pith. sign in

arxiv: 2606.03516 · v1 · pith:V5TKMRGUnew · submitted 2026-06-02 · 🧮 math.NA · cs.NA

Linear Convergence of Parareal Algorithm for Semilinear Parabolic Equations

Guanglian Li , Qingle Lin , Shu-lin Wu , Zhi Zhou This is my paper

Pith reviewed 2026-06-28 09:02 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords parareal algorithmsemilinear parabolic equationslinear convergenceparallel-in-time methodsH^2 initial datarational approximationserror analysis
0
0 comments X

The pith

The parareal algorithm converges linearly for semilinear parabolic equations with H^2 initial data using stable rational approximations and first-order linearization as coarse propagators.

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

The paper shows that the parareal algorithm converges linearly for semilinear parabolic equations with initial data in H^2. It uses stable rational approximations and first-order linearization as the coarse propagator to obtain this rate. The analysis links the result to the linear case through error splitting and optimal error estimates adapted to the data regularity. This allows parallel time integration for nonlinear problems that previously lacked convergence theory.

Core claim

We establish the linear convergence of the parareal algorithm and provide a sharp estimate for the convergence factor for semilinear parabolic equations with H^2 initial data using stable rational approximations and first-order linearization as coarse propagators. The analysis combines the error-splitting technique from the superlinear convergence analysis of the parareal method, a refined linear convergence theory for linear parabolic equations, and a priori error estimates that are optimal with respect to the regularity of the problem data. The analysis shows the close connection between the convergence behavior of nonlinear models and their linear counterparts.

What carries the argument

Stable rational approximations combined with first-order linearization as coarse propagators that allow linear convergence to be established for the nonlinear problem.

If this is right

  • The convergence factor is sharp and can be estimated explicitly.
  • The linear rate holds for initial data with H^2 regularity.
  • The convergence behavior is closely connected to that of the corresponding linear parabolic equation.
  • Numerical experiments confirm the linear convergence rate.

Where Pith is reading between the lines

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

  • This method may apply to other classes of semilinear evolution equations.
  • It opens the door to parallel-in-time methods for problems with limited regularity data in applications like heat transfer with nonlinear sources.
  • Higher order linearizations could be explored to potentially improve the convergence factor.

Load-bearing premise

The coarse propagator is a stable rational approximation combined with first-order linearization and the initial data belongs to H^2.

What would settle it

Numerical computation showing that the observed convergence factor for parareal iterations on a semilinear parabolic problem with H^2 data exceeds the predicted sharp estimate when using the specified coarse propagators.

Figures

Figures reproduced from arXiv: 2606.03516 by Guanglian Li, Qingle Lin, Shu-lin Wu, Zhi Zhou.

Figure 1
Figure 1. Figure 1: denote the lower bounds. The inequality (2.10) is used in Step (iii) of the proof of Theorem 1 to locally bound R(s) and 1 − R(s) using linear approximations. Condition (iv) is used in Step (ii) to ensure that h(r) ̸= 0 implies γ ′ (s) ̸= 0 when ∂sQ(s, r) = 0 in (B.6). The condition R(s) > 0 ensures that γ ′ (s) is well-defined. These conditions can be readily verified for specific CPs. 1 1.5 2 2.5 3 0 0.0… view at source ↗
Figure 2
Figure 2. Figure 2: The graph of S(j) for three CPs: BE, two-stage Lobatto IIIC and OCP [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The maximum L 2 error for three CPs versus the iteration k over the setting cL ∈ {1, 5, 10} and ∆T ∈ {0.05, 0.0125, 0.003125}. Left: BE; Middle: Two-stage Lobatto IIIC method; Right: OCP. 5 Conclusion In this work, we have provided a novel convergence analysis of the parareal algorithm for solving semilinear parabolic equations with the H2 initial data. The algorithm employs a single-step coarse propagator… view at source ↗
read the original abstract

Long-time simulations of evolution equations present substantial computational challenges due to the inherently sequential nature of conventional time-stepping schemes. The parareal method, a leading parallel-in-time (PinT) algorithm, offers a promising approach to overcome the challenge by introducing concurrency in the time domain. While its convergence theory is well-established for linear problems, extending the theory to nonlinear problems, particularly when the problem data have only limited regularity, remains a significant challenge. In this work, we provide the convergence analysis of the parareal algorithm for solving semilinear parabolic equations with an $H^2$ initial data. We employ stable rational approximations and first-order linearization as coarse propagators, establish the linear convergence of the parareal algorithm and provide a sharp estimate for the convergence factor. The analysis combines the error-splitting technique from the superlinear convergence analysis of the parareal method, a refined linear convergence theory for linear parabolic equations, and \textsl{a priori} error estimates that are optimal with respect to the regularity of the problem data. The analysis shows the close connection between the convergence behavior of nonlinear models and their linear counterparts. Numerical experiments fully support the theoretical findings.

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

0 major / 4 minor

Summary. The manuscript establishes linear convergence of the parareal algorithm for semilinear parabolic equations with H^2 initial data. Stable rational approximations combined with first-order linearization serve as the coarse propagator. The proof combines the error-splitting technique, a refined linear convergence theory for parabolic equations, and optimal a priori estimates with respect to the given regularity; a sharp bound on the convergence factor is derived, the connection to the linear case is shown, and numerical experiments are presented in support.

Significance. If the central claims hold, the work meaningfully extends parareal convergence theory from the linear to the semilinear setting under limited (H^2) regularity, which is relevant for practical long-time integration of nonlinear evolution equations. The explicit combination of error-splitting with refined linear estimates and optimal a priori bounds is a technical strength, as is the provision of a sharp convergence-factor estimate and the numerical verification. The manuscript successfully controls the first-order linearization error under the stated H^2 regularity without forcing a sublinear rate, so the stress-test concern does not land as an unresolved issue.

minor comments (4)
  1. Abstract: the LaTeX command "\textsl" appears before "a priori"; this should be removed or replaced with proper formatting in the final version.
  2. Section 2: the precise assumptions on the nonlinearity (global vs. local Lipschitz, growth conditions) are stated only implicitly; an explicit list of hypotheses would improve readability.
  3. Figure 1 and Figure 2: axis labels and legend entries are too small for print; increasing font size would aid clarity.
  4. Notation: the symbol for the coarse propagator is introduced in two slightly different forms (one in the text, one in the algorithm box); consistent notation throughout would help.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the careful reading, positive assessment of the significance of the work, and the recommendation for minor revision. The referee's summary correctly identifies the main contributions, including the use of error-splitting, refined linear estimates, optimal a priori bounds under H^2 regularity, and the sharp convergence factor. No specific major comments were provided in the report.

Circularity Check

0 steps flagged

Extends established linear parareal theory to semilinear case via error-splitting and a priori estimates without self-definitional reduction

full rationale

The derivation combines error-splitting from prior superlinear analyses, refined linear convergence theory for parabolic equations, and optimal a priori estimates under H^2 regularity. The linear convergence claim and sharp factor for the semilinear problem are obtained by these combinations applied to the chosen coarse propagator (stable rational approx + first-order linearization). No step reduces a prediction to a fitted input by construction, nor does the central result depend on a load-bearing self-citation chain that itself lacks independent verification. Prior linear results serve as external support. This is the normal non-circular outcome.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The analysis rests on standard assumptions for semilinear parabolic equations and properties of rational approximations; no free parameters are introduced or fitted in the abstract description.

axioms (2)
  • domain assumption The problem is a semilinear parabolic equation with initial data in H^2
    Invoked to obtain optimal a priori error estimates that respect the limited regularity.
  • domain assumption Stable rational approximations combined with first-order linearization serve as suitable coarse propagators
    Required for the error-splitting technique and the connection to linear convergence theory.

pith-pipeline@v0.9.1-grok · 5743 in / 1351 out tokens · 22278 ms · 2026-06-28T09:02:56.796099+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

34 extracted references · 18 canonical work pages

  1. [1]

    Elsevier, Amsterdam (2003)

    Adams, R.A., Fournier, J.J.: Sobolev Spaces. Elsevier, Amsterdam (2003)

    2003

  2. [2]

    Ariel, G., Kim, S.J., Tsai, R.: Parareal multiscale methods for highly oscillatory dynamical systems. SIAM J. Sci. Comput.38(6), A3540–A3564 (2016). DOI 10.1137/15M1011044. URLhttps://doi.org/10. 1137/15M1011044

    work page doi:10.1137/15m1011044 2016

  3. [3]

    In: Domain Decomposition Methods in Science and Engineering, pp

    Bal, G.: On the convergence and the stability of the parareal algorithm to solve partial differential equations. In: Domain Decomposition Methods in Science and Engineering, pp. 425–432. Springer (2005)

    2005

  4. [4]

    parareal

    Bal, G., Maday, Y .: A “parareal” time discretization for non-linear PDE’s with application to the pricing of an American put. In: Recent Developments in Domain Decomposition Methods (Z ¨urich, 2001),Lect. Notes Comput. Sci. Eng., vol. 23, pp. 189–202. Springer, Berlin (2002). DOI 10.1007/978-3-642-56118-4\ 12. URL https://doi.org/10.1007/978-3-642-56118-4_12

    work page doi:10.1007/978-3-642-56118-4 2001

  5. [5]

    Bossuyt, I., Vandewalle, S., Samaey, G.: Monte Carlo-moments micro-macro Parareal method for unimodal and bimodal scalar McKean-Vlasov SDEs. SIAM J. Sci. Comput.47(6), A3239–A3275 (2025). DOI 10.1137/ 23M1609142. URLhttps://doi-org.ezproxy.lb.polyu.edu.hk/10.1137/23M1609142

    work page doi:10.1137/23m1609142 2025

  6. [6]

    Brehier, C.E., Wang, X.: On parareal algorithms for semilinear parabolic stochastic PDEs. SIAM J. Nu- mer. Anal.58(1), 254–278 (2020). DOI 10.1137/19M1251011. URLhttps://doi.org/10.1137/ 19M1251011

    work page doi:10.1137/19m1251011 2020

  7. [7]

    SIAM Journal on Scientific Computing35(1), A52–A78 (2013)

    Dai, X., Maday, Y .: Stable parareal in time method for first-and second-order hyperbolic systems. SIAM Journal on Scientific Computing35(1), A52–A78 (2013)

    2013

  8. [8]

    De Sterck, H., Falgout, R.D., Friedhoff, S., Krzysik, O.A., MacLachlan, S.P.: Optimizing multigrid reduction- in-time and parareal coarse-grid operators for linear advection. Numer. Linear Algebra Appl.28(4), Paper No. e2367, 22 (2021). DOI 10.1002/nla.2367. URLhttps://doi.org/10.1002/nla.2367

    work page doi:10.1002/nla.2367 2021

  9. [9]

    Dobrev, V .A., Kolev, T., Petersson, N.A., Schroder, J.B.: Two-level convergence theory for multigrid reduction in time (MGRIT). SIAM J. Sci. Comput.39(5), S501–S527 (2017). DOI 10.1137/16M1074096. URLhttps: //doi-org.ezproxy.lb.polyu.edu.hk/10.1137/16M1074096

    work page doi:10.1137/16m1074096 2017

  10. [10]

    Multiscale Model

    Engblom, S.: Parallel in time simulation of multiscale stochastic chemical kinetics. Multiscale Model. Simul. 8(1), 46–68 (2009). DOI 10.1137/080733723. URLhttps://doi.org/10.1137/080733723

    work page doi:10.1137/080733723 2009

  11. [11]

    Fang, L., Vandewalle, S., Meyers, J.: A parallel-in-time multiple shooting algorithm for large-scale PDE- constrained optimal control problems. J. Comput. Phys.452, Paper No. 110,926, 24 (2022). DOI 10.1016/j.jcp.2021.110926. URLhttps://doi-org.ezproxy.lb.polyu.edu.hk/10.1016/j. jcp.2021.110926

    work page doi:10.1016/j.jcp.2021.110926 2022

  12. [12]

    Friedhoff, S., Southworth, B.S.: On “optimal”h-independent convergence of parareal and multigrid-reduction- in-time using Runge-Kutta time integration. Numer. Linear Algebra Appl.28(3), e2301, 30 (2021). DOI 10.1002/nla.2301. URLhttps://doi.org/10.1002/nla.2301

    work page doi:10.1002/nla.2301 2021

  13. [13]

    In: Multiple Shooting and Time Domain Decomposition Methods: MuS-TDD, Heidelberg, May 6-8, 2013, pp

    Gander, M.J.: 50 years of time parallel time integration. In: Multiple Shooting and Time Domain Decomposition Methods: MuS-TDD, Heidelberg, May 6-8, 2013, pp. 69–113. Springer (2015)

    2013

  14. [14]

    In: Domain Decomposition Methods in Science and Engineering XVII, pp

    Gander, M.J., Hairer, E.: Nonlinear convergence analysis for the parareal algorithm. In: Domain Decomposition Methods in Science and Engineering XVII, pp. 45–56. Springer (2008)

    2008

  15. [15]

    Gander, M.J., Jiang, Y .L., Song, B., Zhang, H.: Analysis of two parareal algorithms for time-periodic problems. SIAM J. Sci. Comput.35(5), A2393–A2415 (2013)

    2013

  16. [16]

    Gander, M.J., Kulchytska-Ruchka, I., Niyonzima, I., Scho ¨ops, S.: A new parareal algorithm for problems with discontinuous sources. SIAM J. Sci. Comput.41(2), B375–B395 (2019)

    2019

  17. [17]

    Gander, M.J., Kwok, F., Salomon, J.: ParaOpt: a parareal algorithm for optimality systems. SIAM J. Sci. Comput.42(5), A2773–A2802 (2020). DOI 10.1137/19M1292291. URLhttps://doi-org.ezproxy. lb.polyu.edu.hk/10.1137/19M1292291

    work page doi:10.1137/19m1292291 2020

  18. [18]

    Golub and Charles F

    Gander, M.J., Lunet, T.: Time Parallel Time Integration. SIAM, Philadelphia, PA (2024). DOI 10.1137/1. 9781611978025

    work page doi:10.1137/1 2024

  19. [19]

    Gander, M.J., Vandewalle, S.: Analysis of the parareal time-parallel time-integration method. SIAM J. Sci. Com- put.29(2), 556–578 (2007). DOI 10.1137/05064607X. URLhttps://doi.org/10.1137/05064607X

    work page doi:10.1137/05064607x 2007

  20. [20]

    Acta Nu- mer.34, 385–489 (2025)

    Gander, M.J., Wu, S.L., Zhou, T.: Time parallelization for hyperbolic and parabolic problems. Acta Nu- mer.34, 385–489 (2025). DOI 10.1017/S0962492924000072. URLhttps://doi.org/10.1017/ S0962492924000072

    work page doi:10.1017/s0962492924000072 2025

  21. [21]

    Hessenthaler, A., Southworth, B.S., Nordsletten, D., R ¨ohrle, O., Falgout, R.D., Schroder, J.B.: Multilevel con- vergence analysis of multigrid-reduction-in-time. SIAM J. Sci. Comput.42(2), A771–A796 (2020)

    2020

  22. [22]

    Jin, B., Lin, Q., Zhou, Z.: Optimizing coarse propagators in parareal algorithms. SIAM J. Sci. Comput.47(2), A735–A761 (2025)

    2025

  23. [23]

    Legoll, F., Lelievre, T., Samaey, G.: A micro-macro parareal algorithm: application to singularly perturbed ordinary differential equations. SIAM J. Sci. Comput.35(4), A1951–A1986 (2013)

    2013

  24. [24]

    Li, G.: Wavelet-based edge multiscale parareal algorithm for subdiffusion equations with heterogeneous coeffi- cients in a large time domain. J. Comput. Appl. Math.440, 115,608, 24 pp. (2024). DOI 10.1016/j.cam.2023. 115608. URLhttps://doi.org/10.1016/j.cam.2023.115608 24 G. Li, Q. Lin, S-L. Wu, and Z. Zhou

    work page doi:10.1016/j.cam.2023 2024

  25. [25]

    parar ´eel

    Lions, J.L., Maday, Y ., Turinici, G.: R ´esolution d’EDP par un sch ´ema en temps “parar ´eel”. C. R. Acad. Sci. Paris S ´er. I Math.332(7), 661–668 (2001). DOI 10.1016/S0764-4442(00)01793-6. URLhttps://doi. org/10.1016/S0764-4442(00)01793-6

    work page doi:10.1016/s0764-4442(00)01793-6 2001

  26. [26]

    Nirenberg, L.: An extended interpolation inequality. Ann. Scuola Norm. Sup. Pisa Cl. Sci. (3)20, 733–737 (1966)

    1966

  27. [27]

    Pag `es, G., Pironneau, O., Sall, G.: The parareal algorithm for American options. C. R. Math. Acad. Sci. Paris 354(11), 1132–1138 (2016). DOI 10.1016/j.crma.2016.09.010. URLhttps://doi.org/10.1016/j. crma.2016.09.010

    work page doi:10.1016/j.crma.2016.09.010 2016

  28. [28]

    SIAM Journal on Scientific Computing45(5), A2657–A2678 (2023)

    Pentland, K., Tamborrino, M., Sullivan, T.J.: Error bound analysis of the stochastic parareal algorithm. SIAM Journal on Scientific Computing45(5), A2657–A2678 (2023)

    2023

  29. [29]

    AMS, Providence, RI (2014)

    Spencer, J.: Asymptopia. AMS, Providence, RI (2014)

    2014

  30. [30]

    Springer, Dordrecht (2013)

    Suhubi, E.: Functional analysis. Springer, Dordrecht (2013)

    2013

  31. [31]

    Springer-Verlag, Berlin (2006)

    Thom ´ee, V .: Galerkin Finite Element Methods for Parabolic Problems, second edn. Springer-Verlag, Berlin (2006)

    2006

  32. [32]

    Wu, S.L.: Convergence analysis of some second-order parareal algorithms. IMA J. Numer. Anal.35(3), 1315– 1341 (2015)

    2015

  33. [33]

    Wu, S.L., Zhou, T.: Convergence analysis for three parareal solvers. SIAM J. Sci. Comput.37(2), A970– A992 (2015). DOI 10.1137/140970756. URLhttps://doi-org.ezproxy.lb.polyu.edu.hk/10. 1137/140970756

    work page doi:10.1137/140970756 2015

  34. [34]

    Xu, Q., Hesthaven, J.S., Chen, F.: A parareal method for time-fractional differential equations. J. Comput. Phys.293, 173–183 (2015). DOI 10.1016/j.jcp.2014.11.034. URLhttps://doi.org/10.1016/j.jcp. 2014.11.034

    work page doi:10.1016/j.jcp.2014.11.034 2015