On the inverse scattering transform for the KdV equation with summable initial data
Alexei Rybkin · 2026
Read On the inverse scattering transform for the KdV equation with summable initial data by Alexei Rybkin (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-14412.txt.
Introduction
Rybkin treats the Korteweg-de Vries equation on the half-line for real initial data that are integrable and square-integrable. With Hankel operators and the left reflection coefficient on Hardy space, he derives a trace-type formula for the wave. Approximation by compact potentials, uniform convergence of reflection coefficients away from zero, and continuity of the Hankel operators carry the proof. The result gives an inverse-scattering construction beyond the usual short-range setting.
Catalog record
| CREATOR | Alexei Rybkin |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38577 |
| LICENSE | http://creativecommons.org/publicdomain/zero/1.0/ |
| RIGHTS | CC0 |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.14412
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On the inverse scattering transform for the KdV equation with summable initial data Thanks: The author is supported in part by the NSF grant DMS-2307774. Alexei Rybkin Abstract . We consider the Cauchy problem for the Korteweg–de Vries equation with real initial data q ∈ L 1 ∩ L 2 q\in L^{1}\cap L^{2} supported on ( 0 , ∞ ) (0,\infty) . Using the left reflection coefficient and Hankel operators on the Hardy space H 2 ( ℂ + ) H^{2}(\mathbb{C}^{+}) , we derive a trace-type representation for the corresponding solution. The proof is based on approximation by compactly supported potentials, uniform convergence of the associated reflection coefficients away from the origin, and continuity properties of the resulting Hankel operators. This yields a rigorous inverse scattering construction for a class of summable half-line supported initial data beyond the standard short-range setting. 1991 Mathematics Subject Classification 34L25, 37K15, 47B35 1. Introduction We are concerned with the inverse scattering transform (IST) for the Cauchy problem for the Korteweg-de Vries equation \begin{cases}\partial_{t}q-6q\partial_{x}q+\partial_{x}^{3}q=0,\ \ \ x\in\mathbb{R},t\geq 0\\ q(x,0)=q(x).\end{cases} (1.1) This is of course a classical well-known problem solved in the seminal 1967 Gardner-Greene-Kruskal-Miura paper [ 10 ] who solved it for Schwarz initial data q . Conceptually, the IST is similar to the Fourier method but is based on the direct/inverse scattering 1 1 1 If q is periodic then in place of the IST one needs to use the inverse spectral transform which is also well-developed. theory for the 1D Schrödinger operator \mathbb{L}_{q}=-\partial_{x}^{2}+q(x) . More specifically, solution to ( 1.1 ) is given in terms of the scattering data S_{q} (see below) via a three step procedure. The crucial use of the inverse scattering problem imposes a severe limitation on q - it has to satisfy \int_{\mathbb{R}}\left(1+x\right)\left|q\left(x\right)\right|\mathrm{d}x<\infty\text{ (short-range).} (1.2) Any relaxation of ( 1.2 ) meets serious complications. In [ 1 ] two potentials q_{\pm} , supported on \left(\pm\infty,0\right) , and behaving like q\left(x\right)=O\left(\left|x\right|^{-2}\right) , x\rightarrow\pm\infty , respectively, are explicitly constructed , that share the same reflection coefficients. I.e. the scattering data S_{q} no longer determine q uniquely and scattering data are no longer data. This of course means that the IST method, as we know it, break down. The uniqueness in [ 1 ] can be easily restore by specifying the support of q Moreover the situation remains manageable if q is assumed to be short-range only at +\infty . Then there exists a suitable set of scattering data from the right that allows to push the IST method to essentially any q such that the spectrum of \mathbb{L}_{q} is bounded below. Such q is commonly referred as to step-like or step-type (see [ 13 ] and the literature cited therein). The decay of type ( 1.2 ) is still assumed at +\infty and this assumption is crucial for the technics of [ 13 ] to work. The main goal of our contribution is to address this problem when ( 1.2 ) is violated at +\infty . Of course, due to unidirectional nature of the KdV equation, the simple trick of considering q\left(-x,-t\right) (which clearly solves KdV) does not work. Indeed, while +\infty and -\infty do switch, but the technics of [ 13 ] dramatically fail for t<0 . The main reason is that while scattering matrix is well-defined for any L^{1} potential but continuity is lost at the zero energy. In other words, zero energy (momentum) is a spectral singularity which is notoriously hard to control. By the same token, we don’t know if it is possible at all to associate with a zero spectral singularity a norming constant that would restore uniqueness. In this paper, we develop a systematic framework in which detour this circumstance by imposing a condition that q\in L^{1} are supported on \left(0,\infty\right) . This does not remove a spectral singularity at 0 but use the techniques of Hankel operators developed in [ 15 ] , [ 13 ] , and [ 31 ] are robust enough to circumvent the issues. We note that in the context of Wigner-von Neumann initial data certain data behaving like q\left(x\right)=0(1/x) as x\rightarrow+\infty is recently considered in [ 14 ] , [ 30 ] . We believe that putting L^{1} potentials into IST setting is approached on the rigorous bases for the first time here. We discuss challenges below in detail once we set up our framework. The structure of the paper is as follows. In Section 2 we fix notation. Section 3 reviews the Hardy-space and Hankel-operator tools used throughout the paper. In Section 4 we briefly recall the relevant facts from one-dimensional scattering theory and the associated IST for compactly supported potentials, with particular emphasis on the left reflection coefficient and on the approximation by truncated potentials. Section 5 contains the proof of the main trace formula for KdV solutions corresponding to real L^{1}\cap L^{2} initial data supported on (0,\infty) . The appendix collects several auxiliary facts used in the limiting arguments. 2. Notations Our notations are quite standard: • \mathrm{1}_{S} is the characteristic function of a (measurable) set S . • L^{p}\left(\mathbb{R}\right)=L^{p} , 1\leq p\leq\infty ; \left\|\cdot\right\|_{p} is the L^{p} norm. • We write x\lesssim_{a}y if x,y\geq 0 and x\leq C\left(a\right)y with a positive C dependent on a . We drop a if C is a universal constant. • As always for a Hilbert space operator T we T\geq 0 if \left\langle Tf,f\right\rangle\geq 0 for all f (a positive operator). 3. Hardy spaces and Hankel operators To fix our notation we review some basics of Hardy spaces and Hankel operators following [ 23 ] , [ 24 ] A function f analytic in \mathbb{C}^{\pm}=\left\{z\in\mathbb{C}:\pm\operatorname{Im}z>0\right\} is in the Hardy space H_{\pm}^{p} for some 0<p\leq\infty if \|f\|_{H_{\pm}^{p}}^{p}\overset{\operatorname*{def}}{=}\sup_{y>0}\|f(\cdot\pm\mathrm{i}y)\|_{p}<\infty. We set H^{p}=H_{+}^{p}. Every f\in H_{\pm}^{p} has non-tangential boundary values f\left(x\pm\mathrm{i}0\right) for almost every (a.e.) x\in\mathbb{R} and \|f\|_{H_{\pm}^{p}}=\|f\left(\cdot\pm\mathrm{i}0\right)\|_{L^{p}}=\left\|f\right\|_{L^{p}}. (3.1) Recall that H_{\pm}^{2} are Hilbert spaces with the inner product induced from L^{2} . It is well-known that L^{2}=H^{2}\oplus H_{-}^{2}, the orthogonal (Riesz) projection \mathbb{P}_{\pm} onto H_{\pm}^{2} being given by (\mathbb{P}_{\pm}f)(x)=\pm\frac{1}{2\pi\mathrm{i}}\lim_{\varepsilon\rightarrow 0+}\int_{\mathbb{R}}\frac{f(s)\mathrm{d}s}{s-(x\pm\mathrm{i}\varepsilon)}=:\pm\frac{1}{2\pi\mathrm{i}}\int_{\mathbb{R}}\frac{f(s)\ \mathrm{d}s}{s-(x\pm\mathrm{i}0)}. (3.2) The class H^{\infty} is the algebra of functions uniformly bounded in \mathbb{C}^{+} . Define the Hankel operator on H^{2} . Let (\mathbb{J}f)(x)=f(-x) be the operator of reflection. Given \varphi\in L^{\infty} the operator \mathbb{H}(\varphi):H^{2}\rightarrow H^{2} given by the formula \mathbb{H}(\varphi)f=\mathbb{JP}_{-}\varphi f,\ \ \ f\in H^{2}, (3.3) is called the Hankel operator with symbol \varphi . Clearly \left\|\mathbb{H}(\varphi)\right\|\leq\left\|\varphi\right\|_{L^{\infty}} , \mathbb{H}(\varphi) is self-adjoint if (\mathbb{J}\varphi)(x)=\overline{\varphi\left(x\right)} (this is always our case), \mathbb{H}(\varphi)=0 if \varphi is a constant, and if \varphi\in L^{2}\cap L^{\infty} \mathbb{H}(\varphi)=\mathbb{H}(\mathbb{P}_{-}\varphi)\text{.} (3.4) We will also need the Nehari theorem saying that \left\|\mathbb{H}(\varphi)\right\|\lesssim\left\|\mathbb{P}_{-}\varphi\right\|_{BMO}, where \left\|f\right\|_{BMO}=\sup_{I\in\mathbb{R}}\frac{1}{\left|I\right|}\int_{I}\left|f-f_{I}\right|,\ \ \ f_{I}:=\frac{1}{\left|I\right|}\int_{I}f is the BMO norm (boundary mean oscillation) of f . Finally, the famous Hartman theorem says that \mathbb{H}(\varphi) is a compact operator if and only if \varphi\in H^{\infty}+C (aka the Sarason algebra), where C is the set of continuous functions on the one point compactification of \mathbb{R} . The relevance of the Hankel operator in the IST setting is obvious as the classical Marchenko operator is a Hankel operator on L^{2}\left(0,\infty\right) . In our approach, however, Hankel operators on H^{2} are more convenient. Finally we emphasize that the use of Hankel-operator methods in the study of integrable systems has recently gained momentum. In addition