New Lower Bounds of Spatial Analyticity Radius for the Kawahara Equation
Abstract
In this paper, an algebraic decay rate for the radius of spatial analyticity of solutions to the Kawahara equation is investigated. With given analytic initial data having a fixed radius of analyticity σ0, we derive an algebraic decay rate σ(t) ~ |t|−1/2 for the uniform radius of spatial analyticity of solutions to the Kawahara equation. This improves a recent result due to Ahn et al.’s study, where they demonstrated a decay rate of order |t|−1. Our strategy mainly relies on an approximate conservation law in a modified Gevrey space and bilinear estimate in Bourgain space.
1. Introduction
The low regularity well-posedness of the IVP (1.1) has been investigated by Zhang et al. [3]. Following the ideas of the authors in [4, 5], first, the authors showed that the local well-posedness is established for the initial data with s ≥ −7/4. Then, using these results and conservation laws, they also proved that the IVP is globally well-posed for the initial data with s = 0. Recently, Zhang et al. [6] investigated the IVP of the higher order nonlinear dispersive equation for the initial data in the Sobolev space . For further survey, we refer the readers to [7, 8] and references therein for additional information about the well-posedness of (1) in the Sobolev space.
According to the Paley–Wiener theorem [12], any function f ∈ Gσ,s with σ > 0 can be extended to an analytic function in the strip . This is the reason why one can take initial data in the Gevrey space Gσ,s.
In the paper [35], Zhang et al. investigated the Cauchy problem for the higher order nonlinear dispersive equation with initial data in Gevrey space Gσ,s. First, the authors used Tao’s [k, Z]− multiplier method to establish the basic estimate on dyadic blocks. Also, using the Fourier restriction norm method, they established the bilinear estimate and approximate conservation law. Then, using the contraction mapping principle, iteration technique, and the bilinear estimate, they proved the local well-posedness for the initial data u0 ∈ Gσ,s with s ≥ −11/4. Finally, based on the local well-posedness and approximate conservation law, the authors obtained that the analytic radius dores not decay faster than t−4/11 as time t goes to infinity.
In this paper, we follow the idea from the authors in [35] and use an approximate conservation law in the modified Gevrey space to improve the required lower bound.
As a consequence of the embedding (8) for σ′ = 0, and the existing global well-posedness theory in Hs for s ≥ −7/4 (see [6]), we deduce that the IVP (1) has a unique, smooth solution for all time, given initial data for any σ0 > 0 and .
We now state our main result on the algebraic decay rate.
Theorem 1. Let for some σ0 > 0 and for . Then, the global C∞ solution u of the IVP (1) satisfies
The organization of this paper is as follows: In Section 2, we introduce some function spaces. We also present bilinear estimates and energy estimates in the restriction of the Bourgain space. In Section 3, we prove the local well-posedness result using energy inequalities and contraction mapping principles. In Section 4, we prove an approximate conservation law by making use of the bilinear estimate in the Bourgain space. Finally, in Section 5, we prove our main result Theorem 1 by iterating the local result on the approximate conservation law.
1.1. Notations
We use C to denote various space- and time-independent constants whose exact values may vary from line to line. For two positive quantities a and b, we use a≲b to mean a ≤ Cb, while a ~ b means a≲b and b≲a. Moreover, we use m± to mean m ± ϵ for a sufficiently small constant ϵ > 0 and any real number m.
2. Preliminary Results
In this section, we introduced the function spaces, bilinear estimates, and energy estimates necessary for our proofs.
2.1. Function Spaces
2.2. Bilinear Estimates and Energy Estimates
This section is dedicated to present bilinear estimates and energy estimates, which will use in proving the local existence result and approximate conservation law.
Lemma 1. Let δ > 0 and 1/2 < b ≤ b′ < 1. Let |D| be the Fourier multiplier with symbol |ξ|. Then, we have
Proof 1. Jia and Huo [[38], Theorem 3.5] proved that
An application of Plancherel’s theorem yields
Since all estimates of Xs,b can apply to its restriction, the desired estimate follows from (26) and (25).
Lemma 2. Let s > −7/4, δ > 0 and 1/2 < b ≤ b′ < 1. Then,
Proof 2. Zhang and Huang ([5], Lemma 5.1) (see also [[37], Lemma 3.2] and [[9], Lemma 3.1]) proved that
As estimates of Xs,b can hold for its restrictions, (29) in combination with (28) yields the estimate what we claimed.
3. Local Well-Posedness Result
In this section, we prove existence of the local result in Hσ,s for σ > 0 and s > −7/4.
Theorem 2. Let s > −7/4 and σ > 0. Then, for any given Ψ ∈ Hσ,s, there exists a time δ > 0 depending on and a unique solution u of (1) on the time interval (−δ, δ) such that
In proving this theorem, the objective is to perform energy estimates, Sobolev embedding, and contraction mapping principle.
Indeed, we claim the following nonlinear estimate.
Lemma 3. Let δ > 0, σ ≥ 0, s > −7/4 and 1/2 < b ≤ b′ < 1. Then, for any , we have
Proof 3. From the left-hand side of (35), we have the following estimate (see [39], Lemma 2.11], [38], Lemma 2.15])
We have |ξ| ≤ |ξ1| + |ξ2| by the triangle inequality. As a consequence of it, we then have
Setting Uσ≔cosh(σ|D|)u. Then, by using Plancherel’s theorem and (37), the right-hand side of (36) is estimated as
Denoting , and then by Plancherel’s theorem, Lemma 2, and Sobolev embedding, we have
It follows from (34) and (35) that the energy inequality for the IVP (1) is
Now, by combining (40) with the contraction mapping principle in the space , we obtain a unique solution for (1), with existence time
The other key ingredient that we will use to derive the lower bound estimate for solutions to (1) is an approximate conservation law.
4. Approximate Conservation Law
In taking the limit σ⟶0, we recover the conservation for all t ∈ [0, δ].
The integral (49) satisfies the following a priori estimate.
Lemma 4. Let δ > 0. Then, for all and 0 ≤ s ≤ δ, we have
Finally, inserting (54) into (50) and then using the result in (48) gives the desired estimate (46).
4.1. Proof of (50)
In addition to estimate (58), we need to recall the following estimate:
Lemma 5 ([35], Lemma 2.3]). Assume that and δ > 0. Then, for any time interval I ⊂ [−δ, δ], we have the estimate
5. Proof of Theorem 1
We closely follow the argument in [35] to derive the claimed decay rate on the lower bound for the radius of analyticity.
Now, we first consider the case s = 0 by taking the advantage of the almost conservation law in . By the embedding property (8), the general case, , will essentially reduce to s = 0 which is shown in the next subsection.
5.1. Case: s = 0
If we do this, becomes finite for all t ∈ [0, T], which in turn imply that u ∈ C([0, T]; Hσ,0), and thus the proof of (34) is completed for s = 0.
Now, it remains to prove (66). We choose so that T ∈ [nδ, (n + 1)δ].
This implies that (68) holds true for k = 1.
This completes the proof of our goal Theorem 1 for s = 0.
5.2. General Case:
Conflicts of Interest
The author declares no conflicts of interest.
Funding
The author received no financial support for this manuscript.
Endnotes
1Since we may assume that Uσ(x, t) and all its spatial derivatives vanish as |x|⟶∞, integration by parts is justified [27, 35].
Open Research
Data Availability Statement
Data sharing is not applicable to this article as no new data were created or analyzed in this study.