On the well-posedness, ill-posedness and norm-inflation for a higher order water wave model on a periodic domain

Mahendra Panthee; Ricardo Pastran; Xavier Carvajal

REVIEW 4 major objections 4 minor 20 references

On the well-posedness, ill-posedness and norm-inflation for a higher order water wave model on a periodic domain

T0 review · 4 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read This paper claims the periodic fifth-order KdV-BBM water-wave model is locally well-posed in $H^s(\mathbb{T})$ for $s\ge 1$, globally well-posed when $\gamma=7/48$, and that below $s=1$ the solution map is discontinuous and exhibits norm…

desk verdict Solid periodic local well-posedness and ill-posedness results, but the norm-inflation proof collapses: the chosen time scale makes the supposedly dominant term vanish, so Theorem 1.4 is unproven. read the letter →

arxiv 1908.07508 v1 pith:SLKNAF5G submitted 2019-08-20 math.AP

Xavier Carvajal , Mahendra Panthee , Ricardo Pastran This is my paper · ORCID

classification math.AP MSC 35A0135Q53

keywords KdVequationBBMInitialvalueproblemLocalwell-posednessGlobalIll-posednessNorm-inflationPeriodicdomain

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

This paper studies the periodic initial-value problem for a fifth-order KdV-BBM water-wave model, a unidirectional water-wave equation with linear dispersion up to five derivatives and quadratic and cubic nonlinearities. It claims that the problem is locally well-posed in the Sobolev space $H^s(\mathbb{T})$ for every $s\ge 1$, with the solution depending analytically on the data. For the special coefficient choice $\gamma=7/48$, the paper claims the solution extends globally in time for all $s\ge 1$. It further claims the threshold $s=1$ is sharp: for $s<1$ the data-to-solution map is discontinuous at the origin, and there are smooth data arbitrarily small in $H^s$ whose solutions develop arbitrarily large $H^s$ norm in arbitrarily short time. If true, this pins down the exact regularity needed to run a contraction-mapping theory on the torus.

What carries the argument

The argument is carried by frequency-localized multilinear estimates for the Fourier multiplier operators derived from the linear phase. The central estimate, Proposition 2.1, says that $\|\omega(\partial_x)(uv)\|_{H^s} \lesssim \|u\|_{H^s}\|v\|_{H^s}$ for $s \ge 0$, where $\omega(\partial_x)$ has symbol $|k|/(1+k^2)$; this is exactly the regularizing factor that lets the quadratic and cubic terms be controlled in $H^s$. The ill-posedness and norm-inflation parts use the second Picard iterate $I_2$ on initial data concentrated near frequency $\pm N$, and the norm-inflation construction uses trigonometric data $\sin(k_1 x) + \sin(k_2 x)$ with $k_2 = k_1 + 1$ after a change of variables that removes the fifth-order linear term.

What would settle it

Directly compute, for the trigonometric data in Theorem 1.4, the $H^s$-norm of the second Picard iterate at time $T_j = k_1^{-\theta \sigma}$; equation (5.35) gives $\|\eta_1(T_j)\|_{H^s} \sim k_1^{(2-\theta)\sigma}$, which tends to zero when $\theta>3$, so a direct evaluation would settle whether the claimed growth occurs. For the global result, verifying whether the missing bound (3.21) holds on the torus would settle the splitting argument.

Watch

Extended reading notes

Core claim

On the periodic domain $\mathbb{T}$, the initial-value problem $\eta_t + \eta_x - \gamma_1 \eta_{xxt} + \gamma_2 \eta_{xxx} + \delta_1 \eta_{xxxxt} + \delta_2 \eta_{xxxxx} + \tfrac{3}{2} \eta \eta_x + \gamma (\eta^2)_{xxx} - \tfrac{7}{48}(\eta_x^2)_x - \tfrac{1}{8}(\eta^3)_x = 0$, $\eta(x,0)=\eta_0(x)$, is locally well-posed for $\eta_0 \in H^s(\mathbb{T})$, $s\ge 1$, by a contraction mapping built on multilinear estimates for the Fourier multipliers coming from the linear symbol. When $\gamma = \tfrac{7}{48}$, the energy $E(\eta) = \frac{1}{2}\int (\eta^2 + \gamma_1 \eta_x^2 + \delta_1 \eta_{xx}^2 )\, dx$ is conserved, and a low-frequency/high-frequency splitting extends the local solution to arbitrary time intervals for all $s\ge 1$. The regularity threshold is sharp: for $s<1$ the flow map $\eta_0 \mapsto \eta(t)$ is discontinuous at the origin from $H^s(\mathbb{T})$ to periodic distributions, and a sequence of smooth data converging to zero in $H^s(\mathbb{T})$ produces solutions whose $H^s(\mathbb{T})$-norm exceeds any prescribed bound at times tending to zero.

Load-bearing premise

The proofs rely on the assumption that a single short time scale $T_j = k_1^{-\theta \sigma}$ (with $\theta>3$) can make the second Picard iterate dominate in $H^s$ while the remainder stays bounded, and, for the global result, that the real-line iteration estimates from the earlier work transfer to the periodic setting.

Editorial extensions

If this is right

If Theorem 1.1 is correct, the periodic fifth-order KdV-BBM equation is locally well-posed exactly in $H^s(\mathbb{T})$ for $s \ge 1$, with real-analytic dependence of solutions on data.
If Theorem 1.2 is correct, the special coefficient value $\gamma=7/48$ gives global solutions for all $s \ge 1$ on the torus, matching the real-line result.
If Theorem 1.3 is correct, any well-posedness theory for $s<1$ must abandon continuity of the data-to-solution map at the origin.
If Theorem 1.4 is correct, smooth data with arbitrarily small $H^s$ norm, $s<1$, can develop arbitrarily large $H^s$ norm in arbitrarily short time.

Reading between the lines

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

The same frequency-localized failure of the second iterate is likely to produce ill-posedness for other fifth-order dispersive equations with BBM-type regularizing denominators, since only the shape of the multiplier symbol matters.
Because the $H^s$ norm is invariant under the translation change of variables used in Section 5, the norm-inflation statement for the transformed equation transfers verbatim to the original equation, which the paper itself exploits.
A testable extension would be to replace the two-frequency initial data $\sin(k_1x)+\sin(k_2x)$ by a single frequency $\sin(k_1x)$; if the resonance terms still dominate, the construction would simplify.
If the global splitting argument is sound, the growth bound $\|\eta(t)-S(t)\eta_0\|_{H^2} \lesssim (1+T)^{2-s}$ implies that the high-frequency part of the solution remains essentially linear in $H^2$, a property that could be checked numerically for the periodic model.

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.

Desk Editor's Note

read the letter

Here's my read on the Carvajal–Panthee–Pastran paper. The periodic local well-posedness (s≥1) and the ill-posedness below s=1 look like solid work, and those parts deserve credit. But the norm-inflation theorem, which is the headline contribution, is not proven: the chosen time scale kills the very term that is supposed to inflate the norm, and no alternative scaling repairs the remainder bootstrap. The global well-posedness also has a gap for s>2 and an unstated parameter restriction.

What's actually new: the periodic setting for fifth-order KdV-BBM, with discrete multilinear estimates and a resonance analysis for ill-posedness. The estimates in §2 are clean, and the ill-posedness construction based on the second Picard iterate is convincing—it transfers the real-line idea to the circle with the right frequency blocks. The norm-inflation technique is also new for this model, but that's where it falls apart.

The specific problem: in §5, (5.35) gives ||η1(t)|| ~ k1^{2σ} t. The authors then choose T_j = k1^{-θσ} with θ>3. Substituting gives ||η1(T_j)|| ~ k1^{(2-θ)σ} → 0. The linear term tends to 0 too, and the remainder is bounded by 2C k1^{(3-θ)σ} → 0. So at the claimed times the solution norm is actually small. If you try θ<2 to make η1 grow, the remainder bound (5.41) fails because α ~ k1^{(3-θ)σ} blows up. There is no θ interval that makes both work. This isn't a minor typo; it's the central estimate of Theorem 1.4.

Other issues: Theorem 3.1 claims global well-posedness for s≥2 using only the H^2 conserved quantity. That doesn't control H^s for s>2, and no higher-energy estimate is supplied. The splitting argument for 1≤s<2 is sketched with a reference to the real-line case, but the key bound (3.21) is asserted rather than proved; the proof of Lemma 3.1 has a suspicious line where the δ≥s range is used and then δ=1 is taken. Also, the norm-inflation section introduces δ3 = 1 - δ2/δ1 > 0, but the paper's parameter constraints don't guarantee this; so Theorem 1.4 only could hold for a restricted parameter set, unstated in the theorem.

Bottom line: the local and ill-posedness results look salvageable and probably correct. The norm-inflation theorem as stated is unsupported, and the global theorem has gaps. The paper deserves a serious referee because the local/ill-posedness part is genuine, but it needs substantial revision before it's publishable. I'd send it back with clear requests: fix the norm-inflation time scale or recast the claim, prove the s>2 global step, and state the δ3>0 assumption.

Referee Report

4 major / 4 minor

Summary. The paper studies the periodic initial-value problem for a fifth-order KdV-BBM water-wave model (1.1). It claims local well-posedness in H^s(T) for s >= 1 via multilinear estimates and a contraction argument (Theorem 1.1); global well-posedness for gamma = 7/48 and s >= 1 using a conserved H^2 energy and a low-high splitting argument (Theorems 1.2, 3.1, 3.3); ill-posedness for s < 1 by discontinuity of the flow map (Theorem 1.3); and norm inflation in H^s(T) for s < 1 (Theorem 1.4). The abstract frames these results as a sharp regularity threshold for the periodic model, in parallel with the real-line results in [4] and [12].

Significance. If all four theorems were established, the paper would provide a complete periodic analogue of the sharp well-posedness theory for this higher-order water-wave model, and the norm-inflation result would substantially strengthen the ill-posedness statement. The local well-posedness proof rests on clean multilinear estimates, and the ill-posedness argument is a concrete Picard-iterate discontinuity; these parts are convincing and are a real contribution. However, the global well-posedness argument for s >= 2 and the norm-inflation proof contain load-bearing gaps, and the norm-inflation argument as written contains an internal contradiction. Because Theorems 1.2 and 1.4 are central claims of the paper, the manuscript is not acceptable in its present form.

major comments (4)

[Section 5, Eqs. (5.35), (5.38)-(5.41)] The norm-inflation proof has a fatal time-scale inconsistency. The choice T_j = k1^{-theta sigma} with theta > 3 is substituted only into the remainder bound (5.41), but not into the dominant term (5.35). Substituting t = T_j into (5.35) gives ||eta1(.,T_j)||_{H^s} ~ k1^{(2-theta)sigma}, which tends to 0 as k1 -> infinity because (2 - theta)sigma < 0. The linear term has size O(k1^{sigma-1+s}) and also tends to 0, and (5.41) gives ||zeta|| = O(k1^{(3-theta)sigma}) -> 0. Thus at the asserted times T_j the solution norm cannot grow, let alone exceed j. The sentence after (5.35) that the norm 'can be made as big as we wish by choosing k1 large' is only valid for fixed t, not for t = T_j. Moreover, the two requirements are incompatible: choosing theta < 2 would make the eta1-term grow at T_j, but then the alpha-term in (5.38) becomes k1^{(3-theta)sigma} -> infinity and the remainder bootstrap (5.39)-(5.42) fails. Hence Theorem 1.4 is not proven as written.
[Section 3.2, Theorem 3.3 and Lemma 3.1] The extension from local to global well-posedness for 1 <= s < 2 is not completed. Lemma 3.1 estimates one step of the splitting, but the induction argument for continuing the process is delegated to the real-line paper with the phrase 'as in the real line case (see [12])'. In particular, the new low-frequency object u1 = u(t0) + h(t0) in (3.27) is not shown to satisfy the frequency localization or the growth bounds required for the next splitting step. The asserted bound (3.21) in Theorem 3.3 is stated but never derived; the proof ends with (3.32) and the sentence 'this completes the proof'. Since Theorem 1.2 depends on this iteration, the global claim for the range 1 <= s < 2 is unsupported.
[Section 3.1, Theorem 3.1] The proof of global well-posedness for s > 2 is only described as 'a standard argument' using the conserved H^2 energy. For s = 2 the conserved energy controls the H^2 norm, but for s > 2 the local existence time in (2.26) depends on ||eta0||_{H^s}, and no H^s a priori estimate or higher-energy argument is provided. Since Theorem 1.2 claims global well-posedness for all s >= 1, this missing estimate is a load-bearing gap, not a routine detail.
[Section 5, Eq. (5.1)] The proof of Theorem 1.4 assumes delta3 = 1 - delta2/delta1 > 0 after the change of variables, but this condition is not stated in the introduction, in the parameter conditions after (1.3), or in Theorem 1.4. If it is an additional hypothesis on the coefficients, it must be made explicit; if it is always satisfied by the abcd-system parameters, that needs proof.

minor comments (4)

[Section 4, after Eq. (4.6)] The sentence 'inserting (4.6) in (5.6)' should refer to (4.5), since (5.6) is defined later and in a different section.
[Throughout] There are several typographical errors, including 'solucion' in Theorem 1.1, 'Similarty' in the paragraph after (2.3), and 'inducted by' in the proof of Theorem 1.3; these should be corrected.
[Theorem 1.4 and Section 5] Theorem 1.4 is stated for the homogeneous space dot H^s(T), but the proof estimates the H^s(T) norm. For the mean-zero initial data constructed in (5.11)-(5.12) the two norms coincide, but this equivalence should be mentioned explicitly.
[Section 5, Eq. (5.36)] The remainder estimate (5.36) is written with powers of ||eta1|| and ||S(t)eta0|| that are not derived from (2.15)-(2.20); for example, the cubic and mixed terms in the displayed inequality appear to be overcounted. The subsequent bootstrap may still be valid, but the displayed estimate needs justification or correction.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: the periodic well-posedness and norm-inflation proofs are self-contained, with only a non-load-bearing method citation to the authors' own real-line paper [12].

full rationale

The derivation chain does not reduce its conclusions to its inputs. The local result (Theorem 1.1) is proved from new multilinear estimates (Propositions 2.1-2.4) and a contraction argument; the global result (Theorem 1.2) is based on the conserved quantity (3.3), the splitting estimates in Lemma 3.1, and an explicitly worked induction step for k=1 in (3.28)-(3.30). The repeated phrase 'as in the real line case (see [12])' imports a proof scheme from the authors' earlier paper, not the target periodic theorem; the cited result is therefore not a load-bearing circular premise. The ill-posedness and norm-inflation results are constructed from explicit Fourier data (4.1), (5.11)-(5.12) and explicit Picard iterates (4.4), (5.5), with no fitted parameter renamed as a prediction. The apparent scaling problem in the norm-inflation proof (T_j = k1^{-theta sigma} with theta>3 makes the computed eta1 term in (5.35) tend to zero) is a correctness defect in the proof as written, not an equivalence-by-definition between input and claimed conclusion. Hence no circular step meeting the 'quote and reduce' standard is present.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claims rest on standard harmonic analysis, on parameter restrictions inherited from the model, and on the proof-device cutoff N. The norm-inflation part introduces exponents sigma and theta whose compatibility with the time scale is not established. No new physical entities are introduced.

free parameters (2)

frequency cutoff N = N = T (Theorem 3.3)
The low-high frequency splitting rate and all subsequent bounds in Lemma 3.1 and (3.30) depend on this choice. It is chosen by hand to control the accumulated error over T N^{2(2-s)} steps, rather than fitted to data.
norm-inflation exponents sigma and theta = 0 < sigma < 1-s; theta > 3 (Section 5)
The construction of the initial data and the time scale T_j = k1^{-theta sigma} depend on these choices. The incompatibility of sigma and theta is the source of the norm-inflation gap.

assumptions (4)

standard math H^s(T) is a Banach algebra for s > 1/2.
Used in Propositions 2.3 and 2.4 to bound products of H^s functions in the multilinear estimates.
domain assumption The Fourier multiplier symbols tau(k), psi(k) decay like |k|^{-1} and |k|^{-3} respectively because delta1, gamma1 > 0.
The positivity of delta1 and gamma1 makes phi(k)=1+gamma1 k^2 + delta1 k^4 strictly positive and gives the decay used throughout the local theory.
domain assumption The Hamiltonian parameter restriction gamma = 7/48 is imposed for the conserved energy (3.3).
Global well-posedness in Theorems 1.2, 3.1 and 3.3 uses the conservation law (3.3), which holds only at gamma = 7/48.
ad hoc to paper The change of variables in Section 5 assumes delta3 = 1 - delta2/delta1 > 0.
This assumption contradicts the parameter relations in Section 1: with the stated definitions, gamma1 = 1/12 and delta2 - delta1 = 7/180 > 0, so delta2/delta1 > 1 and delta3 is negative. The norm-inflation proof as written relies on this asserted positivity.

how reviews work

0 comments

Cite this review

Pith. "Pith review of On the well-posedness, ill-posedness and norm-inflation for a higher order water wave model on a periodic domain." pith.science (2026). https://pith.science/paper/SLKNAF5G

@misc{pith2026190807508,
  author       = {Pith},
  title        = {Pith review of: On the well-posedness, ill-posedness and norm-inflation for a higher order water wave model on a periodic domain},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SLKNAF5G}},
  note         = {Machine review of arXiv:1908.07508}
}

abstract

In this work we are interested in the well-posedness issues for the initial value problem associated with a higher order water wave model posed on a pe\-rio\-dic domain $\mathbb{T}$. We derive some multilinear estimates and use them in the contraction mapping argument to prove local well-posedness for initial data in the periodic Sobolev space $H^s(\mathbb{T})$, $s\geq 1$. With some restriction on the parameters appeared in the model, we use the conserved quantity to obtain global well-posedness for given data with Sobolev regularity $s\geq 2$. Also, we use splitting argument to improve the global well-posedness result in $H^s(\mathbb{T})$ for $1\leq s< 2$. Well-posedness result obtained in this work is sharp in the sense that the flow-map that takes initial data to the solution cannot to be continuous for given data in $H^s(\mathbb{T})$, $s< 1$. Finally, we prove a norm-inflation result by showing that the solution corresponding to a smooth initial data may have arbitrarily large $H^s(\mathbb{T})$ norm, with $s<1$, for arbitrarily short time.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 20 canonical work pages

[4]

J. L. Bona, X. Carvajal, M. Panthee, M. Scialom; Higher-Order Hamiltonian Model for Unidirectional Water Waves , J. Nonlinear Sci. 28 (2018) 543–577

work page 2018
[12]

Carvajal, M

X. Carvajal, M. Panthee; On sharp global well-posedness and Ill-posedness for a ﬁfth- order KdV-BBM type equation , Jr. Math. Anal. Appl. 479 (2019) 688–702

work page 2019
[1]

Bejenaru, T

I. Bejenaru, T. Tao, Sharp well-posedness and ill-posedness results for a quadr atic non-linear Schr¨ odinger equation, Jr. Functional Analysis, 233 (2006) 228–259

work page 2006
[2]

T. B. Benjamin, J. L. Bona and J. J. Mahony; Model equations for long waves in nonlinear dispersive media, Philos. Trans. Royal Soc. London, Series A 272 (1972), 47–78

work page 1972
[3]

J. L. Bona, H. Chen, C. Guillop´ e, Further Theory for a Higher-order Water Wave Model , to appear in Jr. Pure and Applied Functional Analysis

work page
[5]

J. L. Bona, M. Dai, Norm-inﬂation results for the BBM equation , Jr. math. Anal. Appl. 446 (2017) 879–885

work page 2017
[6]

J. L. Bona, N. Tzvetkov, Sharp well-posedness results for the BBM equation , Discrete and Continuous Dynamical Systems 23 4 (2009), 1241–1252

work page 2009
[7]

J. L. Bona, M. Chen and J. C. Saut; Boussinesq equations and other systems for small- amplitude long waves in nonlinear dispersive media I. Deriva tion and linear theory , J. Non- linear Sci. 12 (2002) 283–318

work page 2002

Show all 20 references

[8]

J. L. Bona, M. Chen and J. C. Saut; Boussinesq equations and other systems for small- amplitude long waves in nonlinear dispersive media II. The nonl inear theory, Nonlinearity 17 (2004) 925–952

2004
[9]

J. L. Bona and R. Smith; The initial-value problem for the Korteweg-de Vries equatio n, Philos. Trans. Royal Soc. London, Series A, 278 (1975) 555–601

1975
[10]

Bourgain; Periodic Korteweg de Vries equation with measures as initia l data , Sel

J. Bourgain; Periodic Korteweg de Vries equation with measures as initia l data , Sel. math., New ser. 3 (1997) 115–159

1997
[11]

Bourgain; Reﬁnements of Strichartz inequality and applications to 2D -NLS with critical nonlinearity, IMRN 5 (1998) 253–283

J. Bourgain; Reﬁnements of Strichartz inequality and applications to 2D -NLS with critical nonlinearity, IMRN 5 (1998) 253–283

1998
[13]

Chen; Well-posedness for a higher-order, nonlinear, dispersive equation on a quarter plane, Discrete Cont

H. Chen; Well-posedness for a higher-order, nonlinear, dispersive equation on a quarter plane, Discrete Cont. Dynamical Systems, Ser. A, 38 (2018) 397–429

2018
[14]

Fonseca, F

G. Fonseca, F. Linares, G. Ponce; Global well-posedness for the modiﬁed Korteweg-de Vries equation, Communications in Partial Diﬀerential Equations 24 (1999) 683–70

1999
[15]

Fonseca, F

G. Fonseca, F. Linares, G. Ponce; Global existence for the critical generalized KdV equation , Proc. AMS 131 (2002) 1847–1855

2002
[16]

Molinet, Sharp ill-posedness result for the periodic Benjamin-Ono e quation J

L. Molinet, Sharp ill-posedness result for the periodic Benjamin-Ono e quation J. Funct. Anal., 257 (2009) 3488–3516

2009
[17]

Molinet, J

L. Molinet, J. C. Saut, N. Tzvetkov, Ill-posedness issues for the Benjamin-Ono and related equations, SIAM J. Math. Anal. 33 (2001) 982–988. 26 X. CAR V AJAL, M. PANTHEE, AND R. PASTR ´AN

2001
[18]

Molinet, J

L. Molinet, J. C. Saut, N. Tzvetkov, Well-posedness and ill-posedness results for the Kadomtsev-Petviashvili-I equation , Duke Math. J., 115 (2002), 353–384

2002
[19]

Molinet, S

L. Molinet, S. Vento, Sharp ill-posedness and well-posedness results for the KdV -Burgers equation: the real line case , Annali Scuola Normale Superiore di Pisa 10 (2011) 531–560

2011
[20]

Molinet, S

L. Molinet, S. Vento, Sharp ill-posedness and well-posedness results for the KdV -Burgers equation: the periodic case , Transactions of the AMS 365 (2013) 123–141. Instituto de Matem ´atica, UFRJ, 21941-909, Rio de Janeiro, RJ, Brazil E-mail address : carvajal@im.ufrj.br Depar...

2013

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [4]

J. L. Bona, X. Carvajal, M. Panthee, M. Scialom; Higher-Order Hamiltonian Model for Unidirectional Water Waves , J. Nonlinear Sci. 28 (2018) 543–577

work page 2018

[2] [12]

Carvajal, M

X. Carvajal, M. Panthee; On sharp global well-posedness and Ill-posedness for a ﬁfth- order KdV-BBM type equation , Jr. Math. Anal. Appl. 479 (2019) 688–702

work page 2019

[3] [1]

Bejenaru, T

I. Bejenaru, T. Tao, Sharp well-posedness and ill-posedness results for a quadr atic non-linear Schr¨ odinger equation, Jr. Functional Analysis, 233 (2006) 228–259

work page 2006

[4] [2]

T. B. Benjamin, J. L. Bona and J. J. Mahony; Model equations for long waves in nonlinear dispersive media, Philos. Trans. Royal Soc. London, Series A 272 (1972), 47–78

work page 1972

[5] [3]

J. L. Bona, H. Chen, C. Guillop´ e, Further Theory for a Higher-order Water Wave Model , to appear in Jr. Pure and Applied Functional Analysis

work page

[6] [5]

J. L. Bona, M. Dai, Norm-inﬂation results for the BBM equation , Jr. math. Anal. Appl. 446 (2017) 879–885

work page 2017

[7] [6]

J. L. Bona, N. Tzvetkov, Sharp well-posedness results for the BBM equation , Discrete and Continuous Dynamical Systems 23 4 (2009), 1241–1252

work page 2009

[8] [7]

J. L. Bona, M. Chen and J. C. Saut; Boussinesq equations and other systems for small- amplitude long waves in nonlinear dispersive media I. Deriva tion and linear theory , J. Non- linear Sci. 12 (2002) 283–318

work page 2002