionequationusinglow-rankapproxima- tionofthegreen’sfunction

    J.R.Bull. Afastdirectsolverfortheadvection-diffusionequationusinglow-rankapproxima- tionofthegreen’sfunction. InSpectralandHighOrderMethodsforPartialDifferentialEquations ICOSAHOM 2016, pages 423–435. Springer International Publishing, 2017

    work page 2016

  3. [3]

    Modellinguncertainties in the diffusion-advection equation for radon transport in soil using interval arithmetic

    S.Chakraverty,B.K.Sahoo,T.D.Rao,P.Karunakar,andB.K.Sapra. Modellinguncertainties in the diffusion-advection equation for radon transport in soil using interval arithmetic. Journal of Environmental Radioactivity, 182:165–171, 2018

    work page 2018

  4. [4]

    B. M. Chen-Charpentier and H. V. Kojouharov. An unconditionally positivity preserving scheme for advection–diffusion reaction equations.Mathematical and Computer Modelling, 57(9):2177–2185, 2013. System Dynamics in Project Management and Applied Mathematics and Computational Science and Engineering—Selected Papers of the Seventh PanAmerican Workshop — June...

    work page 2013

  5. [5]

    Fastexplicitoperatorsplittingmethodforconvec- tion–diffusion equations.International Journal for Numerical Methods in Fluids, 59(3):309–332, 2009

    A.Chertock,A.Kurganov,andG.Petrova. Fastexplicitoperatorsplittingmethodforconvec- tion–diffusion equations.International Journal for Numerical Methods in Fluids, 59(3):309–332, 2009

    work page 2009

  6. [6]

    Asummaryofnumericalmethodsfortime-dependentadvection- dominated partial differential equations.Journal of Computational and Applied Mathematics, 128(1):423–445, 2001

    R.E.EwingandH.Wang. Asummaryofnumericalmethodsfortime-dependentadvection- dominated partial differential equations.Journal of Computational and Applied Mathematics, 128(1):423–445, 2001. Numerical Analysis 2000. Vol. VII: Partial Differential Equations

    work page 2001

  7. [7]

    Gharib, M

    M. Gharib, M. Khezri, and S. J. Foster. Meshless and analytical solutions to the time- dependent advection-diffusion-reaction equation with variable coefficients and boundary conditions. Applied Mathematical Modelling, 49:220–242, 2017

    work page 2017

  8. [8]

    A.JafarimoghaddamandM.Soler.Perturbed-analyticdirecttranscriptionforoptimalcontrol (PADOC).Optimal Control Applications and Methods, 44(4):2022–2054, 2023

    work page 2022

Show all 27 references
  1. [9]

    Theoptimaldecompositionmethod(ODM) for nonlinear problems.Journal of Computational Science, 62:101690, 2022

    A.Jafarimoghaddam,M.Soler,andA.Simorgh. Theoptimaldecompositionmethod(ODM) for nonlinear problems.Journal of Computational Science, 62:101690, 2022

    2022

  2. [10]

    D.K.Jaiswal,A.Kumar,N.Kumar,andR.R.Yadav. Analyticalsolutionsfortemporallyand spatially dependent solute dispersion of pulse type input concentration in one-dimensional semi-infinite media.Journal of Hydro-environment Research, 2(4):254–263, 2009. 28

    2009

  3. [11]

    H. K. Jalghaf, E. Kovacs, J. Majar, Á. Nagy, and A. H. Askar. Explicit stable finite difference methods for diffusion-reaction type equations.Mathematics, 9(24), 2021

    2021

  4. [12]

    A. S. Kim. Complete analytic solutions for convection-diffusion-reaction-source equations without using an inverse laplace transform.Scientific Reports, 10(8040), 2020

    2020

  5. [13]

    E. Kovacs. A class of new stable, explicit methods to solve the non-stationary heat equation. Numerical Methods for Partial Differential Equations, 37(3):2469–2489, 2021

    2021

  6. [14]

    Unconditionallypositive,explicit,fourthordermethodfor the diffusion- and nagumo-type diffusion–reaction equations.J Sci Comput, 98(39), 2024

    E.Kovacs,J.Majar,andM.Saleh. Unconditionallypositive,explicit,fourthordermethodfor the diffusion- and nagumo-type diffusion–reaction equations.J Sci Comput, 98(39), 2024

    2024

  7. [15]

    Kumar, D

    A. Kumar, D. K. Jaiswal, and N. Kumar. Analytical solutions to one-dimensional advec- tion–diffusionequationwithvariablecoefficientsinsemi-infinitemedia.JournalofHydrology , 380(3):330–337, 2010

    2010

  8. [16]

    Lee and B

    J. Lee and B. Fornberg. A split step approach for the 3-d maxwell’s equations.Journal of Computational and Applied Mathematics, 158(2):485–505, 2003

    2003

  9. [17]

    Mojtabi and M

    A. Mojtabi and M. O. Deville. One-dimensional linear advection–diffusion equation: Ana- lytical and finite element solutions.Computers and Fluids, 107:189–195, 2015

    2015

  10. [18]

    Limitationsofananalyticalsolutionforadvection–diffusionwithspatiallyvariable coefficients

    S.Néelz. Limitationsofananalyticalsolutionforadvection–diffusionwithspatiallyvariable coefficients. Communications in Numerical Methods in Engineering, 22(5):387–396, 2006

    2006

  11. [19]

    Saleh, E

    M. Saleh, E. Kovacs, and I. F. Barna. Analytical and numerical results for the transient diffusion equation with diffusion coefficient depending on both space and time.Algorithms, 16(4), 2023

    2023

  12. [20]

    M. J. Simpson and L. C. Morrow. Analytical model of reactive transport processes with spatially variable coefficients.Royal Society Open Science, 2(5):140348, 2015

    2015

  13. [21]

    G. Strang. On the construction and comparison of difference schemes.SIAM Journal on Numerical Analysis, 5(3):506–517, 1968

    1968

  14. [22]

    PhysicsLettersA ,150(5):262– 268, 1990

    H.Yoshida.Constructionofhigherordersymplecticintegrators. PhysicsLettersA ,150(5):262– 268, 1990

    1990

  15. [23]

    Zamani and F

    K. Zamani and F. A. Bombardelli. Analytical solutions of nonlinear and variable-parameter transport equations for verification of numerical solvers.Environmental Fluid Mechanics, 14:711–742, 2014. 29

    2014

  16. [24]

    Zoppou and J

    C. Zoppou and J. H. Knight. Analytical solution of a spatially variable coefficient advec- tion–diffusionequationinuptothreedimensions. AppliedMathematicalModelling ,23(9):667– 685, 1999. Supplementary Information: A Directional-ODE Framework for Discretization of Advection-Diffus...

    1999

  17. [25]

    Implicit discretization without operator splitting technique: In this case, the nonlinear term ( ∂u ∂x )2 is initiallylinearizedas ( ∂u ∂x )2 ≈ 2 ( ∂u ∂x ) |n( ∂u ∂x ) |n+1 − ( ∂u ∂x |n)2. Therefore,wewritethefullimplicitscheme (using the central difference scheme) as: un+1 i −...

  18. [26]

    In this case, we initially write: ∂u ∂t − Vx ∂u ∂x = ¯d ∂2u ∂x2 , Vx := ( ∂ ¯d ∂u )( ∂u ∂x )

    Implicit discretization with operator splitting technique. In this case, we initially write: ∂u ∂t − Vx ∂u ∂x = ¯d ∂2u ∂x2 , Vx := ( ∂ ¯d ∂u )( ∂u ∂x ) . (49) 32 Next, we apply a single-step advection-diffusion splitting algorithm to solve the advection- diffusion equation (ADE)...

  19. [27]

    = ¯d ¯a · U n+1 2 . This substantiates the proposition.□ 9 Temporal-ODEDiscretizationSchemewithUncertain/Stochastic Diffusion Coefficient:P t h-Order Solution Werevisitthetemporal-ODEschemeinthepresenceofanuncertaindiffusioncoefficient,denoted by ˜¯d. TheP t h-orderrepresentativeODE...

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.