pith. sign in

arxiv: 2607.00384 · v1 · pith:L5G7BOKUnew · submitted 2026-07-01 · 🧮 math.NA · cs.NA

Exponential Low-Regularity Parareal Algorithms for Nonlinear Schr\"odinger Equations

Qingle Lin , Zhi Zhou This is my paper

Pith reviewed 2026-07-02 08:22 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords parareal algorithmnonlinear Schrödinger equationexponential integratorslow-regularity methodsparallel-in-time methodsconvergence analysisnumerical PDE methods
0
0 comments X

The pith

Exponential low-regularity integrators enable linearly convergent parareal algorithms for NLS equations even with limited regularity solutions.

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

The paper develops parareal algorithms for the nonlinear Schrödinger equation that use an exact fine propagator paired with specially chosen coarse propagators. It supplies a general convergence theory that applies to solutions of limited regularity and rests on stability plus local truncation error conditions for the coarse step. These conditions are checked for exponential low-regularity integrators on one-dimensional quadratic and cubic NLS, producing the first proof of linear convergence whose contraction factor is proportional to the coarse time-step size. Experiments on quadratic, cubic, and quintic cases show faster iteration counts than splitting or classical exponential Runge-Kutta coarse methods.

Core claim

A convergence framework for parareal methods on NLS guarantees linear convergence with contraction factor proportional to the coarse time-step size for solutions of limited regularity whenever the coarse propagator satisfies the stated stability and local truncation error assumptions; the assumptions hold for the selected exponential low-regularity integrators on one-dimensional quadratic and cubic NLS.

What carries the argument

General convergence framework resting on stability and local truncation error assumptions for the coarse propagator, verified for exponential low-regularity integrators.

Load-bearing premise

The chosen exponential low-regularity integrators must satisfy the required stability and local truncation error bounds for the coarse propagator.

What would settle it

Numerical runs in which the observed contraction factor fails to scale linearly with the coarse time-step size or in which the iteration diverges for low-regularity initial data would disprove the claimed convergence.

Figures

Figures reproduced from arXiv: 2607.00384 by Qingle Lin, Zhi Zhou.

Figure 1
Figure 1. Figure 1: L 2 error versus coarse time step τ ∈ {0.1, 0.05, 0.025, 0.02, 0.01} for the parareal applied to quadratic nonlinear Schrodinger equations of type ( ¨ 3.1) (H 1 2 initial data, µ = −1) with three CPs: ELRI (3.2) (left), Strang splitting (middle), and ERK3 (right). Remark 3.2. We note that, in the proof of Theorem 3.1, the key steps are to establish the stability estimate (3.8) and the consistency estimate … view at source ↗
Figure 2
Figure 2. Figure 2: L 2 error versus iteration k for the parareal applied to quadratic nonlinear Schrodinger equations ¨ of type (3.1) (H2− initial data, µ = 1) with five CPs, and τ ∈ {0.1, 0.05, 0.025}. the exponential low-regularity parareal algorithm converges linearly, with a contraction factor that decreases as τ decreases, thereby fully supporting Theorem 3.1. 4 Exponential low-regularity parareal algorithms Now, we are… view at source ↗
Figure 3
Figure 3. Figure 3: illustrates the results of Lemma 4.4 and Theorem 4.1. The same three CPs as before are tested: the ELRI (4.11), Strang splitting, and ERK3. The parareal method with the ELRI achieves the desired order predicted by Theorem 4.1, whereas the classical schemes again stagnate after the first iteration. 0.01 0.05 0.1 10-15 10-10 10-5 100 0.01 0.05 0.1 10-15 10-10 10-5 100 0.01 0.05 0.1 10-15 10-10 10-5 100 [PIT… view at source ↗
Figure 4
Figure 4. Figure 4: L 2 errors of parareal iterations for 1-D cubic NLS with H3/2− and H2− initial data, µ = 1. Next, we present numerical results for the cubic NLS (1.1) on T 2 . Here, the ELRI (2.5) is used as the CP 20 [PITH_FULL_IMAGE:figures/full_fig_p020_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: L 2 errors of parareal iterations for 1-D cubic NLS with H3/2− and H2− initial data, µ = −1. in the parareal algorithm. We note that the convergence theory for the parareal method in this setting remains open, since the scheme may not satisfy condition (ii) in Assumption 4.1. Nevertheless, these preliminary numerical experiments demonstrate its potential and motivate further study. We set µ = −1 and T = 5 … view at source ↗
Figure 6
Figure 6. Figure 6: L 2 errors of parareal iterations for 2-D cubic NLS with H2− initial data, µ = −1. follow a common pattern: they initially increase before eventually decreasing at a superlinear rate, but their overall performance remains poor. A rigorous theoretical investigation of this setting is left for future work. 0 5 10 15 20 10-6 10-4 10-2 100 ERK1: blow up ERK3: blow up 0 5 10 15 20 10-6 10-4 10-2 100 ERK1: blow … view at source ↗
Figure 7
Figure 7. Figure 7: L 2 errors of parareal iterations for quintic NLS with H2− initial data, µ = −4, and T = 16. 6 Concluding remarks In this work, we developed and analyzed parareal algorithms for solving NLS equations, assuming an exact fine propagator and focusing on the design of suitable coarse propagators. It is well-known that standard parareal methods may converge slowly or even become unstable for NLS equations, larg… view at source ↗
read the original abstract

The parareal algorithm is one of the most widely studied parallel-in-time methods for the numerical approximation of time-dependent problems. For non-diffusive equations, however, standard parareal methods may converge slowly or even become unstable due to the absence of damping, while nonlinear interactions can transfer and amplify phase errors across Fourier modes. In this work, we consider the nonlinear Schr\"odinger equation (NLS) as a representative non-diffusive model and analyze parareal algorithms with an exact fine propagator, with particular emphasis on the design of suitable coarse propagators. We establish a general convergence framework, valid for solutions with limited regularity, under stability and local truncation error assumptions on the coarse propagator. These assumptions are verified for selected exponential low-regularity integrators designed for one-dimensional quadratic and cubic NLS equations, which achieve optimal approximation orders without derivative loss. To the best of our knowledge, this is the first construction of parareal algorithms for NLS equations that are provably linearly convergent, with a contraction factor proportional to the coarse time-step size even for solutions of limited regularity. Numerical experiments on quadratic, cubic, and quintic NLS equations demonstrate rapid convergence and improved performance over parareal variants using classical coarse propagators, including Lie and Strang splitting methods and first- and third-order exponential Runge--Kutta integrators.

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 / 2 minor

Summary. The paper constructs parareal algorithms for the nonlinear Schrödinger equation that employ exponential low-regularity integrators as coarse propagators. It derives a general linear convergence framework under independent stability and local truncation error assumptions on the coarse propagator (valid for limited-regularity solutions), verifies those assumptions for selected integrators on one-dimensional quadratic and cubic NLS, and reports numerical experiments on quadratic, cubic, and quintic NLS demonstrating faster convergence than variants using Lie/Strang splitting or classical exponential Runge–Kutta coarse propagators.

Significance. If the framework and verifications hold, the result supplies the first provably linearly convergent parareal method for NLS with contraction factor proportional to the coarse step size, even for low-regularity data. The separation of the general convergence argument from the specific integrator analysis, together with the numerical confirmation across multiple nonlinearities, constitutes a substantive advance for parallel-in-time methods on non-diffusive nonlinear PDEs.

minor comments (2)
  1. [§3.2] §3.2: the statement of the local truncation error assumption would be clearer if the precise dependence on the coarse step ΔT and the regularity index s were written explicitly rather than left implicit in the O(·) notation.
  2. [Figure 4] Figure 4 caption: the legend does not distinguish the three different NLS nonlinearities shown in the convergence plots; adding a short parenthetical would improve readability.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive review, the recognition of the novelty of the convergence framework, and the recommendation to accept the manuscript.

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper presents a general convergence framework for parareal methods on NLS that rests on independent stability and local truncation error assumptions for the coarse propagator; these assumptions are stated separately and then verified directly for the selected exponential low-regularity integrators on 1D quadratic and cubic cases. No load-bearing step reduces by construction to a self-definition, a fitted parameter renamed as prediction, or a self-citation chain. The central claim of linear convergence with contraction factor proportional to coarse step size follows from the framework plus the verified assumptions, which are external to the target result and do not presuppose it. The derivation is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Based solely on abstract; no explicit free parameters, ad-hoc axioms, or invented entities are described.

pith-pipeline@v0.9.1-grok · 5767 in / 1058 out tokens · 28394 ms · 2026-07-02T08:22:00.572035+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

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

  1. [1]

    Audouze, M

    C. Audouze, M. Massot, and S. V olz. Symplectic multi-time step parareal algorithms applied to molec- ular dynamics, Feb. 2009. HAL preprint, hal-00358459

    2009

  2. [2]

    G. Bal. On the convergence and the stability of the parareal algorithm to solve partial differential equations. InDomain decomposition methods in science and engineering, pages 425–432. Springer, 2005

    2005

  3. [3]

    Banjai and D

    L. Banjai and D. Peterseim. Parallel multistep methods for linear evolution problems.IMA J. Numer. Anal., 32(3):1217–1240, 2012

    2012

  4. [4]

    Bejenaru and T

    I. Bejenaru and T. Tao. Sharp well-posedness and ill-posedness results for a quadratic non-linear Schr¨odinger equation.J. Funct. Anal., 233(1):228–259, 2006

    2006

  5. [5]

    L. Berg ´e. Wave collapse in physics: principles and applications to light and plasma waves.Phys. Rep., 303(5):259–370, 1998

    1998

  6. [6]

    I. Bihari. A generalization of a lemma of Bellman and its application to uniqueness problems of differential equations.Acta Math. Hungar., 7(1):81–94, 1956

    1956

  7. [7]

    Bossuyt, S

    I. Bossuyt, S. Vandewalle, and G. Samaey. Micro-macro parareal, from ordinary differential equations to stochastic differential equations and back again.ANZIAM J., 67:Paper No. e13, 23, 2025

    2025

  8. [8]

    Brehier and X

    C.-E. Brehier and X. Wang. On parareal algorithms for semilinear parabolic stochastic PDEs.SIAM J. Numer. Anal., 58(1):254–278, 2020

    2020

  9. [9]

    Buvoli and M

    T. Buvoli and M. Minion. IMEX Runge-Kutta parareal for non-diffusive equations. In B. Ong, J. Schroder, J. Shipton, and S. Friedhoff, editors,Parallel-in-Time Integration Methods, pages 95–127, Cham, 2021. Springer International Publishing

    2021

  10. [10]

    Buvoli and M

    T. Buvoli and M. Minion. Exponential Runge-Kutta parareal for non-diffusive equations.J. Comput. Phys., 497:Paper No. 112623, 27, 2024

    2024

  11. [11]

    Chabchoub, N

    A. Chabchoub, N. P. Hoffmann, and N. Akhmediev. Rogue wave observation in a water wave tank. Phys. Rev. Lett., 106:204502, May 2011. 23

    2011

  12. [12]

    S. M. Cox and P. C. Matthews. Exponential time differencing for stiff systems.J. Comput. Phys., 176(2):430–455, 2002

    2002

  13. [13]

    Dai and Y

    X. Dai and Y . Maday. Stable parareal in time method for first- and second-order hyperbolic systems. SIAM J. Sci. Comput., 35(1):A52–A78, 2013

    2013

  14. [14]

    V . A. Dobrev, T. Kolev, N. A. Petersson, and J. B. Schroder. Two-level convergence theory for multigrid reduction in time (MGRIT).SIAM J. Sci. Comput., 39(5):S501–S527, 2017

    2017

  15. [15]

    Emmett and M

    M. Emmett and M. L. Minion. Toward an efficient parallel in time method for partial differential equations.Commun. Appl. Math. Comput. Sci., 7(1):105–132, 2012

    2012

  16. [16]

    R. D. Falgout, S. Friedhoff, T. V . Kolev, S. P. MacLachlan, and J. B. Schroder. Parallel time integration with multigrid.SIAM J. Sci. Comput., 36(6):C635–C661, 2014

    2014

  17. [17]

    Y . Feng, G. Maierhofer, and K. Schratz. Long-time error bounds of low-regularity integrators for nonlinear Schr¨odinger equations.Math. Comp., 93(348):1569–1598, 2024

    2024

  18. [18]

    P. F. Fischer, F. Hecht, and Y . Maday. A parareal in time semi-implicit approximation of the Navier- Stokes equations. InDomain decomposition methods in science and engineering, volume 40 ofLect. Notes Comput. Sci. Eng., pages 433–440. Springer, Berlin, 2005

    2005

  19. [19]

    M. J. Gander. 50 years of time parallel time integration. InMultiple shooting and time domain decomposition methods, volume 9 ofContrib. Math. Comput. Sci., pages 69–113. Springer, Cham, 2015

    2015

  20. [20]

    M. J. Gander and S. G ¨uttel. PARAEXP: a parallel integrator for linear initial-value problems.SIAM J. Sci. Comput., 35(2):C123–C142, 2013

    2013

  21. [21]

    M. J. Gander, S. G ¨uttel, and M. Petcu. A nonlinear ParaExp algorithm. InDomain decomposition methods in science and engineering XXIV, volume 125 ofLect. Notes Comput. Sci. Eng., pages 261–

  22. [22]

    Springer, Cham, 2018

    2018

  23. [23]

    M. J. Gander, F. Kwok, and J. Salomon. ParaOpt: a parareal algorithm for optimality systems.SIAM J. Sci. Comput., 42(5):A2773–A2802, 2020

    2020

  24. [24]

    M. J. Gander and S. Vandewalle. Analysis of the parareal time-parallel time-integration method.SIAM J. Sci. Comput., 29(2):556–578, 2007

    2007

  25. [25]

    M. J. Gander and S.-L. Wu. Convergence analysis of aperiodic-likewaveform relaxation method for initial-value problems via the diagonalization technique.Numer. Math., 143(2):489–527, 2019

    2019

  26. [26]

    M. J. Gander, S.-L. Wu, and T. Zhou. Time parallelization for hyperbolic and parabolic problems.Acta Numer., 34:385–489, 2025

    2025

  27. [27]

    E. P. Gross. Structure of a quantized vortex in boson systems.Il Nuovo Cimento, 20(3):454–477, 1961

    1961

  28. [28]

    Hasegawa and F

    A. Hasegawa and F. Tappert. Transmission of stationary nonlinear optical pulses in dispersive dielectric fibers. i. anomalous dispersion.Appl. Phys. Lett., 23(3):142–144, 1973

    1973

  29. [29]

    Hochbruck and A

    M. Hochbruck and A. Ostermann. Exponential integrators.Acta Numer., 19:209–286, 2010

    2010

  30. [30]

    C. E. Kenig, G. Ponce, and L. Vega. Quadratic forms for the1-D semilinear Schr ¨odinger equation. Trans. Amer. Math. Soc., 348(8):3323–3353, 1996. 24

    1996

  31. [31]

    Legoll, T

    F. Legoll, T. Leli `evre, and U. Sharma. An adaptive parareal algorithm: application to the simulation of molecular dynamics trajectories.SIAM J. Sci. Comput., 44(1):B146–B176, 2022

    2022

  32. [32]

    Li and Y

    B. Li and Y . Wu. A fully discrete low-regularity integrator for the 1D periodic cubic nonlinear Schr¨odinger equation.Numer. Math., 149(1):151–183, 2021

    2021

  33. [33]

    Li and Y

    B. Li and Y . Wu. An unfiltered low-regularity integrator for the KdV equation with solutions below H1.Found. Comput. Math., 26(3):1321–1380, 2026

    2026

  34. [34]

    G. Li, Q. Lin, S.-L. Wu, and Z. Zhou. Linear convergence of parareal algorithm for semilinear parabolic equations.preprint, arXiv:2606.03516, 2026

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  35. [35]

    Y . Li, Y . Wu, and F. Yao. Convergence of an embedded exponential-type low-regularity integrators for the KdV equation without loss of regularity.Ann. Appl. Math., 37(1):1–21, 2021

    2021

  36. [36]

    parar´eel

    J.-L. Lions, Y . Maday, and G. Turinici. 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

    2001

  37. [37]

    T. P. Mathew, M. Sarkis, and C. E. Schaerer. Analysis of block parareal preconditioners for parabolic optimal control problems.SIAM J. Sci. Comput., 32(3):1180–1200, 2010

    2010

  38. [38]

    M. L. Minion, R. Speck, M. Bolten, M. Emmett, and D. Ruprecht. Interweaving PFASST and parallel multigrid.SIAM J. Sci. Comput., 37(5):S244–S263, 2015

    2015

  39. [39]

    Nievergelt

    J. Nievergelt. Parallel methods for integrating ordinary differential equations.Comm. ACM, 7:731– 733, 1964

    1964

  40. [40]

    B. W. Ong and J. B. Schroder. Applications of time parallelization.Comput. Vis. Sci., 23(1-4):Paper No. 11, 15, 2020

    2020

  41. [41]

    Ostermann, F

    A. Ostermann, F. Rousset, and K. Schratz. Fourier integrator for periodic NLS: low regularity estimates via discrete Bourgain spaces.J. Eur. Math. Soc. (JEMS), 25(10):3913–3952, 2023

    2023

  42. [42]

    Ostermann and K

    A. Ostermann and K. Schratz. Low regularity exponential-type integrators for semilinear Schr ¨odinger equations.Found. Comput. Math., 18(3):731–755, 2018

    2018

  43. [43]

    Ostermann and F

    A. Ostermann and F. Yao. A fully discrete low-regularity integrator for the nonlinear Schr ¨odinger equation.J. Sci. Comput., 91(1):Paper No. 9, 14, 2022

    2022

  44. [44]

    L. P. Pitaevskii. V ortex lines in an imperfect Bose gas.Sov. Phys. JETP, 13(2):451–454, 1961

    1961

  45. [45]

    Rousset and K

    F. Rousset and K. Schratz. A general framework of low regularity integrators.SIAM J. Numer. Anal., 59(3):1735–1768, 2021

    2021

  46. [46]

    Ruprecht

    D. Ruprecht. Wave propagation characteristics of parareal.Comput. Vis. Sci., 19(1-2):1–17, 2018

    2018

  47. [47]

    Schratz, Y

    K. Schratz, Y . Wang, and X. Zhao. Low-regularity integrators for nonlinear dirac equations.Math. Comp., 90(327):189–214, 2021

    2021

  48. [48]

    Sheen, I

    D. Sheen, I. H. Sloan, and V . Thom´ee. A parallel method for time-discretization of parabolic problems based on contour integral representation and quadrature.Math. Comp., 69(229):177–195, 2000

    2000

  49. [49]

    Sheen, I

    D. Sheen, I. H. Sloan, and V . Thom´ee. A parallel method for time discretization of parabolic equations based on Laplace transformation and quadrature.IMA J. Numer. Anal., 23(2):269–299, 2003. 25

    2003

  50. [50]

    Steiner, D

    J. Steiner, D. Ruprecht, R. Speck, and R. Krause. Convergence of parareal for the Navier-Stokes equations depending on the Reynolds number. InNumerical mathematics and advanced applications— ENUMATH 2013, volume 103 ofLect. Notes Comput. Sci. Eng., pages 195–202. Springer, Cham, 2015

    2013

  51. [51]

    Wang and X

    Y . Wang and X. Zhao. A symmetric low-regularity integrator for nonlinear klein-gordon equation. Math. Comp., 91(337):2215–2245, 2022

    2022

  52. [52]

    S.-L. Wu, T. Zhou, and Z. Zhou. A uniform spectral analysis for a preconditioned all-at-once system from first-order and second-order evolutionary problems.SIAM J. Matrix Anal. Appl., 43(3):1331– 1353, 2022

    2022

  53. [53]

    Wu and F

    Y . Wu and F. Yao. A first-order Fourier integrator for the nonlinear Schr¨odinger equation onTwithout loss of regularity.Math. Comp., 91(335):1213–1235, 2022

    2022

  54. [54]

    Wu and X

    Y . Wu and X. Zhao. Optimal convergence of a second-order low-regularity integrator for the kdv equation.IMA J. Numer. Anal., 42(4):3499–3528, 2022. 26

    2022