Weakly localized states of one dimensional Schrödinger equations have localized energy

Avy Soffer, Gavin Stewart · 2025

Read Weakly localized states of one dimensional Schrödinger equations have localized energy by Avy Soffer, Gavin Stewart (2025) 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-2510-16283.txt.

Introduction

Gavin Stewart and Avy Soffer wrote a 2025 maths paper on how waves spread in one dimension over time. They study the Schrödinger equation, the core rule for quantum waves. The setting uses a pull that fades fast far from the centre. A weakly bound part is a wave part that stays near the centre but holds on loosely.

The pair prove each wave splits into a free wave plus a weakly bound part. They then show the bound part has its slope packed near the centre at all times. The result extends past work to the harder one-line case. It helps show which wave states keep their energy close to home.

Catalog record

CREATORAvy Soffer, Gavin Stewart
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE80 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-16283

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Weakly localized states of one dimensional Schrödinger equations have localized energy Thanks: 2020 Mathematics Subject Classification. 35Q55 Thanks: A.Soffer is supported in part by NSF-DMS Grant number 2205931 Avy Soffer, Gavin Stewart Abstract . We study the asymptotics of the Schrödinger equation with time-dependent potential in dimension one. Assuming that the potential decays sufficiently rapidly as | x | → ∞ |x|\to\infty , we prove that the solution can be written as the sum of a free wave e − i ​ t ​ Δ ​ u + e^{-it\Delta}u_{+} and a weakly bound component u wb ​ ( t ) u_{\textup{wb}}(t) . Moreover, we show that the weakly bound part decomposes as u wb ​ ( t ) = u loc ​ ( t ) + o H ˙ 1 ​ ( 1 ) u_{\textup{wb}}(t)=u_{\textup{loc}}(t)+o_{\dot{H}^{1}}(1) , where ∂ x u loc ​ ( t ) \partial_{x}u_{\textup{loc}}(t) is localized near the origin uniformly in time. Since decay conditions on the potential do not preclude resonances unless d ≥ 5 d\geq 5 , our results can be seen as a natural extension of [ 34 , 31 ] to the lower-dimensional case. 1. Introduction We will study the equation (1) i\partial_{t}u-\Delta u+V(x,t)u=0 where the potential V is assumed to satisfy the conditions (2) \lVert\langle x\rangle^{\sigma}V\rVert_{L^{\infty}_{t,x}}<\infty and (3) \lVert\langle x\rangle^{\sigma+1}\partial_{x}V\rVert_{L^{\infty}_{t,x}}<\infty for some \sigma>2 . Based on the localization of the potential, it would be natural to expect that solutions u to ( 1 ) decompose asymptotically into a localized bound state component u_{\textup{loc}} together with linearly scattering radiation. Currently, such a precise characterization of solutions is only known in dimensions d\geq 5 [ 34 , 35 , 31 ] . In lower dimensions, the best we can say is that u decomposes asymptotically into linearly scattering radiation and a slowly spreading weakly bound state u_{\textup{wb}} : u(x,t)=u_{\textup{wb}}(x,t)+e^{-it\Delta}u_{+}+o_{H^{1}}(1) where u_{\textup{wb}}(t) is becomes orthogonal to any solution to the free Schrödinger equation as t\to\infty and is localized to the region |x|<t^{1/2+} : \lim_{t\to\infty}\lVert u_{\textup{wb}}\rVert_{L^{2}(|x|\geq t^{1/2+})}=0 see [ 30 , 28 , 18 ] . The significance of the threshold at dimension 5 was explained in [ 34 ] in terms of the applicability of the double Duhamel trick, which can also be used to prove the nonexistence of resonances in dimension d\geq 5 for (time-independent) potentials satisfying ( 2 ). Since resonances lie in \dot{H}^{1} but not in L^{2} , it is natural to expect that we can prove localization (and not just slow spreading) for \partial_{x}u_{\textup{wb}} . Our main result in 1 shows that this is indeed the case: The weakly bound component u_{\textup{wb}} decomposes as u_{\textup{wb}}(x,t)=u_{\textup{loc}}(x,t)+o_{\dot{H}^{1}}(1) , where \langle x\rangle^{\theta}\partial_{x}u_{\textup{loc}}\in L^{\infty}_{t}L^{2}_{x} for any \theta\in(0,1) . A similar result was proved by Tao in [ 33 ] for the radial cubic NLS in dimension 3 . 1.1. Previous work In the case where V is time-independent, the asymptotics for ( 1 ) are by now quite well understood, even if we significantly relax the decay assumptions eqs. 2 and 3 : See [ 8 , 36 , 23 ] and the references therein. In this case, the spectral theorem allows us to decompose solutions into two parts: A discrete part, which can be written as a time quasiperiodic sum of eigenfunctions of -\Delta+V ; and a continuous part, which under the assumption ( 2 ) with \sigma>1 approaches (quantum) free flow as t\to\infty . In particular, the long-time asymptotics are given by (4) u(x,t)=\sum e^{-itE_{n}}\psi_{n}(x)+e^{-it\Delta}u_{+}+o_{{H^{1}}}(1) A similar decomposition also holds for coupled systems of Schrödinger equations describing the interaction of multiple particles. There, we must consider multiple scattering channels (corresponding to different particles becoming bound) with the conclusion that solutions asymptotically arrange themselves into bound states and free particles which become widely separated as t\to\infty [ 26 , 9 , 15 , 8 ] . Much less is understood in the case where V is allowed to vary in time. If the time dependence is small, it is often possible to recover the same results as in the time-independent case. For example, for a small, time-dependent potential in dimension 3 , it is possible to prove global Strichartz estimates, which then imply scattering [ 25 ] . There are also a number of results on nonlinear Schrödinger equations with linear potentials [ 12 , 13 , 21 , 4 , 32 , 10 , 1 , 14 , 27 , 5 , 7 , 20 , 2 , 22 , 16 , 17 ] , including notable results on the asymptotic stability of solitons [ 2 , 6 , 3 ] ; the interested reader can find a review in [ 11 ] . When V has a large time-dependent part, it has not yet been possible to prove results as precise as ( 4 ). Currently, the most complete results are those of [ 35 , 31 ] , where solutions to radial nonlinear Schrödinger-type equations in dimensions d\geq 5 were shown to decompose as (5) u(x,t)=u_{\textup{loc}}(x,t)+e^{-it\Delta}u_{+}+o_{H^{1}}(1) for some function u_{\textup{loc}} with \langle x\rangle^{\delta}u_{\textup{loc}}\in L^{\infty}_{t}L^{2}_{x} for some constant \delta depending on the dimension. In the non-radial case, [ 34 ] proved that solutions decompose as u(x,t)=\sum u_{\textup{loc}}^{(j)}(x-x_{j}(t),t)+e^{-it\Delta}u_{+}+o_{H^{1}}(1) for some localized functions u_{\textup{loc}}^{(j)} and x_{j}(t) such that |x_{i}(t)-x_{j}(t)|\to\infty ( i\neq j ). Here, localization means \lim_{R\to\infty}\sup_{t}\lVert u_{\textup{loc}}^{(j)}(x,t)\rVert_{L^{2}(|x|\geq R)}=0 . In lower dimensions, it has not yet been possible to prove such a precise decomposition as ( 5 ). However, we do have the weaker decomposition u(x,t)=u_{\textup{wb}}(x,t)+e^{it\Delta}u_{+}+o_{H^{1}}(1) where u_{\textup{wb}}(t) is concentrated in the region |x|\lesssim t^{1/2+} [ 30 , 18 , 28 ] . Moreover, in dimension 3 for the radial cubic nonlinear Schrödinger equation, Tao proved [ 33 ] that we can further decompose u_{\textup{wb}} as u_{\textup{wb}}(t)=u_{\textup{loc}}(t)+o_{\dot{H}^{1}}(1) , where (6) |\nabla_{x}^{j}u_{\textup{loc}}(x,t)|\lesssim\langle x\rangle^{-3/2-j+} 1.2. Main result and sketch of the proof The main result in our paper states that solutions of ( 1 ) decompose into a linearly scattering radiation term and a term with localized kinetic energy: Theorem 1 . Suppose u solves ( 1 ) with a potential V satisfying (7) \sup_{t}|V(x,t)|+|\langle x\rangle\partial_{x}V(x,t)|\lesssim\langle x\rangle^{-\sigma} and that \sup_{t}\lVert u(t)\rVert_{H^{1}}<\infty , and that \theta\in\left(0,1\right) is a constant. Then, we have the decomposition (8) u(t)=e^{-it\Delta}u_{+}+u_{\textup{loc}}(t)+o_{\dot{H}^{1}}(1) where u_{+}\in H^{1} and u_{\textup{loc}} has a localized derivative in the sense that (9) \langle x\rangle^{\theta}\partial_{x}u_{\textup{loc}}\in L^{\infty}_{t}L^{2}_{x} Moreover, if V satisfies the symbol-type bounds (10) \sup_{k\in\{1,2,\cdots,n\}}\sup_{t}|\langle x\rangle^{k}\partial_{x}^{k}V(x,t)|\lesssim\langle x\rangle^{-\sigma} for n>1 and \theta\in\left(0,1\right) then we can redefine u_{\textup{loc}} in ( 8 ) such that (11) \langle x\rangle^{\theta k}\partial_{x}^{k}u_{\textup{loc}}\in L^{\infty}_{t}L^{2}_{x} for 1\leq k\leq n . Remark 1 . It is not clear whether u_{\textup{loc}} is also localized in L^{2} or not. We note that [ 29 ] showed that in high dimensions, it is possible to adiabatically rescale a potential V that admits a bound state in such a way that u_{\textup{wb}} (and hence u_{\textup{loc}} ) spreads at rate t^{1/2-} , nearly saturating the upper bound of t^{1/2+} . In particular, these self-similar bound states satisfy both ( 9 ) and ( 11 ). To prove 1 , we first extract the linear radiation. To do this, we define the free channel projection operator J_{\textup{free}}(t) as J_{\textup{free}}(t)=e^{-it\Delta}F_{\leq t^{\alpha}}(|x|)e^{it\Delta}F_{\geq t^{-\delta}}(|D|) Since solutions to the free Schrödinger equation concentrate on the phase-space support of J_{\textup{free}} , we define u_{+}=\Omega_{\textup{free}}u_{0}=\slim_{t\to\infty}e^{it\Delta}J_{\textup{free}}(t)u(t) Based on the decay of V

Read the complete text