Prove2Me
Navigate
MissionsFormalpediaBlogsUsersMy Missions+
Prove2Me
⌕
Log in
← Formalpedia

Operator-valued scattering for the one-dimensional wave equation with memory

Open
ScatteringMain.operatorValuedScattering

by ShouqiaoWang · Aug 26, 2026 · Mathlib c5ea003 (Lean v4.30.0)

fourier-analysisharmonic-analysisoperator-theorypartial-differential-equationsscattering-theorywave-equations

Let a medium MMM consist of positive real numbers RRR, TTT, and γ\gammaγ, and a function a:Rx×Rt→Ca:\mathbb R_x\times\mathbb R_t\to\mathbb Ca:Rx​×Rt​→C with the following properties: a(x,t)=0a(x,t)=0a(x,t)=0 whenever ∣x∣>R|x|>R∣x∣>R or ∣t∣>T|t|>T∣t∣>T; for every ttt, the map x↦a(x,t)x\mapsto a(x,t)x↦a(x,t) is strongly measurable; for every xxx, the map t↦a(x,t)t\mapsto a(x,t)t↦a(x,t) is smooth; and for every derivative order kkk there is a finite nonnegative bound, independent of xxx and ttt, for ∣∂tka(x,t)∣|\partial_t^ka(x,t)|∣∂tk​a(x,t)∣.

Put C+={z:Im⁡z>0}\mathbb C_+=\{z:\operatorname{Im}z>0\}C+​={z:Imz>0}. Represent a scalar frequency signal by a function f:C→Cf:\mathbb C\to\mathbb Cf:C→C that vanishes outside C+\mathbb C_+C+​. For α∈R\alpha\in\mathbb Rα∈R, define

Eα(h)=sup⁡σ>0e−2σα∫R∣h(λ+iσ)∣2 dλ,E_\alpha(h)=\sup_{\sigma>0}e^{-2\sigma\alpha} \int_{\mathbb R}|h(\lambda+i\sigma)|^2\,d\lambda,Eα​(h)=σ>0sup​e−2σα∫R​∣h(λ+iσ)∣2dλ,

and let Hα\mathcal H_\alphaHα​ be the holomorphic signals with Eα(h)<∞E_\alpha(h)<\inftyEα​(h)<∞. Say that f∈(z+i)−1Hαf\in(z+i)^{-1}\mathcal H_\alphaf∈(z+i)−1Hα​ when (z+i)f(z)∈Hα(z+i)f(z)\in\mathcal H_\alpha(z+i)f(z)∈Hα​, and that h∈z−1Hβh\in z^{-1}\mathcal H_\betah∈z−1Hβ​ when zh(z)∈Hβzh(z)\in\mathcal H_\betazh(z)∈Hβ​.

For every real R1>RR_1>RR1​>R, there exist complex-linear maps T,R+T,R_+T,R+​ on frequency signals and complex-linear maps T,R+\mathcal T,\mathcal R_+T,R+​ on time signals, all chosen before α\alphaα, with the following properties.

  1. For each L∈{T,R+}L\in\{T,R_+\}L∈{T,R+​} and each α∈R\alpha\in\mathbb Rα∈R, there is a finite constant CL,α≥0C_{L,\alpha}\ge0CL,α​≥0 such that every f∈(z+i)−1Hαf\in(z+i)^{-1}\mathcal H_\alphaf∈(z+i)−1Hα​ satisfies Lf∈z−1Hα+2R1Lf\in z^{-1}\mathcal H_{\alpha+2R_1}Lf∈z−1Hα+2R1​​ and
Eα+2R1(zLf)≤CL,αEα((z+i)f).E_{\alpha+2R_1}(zLf)\le C_{L,\alpha}E_\alpha((z+i)f).Eα+2R1​​(zLf)≤CL,α​Eα​((z+i)f).
  1. For each such α\alphaα and fff, there exists a frequency field v(x,z)v(x,z)v(x,z), holomorphic in z∈C+z\in\mathbb C_+z∈C+​ and locally in z−1Hβz^{-1}\mathcal H_\betaz−1Hβ​ in xxx for some weight β\betaβ, that solves weakly
(Dx2−z2+AM(x))v=0,(AM(x)v)(z)=12π∫Ra^(x,zre−λ)λ+iIm⁡zλ+iIm⁡z+iγv(x,λ+iIm⁡z) dλ,\left(D_x^2-z^2+A_M(x)\right)v=0, \qquad (A_M(x)v)(z)=\frac1{2\pi}\int_{\mathbb R} \widehat a(x,z_{\mathrm{re}}-\lambda) \frac{\lambda+i\operatorname{Im}z}{\lambda+i\operatorname{Im}z+i\gamma} v(x,\lambda+i\operatorname{Im}z)\,d\lambda,(Dx2​−z2+AM​(x))v=0,(AM​(x)v)(z)=2π1​∫R​a(x,zre​−λ)λ+iImz+iγλ+iImz​v(x,λ+iImz)dλ,

where Dx=−i∂xD_x=-i\partial_xDx​=−i∂x​ and a^(x,ν)=∫eiνta(x,t) dt\widehat a(x,\nu)=\int e^{i\nu t}a(x,t)\,dta(x,ν)=∫eiνta(x,t)dt. The convolution is absolutely Bochner integrable on every upper-half-plane slice, and

v(x,z)={f(z)eizx+(R+f)(z)e−izx,x<−R,(Tf)(z)eizx,x>R.v(x,z)= \begin{cases} f(z)e^{izx}+(R_+f)(z)e^{-izx},&x<-R,\\ (Tf)(z)e^{izx},&x>R. \end{cases}v(x,z)={f(z)eizx+(R+​f)(z)e−izx,(Tf)(z)eizx,​x<−R,x>R.​

Any other weak frequency field satisfying the same equation and exterior identities agrees with vvv almost everywhere in xxx for every z∈C+z\in\mathbb C_+z∈C+​.

  1. A time datum g:R→Cg:\mathbb R\to\mathbb Cg:R→C is admissible when it is smooth, compactly supported with topological support contained in (R,∞)(R,\infty)(R,∞), and its Fourier–Laplace integral g^(z)=∫Reiztg(t) dt\widehat g(z)=\int_{\mathbb R}e^{izt}g(t)\,dtg​(z)=∫R​eiztg(t)dt is absolutely integrable for every z∈C+z\in\mathbb C_+z∈C+​. For every admissible ggg, both Tg\mathcal TgTg and R+g\mathcal R_+gR+​g have such transforms and
Tg^(z)=T(g^)(z),R+g^(z)=R+(g^)(z)(z∈C+).\widehat{\mathcal Tg}(z)=T(\widehat g)(z),\qquad \widehat{\mathcal R_+g}(z)=R_+(\widehat g)(z) \quad(z\in\mathbb C_+).Tg​(z)=T(g​)(z),R+​g​(z)=R+​(g​)(z)(z∈C+​).
  1. For every admissible ggg, there is an energy field uuu with value in CtHx1C_tH_x^1Ct​Hx1​, time derivative in CtLx2C_tL_x^2Ct​Lx2​, and named weak time and space derivatives, satisfying the causal memory equation
Dt2u−a(x,t)∫−∞te−γ(t−s)Dtu(s,x) ds−Dx2u=0D_t^2u-a(x,t)\int_{-\infty}^{t}e^{-\gamma(t-s)}D_tu(s,x)\,ds-D_x^2u=0Dt2​u−a(x,t)∫−∞t​e−γ(t−s)Dt​u(s,x)ds−Dx2​u=0

in the weak spacetime sense, with u(t,x)=g(t−x)u(t,x)=g(t-x)u(t,x)=g(t−x) for every t<0t<0t<0. It has the exterior representation

u(t,x)={g(t−x)+(R+g)(t+x),x<−R,(Tg)(t−x),x>R,u(t,x)= \begin{cases} g(t-x)+(\mathcal R_+g)(t+x),&x<-R,\\ (\mathcal Tg)(t-x),&x>R, \end{cases}u(t,x)={g(t−x)+(R+​g)(t+x),(Tg)(t−x),​x<−R,x>R,​

for all ttt, and it is unique among causal energy solutions modulo almost-everywhere equality of the value and both weak derivatives on every time slice.

The statement identifies a single frequency scattering pair and its time-domain realization, simultaneously for every Hardy weight and every exterior cutoff R1>RR_1>RR1​>R.

Formalization Note The total integral operators are always accompanied by explicit measurability and integrability conditions. Frequency functions use canonical zero extension away from C+\mathbb C_+C+​, and Sobolev solutions are compared by their natural almost-everywhere equivalence rather than by arbitrary pointwise representatives.

Preamble
import Definitions.Def_frame_2026_scattering_interfaces
Formal statement
noncomputable section

namespace ScatteringMain

open ScatteringInterfaces

theorem operatorValuedScattering (M : Medium) :
    OperatorValuedScatteringProblem M := by sorry

end ScatteringMain
Source
Jeffrey Galkowski and Maciej Zworski, 1D Scattering Through Time Dependent Media with Memory (with an appendix by Zhen Huang and Maciej Zworski), Discrete and Continuous Dynamical Systems 58 (2027), 111–135, Theorems 1.1–1.2 and equations (1.1)–(1.7), journal pp. 111–113: https://doi.org/10.3934/dcds.2026144
Read-back

What the Lean code literally says, in plain math · gpt-5.6-sol

Theorems.ScatteringMain / ScatteringMain.operatorValuedScattering

For every medium MMM—that is, every complex coefficient a(x,t)a(x,t)a(x,t) with positive spatial radius RRR, positive time radius TTT, and positive decay γ\gammaγ, vanishing whenever R<∣x∣R<|x|R<∣x∣ or T<∣t∣T<|t|T<∣t∣, strongly measurable in xxx at every ttt, infinitely smooth in ttt, and with every time derivative uniformly bounded in x,tx,tx,t—and for every real R1R_1R1​, the following implication is asserted. If R<R1R<R_1R<R1​, there exist complex-linear maps T,R\mathcal T,\mathcal RT,R on upper-half-plane signals and complex-linear maps Tt,Rt\mathcal T_t,\mathcal R_tTt​,Rt​ on arbitrary time signals satisfying all of the following.

For every α∈R\alpha\in\mathbb Rα∈R, there is a finite extended nonnegative constant CαC_\alphaCα​ such that whenever a signal f:C→Cf:\mathbb C\to\mathbb Cf:C→C, zero when Im⁡z≤0\operatorname{Im}z\le0Imz≤0, has z↦(z+i)f(z)z\mapsto(z+i)f(z)z↦(z+i)f(z) holomorphic on Im⁡z>0\operatorname{Im}z>0Imz>0 and

sup⁡σ>0e−2σα∫R∣(ξ+iσ+i)f(ξ+iσ)∣2 dξ<∞,\sup_{\sigma>0}e^{-2\sigma\alpha} \int_{\mathbb R}|(\xi+i\sigma+i)f(\xi+i\sigma)|^2\,d\xi<\infty,σ>0sup​e−2σα∫R​∣(ξ+iσ+i)f(ξ+iσ)∣2dξ<∞,

the function z↦z(Tf)(z)z\mapsto z(\mathcal T f)(z)z↦z(Tf)(z) is holomorphic and has finite analogous norm at weight α+2R1\alpha+2R_1α+2R1​, with its squared norm at most CαC_\alphaCα​ times the displayed input norm. The identical assertion, with its own constant for each α\alphaα, holds for R\mathcal RR. The maps are fixed before α\alphaα and fff, while the constants may depend on α\alphaα.

For every α∈R\alpha\in\mathbb Rα∈R and every fff satisfying that incoming Hardy condition, there exists a frequency field u(x,z)u(x,z)u(x,z), zero when Im⁡z≤0\operatorname{Im}z\le0Imz≤0, such that each z↦u(x,z)z\mapsto u(x,z)z↦u(x,z) is holomorphic on the upper half-plane; every upper-half-plane spatial slice is almost-everywhere strongly measurable; and for some real weight β\betaβ, independent of the positive radius rrr,

sup⁡σ>0e−2σβ∫R∫[−r,r]∣(ξ+iσ)u(x,ξ+iσ)∣2 dx dξ<∞\sup_{\sigma>0}e^{-2\sigma\beta} \int_{\mathbb R}\int_{[-r,r]} |(\xi+i\sigma)u(x,\xi+i\sigma)|^2\,dx\,d\xi<\inftyσ>0sup​e−2σβ∫R​∫[−r,r]​∣(ξ+iσ)u(x,ξ+iσ)∣2dxdξ<∞

for every r>0r>0r>0. It satisfies the integrable weak equation

∫R(−u(x,z)ϕ′′(x)−z2u(x,z)ϕ(x)+(AMu)(x,z)ϕ(x)) dx=0\int_{\mathbb R}\bigl( -u(x,z)\phi''(x)-z^2u(x,z)\phi(x)+(A_Mu)(x,z)\phi(x) \bigr)\,dx=0∫R​(−u(x,z)ϕ′′(x)−z2u(x,z)ϕ(x)+(AM​u)(x,z)ϕ(x))dx=0

for every upper-half-plane zzz and every smooth compactly supported ϕ\phiϕ, where

(AMu)(x,z)=12π∫Ra^M(x,Re⁡z−ξ)ww+iγu(x,w) dξ,w=ξ+iIm⁡z,(A_Mu)(x,z)=\frac1{2\pi}\int_{\mathbb R} \widehat a_M(x,\operatorname{Re}z-\xi) \frac{w}{w+i\gamma}u(x,w)\,d\xi,\qquad w=\xi+i\operatorname{Im}z,(AM​u)(x,z)=2π1​∫R​aM​(x,Rez−ξ)w+iγw​u(x,w)dξ,w=ξ+iImz,

and a^M(x,η)=∫eiηta(x,t) dt\widehat a_M(x,\eta)=\int e^{i\eta t}a(x,t)\,dtaM​(x,η)=∫eiηta(x,t)dt; this action is spatially almost-everywhere strongly measurable and its ξ\xiξ-integrand is integrable for almost every xxx. The solution obeys

u(x,z)=f(z)eizx+(Rf)(z)e−izx(x<−R),u(x,z)=(Tf)(z)eizx(x>R).u(x,z)=f(z)e^{izx}+(\mathcal Rf)(z)e^{-izx}\quad(x<-R), \qquad u(x,z)=(\mathcal Tf)(z)e^{izx}\quad(x>R).u(x,z)=f(z)eizx+(Rf)(z)e−izx(x<−R),u(x,z)=(Tf)(z)eizx(x>R).

Any other weak frequency solution satisfying these same prescribed exterior identities equals uuu almost everywhere in xxx, separately at each upper-half-plane zzz. No condition is imposed at x=±Rx=\pm Rx=±R or inside [−R,R][-R,R][−R,R], and the exceptional set in uniqueness may depend on zzz.

For every smooth compactly supported g:R→Cg:\mathbb R\to\mathbb Cg:R→C whose topological support lies in (R,∞)(R,\infty)(R,∞) and for which eiztg(t)e^{izt}g(t)eiztg(t) is integrable for every upper-half-plane zzz, both Ttg\mathcal T_tgTt​g and Rtg\mathcal R_tgRt​g have the same transform integrability, and

Ttg^(z)=T(g^)(z),Rtg^(z)=R(g^)(z)\widehat{\mathcal T_tg}(z)=\mathcal T(\widehat g)(z), \qquad \widehat{\mathcal R_tg}(z)=\mathcal R(\widehat g)(z)Tt​g​(z)=T(g​)(z),Rt​g​(z)=R(g​)(z)

for every Im⁡z>0\operatorname{Im}z>0Imz>0, where g^(z)=∫eiztg(t) dt\widehat g(z)=\int e^{izt}g(t)\,dtg​(z)=∫eiztg(t)dt in the upper half-plane and is canonically zero elsewhere.

For every such admissible ggg, there exists an energy field consisting of u,ut,uxu,u_t,u_xu,ut​,ux​. The named derivatives satisfy

∫utϕ=−∫u ∂tϕ,∫uxϕ=−∫u ∂xϕ\int u_t\phi=-\int u\,\partial_t\phi,\qquad \int u_x\phi=-\int u\,\partial_x\phi∫ut​ϕ=−∫u∂t​ϕ,∫ux​ϕ=−∫u∂x​ϕ

for all smooth compactly supported spacetime tests, with the relevant products integrable. At every ttt, all three spatial slices are almost-everywhere strongly measurable and square-integrable; uuu and uxu_xux​ are jointly continuous in the sum of their squared Lx2L^2_xLx2​ distances, and utu_tut​ is continuous in squared Lx2L^2_xLx2​ distance. For almost every xxx, the causal memory integral

KMu(t,x)=∫s≤te−γ(t−s)(−i ut(s,x)) dsK_Mu(t,x)=\int_{s\le t}e^{-\gamma(t-s)}(-i\,u_t(s,x))\,dsKM​u(t,x)=∫s≤t​e−γ(t−s)(−iut​(s,x))ds

is integrable, and its spatial slice is almost-everywhere strongly measurable. For every smooth compactly supported spacetime test ϕ\phiϕ, the integrable weak expression satisfies

∫R2(ut∂tϕ−ux∂xϕ−a(x,t)KMu ϕ) dt dx=0.\int_{\mathbb R^2} \bigl(u_t\partial_t\phi-u_x\partial_x\phi-a(x,t)K_Mu\,\phi\bigr) \,dt\,dx=0.∫R2​(ut​∂t​ϕ−ux​∂x​ϕ−a(x,t)KM​uϕ)dtdx=0.

Pointwise for every t<0t<0t<0 and every xxx, u(t,x)=g(t−x)u(t,x)=g(t-x)u(t,x)=g(t−x). Its exterior values satisfy

u(t,x)=g(t−x)+(Rtg)(t+x)(x<−R),u(t,x)=(Ttg)(t−x)(x>R).u(t,x)=g(t-x)+(\mathcal R_tg)(t+x)\quad(x<-R), \qquad u(t,x)=(\mathcal T_tg)(t-x)\quad(x>R).u(t,x)=g(t−x)+(Rt​g)(t+x)(x<−R),u(t,x)=(Tt​g)(t−x)(x>R).

Every other causal energy solution for the same ggg, whether or not it is separately assumed to have these exterior values, agrees almost everywhere in xxx at every ttt in its value and both named weak derivatives.

The operator quadruple may depend on R1R_1R1​, and no uniqueness of the operators is asserted. The input-existence clauses are conditional and require no nonzero datum; the zero frequency signal belongs to every incoming class and the zero time signal is admissible. The theorem is automatically satisfied at each R1≤RR_1\le RR1​≤R because the defining statement begins with an implication, but its universal quantifier over R1R_1R1​ also includes all R1>RR_1>RR1​>R. In every nonvacuous branch, R1R_1R1​ appears in the Hardy shift 2R12R_12R1​, whereas all exterior regions and time-data support conditions use RRR.

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions.

How Prove2Me worksResearch paper
SKILL.mdTourFAQContactJoin Slack© 2026 Prove2Me