1: \documentclass[12pt,reqno]{amsart}
2: \usepackage{graphicx}
3: \usepackage{amssymb,amscd,amsbsy}
4: \usepackage{amsthm}
5: %\input{definitions}
6: \pagestyle{plain}
7: %\date{\mbox{}}
8: \setlength{\topmargin}{.1in}
9: \setlength{\textheight}{8in}
10: \setlength{\textwidth}{5.8in}
11: \setlength{\evensidemargin}{.4in}
12: \setlength{\oddsidemargin}{.4in}
13:
14: %\newmathalphabet*{\bit}{cmm}{b}{it}
15:
16:
17:
18: \newcommand{\lb}{\linebreak}
19: \newcommand{\noi}{\noindent}
20: \newcommand{\nl}{\newline}
21:
22:
23:
24:
25: \newcommand{\Z}{{\mathbb Z}}
26: \newcommand{\C}{{\mathbb C}}
27: \newcommand{\RB}{{\mathbb R}}
28: \newcommand{\RP}{{\mathbb R}{\mathbb P}}
29: \newcommand{\CP}{{\mathbb C}{\mathbb P}}
30:
31: \newcommand\Lg{\mathfrak L}
32: \newcommand\Dg{\mathfrak D}
33: \newcommand{\Rg}{{\mathfrak R}}
34:
35: \newcommand{\HH}{{\mathcal H}}
36: \newcommand{\MM}{{\mathcal M}}
37: \newcommand{\TT}{{\mathcal T}}
38:
39: \def\SL{{\mathcal S}(L_0)}
40: \def\Os{\Omega_0}
41: \def\O{\Omega}
42: \def\F{\Phi}
43: \def\P{\Psi}
44: \def\[{\left[}
45: \def\]{\right]}
46: \def\({\left(}
47: \def\){\right)}
48: \def\t{\theta}
49: \def\l{\lambda}
50: \def\m{\mu}
51: \def\lt{\frac{\lambda}{2}}
52: \def\psib{\overline{\psi}}
53: \def\D{\Delta}
54: \def\d{\delta}
55: \def\ea{\epsilon_a}
56: \def\epm{\epsilon_{\pm}}
57:
58:
59: \def\mi{\mu_{\infty}(\lambda)}
60: \def\pin{p_{\infty} (\lambda)}
61: \def\fr{\f_{\rightarrow}}
62: \def\fl{\f_{\leftarrow}}
63: \def\Gin{\Gamma_\infty}
64: \def\G{\Gamma}
65:
66: \newcommand{\e}{{\boldsymbol e}}
67: \newcommand{\f}{{\boldsymbol f}}
68: \newcommand{\jo}{{\boldsymbol j}}
69: \newcommand{\js}{{\mathfrak j}}
70: \newcommand{\g}{\gamma}
71: \newcommand{\gt}{\tilde{\gamma}}
72: \newcommand{\xit}{\tilde{\xi}}
73: \newcommand{\bb}{\bar{b}}
74:
75: \newcommand{\p}{\partial}
76: \newcommand{\eeq}{\end{equation}}
77: \newcommand{\beq}{\begin{equation}}
78: \newcommand{\bay}{\begin{eqnarray}}
79: \newcommand{\ey}{\end{eqnarray}}
80: \newcommand{\bey}{\begin{eqnarray*}}
81: \newcommand{\eey}{\end{eqnarray*}}
82:
83: %Peller's abbreviations
84:
85: \newcommand{\Rat}{\operatorname{Rat}}
86: \newcommand{\R}{\operatorname{res}}
87: \newcommand{\ph}{\operatorname{ phase}}
88: \newcommand{\df}{\stackrel{\mathrm{def}}{=}}
89: \newcommand{\dist}{\operatorname{dist}}
90: \newcommand{\Ker}{\operatorname{Ker}}
91: \newcommand{\re}{\operatorname{Re}}
92: \newcommand{\spn}{\operatorname{span}}
93: \newcommand{\supp}{\operatorname{supp}}
94: \newcommand{\clos}{\operatorname{clos}}
95: \newcommand{\convh}{\operatorname{convh}}
96: \newcommand{\codim}{\operatorname{codim}}
97: \newcommand{\rank}{\operatorname{rank}}
98: \newcommand{\const}{\operatorname{const}}
99: \newcommand{\tr}{\operatorname{trace}}
100: \newcommand{\sign}{\operatorname{sign}}
101:
102:
103: \newtheorem{thm}{\hspace{\parindent}Theorem}[section]
104: \newtheorem{prop}[thm]{Proposition}
105: \newtheorem{cor}[thm]{\hspace{\parindent}Corollary}
106: \newtheorem{lem}[thm]{\hspace{\parindent}Lemma}
107: \newtheorem{conj}[thm]{Conjecture}
108: \newtheorem{exa}[thm]{Example}
109: \newtheorem{defn}[thm]{Definition}
110:
111:
112:
113:
114:
115:
116:
117:
118: \theoremstyle{remark}
119: \newtheorem{rem}[thm]{Remark}
120: \newtheorem*{rem*}{Remark}
121:
122:
123:
124: \newcommand{\shto}{\raisebox{.5ex}{$\scriptscriptstyle\rightarrow$}\!}
125: \newcommand\cE{\mathcal{E}}
126:
127: %\numberwithin{equation}{section}
128:
129: \begin{document}
130:
131: \newcommand{\vse}{\vspace{.2in}}
132: \numberwithin{equation}{section}
133:
134: \title{\bf Symplectic structures for the cubic Schr\"{o}dinger equation in the
135: periodic and scattering case.}
136: \author{K.L. Vaninsky}
137: \keywords{nonlinear Schr\"{o}dinger, symplectic}
138: \thanks{ 35Q53; 58B99. The work is partially supported by NSF grant DMS-9971834.}
139:
140:
141: \begin{abstract}
142: We develop a unified approach for construction of symplectic forms for 1D integrable
143: equations with the periodic and rapidly decaying initial data. As an example we consider
144: the cubic nonlinear Schr\"{o}dinger equation.
145: \end{abstract}
146: \maketitle
147: \setcounter{section}{0}
148: \setcounter{equation}{0}
149: \section{\bf Introduction}
150:
151:
152:
153: \subsection{General remarks.} The main technical tool for the study of soliton systems is commutator formalism.
154: All fashionable soliton systems like the Korteveg--de--Vriez equation (KdV), the cubic nonlinear Schr\"{o}dinger
155: equation (NLS), the $\sin$--Gordon equation, the Toda lattice, {\it etc}; have such representation. Within the commutator formalism
156: approach the dynamical system appears as a compatibility condition for an over-determined system of equations.
157: As an example, we consider the NLS equation with repulsive nonlinearity
158: \footnote{Prime $'$ signifies derivative in variable $x$ and dot $\bullet$ in time.}
159: $$
160: i \psi^{\bullet}= -\psi'' + 2 |\psi |^2 \psi,
161: $$
162: where $\psi(x,t)$ is a complex function of spatial variable $x$ and time $t$.
163: The flow is a compatibility condition for the commutator
164: $$
165: [\partial_t- V_3, \partial_x-V_2]=0,
166: $$
167: with
168: $$
169: V_2(x,t)=V = - \frac{i \l}{2} \sigma_3 +Y_0 =- \frac{i \l}{2}
170: \(\begin{array}{ccccc}
171: 1 & 0\\
172: 0 & -1 \end{array}\) +
173: \(\begin{array}{ccccc}
174: 0& \overline{\psi} \\
175: \psi & 0 \end{array}\)
176: $$
177: and
178: $$
179: V_3(x,t) = \lt^2i \sigma_3 -\l Y_0 + |\psi|^2 i\sigma_3 -i \sigma_3 Y_0'.
180: $$
181: The corresponding auxiliary linear problem
182: $$
183: (\partial_x -V )\f=0, \qquad\qquad\qquad \f=\(\begin{array}{ccccc} f_1\\
184: f_2\end{array}\)
185: $$
186: can be written in the form of an eigenvalue problem for the Dirac operator
187: $$
188: \Dg \f= \[\(\begin{array}{ccccc} 1& 0 \\ 0 & -1 \end{array}\) i\partial_x +
189: \(\begin{array}{ccccc} 0& -i \psib \\
190: i \psi & 0 \end{array}\)\] \f= \lt \f.
191: $$
192:
193:
194:
195: Another important feature of soliton systems is the Hamiltonian formulation. Here we assume that the potential $\psi(x,t)$ is
196: $2l$--periodic: $\psi(x+2l,t)=\psi(x,t)$.
197: For instance, the NLS flow can be written as
198: $$
199: \psi^{\bullet}= \{\psi, H_3\},
200: $$
201: with Hamiltonian
202: $H_3=\frac{1}{ 2} \int_{-l}^{l} |\psi'|^2 +|\psi|^4 \, dx=${\it energy} and bracket
203: $$
204: \{A,B\}=2i \int_{-l}^{l} \frac{\partial A}{ \partial \psib(x)}\frac{\partial B}{ \partial
205: \psi(x)}- \frac{\partial A}{ \partial \psi(x)}\frac{\partial B}{ \partial \psib(x)}\, dx.
206: $$The bracket is nondegenerate.
207: The corresponding symplectic form (up to a scalar) is:
208: $$
209: \omega_0= 2i< \d \psib \wedge \d \psi >, \qquad\qquad\qquad <\bullet>\;\;= \frac{1}{ 2l} \int_{-l}^{l} dx.
210: $$
211:
212:
213: {\it A priori} it is not clear why the dynamical system, which arises as a compatibility
214: condition has a Hamiltonian formulation. To put it
215: differently, is it possible to obtain Hamiltonian formalism from the spectral problem?
216:
217: Here we would like to make some historical remarks. Originally, the Hamiltonian formulation of basic integrable models
218: was found as an experimental fact. For the KdV equation
219: the computation of symplectic structure in terms of the scattering data was performed by Faddeev and
220: Zakharov, \cite{FZ}. It involved some
221: nontrivial identities for the products of solutions. Later Kulish and Reiman, \cite{KR}, noted that
222: all higher symplectic structures also can be written in terms of the scattering data. Again,
223: they used the scheme of \cite{FZ} and explicit calculations. Finally, we note that Zakharov and Manakov,
224: \cite{ZM}, for the NLS equation adopted a different approach. Instead of the symplectic structure they worked with the corresponding
225: Poisson bracket. Again, using explicit formulas for the product of solutions they computed the
226: Poisson bracket between the coefficients of the scattering matrix. An appearance of explicit formulas that are the moving force of all
227: these computations seems to be quite mysterious. This was already discussed in the literature
228: \cite{FT}, and described as a ''computational miracle".
229:
230:
231:
232: The standard assumption needed to carry out spectral analysis is that the potential either is periodic
233: or has rapid decay at infinity. We refer to the latter case as scattering.
234: Recently, in connection with the Seiberg-Witten theory \cite{SW1,SW2}, Krichever and Phong, \cite{KP}
235: developed a new approach for the construction of symplectic formalism. The latest exposition of their results can be found in \cite{DKP}.
236: The main idea of the Krichever--Phong approach is to introduce in a universal
237: way the two-form on the space
238: of auxiliary linear operators. This form is written in terms of the operator itself and its
239: eigenfunctions. The goal of this paper is to review the Krichever-Phong approach in the case of 1D periodic NLS
240: and to extend it to the scattering case. Within the unified approach, we reduce the number of formulas and eliminate unnecessary explicit computations. For instance computation of the symplectic form in terms of the spectral data (both in the periodic and the scattering case) becomes
241: an application of the Cauchy residue theorem.
242:
243: \subsection{ The periodic case.} We assume that the potential is periodic with the period $2l:
244: \; \psi(x+2l,t)=\psi(x,t)$. The Krichever--Phong formula, in the NLS context, takes the form
245: $$
246: \omega_0=\sum\limits_{P_{\pm}} \R <\e^* J\; \delta V \wedge \delta \e> d \lambda.
247: $$
248: This formula defines a closed 2--form $\omega_0$ on the space of operators
249: $\partial_x - V(x,\l)$ with $2l$ periodic potential. The set--up for this formula is broadly as follows \footnote{ We refer to Section 2 for detailed discussion.}.
250:
251: The eigenvalue problem $$[J\partial_x -JV(x,\l)]\e(x,\l)=0,\qquad\qquad\qquad J=i\sigma_2= \(\begin{array}{ccccc}
252: 0& 1 \\
253: -1 & 0 \end{array}\)$$
254: has special solutions, so--called Floquet solutions determined by the property $\e(x+2l,\l)= w(\l) \e(x,\l)$.
255: The complex constant $w(\l)$ is called a Floquet multiplier. For each value of the spectral parameter $\l$
256: there are two linearly independent Floquet solutions and two distinct Floquet multipliers.
257: These solutions and correspondingly multipliers become single--valued functions of a point on the two--sheeted covering of the plane of
258: spectral parameter $\l$. The simple points of the periodic/antiperiodic spectrum of the eigenvalue problem constitute
259: branching points of the cover. We assume that there is a finite number of simple points (so--called, finite gap potential).
260: \begin{figure}[ht]
261: \includegraphics[width=0.60\textwidth]{fig211.eps}
262: \caption{Smooth Riemann surface $\Gamma$.}
263: \end{figure}
264:
265:
266:
267: This two sheeted covering constitutes a hyperelliptic Riemann surface $\Gamma$ with two infinities $P_{+}$ and $P_-$ (Fig. 1). Each point $Q=(\l,\pm)$ of $\Gamma$ is specified by the value of spectral parameter $\l$ and the sheet ''+'' or ''-'' which determines the Floquet multiplier $w(Q)$ corresponding to this $\l$. At every point of the curve we also have a Floquet solution $\e(x,Q)$ which becomes a function of the point $Q$ and satisfies the
268: identity $\e(x+2l,Q)= w(Q) \e(x,Q)$. The Floquet solution $\e(x,Q)$ has an exponential singularity at infinities and plays the role of
269: so--called Baker--Akhiezer function for the curve $\Gamma$.
270:
271: At every point of the curve $\Gamma$ we can define another solution $\e^*(x,Q)$. This is the Floquet solution
272: which is brought from a point on the different sheet but with the same value of the spectral
273: parameter $\l$. It is transposed and suitably normalized. The operator $J\partial_x -JV(x,\l)$ acts on the solution $\e^*(x,Q)$ as an adjoint {\it i.e.} on the right:
274: $$
275: \e^*(x,Q)\[J\partial_x-J V\]=0.
276: $$
277:
278:
279: It is assumed that the phase space consists of smooth
280: $2l$--periodic functions $\psi(x)$ or equivalently operators $\partial_x - V(x,\l)$ with $2l$--periodic potential.
281: The NLS flow acts on this space as it acts on the space of functions $\psi(x)$.
282: All notions of differential geometry with obvious conventions can be applied to this space of operators.
283: On the space of potentials we have a variation $\delta \psi(x)$.
284: Thus for a fixed value of the spectral parameter $\l$ we have well defined variation $ \delta V(x,\l)$. The variations $ \delta \e(x,Q),
285: \delta \e^*(x,Q)$ are defined correctly when $\l=\l(Q)$ is fixed. Therefore, at each point $Q$ of the surface $\Gamma$ we have well defined meromorphic in $Q$ the two--form
286: $$
287: <\e^* J\delta V \wedge \delta \e> d \lambda.
288: $$
289: It takes values in the space of skew--symmetric two--forms on the space of operators
290: $\partial_x - V$. The result of Krichever and Phong states that the sum of residues of this form at infinities $P_{\pm}$ is nothing but the symplectic
291: form $\omega_0$.
292:
293:
294:
295:
296:
297: The formula has a lot of good properties. First, it produces all higher symplectic structures by introducing the weight $\l^n$ under the residue
298: $$
299: \omega_n=\sum\limits_{P_{\pm}} \R \; \l^n\,<\e^* J\; \delta V \wedge \delta \e> d\,\lambda,\qquad\qquad\qquad\qquad n=1,2,\hdots.
300: $$
301: Second, it easily leads to the Darboux coordinates, or in physics terminology the separation variables, see Sklyanin \cite{S}.
302: These are local coordinates where the symplectic form $\omega_0$ takes the simple canonical form
303: $$
304: \omega_0=\frac{2}{i}\sum\limits_{k} \d p(\gamma_k) \wedge \d \l (\gamma_k).
305: $$
306: This merits special explanation.
307: It is well-known since the work of Flashka--McLaughlin, \cite{FM},
308: that the poles $\gamma_k$ of Floquet solutions lead to the Darboux coordinates for symplectic
309: forms \footnote{See also Novikov--Veselov, \cite{NV},
310: for general discussion.}.
311: Recently, a lot of work was performed, \cite{SNK}, to construct such variables for the Ruijsenaars--Sneider
312: and the Moser--Calogero systems. This required formidable technical machinery and extensive
313: computations. At the same time, as it was demonstrated by Krichever, \cite{K2}, the formula leads
314: to the same result only by applying the Cauchy residue theorem.
315:
316: \subsection{ The Scattering case.} The main goal of the present paper is to show that suitably interpreted
317: the new approach can be adopted for soliton systems with rapidly decaying initial data
318: on the entire line. This is so--called the scattering case\footnote{ We refer to Section 3 for detailed definitions.}.
319:
320:
321: For such potentials one can define so--called Jost solutions $J_{\pm}(x,\l)$.
322: These are matrix solutions of the auxiliary linear problem $J_{\pm}'= V J_{\pm}$ with the asymptotics
323: $J_{\pm}'(x,\l)= \exp{(-{i\lt x} \sigma_3)} + o(1)$, as $x\rightarrow \pm \infty$.
324: Their columns $J_{\pm}=\[\jo^{(1)}_{\pm},\jo^{(2)}_{\pm}\]$ are analytic in the corresponding
325: upper/lower half--plane.
326:
327: \begin{figure}[htb]
328: \includegraphics[width=0.60\textwidth]{fig212.eps}
329: \caption{ Singular Riemann surface $\Gin$.}
330: \end{figure}
331:
332:
333:
334: Our construction of the associated Riemann surface $\Gin$ is a geometrical interpretation of what
335: is called the Riemann-Hilbert approach to the scattering problem, see \cite{FT}.
336: A singular curve $\Gin$ is obtained by taking two copies of the complex plane and gluing them to each other along the real line (Fig. 2).
337: The curve $\Gin$ has two infinities $P_+$ and $P_-$ and continuum set of singular points
338: above the real line. The standard Jost solutions are lifted on
339: $\Gin$ and become the single valued function of a point on the curve.
340: Different branches of BA function are connected along the real line by the
341: scattering matrix $S$:
342: $$
343: S(\l) = \frac{1}{ a} \[\begin{array}{ccccc} \;\; {1} & {\overline{b}}\\
344: -{b}& {1}
345: \end{array}\].
346: $$
347: The Jost solution
348: has exponential singularity at infinities and plays the role of the Baker--Akhiezer function for
349: the curve $\Gin$. This construction is explained in detail in Section 3.
350:
351: The formula of Theorem 3.4 looks similar to the periodic case
352: $$
353: \omega_0=
354: \tr\; \R\; \frac{1}{ 2} \[ <H_+^* J\delta V \wedge \delta H_+> +
355: <H_-^* J\delta V \wedge \delta H_->\] d \lambda.
356: $$
357: The only difference now is that we work with the matrix solutions
358: $$
359: H_+(\l)=\[ \jo_-^{(1)}(\l),\jo_+^{(2)}(\l)\] \quad {\rm and}\quad\quad H_-(\l)=\[ \jo_
360: +^{(1)}(\l),\jo_-^{(2)} (\l) \]
361: $$
362: and $H_+^+(\l)=\sigma_1 H_+^T$, $H_-^+(\l)=\sigma_1 H_-^T$, with
363: $$
364: \sigma_1=\left(\begin{array}{ccccc}
365: 0& 1\\
366: 1 & 0
367: \end{array} \right).
368: $$
369: The averaging now corresponds to the integration on the entire line
370: $$
371: <\bullet>\;\;= \int_{-\infty}^{+\infty} dx.
372: $$
373: The residue can be computed explicitly $\omega_0=2i<\d \psib\wedge \d \psi>$.
374: Theorem 3.6 states that the symplectic structure can be put in the Darboux form
375: $$
376: \omega_0= \frac{1}{ \pi i} \int\limits_{-\infty}^{+\infty}
377: \frac{\d \bar b(\l)\wedge \d b(\l)}{ |a(\l)|^2}\, d\l,
378: $$
379: where $a$ and $b$ are coefficients of the scattering matrix $S$.
380: Again identically to the periodic case this result is obtained by applying the Cauchy residue theorem.
381: Only now the sum of the residues in the affine part of the curve is replaced by its' continuous
382: analog. This is the integral which stays in the right hand side of the formula.
383:
384:
385: The unified approach to construction of symplectic forms produces an interesting problem. As we see,
386: the symplectic form constructed in the periodic case has two systems of Darboux coordinates.
387: One system is associated with poles of the Floquet solution. It is
388: the divisor--quasimomentum Darboux coordinates. Another system of Darboux coordinates is the
389: action-angle variables. At the same time in the scattering case we know only one system of
390: Darboux coordinates. These are action-angle variables. What is the correct analog of the
391: divisor--quasimomentum in the scattering case? This is a subject of future publication, \cite{V2}.
392:
393:
394: We conclude this introduction by expressing thanks to A. Its, H. McKean and I. Krichever for
395: stimulating discussions. We are also greatful to anonymous referee for remarks that helped to improve the presentation.
396:
397:
398: \section{ The Periodic Case.}
399: \subsection{The Direct Spectral Problem}
400:
401: We provide here information needed in the next section for construction of symplectic forms.
402: We refer
403: to classical books \cite{MO,NMP} for standard facts of spectral theory and algebraic--geometrical
404: approach to solitons.
405:
406: The NLS equation
407: \bay\label{nls}
408: i \psi^{\bullet}= -\psi'' + 2 |\psi |^2 \psi,
409: \ey
410: where $\psi(x,t)$ is a smooth complex function $2l$--periodic in $x$,
411: is a Hamiltonian system
412: $$
413: \psi^{\bullet}= \{\psi, H\},
414: $$
415: with the Hamiltonian
416: $H=\frac{1}{ 2} \int_{-l}^{l} |\psi'|^2 +|\psi|^4 \, dx=${\it energy} and the bracket
417: \bey
418: \{A,B\}=2i \int_{-l}^{l} \frac{\partial A}{ \partial \psib(x)}\frac{\partial B}{ \partial
419: \psi(x)}- \frac{\partial A}{ \partial \psi(x)}\frac{\partial B}{ \partial \psib(x)}\, dx.
420: \eey
421: The NLS Hamiltonian $H=H_3$ is one in the infinite series of conserved integrals of motion.
422: \bey
423: H_1&=&\frac{1}{ 2} \int_{-l}^{l} |\psi|^2 dx,\\
424: H_2&=&\frac{1}{ 2i} \int_{-l}^{l} \psi \bar{\psi}' dx,\\
425: H_3&=&\frac{1}{ 2} \int_{-l}^{l} |\psi'|^2 +|\psi|^4 \, dx,\quad etc.
426: \eey
427: These Hamiltonians produce an infinite hierarchy of flows $e^{tX_m},\; m=1,2,\ldots$.
428:
429:
430:
431:
432:
433: The first in the hierarchy is the phase flow $e^{tX_1}$ generated by the vector field
434: $$
435: X_1:\qquad\qquad \psi^{\bullet}=\{\psi, H_1\}= -i \psi.
436: $$
437: The phase flow is a compatibility condition for
438: \bay\label{com}
439: [\partial_t- V_1, \partial_x-V_2]=0,
440: \ey
441: with\footnote{
442: Here and below $\sigma$ denotes the {\it Pauli matrices}
443: $$
444: \sigma_1=\left(\begin{array}{ccccc}
445: 0& 1\\
446: 1 & 0
447: \end{array} \right), \nonumber \quad
448: \sigma_2=\left(\begin{array}{ccccc}
449: 0 & -i\\
450: i & 0
451: \end{array} \right), \nonumber \quad
452: \sigma_3=\left(\begin{array}{ccccc}
453: 1 & 0\\
454: 0 & -1
455: \end{array} \right). \nonumber
456: $$
457: }
458: $V_1=\frac{i}{ 2} \sigma_3$ and
459: $$
460: V_2 = - \frac{i \l }{ 2} \sigma_3 +Y_0 =
461: \(\begin{array}{ccc}
462: - \frac{i\l}{ 2} & 0\\
463: 0 & \frac{i\l}{ 2} \end{array}\) +
464: \(\begin{array}{ccccc}
465: 0& \overline{\psi} \\
466: \psi & 0 \end{array}\).
467: $$
468: We often omit the subscript $V=V_2$.
469: The second, translation flow $e^{tX_2}$ generated by
470: $$
471: X_2:\qquad\qquad \psi^{\bullet}=\{\psi, H_2\}= \psi'
472: $$
473: is equivalent to \ref{com} with $V_1$ replaced by $V_2$.
474: Finally, the third, original NLS flow \ref{nls} is a compatibility condition for \ref{com}
475: with $V_1$ replaced by
476: $$
477: V_3 = \frac{\l^2}{ 2}i \sigma_3 -\l Y_0 + |\psi|^2 i\sigma_3 -i \sigma_3 Y_0'.
478: $$
479:
480: All flows of infinite hierarhy $e^{tX_m},\; m=1,2,\ldots$ commute with each other
481: $$
482: [\partial_{\tau_m} - V_m, \partial_{\tau_n} -V_n]=0.
483: $$
484: The first times $\tau_1,\tau_2$ and $\tau_3$ correspond to the first three flows.
485:
486:
487:
488:
489: We introduce a $2\times 2$ transition matrix $M(x,y,\l),\; x \geq y$; that satisfies
490: $$
491: M'(x,y,\l)=V(x,\l) M(x,y,\l),\qquad\qquad\qquad M(y,y,\l)=I.
492: $$
493: The solution is given by the formula
494: $$
495: M(x,y,\l)=\exp\int_{y}^{x} V(\xi,\l) d\xi.
496: $$
497: The matrix $M(x,y,\l)$ is unimodular because $V$ is traceless.
498:
499: The symmetry
500: $$
501: \sigma_1 \overline{V}(x,\l)\sigma_1= V(x,\bar{\l})
502: $$
503: produces the same relation for the transition matrix
504: \bay\label{real}
505: \sigma_1 \overline{M}(x,y,\l)\sigma_1= M(x,y,\bar{\l}).
506: \ey
507: Another symmetry
508: $$
509: V^T(x,\l) J=-J V(x,\l),
510: $$
511: where $J=i\sigma_2$, implies
512: \beq\label{nsy}
513: M^T(x,y,\l)^{-1}J=J M(x,y,\l).
514: \eeq
515:
516:
517:
518:
519: The quantity $\Delta(\l)=\frac{1}{ 2} \text{trace} M(l,-l,\l)$ is called a discriminant.
520: The formula \ref{real} implies $\overline{\Delta}(\l)=\Delta(\bar{\l})$ and $\Delta(\l)$ is real for real $\l$.
521: The eigenvalues of the monodromy matrix have a name of Floquet multipliers and
522: they are roots of the quadratic equation
523: \beq\label{qe}
524: w^2 -2\Delta w +1=0.
525: \eeq
526: The Floquet multipliers are given by the formula $w=\Delta\pm \sqrt{\Delta^2-1}$. The values of $\l: w(\l)=\pm 1$ constitute the points of the periodic/antiperiodic
527: spectrum. The corresponding auxiliary linear problem
528: $$
529: (\partial_x -V )\f=0, \qquad\qquad\qquad \f^T=(f_1,f_2);
530: $$
531: can be written in the form of an eigenvalue problem for the self--adjoint Dirac operator
532: $$
533: {\Dg} \f= \[\(\begin{array}{ccccc} 1& 0 \\ 0 & -1 \end{array}\) i\partial_x +
534: \(\begin{array}{ccccc}
535: 0& -i \overline{\psi} \\
536: i \psi & 0 \end{array}\)\] \f= \lt \f.
537: $$
538: The self--adjointness implies that points of the spectra are real.
539:
540:
541:
542: \noi
543: {\it Example.} Let $\psi \equiv 0$. The corresponding monodromy matrix can be easily computed
544: $M(x,y,\l)=e^{-{i\lt}\sigma_3 (x-y)}$.
545: We have $\Delta(\l)=\cos \l l$ and double
546: eigenvalues at the points $\l_n^{\pm}=\frac{\pi n}{ l}.$
547: If n is even/odd, then the corresponding $\l_n^{\pm}$ belongs to the periodic/
548: anti-periodic spectrum.
549:
550:
551:
552: For a generic potential the double points $\l_n^{\pm}$ of the periodic/anti-periodic
553: spectrum split, but they always stay real.
554: The size of the spectral gap is determined, roughly speaking, by
555: the corresponding Fourier coefficients of the potential. In our considerations we assume that
556: there is a finite number of $g+1$ open gaps in the spectrum
557: $$
558: \hdots<\l_{n-1}^{-}=\l_{n-1}^{+}<\l_{n}^{-}<\l_{n}^{+}<\hdots< \l_{n+g}^{-}<\l_{n+g}^+< \l_{n+g+1}^{-}=\l_{n+g+1}^{+}<\hdots
559: $$
560: These are so--called finite gap potentials which are dense among all potentials.
561:
562: The Floquet multipliers become single--valued on
563: the Riemann surface:
564: $$
565: \Gamma=\{Q=(\l,w)\in { \C}^2:\quad R(\l,w)=\det\[M(l,-l,\l)-wI\]=0\}.
566: $$
567: The Riemann surface consists of two sheets covering the plane of the spectral parameter $\l$.
568:
569: \noi
570: {\it Example.} Let $\psi \equiv 0$.
571: We have $\Delta(\l)=\cos \l l$ and quadratic equation \ref{qe} has the solutions $w(\l)=e^{\pm il \l}$.
572: The Riemann surface $\G=\G_+ + \G_-$ is reducible and consists of two copies of the complex plane $\C$ that intersect each other at the points of the double spectrum $\l_n^{\pm}$. Each part $\G_+$ or $\G_-$ contains the corresponding infinity $P_+$ or $P_-$.
573: The Floquet multipliers are single valued on $\G$:
574: \bey
575: w(Q)&=&e^{+i\l l},\qquad\qquad\qquad\qquad Q\in \G_+;\\
576: w(Q)&=&e^{-i\l l},\qquad\qquad\qquad\qquad Q\in \G_-.
577: \eey
578:
579:
580:
581:
582:
583:
584:
585: For a finite--gap potential the Riemann surface $\G$ is irreducible.
586: There are three types of important points on $\Gamma$. These are the singular points, the points above $\l=\infty$ and the branch points which we
587: discuss now in detail.
588: \begin{itemize}
589: \item The singular points are determined by the condition
590: $$
591: \partial_{\l}R(\l,w)=\partial_{w}R(\l,w)=0.
592: $$
593: These are the points $(\l^{\pm},\pm 1)$ of the double spectrum. At these points two sheets of the curve intersect.
594: \item There are two nonsingular points $P_+$ and $P_-$ above $\l=\infty$. At these points\footnote{The notation $Q\in (P)$ means that the point $Q$ is in the vicinity of the point $P$.}
595: \bay
596: w(Q)&=&e^{+i\l l}\(1+O\({1/\l}\)\),\qquad\qquad Q\in (P_+);\label{asf}\\
597: w(Q)&=&e^{-i\l l}\(1+O\({1/\l}\)\),\qquad\qquad Q\in (P_-).\label{assf}
598: \ey
599: \item The branch points are specified by the condition
600: $$
601: \partial_{w}R(\l,w)=0.
602: $$
603: They are different from the singular points and correspond to the simple periodic/antiperiodic spectrum. We denote these points by
604: $s_k^{\pm}=(\l^{\pm}_k,(-1)^k),$ $\;k=n,\hdots,n+g$. There are $2(g+1)$ of them, each
605: has a ramification index 2.
606: \end{itemize}
607: \noi
608: The desingularized curve $\G$ is biholomorphicaly equivalent to a hyperelliptic curve with branch points at the points of the simple spectrum. We also
609: denote the hyperellitic curve by $\G$.
610: The Riemann-Hurwitz formula for the genus of $\G$ implies
611: $$
612: \text{genus}=\frac{R}{2}-n+1,
613: $$
614: where $R$ is a total ramification index and $n$ is the number of sheets. Each branch point has a ramification index $1$ and therefore $R=2(g+1)$ and
615: $n=2$. Therefore, the genus of $\G$ is $g$, one off the number of open gaps in the spectrum.
616:
617:
618:
619:
620: Let $\epm$ be a
621: holomorphic involution on the curve $\Gamma$ permuting sheets
622: $$\epm: \; (\l,w)\longrightarrow (\l,1/w).$$
623: The fixed points of $\epm$ are the branch points of $\G$. The involution $\epm$ permutes infinities $\epm: P_-\longrightarrow P_+$.
624: Let us also define on $\Gamma$ an antiholomorphic involution
625: $$\epsilon_a: (\l, w ) \longrightarrow (\bar{\l}, \bar{w}).$$ The involution $\ea$ also permutes infinities and commutes with $\epm$.
626: Points of the curve above gaps $\[\l_n^-, \l_n^+\]$ where $|\Delta(\l)| \geq 1$ form $g+1$ fixed ``real'' ovals of $\epsilon_a$.
627: We call them $a$--periods.
628:
629: The quasimomentum $p(Q)$ is a multivalued function on the curve $\Gamma$. It is introduced
630: by the formula $w(Q)=e^{i p(Q)2l}$. Evidently, it is defined up to $\frac{\pi n}{ l}$, where $n$ is an integer.
631: The asymptotic expansion for $p(Q)$ at infinities can be easily computed
632: $$
633: \pm p(\l)= \lt - p_{0}^{\pm} - \frac{p_{1}}{ \l} - \frac{p_{2}}{ \l^2}\ldots,
634: \quad \quad \quad \quad Q\in(P_{\pm}),\;\, \l=\l(Q);
635: $$
636: where $p_{0}^{\pm} = \frac{\pi k_{\pm}}{ l}, \; k_{\pm}$ is an integer and
637: $$
638: p_{1} = \frac{1}{ l} H_1, \quad
639: p_{2} = \frac{1}{ l} H_2, \quad
640: p_{3} = \frac{1}{ l} H_3, \quad etc.
641: $$
642: Moreover, the function $w(Q)+ w(\epsilon_\pm Q)$ does not depend on the sheet and is equal to $2\Delta(\l)$.
643: Thus $\Delta(\l(Q))=\cosh i p(Q) 2l$ and the formula
644: $$
645: dp= \pm \frac{1}{ i 2l} d \cosh^{-1} \Delta(\l) = \pm \frac{1}{ i 2l}\frac{\Delta^{\bullet}(\l)\,
646: d\l}{ \sqrt{\Delta^2-1}}
647: $$
648: implies that differential $dp$ is of the second kind with double poles at the infinities:
649: $\pm dp=d\(\lt +O(1)\)$. The same formula implies that the differential $dp$ is pure complex on the real ovals. At the same time, the condition $w(s_k^-)=w(s_k^+)$ requires the increment $p(s_k^+)-p(s_k^-)$ to be real.
650: Therefore, $dp$ has zero $a$-periods
651: $$
652: \int_{a_k} dp=0.
653: $$
654: Since the Floquet multiplies are single--valued on $\G$ for the $b$--periods we have
655: \beq\label{pc}
656: \int_{b_k} dp=\frac{\pi n_{b_k}}{ l},\;\qquad\qquad\qquad n_{b_k} \in \Z,\;\quad k=1,\hdots,g.
657: \eeq
658: These are so--called the {\it periodicity conditions}, \cite{NMP}.
659:
660:
661:
662: The Floquet solution is the vector--function
663: $$
664: \e(x,Q)= \[\begin{array}{ccccc}
665: e^1(x,Q)\\
666: e^2(x,Q)\end{array} \]
667: $$
668: which is a solution of the auxiliary spectral problem $\e'= V \e$ with the property
669: \beq\label{mp}
670: \e(x+2l, Q) = M(l,-l,\l) \e(x,Q)= w(Q) \e(x, Q)
671: \eeq
672: and normalized by the condition
673: \bay\label{norm}
674: e^1(-l,Q) +e^2(-l,Q)=1.
675: \ey
676:
677: \noi
678: {\it Remark.} If $\f(x,\l)$ is a solution of the auxiliary problem
679: $$
680: (\partial_x -V(x,\l) )\f=0
681: $$
682: corresponding to $\l$, then $\hat \f=\sigma_1 \bar{\f}$ is a solution of $(\partial_x -V(x,\overline{\l}) )\hat\f=0$ corresponding to $\overline{\l}$.
683:
684: \noi
685: {\it Example.} Let $\psi=0$. The Floquet solution is
686: given by the formula
687: $$
688: \e(x,Q) = e^{+i \lt(x+l)} \e_0 = e^{+i \lt(x+l)}
689: \[\begin{array}{ccccc} 0\\ 1\end{array}\], \qquad\qquad Q\in \G_+,
690: $$
691: $$
692: \e(x,Q) = e^{-i \lt(x+l)} \hat \e_0 = e^{-i \lt(x+l)} \[\begin{array}{ccccc}
693: 1\\
694: 0\end{array} \], \qquad\qquad Q\in \G_-.
695: $$
696: It has no poles in the affine part of the curve.
697:
698:
699: For a general finite gap potential the situation is more complicated.
700: \begin{lem}
701: \label{lfloc}The Floquet solution satisfies the identity
702: $$
703: e(x,\ea Q)=\sigma_1 \overline{\e(x,Q)}.
704: $$
705: The Floquet solution $\e(x,Q)$ has poles common for both components at the points $$\gamma_1,\gamma_2,\hdots,\gamma_{g+1}.$$
706: Projections of poles $\mu_k= \l(\gamma_k)$ are real. Each $\gamma_k$ lies on the real oval above the corresponding open
707: gap $[\l_k^-,\l^+_k]$.
708: Each component $e^i(x,Q)$ has $g+1$ zeros $$\sigma^i_1(x),\; \sigma^i_2(x),\hdots,\sigma^i_{g+1}(x);\qquad\qquad\qquad\; i=1,2.$$ These zeros depend on the parameter $x$.
709: In the vicinity of infinities the function $\e(x,Q)$ has the asymptotics
710: $$
711: \e(x,Q) = e^{\pm i \lt(x+l)} \[ \e_0 / \hat \e_0 + o(1)\], \qquad \qquad
712: Q\in (P_{\pm}).
713: $$
714: \end{lem}
715:
716: Before proceeding to the proof of the Lemma we note that the differential equation for the monodromy matrix
717: $$
718: M'(x,y,\l)=\[- \frac{i \l }{ 2} \sigma_3 +Y_0\] M(x,y,\l),\qquad\qquad\qquad M(y,y,\l)=I,
719: $$
720: multiplied (gauged) on the left and right by the matrices
721: $$
722: C=\(\begin{array}{ccccc} 1& 1 \\ i & -i \end{array}\)\qquad\text{and}\qquad
723: C^{-1}=\frac{1}{2}\(\begin{array}{ccccc} 1& -i \\ 1 & i \end{array}\)
724: $$
725: transforms into
726: $$
727: M^{\RB}(x,y,\l)=\[- \frac{i \l }{ 2} \sigma_2 +Y_0^{\RB}\] M^{\RB}(x,y,\l),\qquad\qquad\qquad M^{\RB}(y,y,\l)=I;
728: $$
729: with
730: $$
731: Y_0^{\RB}=\(\begin{array}{ccccc} q& p \\ p & -q \end{array}\),\qquad\qquad\qquad \psi=q+ip.
732: $$
733: This is a real version of the eigenvalue problem which is more convenient in some situations, see \cite{MV}. The Floquet solution
734: $\e^{\RB}(x,Q)$ corresponding to the real version of the eigenvalue problem is related to $\e(x,Q)$ by the formula
735: $$\e^{\RB}(x,Q)=C \e(x,Q).
736: $$
737: Therefore the result of the Lemma for $\e(x,Q)$ follows from the corresponding result for $\e^{\RB}(x,Q)$ given in \cite{MV}. We prefer to give
738: a direct proof, though the gauge transformation is behind all arguments.
739:
740:
741: \noi
742: {\it Proof.} The proof is based on the explicit formula for the Floquet solution. Let
743: $$
744: M(x,-l,\l)= \[ \begin{array}{ccccc}
745: m_{11} & m_{12}\\
746: m_{21} & m_{22} \end{array}\] (x,-l,\l);
747: $$
748: and $M_{11}, M_{12},$ {\it etc.} be the elements of the matrix $M(l,-l,\l)$.
749: The Floquet solution $\e(x,Q)$ is given by the formula
750: \bay\label{floq}
751: \e(x,Q)=A(Q) \[\begin{array}{cccc}
752: m_{11}\\
753: m_{21} \end{array} \](x,-l,\l) + (1-A(Q)) \[\begin{array}{cccccc}
754: m_{12}\\
755: m_{22} \end{array} \](x,-l,\l),
756: \ey
757: where $\l= \l(Q)$ and the coefficient $A(Q)$ is
758: \bay\label{coef}
759: A(Q)=\frac{M_{12}}{ M_{12}-M_{11} +w(Q)} \quad \quad \text{or} \quad \quad
760: A(Q)=\frac{w(Q)- M_{22}}{ M_{21}- M_{22} +w(Q)}.
761: \ey
762:
763:
764: To prove the formula note that the Floquet solution is a linear combination of columns of the monodromy matrix $M(x,-l,\l)$:
765: $$
766: \e(x,Q)=A(Q) \[\begin{array}{cccc}
767: m_{11}\\
768: m_{21} \end{array} \](x,-l,\l) + A'(Q) \[\begin{array}{cccccc}
769: m_{12}\\
770: m_{22} \end{array} \](x,-l,\l), \quad \quad \l= \l(Q).
771: $$
772: The normalization condition \ref{norm} implies $A'(Q)=1-A(Q)$. At the same time the Floquet solution is an eigenvector of the monodromy matrix
773: $$
774: M(l,-l,\l) \[\begin{array}{cccc}
775: A(Q)\\
776: 1-A(Q) \end{array} \]= w(Q) \[\begin{array}{cccc}
777: A(Q)\\
778: 1-A(Q) \end{array} \].
779: $$
780: This leads to two equations
781: $$
782: M_{11}A(Q) + M_{12} (1-A(Q))=w(Q) A(Q),
783: $$
784: or
785: $$
786: M_{21}A(Q)+ M_{22}(1-A(Q))=w(Q) (1-A(Q)).
787: $$
788: Each equation implies the corresponding formula for $A(Q)$.
789:
790: The formulas \ref{real} and \ref{coef} imply
791: $$
792: 1-A(\ea Q)=\overline{A(Q)}.
793: $$
794: This and \ref{real}, \ref{floq} imply the stated identity for the Floquet solution.
795:
796: The relation $M^{\RB}=CMC^{-1}$ implies
797: \bey
798: M^{\RB}&=& \[\begin{array}{ccccc}
799: M^{\RB}_{11} & M^{\RB}_{12}\\
800: M^{\RB}_{21}& M^{\RB}_{22} \end{array}\]\\
801: &=&\frac{1}{2} \[ \begin{array}{ccccc}
802: M_{11}+M_{12}+M_{21}+M_{22}& i(M_{12}+M_{22}-M_{11}-M_{21})\\
803: i(M_{11}+M_{12}-M_{21}-M_{22})& M_{11}+M_{22}-M_{12}-M_{21} \end{array}\].
804: \eey
805: Due to \ref{real} $M^{\RB}(\l)$ is real for real $\l$. Consider the function $M^{\RB}_{12}(\l)$ and look at the roots
806: $\mu_n:\;M^{\RB}_{12}(\mu_n)=0$. For $\psi\equiv 0$ we have
807: $M^{\RB}_{12}(\l)=-\sin\frac{\l l}{2}$ with roots at the points $\mu_n=\frac{2\pi n}{l},\;
808: n\in \Z$. When we add the potential the roots $\mu_n$ move but stay real. They are caught by open gaps or match double
809: periodic/antiperiodic spectrum. Indeed at $\mu_n$ the matrix $M^{\RB}$ is lower triangular and real entries $M^{\RB}_{11}$ and $M^{\RB}_{22}$
810: coincide with Floquet multipliers. Since $M^{\RB}_{11}M^{\RB}_{22}(\mu_n)=1$ we have $|\Delta(\mu_n)|=\frac{1}{2}|M^{\RB}_{11}+M^{\RB}_{22}|
811: \geq 1$.
812:
813:
814: The points of the divisor $\gamma_k\in \Gamma$ lie above the points $\mu_k$ on the sheet with $w(Q)=M^{\RB}_{22}(\mu_n)$. At these points
815: the denominator in \ref{coef}
816: $$
817: M_{12}-M_{11} +w(Q)\qquad\text{or}\qquad M_{21}- M_{22} +w(Q)
818: $$
819: vanishes. Indeed from $M^{\RB}_{12}(\mu_n)=0$ we have
820: $$ M_{12}-M_{11}(\mu_n)=M_{21}-M_{22}(\mu_n)$$
821: and $w(Q)=M^{\RB}_{22}(\mu_n)=M_{11}-M_{12}(\mu_n)=M_{22}-M_{21}(\mu_n)$. Moreover $M_{22}\neq M^{\RB}_{22}(\mu_n)$. These
822: produce a pole of the Floquet solution when $\mu_n$ lies in the open gap. When $\mu_n$ is caught by the periodic/antiperiodic spectrum the matrix
823: $
824: M(l,-l,\mu_n)=\pm I
825: $
826: and the zero of denominator is annihilated by the zero of numerator in \ref{coef}.
827:
828:
829:
830:
831:
832: The asimptotics of the Floquet solution follows from the formula \ref{floq} and
833: $$
834: M(x,y,\l)=e^{-{i\lt}\sigma_3 (x-y)}+ o(1),
835: \qquad\qquad \text{when}\qquad \l \rightarrow \infty.
836: $$
837: \qed
838:
839:
840: The Floquet solution $\e(x,Q)$ near infinities can be expanded into the asymptotic series
841: $$
842: \e(x,Q)=e^{+i \lt(x+l)} \sum\limits_{s=0}^{\infty} \e_s(x) \l^{-s} =
843: e^{+i \lt(x+l)} \sum\limits_{s=0}^{\infty} \[\begin{array}{ccccccc}
844: b_s \\
845: d_s
846: \end{array} \] \l^{-s}, \quad Q\in (P_{+}),
847: $$
848: $$
849: \e(x,Q)=e^{-i \lt(x+l)} \sum\limits_{s=0}^{\infty} \hat \e_s(x) \l^{-s}
850: = e^{-i \lt(x+l)} \sum\limits_{s=0}^{\infty} \[\begin{array}{ccccc}
851: \overline{d}_s \\
852: \overline{b}_s
853: \end{array} \] \l^{-s} , \quad Q\in (P_{-}),
854: $$
855: and $b_0= 0,\quad d_0=1$.
856: The coefficients $b_s, d_s$ can be computed from the relation
857: \bay\label{rec}
858: -\[\begin{array}{ccccccc}
859: b_s'\\
860: d_s'
861: \end{array} \] + Y_{0}
862: \[\begin{array}{ccccccc}
863: b_s\\
864: d_s
865: \end{array} \]= \frac{i}{ 2} \(I + \sigma_3 \)
866: \[\begin{array}{ccccccc}
867: b_{s+1}\\
868: d_{s+1}
869: \end{array} \], \quad\quad s=0,1,\ldots,
870: \ey
871: due to the diagonal form of the matrix
872: $$
873: \frac{i}{ 2} \(I + \sigma_3 \)= \[\begin{array}{ccccccc} i& 0\\ 0 & 0
874: \end{array} \].
875: $$
876: Indeed, relation \ref{rec} leads to the identities
877: \bay\label{fir}
878: -b_s' + \psib d_s = ib_{s+1},
879: \ey
880: and
881: \bay\label{sec}
882: -d_s' + \psi b_s = 0.
883: \ey
884: These are supplemented by the boundary condition
885: \bay\label{bo}
886: b_s(x) + d_s(x)|_{x=-l}=0,\quad \quad \quad s \geq 1.
887: \ey
888:
889:
890:
891:
892:
893:
894:
895: When $s=0$ using $b_0= 0,\;\; d_0=1$ from \ref{fir} we obtain
896: $$
897: b_1=-i\overline{\psi}(x).
898: $$
899: The identities \ref{sec} and \ref{bo} imply
900: $$
901: d_1=i\overline{\psi}_0 - i \int\limits_{-l}^{x} |\psi|^2 dx',\qquad\qquad\qquad \psi_0=\psi(-l).
902: $$
903: Similar, we compute
904: \begin{eqnarray*}
905: b_2& =& \overline{\psi}' + \overline{\psi}\, \overline{\psi}_0 - \overline{\psi}
906: \int\limits_{-l}^x |\psi|^2 dx', \\
907: d_2&=& -\overline{\psi}_0' - \overline{\psi}_0^2 + \int\limits_{-l}^{x}
908: \[ \psi \overline{\psi}' +|\psi|^2 \overline{\psi}_0-|\psi|^2 \int\limits_{-l}^{x'}
909: |\psi|^2 dx''\] dx'.
910: \end{eqnarray*}
911: These formulae will be used for explicit computation of symplectic forms.
912:
913:
914:
915:
916:
917: The Floquet solution $\e(x,Q)$ satisfies the identity
918: $$
919: \[J\partial_x- J V\] \e(x,Q) =0, \qquad\qquad\qquad J=i\sigma_2;
920: $$
921: which is just another way to write the spectral problem.
922: Let us define the dual Floquet solution $\e^+(x,Q)=\[e^{1+}(x,Q), e^{2+}(x,Q)\]$ at the point
923: $Q$ as
924: $$
925: \e^+(x,Q)=\e(x,\epsilon_{\pm }Q )^T.
926: $$
927: It can be verified by a direct computation that the dual Floquet solution
928: $\e^+(x,Q)$ satisfies\footnote{The action of the differential operator
929: $D=\sum\limits_{j=0}^{k} \omega_{j}
930: \partial^{j}$ on the row vector $f^{+}$ is defined as
931: $f^+ D=\sum\limits_{j=0}^{k} (-\partial)^{j}(f^{+} \omega_j).$}
932: $$
933: \e^+(x,Q)\[J\partial_x-J V\]=0.
934: $$
935: The fact that the Wronskian $\e^+(x,Q)J \e(x,Q)$ does not depend on $x$ can be verified by differentiation.
936: Introducing the function
937: $$
938: \F(Q)=\e^+(x,Q)J \e(x,Q)=<\e^+(x,Q)J\e(x,Q)>= \frac{1}{ 2l} \int_{-l}^{l}\e^+(x,Q)J\e(x,Q) dx,
939: $$
940: we define another dual Floquet solution $\e^*(x,Q)$ by the formula
941: $$
942: \e^*(x,Q)=\frac{\e^+(x,Q)}{ \F(Q)}.
943: $$
944: Evidently, $\e^*(x,Q) J \e(x,Q)=1$. The symmetry \ref{nsy} produces an analog of monodromy property \ref{mp} for the
945: function $\e^*$:
946: \beq\label{mps}
947: \e^*(x+2l, Q) = \e^*(x, Q) JM^{-1}(l,-l,\l)J^{-1} = w^{-1}(Q) \e^*(x, Q).
948: \eeq
949: \begin{lem}
950: \label{lsfloq}
951: The function $\e^*(x,Q)$ has simple poles at the branch points
952: $s_k^{\pm}$. It has fixed zeros at $\gamma_1,\ldots, \gamma_{g+1}$.
953: The other zeros for each component of the vector function $\e^*$ lie on every real oval and depend on the parameter $x$. The function
954: $\e^*$ has the asymptotics at infinities
955: $$
956: \e^*(x,Q)=\pm e^{\mp \frac{i\l}{ 2}(x+l)} \[ {\hat \e}_0^T / \e_0^T +o(1)\],
957: \quad\quad Q\in(P_+/P_-).
958: $$
959: \end{lem}
960:
961: \noindent
962: {\it Proof.} The function $\F(Q)$ is meromorphic with $2(g+1)$ poles on both sheets above points $\mu_n$ lying in open gaps and $2(g+1)$
963: zeros at the branching points $s_k^{\pm}$. At infinities it has the asymptotics
964: $$
965: \F(Q)=\pm 1 +o(1),\qquad\qquad\qquad\qquad Q\in (P_{\pm}).
966: $$
967: Now it is easy to prove properties of the function $\e^*(x,Q)$. It has poles at the branch points $s_k^{\pm}$ which arise from zeros
968: of $\F(Q)$. It has zeros at $\gamma_1,\ldots, \gamma_{g+1}$, the poles of $\F(Q)$. Other poles of $\F(Q)$ are annihilated by the
969: poles of $\e^+(x,Q)$. Other $g+1$ zeros of $\e^*(x,Q)$ which depend on $x$ are produced by the corresponding zeros of $\e^+(x,Q)$.
970:
971:
972:
973: The asymptotics follows from the asymptotics for $\e(x,Q)$ and $\F(Q)$.
974: \qed
975:
976: Consider periodic variations of the matrix $V(x,\l): \tilde{V}= V+ \d V$. Then $\tilde{p}(Q)=p(Q)+ \d p(Q)+ \hdots$.
977: We need a standard formula connecting the variations $ \d p(Q)$ and $\d V$.
978: \begin{lem} The following identity holds
979: $$
980: i\d p(Q)= < \e^*(x,Q)J \d V \e(x,Q)>.
981: $$
982: \end{lem}
983:
984: \noindent
985: {\it Proof.} Let $\tilde{\e}(x,Q)$ be a Floquet solution corresponding to the deformed potential $\tilde{V}$.
986: From the definition
987: $$
988: \e^+(x,Q)\([J\partial_x -J \tilde{V}]\tilde{\e}(x,Q)\)=0
989: $$
990: and
991: $$
992: \(\e^+(x,Q)[J\partial_x -J {V}]\)\tilde{\e}(x,Q)=0.
993: $$
994: Subtracting one identity from another, we have
995: $$
996: \e^+(J\partial_x \tilde{\e})-(\e^+J\partial_x) \tilde{\e}=\e^+J \tilde{V}\tilde{\e} - \e^+J{V}\tilde{\e}= \e^+J\d {V}\tilde{\e}.
997: $$
998: Integrating both sides, we have
999: $$
1000: \int_{-l}^{+l} \e^+J \tilde{\e}'+ \e^{+'}J \tilde{\e} \;dx=\e^+J \tilde{\e}|_{-l}^{+l}.
1001: $$
1002: Using the identities
1003: $$
1004: \e^+(l,Q)=e^{-ip(Q)2l} \e^+(-l,Q)
1005: $$
1006: and
1007: $$
1008: \tilde{\e}(l,Q)=e^{i\tilde{p}(Q)2l} \;\,\tilde{\e}^+(-l,Q)
1009: $$
1010: for the LHS, we have
1011: $$
1012: \( e^{i\tilde{p}(Q)2l}e^{-ip(Q)2l}-1\) \e^+(x,Q)J \tilde{\e}(x,Q)= i\d p(Q)2l \F(Q)+ \text{lower order terms}.
1013: $$
1014: The RHS is equal to
1015: $$
1016: 2l < \e^+(x,Q)J \d V \e(x,Q)>+ \text{lower order terms}.
1017: $$
1018: Collecting leading terms, we obtain the stated identity.
1019: \qed
1020:
1021:
1022:
1023:
1024: Consider a real hyperelliptic spectral curve $\Gamma$ of finite genus corresponding to some periodic
1025: potential $\psi$.
1026: Let us introduce the Baker-Akhiezer function $\e(\tau,x,t,Q)$ which depends on three parameters (times)
1027: $\tau,\;x$ and $t$ and has the asymptotics at infinities
1028: $$
1029: \e(\tau,x,t,Q)=e^{\pm i\(-\frac{1}{ 2}\tau +{\lt}x-{\lt^2}t\)} \times\[
1030: \e_0/ \hat \e_0 +o(1) \],\qquad \qquad Q\in (P_+/P_-).
1031: $$
1032: The BA function has poles at the points $\gamma$'s, located on the real ovals. These
1033: properties define the BA function uniquely. The BA function can be written explicitly in
1034: terms of theta-functions of the curve $\Gamma$, \cite{K1}.
1035: The BA function has Bloch property in $x$-variable
1036: $\e(\tau, l,t,Q)=w(Q) \e(\tau, -l,t,Q)$ and satisfies the identities
1037: $$
1038: \begin{array}{cccccc}
1039: &\[J\partial_{\tau}-J V_1(\tau,x,t)\]\e(\tau,x,t,Q)=0,\\
1040: &\[J\partial_{x}-J V_2(\tau,x,t)\]\e(\tau,x,t,Q)=0,\\
1041: &\[J\partial_{t}-J V_3(\tau,x,t)\]\e(\tau,x,t,Q)=0.
1042: \end{array}
1043: $$
1044: The three matrices $V_1, V_2$ and $V_3$ are given at the beginning of this section.
1045:
1046:
1047:
1048: Let us define the dual BA function $\e^{+}(\tau, x,t,Q)$ at the point $Q$ as
1049: $$
1050: \e^{+}(\tau, x,t,Q)\equiv \e(\tau, x,t,\epsilon_\pm Q)^{T}.
1051: $$
1052: The identity $w(Q)w(\epsilon_\pm Q)=1$ implies $\e^{+}(\tau,l,t,Q)=w(Q)^{-1}
1053: \e^{+}(\tau,-l,t,Q)$.
1054: The dual BA function $\e^{+}(\tau,x,t,Q)$ satisfies dual identities
1055: $\e^{+}(\tau,x,t,Q)\[J\partial_{\tau}-J V_1\]=0,$ {\it etc}.
1056:
1057:
1058:
1059:
1060:
1061: \subsection{ Symplectic Structures. }
1062:
1063:
1064:
1065: We assumed in the previous section that the phase space consists of smooth
1066: $2l$--periodic functions $\psi(x)$. Instead we can change the language and think about the phase space
1067: as a space of operators $\partial_x - V_2$ with $2l$--periodic potential.
1068: The flows of the NLS hierarchy act on this space as well they act on the space of functions $\psi$.
1069: All notions of differential geometry can be applied to this space of operators with
1070: obvious conventions.
1071:
1072:
1073: \begin{lem} The formula
1074: $$
1075: \omega_0=\sum\limits_{P_{\pm}} \R \; <\e^* J\delta V \wedge \delta \e> d \lambda,
1076: $$
1077: defines a closed 2--form $\omega_0$ on the space of operators
1078: $\partial_x - V_2$ with periodic potential.
1079: The flows $e^{tX_m},\; m=1,2,\ldots$ on the space of operators defined by the
1080: formula
1081: $$
1082: [\partial_{\tau_m} - V_m, \partial_x -V_2]=0,
1083: $$
1084: are Hamiltonian with the symplectic structure $\omega_0$ and the Hamiltonian
1085: function $H_m$ (up to unessential constant factor).
1086: \end{lem}
1087:
1088: \noi
1089: {\it Remark.} The formula
1090: $$
1091: \omega_n=\sum\limits_{P_{\pm}} \R \;\l^n <\e^* J\delta V \wedge \delta \e> d \lambda,
1092: \quad \quad \quad \quad n=0, 1 \ldots,
1093: $$
1094: defines a closed 2--form $\omega_n$ on the space of operators
1095: $\partial_x - V_2$ with periodic potential which satisfy the constrains
1096: $H_k=const,\; k=1,\ldots,n$; see for details \cite{KP}.
1097:
1098: Before proceeding to the proof of the Lemma, we compute the first two symplectic structures using the formula
1099: $$
1100: \omega_n=\sum\limits_{P_{\pm}} \R \frac{\l^n}{ \F(Q)}
1101: <\e^+ J\delta V \wedge \delta \e> d \lambda,\qquad\qquad\qquad n=1,2.
1102: $$
1103: The result is
1104: \begin{eqnarray}
1105: \omega_0&=& 2i <\d \psib \wedge \d \psi>, \label{sy0}\\
1106: \omega_1& =& \omega_1= <\d \psi \wedge \d \psib' + \d \psib \wedge \d \psi' +
1107: 2\d \partial^{-1} |\psi|^2\wedge \d |\psi|^2>.\label{sy1}
1108: \end{eqnarray}
1109: subject to the constraint $H_1=const$.
1110: We present the computaion divided in small steps.
1111:
1112:
1113: \noi
1114: {\it Step 1.} Identity $\F(\tau_{\pm}Q)=-\F(Q)$ implies
1115: \begin{eqnarray*}
1116: \frac{1}{ \F(Q)} & =\;\; \phi_0 + \frac{\phi_1}{ \l} + \frac{\phi_2}{ \l^2} +\cdots,
1117: \quad \quad \quad \quad
1118: Q\in (P_+), \\
1119: \frac{1}{ \F(Q)} & =-\phi_0 - \frac{\phi_1}{ \l} - \frac{\phi_2}{ \l^2} - \cdots,
1120: \quad \quad \quad \quad
1121: Q\in (P_-).
1122: \end{eqnarray*}
1123: Using definition of $\e(x,Q)$ and $\e^+(x,Q)$ from previous section, we have
1124: \begin{eqnarray*}
1125: \phi_0& = &1,\\
1126: \phi_1& = & -<d_1+ \overline{d}_1>.\\
1127: \end{eqnarray*}
1128:
1129:
1130: \noi
1131: {\it Step 2.} Near $P_+$ we obtain
1132: $$
1133: <\e^+ J\d V \wedge \d \e> = \frac{s_1}{ \l} + \frac{s_2}{ \l^2} +
1134: \frac{s_3}{ \l^3} + \cdots,
1135: $$
1136: where
1137: \begin{eqnarray*}
1138: s_1& = &<\hat \e_0^T J \d V \wedge \d \e_1>,\\
1139: s_2& = & <\hat \e_0^T J \d V \wedge \d \e_2> +
1140: <\hat \e_1^T J \d V\wedge \d \e_1>.
1141: \end{eqnarray*}
1142: Similar, at $P_-$, we obtain
1143: $$
1144: <\e^+ J\d V \wedge \d \e> = - \frac{\bar{s}_1}{ \l} - \frac{\bar{s}_2}{ \l^2} -
1145: \frac{\bar{s}_3}{ \l^3} - \cdots.
1146: $$
1147: Using the expansion for $\F(Q)$ from Step 1, we derive
1148: \begin{eqnarray*}
1149: \omega_0& = & \phi_0 (s_1 + \bar{s}_1),\\
1150: \omega_1& = & \phi_0 (s_2 + \bar{s}_2) + \phi_1 (s_1 + \bar{s}_1).
1151: \end{eqnarray*}
1152: \noi
1153: {\it Step 3.} Computing $s_1$, we have
1154: $$
1155: s_1=<\hat \e_0^T J \d V\wedge \d \e_1> =<\d \psi \wedge \d b_1>.
1156: $$
1157: Using the formula $b_1=-i \psib$, we obtain $s_1= -i <\d \psi \wedge \d \psib> $
1158: and \ref{sy0}.
1159:
1160: \noi
1161: {\it Step 4.} The first term in the formula for $s_2$ produces
1162: $$
1163: <\hat \e_0^T J\d V \wedge \d \e_2>=< \d \psi \wedge \d b_2> .
1164: $$
1165: Using recurrence relation \ref{fir}: $ b_2=ib_1' -i \psib d_1 $, we obtain
1166: $$
1167: < \d \psi \wedge i \d b_1' - i \d \psib d_1 - i \psib \d d_1>.
1168: $$
1169: The second term in the formula for $s_2$ produces
1170: $$
1171: <\hat \e_1^T J \d V \wedge \d \e_1> = <- \overline{b}_1 \d \psib \wedge \d d_1 +
1172: \overline{d}_1 \d \psi \wedge \d b_1>.
1173: $$
1174: Finally,
1175: $$
1176: s_2= <\d \psi \wedge \d \psib' -i(d_1+\overline{d}_1) \d \psi \wedge \d \psib - i
1177: \d |\psi|^2 \wedge \d d_1>.
1178: $$
1179: Using the formula for $s_1$,
1180: $$
1181: \omega_1= <\d \psi \wedge \d \psib' + \d \psib\wedge \d \psi' - i \delta |\psi |^2\wedge
1182: \d(d_1-\overline{d}_1)>.
1183: $$
1184: The constraint $H_1=const$ implies $<\d |\psi|^2>=0$ and using the explicit formula for
1185: $d_1$, we obtain \ref{sy1}.
1186:
1187:
1188: \noi
1189: {\it Proof.} Closeness of the form $\omega_0$ follows from the result of next Lemma or from explicit formula \ref{sy0}. For the second statement
1190: we present a complete proof only for $m=0$. Higher flows can be treated similarly.
1191: In order to prove that the first flow is Hamiltonian one has to establish, \cite{A}:
1192: $$
1193: i_{\partial_t} \omega_0= - \d\, 2H_1,
1194: $$
1195: where $i_{\partial_t}$ is the contraction operator produced by the vector field $X_1$.
1196: Using time--dependent BA functions $i_{\partial_t} \delta \e = \e^{\bullet},\;\; i_{\partial_t} \delta V= V^{\bullet}$
1197: we have
1198: $$
1199: i_{\partial_t} \omega_0= \sum\limits_{P_{\pm}} \R < \e^* J V^{\bullet} \delta \e> d\l -
1200: \R <\e^* J \delta V \e^{\bullet}>d\l.
1201: $$
1202:
1203: Let us compute the residue at $P_+$.
1204: From the computation preceding the proof
1205: \begin{eqnarray*}
1206: \frac{1}{ \F(Q)} & = & \;\; \phi_0 + \frac{\phi_1}{ \l} + \ldots,
1207: \quad \quad \quad \quad
1208: Q\in (P_+), \\
1209: \frac{1}{ \F(Q)} & = & -\phi_0 - \frac{\phi_1}{ \l} - \ldots,
1210: \quad \quad \quad \quad
1211: Q\in (P_-).
1212: \end{eqnarray*}
1213: Then, using $V^{\bullet}=\[ \frac{i }{ 2} \sigma_3, V\]=i \sigma_3 V$, we obtain
1214: $$
1215: \R < \e^* J V^{\bullet} \delta \e> d\lambda =
1216: i \phi_0 <\hat \e_0^T J \sigma_3 Y_0 \delta \e_1>.
1217: $$
1218: Similarly, using $\e^{\bullet} = \frac{i}{ 2} \sigma_3 \e$ we have:
1219: \begin{eqnarray*}
1220: \R < \e^* J \delta V \e^{\bullet}> d\l
1221: & = &{i}{ 2 } \phi_1 <\hat{\e}_0^T J \d Y_0 \sigma_3 \e_0> \\
1222: & + & {i}{ 2}\phi_0 \[ < \hat {\e}_0^T J \delta Y_0 \sigma_3 \e_1 > +
1223: < \hat {\e}_1^T J \delta Y_0 \sigma_3 \e_0 >\].
1224: \end{eqnarray*}
1225: The first term vanishes and
1226: \begin{eqnarray*}
1227: i_{\partial_t} \R <\e^* J\delta V \wedge \delta \e>d \lambda &=&
1228: i \phi_0 <\hat{\e}_0^T J \sigma_3 Y_0 \delta \e_1> \\
1229: & - &\frac{i }{ 2 } \phi_0 \[ <\hat{\e}_0^T J \d Y_0 \sigma_3 \e_1> +
1230: < \hat{\e}_1^T J \d Y_0 \sigma_3 \e_0 >\].
1231: \end{eqnarray*}
1232:
1233: Similarly at $P_-$
1234: \begin{eqnarray*}
1235: i_{\partial_t} \R <\e^* J \delta V \wedge \delta \e> d \lambda &=&
1236: -i \phi_0 <\e_0^T J \sigma_3 Y_0 \d \hat{\e}_1> \\
1237: & +& \frac{i}{ 2 } \phi_0 \[ <\e_0^T J \d Y_0 \sigma_3 \hat{\e}_1> +
1238: < \e_1^T J \d Y_0 \sigma_3 \hat{\e}_0 >\].
1239: \end{eqnarray*}
1240: Finally, we obtain
1241: $$
1242: i_{\partial_t} \omega_0 = i \phi_0 \[ <\hat{\e}_0^T J \sigma_3 Y_0 \d {\e}_1> -
1243: <\e_0^T J \sigma_3 Y_0 \d \hat{\e}_1>\] =- \d \, 2H_1.
1244: $$
1245: \qed
1246:
1247: \subsection{Darboux coordinates}
1248:
1249: The formulas \ref{sy0}--\ref{sy1} give examples of symplectic forms. All these forms can be put in the
1250: Darboux form in the coordinates associated with poles of the Baker--Akhiezer function.
1251: \begin{lem} The formula
1252: $$
1253: \xi_0(Q)=<\e^* J\d V \wedge \d \e> d \l
1254: $$
1255: defines meromorphic in $Q$ differential 2-form on $\Gamma$ with poles
1256: at $\gamma_1,\ldots, \gamma_{g+1}$ and $P_{+}, P_{-}$.
1257: The symplectic 2-form defined by the formula
1258: $$
1259: \omega_0= \sum\limits_{P_{\pm}} \R\; \xi_0(Q)
1260: $$
1261: can be written as
1262: $$
1263: \omega_0= \frac{2}{i} \sum\limits_{k=1}^{g+1} \d p(\gamma_k) \wedge \d \l (\gamma_k).
1264: $$
1265: \end{lem}
1266: \noi
1267: {\it Remark 1.} The meaning of the right-hand side of this formula is the following. The curve $\Gamma$ (or its cover $\hat{\Gamma}$) is equipped with two meromorphic functions $\l(Q)$ and $p(Q)$. Their variation $\delta p(Q)$ and $\delta \l(Q)$ at the points of the divisor is computed for variation of the potential $\psi(x),\; \psib(x);\, -l\leq x\leq l$. The RHS of the formula is the sum of an exterior products of these variations.
1268:
1269:
1270: \noi
1271: {\it Remark 2.} In fact for a general smooth potential the divisor $\l(\g_k)$ and values of the quasimomentum $\cosh^{-1} \Delta(\l(\g_k))$ with
1272: suitably chosen sign (=sheet)
1273: determine the potential. In other words they are global coordinates on the phase space. First note that the discriminant $\Delta$ can be reconstructed from this data using Shannon interpolation, see \cite{MV}. Thus the curve $\G$ is known. The potential can be effectively recovered from the divisor via trace formulas, see \cite{MC2}.
1274:
1275:
1276:
1277: \noi
1278: {\it Proof.} Note $\d V= \d Y_0$ does not
1279: depend on $\l$. Essential singularity of the Floquet solutions at $P_{\pm}$ cancels out, and infinities are simple poles for the form
1280: $\xi_0(Q)$. In the finite part of the curve $\xi_0(Q)$ has two sets of poles. One is the poles $\gamma_1,\ldots, \gamma_{g+1}$ of the Baker-Akhiezer function. Another is the branch points of the curve $\G$.
1281: By the Cauchy theorem
1282: $$
1283: \sum\limits_{P_{\pm}} \R \; \xi_0(Q) =- \sum\limits_{\gamma_k} \R\; \xi_0(Q)- \sum\limits_{s_k} \R\; \xi_0(Q).
1284: $$
1285:
1286: Let us compute contribution of the first set of poles.
1287: Near $\gamma_k$ we have
1288: $$
1289: \e=\frac{\R \; \e}{\l-\l(\gamma_k)} +O(1).
1290: $$
1291: Therefore,
1292: $$
1293: \d \e= \frac{\R \,\e}{(\l - \l(\gamma_k))^2} \d \l(\gamma_k) +O\(\frac{1}{
1294: \l-\l(\gamma_k)}\),
1295: $$
1296: and
1297: \beq\label{var}
1298: \d \e= \frac{\e }{ \l - \l(\gamma_k)} \delta\l(\gamma_k) + O\(\frac{1}{ \l- \l(\gamma_k)}\).
1299: \eeq
1300: Note that $\e^*(x,\gamma_k)\equiv 0$ and from Lemma 2.3 we obtain
1301: $$
1302: \R_{\gamma_k}\; \xi_0(Q)= <\e^* J\d V \e> (\gamma_k)\wedge \, \d \l(\gamma_k) \; \R_{\gamma_k}\;\[ \frac{d \l }{ \l - \l(\gamma_k)}\]=
1303: i \d p(\gamma_k)\wedge \d\l (\gamma_k).
1304: $$
1305:
1306: Now consider branch points of the curve. They produce a nontrivial contribution, though the pole of $\e^*$ at the
1307: branch point is annihilated by the zero of the differential $d\l$. Nevertheless, the variation $\delta \e(x,Q)$ has a simple pole at $s_k$.
1308: First, let us make a general remark.
1309:
1310: Consider, the variation of a function $f(Q,\psi,\psib)$ under variation of the potential $\psi(x),\; \psib(x),\, -l\leq x\leq l;$ taken for $Q$ in the vicinity of the branch point $s_k$ and a fixed value of $\l$. Such variation will have a pole at the branch point itself. At the branch point $\l$ fails to be a local parameter, but $w$ is fine due to the fact $\partial_{\l} R(\l,w)|_{s_k}\neq 0$. Now, consider a function $f(Q,\psi,\psib)=f(w,\psi,\psib)$, and define its variation $\delta_0$ for a fixed value of $w$. Then,
1311: $$
1312: \delta f=\delta_0 f + \frac{d f}{d w}\; \delta w.
1313: $$
1314: Take, for example $f(Q)=\l(Q)$, then
1315: \beq\label{so}
1316: 0=\delta_0 \l +\frac{d\l}{d w} \;\delta w.
1317: \eeq
1318: Therefore, for a general $f$ we have
1319: \beq\label{sov}
1320: \delta f=\delta_0 f -\frac{d f}{d \l}\; \delta_0 \l=-\frac{d f}{d \l}\; \delta_0 \l+O(1).
1321: \eeq
1322: The zero of the differential $d\l$ at the branch point produces the pole of $\delta f$.
1323:
1324: We can proceed to the computation of the residues of $\xi_0(Q)$ at $s_k$.
1325: In the local parameter
1326: $(\l-\l(s_k))^{1/2}\sim w-1,\;\l=\l(Q)$:
1327: $$
1328: \delta\e(x,Q)=-\frac{\e_1(x)}{2}\frac{\delta\l(s_k)}{(\l-\l(s_k))^{1/2}}+\hdots,
1329: $$
1330: where
1331: $$
1332: \e(x,Q)=\e_0(x)+\e_1(x)(\l-\l(s_k))^{1/2}+\hdots,\qquad\qquad Q\in(s_k).
1333: $$
1334: Similarly,
1335: $$
1336: d\e(x,Q)=\frac{\e_1(x)}{2}\frac{d\l}{(\l-\l(s_k))^{1/2}}+\hdots,
1337: $$
1338: and we have
1339: $$
1340: \delta\e(x,Q)=-\frac{d\e(x,Q)}{d\l} \d\l(s_k)+O(1).
1341: $$
1342: The leading term is the same as in general formula \ref{sov}. Therefore, \ref{so} implies
1343: $$
1344: \R_{s_k}\; \xi_0(Q)=-\R_{s_k}\[<\e^* J\d V d\e>\]\wedge \, \d\l(s_k)=\R_{s_k}\[<\e^* J\d V d\e>\wedge \, \frac{d\l\;\d w}{d w}\] .
1345: $$
1346:
1347: Now using
1348: \bey
1349: \e^*(x,Q)&=&\e^*(l,Q)M^T(l,x,\l)^{-1},\\
1350: d\e(x,Q)&=&dM(x,-l,\l)\e(-l,Q)+M(x,-l,\l)d\e(-l,Q)
1351: \eey
1352: with the help $dM(x,y,\l)|_{s_k}=0$ we obtain,
1353: $$
1354: \R_{s_k}\; \xi_0(Q)=\R_{s_k}\[\e^*(l,Q)<M^T(l,x,\l)^{-1} J\d V M(x,-l,\l)>d\e(-l,Q)\wedge \, \frac{d\l\;\d w}{d w}\] .
1355: $$
1356: Symmetry \ref{nsy} of the monodromy matrix implies
1357: $$
1358: <M^T(l,x,\l)^{-1} J\d V M(x,-l,\l)>=\frac{1}{2l}J\d M(l,-l,\l),
1359: $$
1360: and using skew-symmetry of the wedge product
1361: \bey
1362: \R_{s_k}\; \xi_0(Q)&=&\frac{1}{2l}\R_{s_k}\[\e^*(l,Q)J\d M(l,-l,\l)d\e(-l,Q)\wedge \, \frac{d\l\;\d w}{d w}\]\\
1363: &=&\frac{1}{2l}\R_{s_k}\[\e^*(l,Q)J(\d M(l,-l,\l)-\d w)d\e(-l,Q)\wedge \, \frac{d\l\;\d w}{d w}\].
1364: \eey
1365: Identities \ref{mp} and \ref{mps} imply
1366: \bey
1367: \e^*J(\d M-\d w)&=&\d\e^*J(w- M),\\
1368: J( w-M)d\e&=&J( dM- dw)\e.
1369: \eey
1370: Therefore,
1371: $$
1372: \R_{s_k}\; \xi_0(Q)=\frac{1}{2l}\R_{s_k}\[\d\e^*(l)J(dM(l,-l,\l)-dw)\e(-l)\wedge\; \frac{d\l\;\d w}{d w}\] .
1373: $$
1374: Since $\e^*(l)J\e(-l)=w^{-1}$, we have
1375: $$
1376: \R_{s_k}\; \xi_0(Q)=\frac{1}{2l}\R_{s_k}\[\e^*(l)J\d \e(-l)\wedge\,\d w\, d\l\].
1377: $$
1378: The one form
1379: $$
1380: \e^*(l,Q)J\d \e(-l,Q)\wedge\,\d w(Q)\, d\l(Q)
1381: $$
1382: is holomorphic (in the parameter $\l$) outside of the poles $\g_k$ and the branch points. At infinity the essential singularity cancels out and
1383: due to \ref{asf}-\ref{assf}, Lemmas \ref{lfloc} and \ref{lsfloq}
1384: $$
1385: e^*(l,Q)J\d \e(-l,Q)\wedge\,\d w(Q)=o\(\frac{1}{\l}\).
1386: $$
1387: This implies
1388: $$
1389: \R_{P_{\pm}}\[\e^*(l,Q)J\d \e(-l,Q)\wedge\,\d w(Q)\, d\l(Q)\] =0.
1390: $$
1391: By the Cauchy theorem,
1392: $$
1393: \sum_{s_k}\R_{s_k}\[\e^*(l)J\d \e(-l)\wedge\; d\l\;\d w\]=-\sum_{\gamma_k}\R_{\gamma_k}\[\e^*(l)J\d \e(-l)\wedge\; d\l\;\d w\]
1394: $$
1395: Therefore, using $\e^*(l)=w^{-1}\e^*(-l)$, we have
1396: \bey
1397: \sum_{s_k}\R_{s_k}\; \xi_0(Q)&=&-\frac{1}{2l}\sum_{\gamma_k}\R_{\gamma_k}\[\e^*(l)J\d \e(-l)\wedge\; d\l\;\d w\]\\
1398: &=&-\frac{1}{2l}\sum_{\gamma_k}\R_{\gamma_k}\[\e^*(-l)J\d \e(-l)\wedge\; d\l\;\frac{\d w}{w}\]
1399: \eey
1400: Using the formula \ref{var} we finally obtain
1401: \bey
1402: \sum_{s_k}\R_{s_k}\; \xi_0(Q)&=&\sum_{\gamma_k}i\d p(\gamma_k)\wedge \R_{\gamma_k}\[\e^*(-l)J\d \e(-l) d\l\]\\
1403: &=&\sum_{\gamma_k}i\d p(\gamma_k)\wedge \d \l(\gamma_k)\R_{\gamma_k}\[ \frac{d\l}{\l-\l(\gamma_k)}\]\\
1404: &=&\sum_{\gamma_k}i\d p(\gamma_k)\wedge \d \l(\gamma_k).
1405: \eey
1406: \qed
1407:
1408: \noi
1409: {\it Remark.} For the higher symplectic structures an analogous result holds
1410: $$
1411: \omega_n=\frac{2}{i} \sum\limits_{k=1}^{g+1} \l^n \d p(\gamma_k) \wedge \d \l (\gamma_k),
1412: \quad \quad \quad \quad n=0, 1 \ldots;
1413: $$
1414: subject to the constrains
1415: $H_k=const,\; k=1,\ldots,n$.
1416: \subsection{Action--angle variables}
1417: Here we describe briefly another system of Darboux coordinates. We refer to the paper \cite{MV} for details.
1418:
1419: The actions $I_k,\;k=1,\hdots, g+1;$ are defined by the formula
1420: $$
1421: I_k=\frac{1}{4\pi} \int_{a_k} p(\l) d\l.
1422: $$
1423: In the formula above the multivalued function $p(\l)$ is normalized such that $p(s_k)=0,\;k=1,\hdots, g+1$.
1424: The angles $\t_k,\;n=1,\hdots, g+1;$ are
1425: $$
1426: \t_k=\sum_{n=1}^{g+1} \int_{s_n}^{\gamma_n} \alpha_k.
1427: $$
1428: The differentials $\alpha_k,\;n=1,\hdots, g+1;$ are of the third kind with poles at the infinities $P_{\pm}$
1429: normalized in such a way that\footnote{$\d_n^k$ is Kronecker delta.}
1430: $$
1431: \int_{a_k}\alpha_n=2 \pi \d_n^k.
1432: $$
1433: As it is proved in \cite{MV} by a direct computation
1434: $$
1435: \{\t_n,I_k\}=\d_n^k.
1436: $$
1437: All other brackets vanish
1438: $$
1439: \{I_n,I_k\}=0,\qquad\qquad \{\t_n,\t_k\}=0.
1440: $$
1441: These formulas imply the identity for symplectic forms.
1442: $$
1443: \omega_0=2 \sum_{n=1}^{g+1} \d I_n\wedge \d \t_n.
1444: $$
1445:
1446:
1447: We conclude this section with a few remarks.
1448:
1449: \noi
1450: {\it Remark 1.}
1451: Another way to prove the identity for symplectic forms without employing the Poisson bracket is found by Krichever, \cite{K2}.
1452:
1453: \noi
1454: {\it Remark 2.} McKean, \cite{MC1}, proved various identities for 1--forms.
1455:
1456:
1457: \noi
1458: {\it Remark 3.} In the finite--gap case the curve $\G$ is specified by $2(g+1)$ branch points.
1459: The actions $I_k,\;k=1,\hdots,g+1;$ together with the $g$ other periods (see \ref{pc}) of the differential $dp$ and the
1460: constant $p_0^{+}$ determine the curve. This fact is due to Krichever, \cite{K3} (see also \cite{BK}).
1461: In the infinite--gap case it is shown in \cite{MV} that $I-\t$'s are global coordinates on the phase space for a general square integrable potential.
1462: This property holds for a finite gap potential as well.
1463:
1464:
1465:
1466:
1467:
1468: \section{The Scattering Case}
1469: \subsection{Jost solutions}
1470: In the next sections we consider the scattering problem for the Dirac operator on the entire line with
1471: rapidly decaying potential.
1472: The Riemann-Hilbert approach to the scattering theory for canonical systems
1473: with summable potential was constructed by M.G. Krein and P.E. Melik--Adamian,
1474: \cite{K, KMA, MA}. This approach was used many times in soliton theory, \cite{FT}.
1475:
1476:
1477: To simplify the estimates we assume that the potential $\psi$ is from the Schwartz'
1478: space $S(\RB)$ of complex rapidly decreasing infinitely differentiable functions on the line such that
1479: $$
1480: \sup_{x} |(1+x^2)^n \, \psi^{(m)}(x)|\, < \infty\qquad\qquad\qquad\qquad m,\,n=0,1,\ldots.
1481: $$
1482: Let us introduce the reduced transition matrix $T(x,y,\l),\, x\geq y;$ by the formula
1483: \bay\label{red}
1484: T(x,y,\l)=E^{-1}\(\frac{\l x}{ 2}\) M(x,y,\l) E^{-1}\(-\frac{\l y}{ 2}\),
1485: \ey
1486: where $E(\frac{\l x}{ 2})= \exp{(-\frac{i\lambda x}{ 2} \sigma_3)}$ is a solution of
1487: the free equation ($\psi\equiv 0$). The matrix $T(x,y,\l)$ solves the equation
1488: $$
1489: T'(x,y,\l) = Y_0(x) E(\l x) T(x,y,\l), \qquad\qquad
1490: \quad T(y,y,\lambda)= I.
1491: $$
1492: The spectral parameter enters multiplicatively into the RHS of the differential equation.
1493: The solution is given by the formula
1494: \bay\label{exp}
1495: T(x,y,\lambda)= \exp \, \int_{y}^x Y_0(\xi) E(\l\xi) d\xi.
1496: \ey
1497: The symmetry of the matrix
1498: $Y_0:\, \sigma_1 Y_0(x)\sigma_1= \overline{Y_0(x)}$
1499: is inherited by unimodular matrix
1500: T:$$\; \sigma_1T(x,y,\overline{\lambda})\sigma_1=
1501: \overline{T(x,y,\lambda)}.$$
1502: For real $\l$ the formula \ref{exp} and the rapid decay of the potential imply an existence of the limit
1503: $$
1504: T(\lambda)= \lim T(x,y,\lambda)= \(\begin{array}{ccccc}
1505: a(\lambda) & \overline{b}(\lambda)\\
1506: b(\lambda) & \overline{a}(\lambda) \end{array} \), \quad
1507: {\rm when} \quad y \rightarrow -\infty \quad\text{and}\quad x
1508: \rightarrow +\infty;
1509: $$
1510: and $|a(\lambda)|^2- |b(\lambda)|^2=1.$
1511: When the potential $\psi\in S(\RB)$, we have $b(\l) \in S(\RB)$.
1512:
1513: We introduce {\it Jost solutions} $J_{\pm}(x,\lambda)$ as a matrix solutions of the differential equation
1514: $$
1515: J_{\pm}'(x,\lambda)=V(x,\lambda)J_{\pm}(x,\lambda),\quad J_{\pm}(x,\l)=
1516: E\(\frac{\lambda x}{ 2}\) + o(1),\quad {\rm when}\quad x\rightarrow \pm \infty.
1517: $$
1518: An existence and analytic properties of the Jost solutions follow from the integral
1519: representations
1520: $$
1521: J_+(x,\lambda)=E\(\frac{x\lambda}{ 2}\)+ \int_x^{+\infty} \Gamma_+(x,\xi)
1522: E\(\frac{\lambda \xi}{ 2}\) d\xi,
1523: $$
1524: $$
1525: J_-(x,\lambda)=E\(\frac{x\lambda}{ 2}\)+ \int_{-\infty}^x \Gamma_-(x,\xi)
1526: E\(\frac{\lambda \xi}{ 2}\) d\xi.
1527: $$
1528: The kernels $\Gamma_{\pm}$ are unique and infinitely smooth in both variables.
1529: Introducing the notation $J_\pm=\[\jo^{(1)}_\pm,\, \jo^{(2)}_\pm\]$ we see from the
1530: integral representations that
1531: $\jo^{(1)}_-(x,\l),$ $ \jo^{(2)}_+(x,\l)$ are analytic in $\l$ in the upper half-plane
1532: and continuous up to the boundary. Also, the columns
1533: $\jo^{(2)}_-(x,\l),\;\; \jo^{(1)}_+(x,\l)$
1534: are analytic in the lower half--plane and continuous up to the boundary.
1535:
1536:
1537:
1538:
1539:
1540: Now we describe analytic properties of the coefficient $a(\l)$ of the matrix $T(\l)$.
1541: The monodromy matrix $M(x,y,\l)$ can be written in the form
1542: $$
1543: M(x,y, \l)=J_+(x)J^{-1}_{+}(y)= J_{-}(x)J^{-1}_{-}(y).
1544: $$
1545: Therefore,
1546: $$
1547: J_{+}^{-1}(x)\, M(x,y,\l)\, J_{-}(y)= J^{-1}_{+}(y)J_{-}(y)=J_{+}^{-1}(x)J_{-}(x).
1548: $$
1549: The variables $x$ and $y$ separate and the above expression does not depend on $x$ or $y$ at all.
1550: By passing to the limit with
1551: $ x\rightarrow +\infty,\, y \rightarrow -\infty$ we have
1552: $$
1553: T(\l)=J^{-1}_{+}(y)J_{-}(y)=J_{+}^{-1}(x)J_{-}(x).
1554: $$
1555: Therefore,
1556: $$a(\l)= {\jo_-^{(1)}}^T(\l)J \jo_+^{(2)}(\l).
1557: $$
1558: The properties of Jost solutions imply that
1559: \begin{itemize}
1560: \item $a(\lambda)$ is analytic in the upper half-plane and continuous up to the boundary;
1561:
1562: \item $a(\lambda)$ is root-free;
1563: \item $|a(\lambda)| \geq 1$
1564: and $|a(\lambda)|^2-1 \in S(\RB)$ for $\l$ real, $a(\lambda) = 1 +o(1)$ as $ |\lambda| \longrightarrow \infty$.
1565: \end{itemize}
1566: This coefficient will be used to construct the scattering curve $\Gin$.
1567:
1568: Let $p_\infty(\lambda)$ be such that $a(\lambda)= \exp(-i2 p_{\infty}(\lambda))$
1569: for $\lambda$ in the upper half-plane. The quantity $\pin$ in analogous to the quasimomentum studied in the periodic case, see \cite{V1}\footnote{The
1570: spectral parameter $\l$ in the paper \cite{V1} has different scaling and should
1571: be replaced by $\lt$.}. From the properties of $a(\lambda)$
1572: it follow that $ \pin $ is analytic in the upper half-plane and
1573: continuous up to the boundary; $\Im \pin \geq 0$ for $ \Im \lambda \geq 0$;
1574: $\pin=o(1)$ for $|\lambda| \rightarrow \infty$;
1575: for real $\lambda$, the density of the measure $d \mi= \Im \pin \, d \lambda$ belongs to
1576: $S(\RB)$. The function $\pin$ can be written in the form
1577: $$
1578: \pin=\frac{1}{ \pi}\int \frac{d \mu_{\infty}(t)}{ t -\lambda}.
1579: $$
1580: Expanding the denominator in inverse powers of $\lambda$, we obtain:
1581: \bay\label{V}
1582: \pin= -\sum^{\infty}_{k=0}\frac{1}{ \lambda^{k+1}} \frac{1}{ \pi}\int_{-\infty}^{+
1583: \infty} t^k d \mu_{\infty}(t)= -\frac{ H_1}{ \lambda}- \frac{ H_2}{ \lambda^{2}}-
1584: \frac{ H_3}{ \lambda^3}+ \dots .
1585: \ey
1586: where $H_1,H_2$ and $H_3$ are the integrals introduced above with $l=+\infty$.
1587: The expansion has an asymptotic character for $\lambda: \; \delta \leq
1588: \arg \lambda \leq \pi - \delta,\;\delta > 0$.
1589:
1590: To describe the asymptotic behavior in $x$ of the Jost solutions $\jo_+^{(2)}(x,\l)$ and $\jo_-^{(1)}(x,\l)$ we assume that $\l$ is real and fixed. Then,
1591: $$
1592: \begin{array}{ccccccccccccccccc}
1593: &\phantom{kkkkkkkkk} &x \rightarrow - \infty & &x\rightarrow + \infty \\
1594: &\jo_+^{(2)} &a(\l) \f_{\rightarrow}(x,\l)-\bar{b}(\l)\f_{\leftarrow}(x,\l)& &
1595: \quad \f_{\rightarrow}(x,\l)\\
1596: &\jo_-^{(1)} &\f_{\leftarrow}(x,\l) &\phantom{h} &
1597: a(\l)\f_{\leftarrow}(x,\l)+
1598: b(\l) \f_{\rightarrow}(x,\l),&
1599: \end{array}
1600: $$
1601: where
1602: $$
1603: \f_{\leftarrow}(x,\l)=\left[ \begin{array}{ccccccccccccccccc}
1604: e^{- i\lt x}\\
1605: 0\end{array} \right], \quad
1606: \f_{\rightarrow}(x,\l)=\left[ \begin{array}{ccccccccccccccccc} 0\\
1607: e^{i\lt x}\end{array} \right]
1608: $$ are solutions of the free equation.
1609:
1610:
1611:
1612: We sketch the derivation of the asymptotics for $\jo^{(2)}_+(x,\l)$, when $x\rightarrow -\infty$.
1613: Let $\psi$ be a potential that vanishes outside the segment $[-L, +L]$.
1614: In this case formula \ref{exp} becomes
1615: $$
1616: T(\lambda)= \exp
1617: \, \int_{-L}^L Y_0(\xi) E(\l\xi) d\xi,
1618: $$
1619: and $T(\l)$ is an entire unimodular function of $\l$ of the form
1620: $$
1621: T(\l)= \[\begin{array}{cccccccc} a(\l) & \overline{b}(\overline{\l})\\
1622: b(\l) & \overline{a}(\overline{\l}) \end{array} \].
1623: $$
1624: From the definition of the matrix $M$:
1625: \bay\label{for}
1626: \jo^{(2)}_+(L,\l)=M(L,-L,\l) \jo^{(2)}_+(-L,\l),
1627: \ey
1628: and from \ref{red}
1629: $$
1630: M(L,-L,\l)=\[\begin{array}{cccccccc} a(\l) e^{-i\l L} & \overline{b}(\overline{\l}) \\
1631: b(\l) & \overline{a}(\overline{\l}) e^{i\l L} \end{array} \].
1632: $$
1633: Obviously, $\jo^{(2)}_+(L,\l)= \fr(L,\l)$ and
1634: $\jo^{(2)}_+ (-L,\l)=c_1 \fr(-L,\l) + c_2 \fl (-L,\l)$ with unknown coefficients $c_1$ and $c_2$. Formula \ref{for} for real $\l$
1635: leads to the linear system
1636: $$
1637: \begin{array}{cccccccc}
1638: c_1 \overline{b}({\l}) + c_2 a(\l) & =0\\
1639: c_1 \overline{a}({\l}) + c_2 b(\l)& =1.
1640: \end{array}
1641: $$
1642: Solving for $c$'s we obtain the stated formula. For a potential with noncompact
1643: support one has to take $L$ sufficiently large to make the error negligible.
1644:
1645: The Riemann surface $\Gin$ is obtained by gluing together along the real line
1646: two copies of the complex plane (see Figure 2). One copy we call "+" and another "-".
1647: Each copy has an infinity $P_{+}$ or $P_{-}$. The point $Q\in \Gin$ is determined by $\l=\l(Q)$
1648: and specification of the sheet $Q=(\l,\pm)$. Let us define for the "+" copy
1649: $\jo(x,Q)$ to be $\jo^{(2)}_+(x,\l)$ if $ \Im \l > 0;$ and
1650: $\jo^{(2)}_-(x,\l)$ if $\Im \l < 0$. For the "-" copy we define $\jo(x,Q)$ to be
1651: $\jo^{(1)}_-(x,\l)$ if $\Im \l > 0;$ and $\jo^{(1)}_+(x,\l)$ if $\Im \l < 0$.
1652: In the vicinity of $P_{\pm}$ the function $\jo(x,Q)$ has asymptotics,
1653: \bay\label{J}
1654: \jo(x,Q) = e^{\pm i\frac{\l}{ 2}x} \[\jo_0/\hat \jo_0 + o\(1\)\],
1655: \ey
1656: where $ \l=\l(Q)$ and
1657: $$
1658: \jo_0= \[\begin{array}{ccc} 0\\1\end{array}\]\qquad\qquad\qquad
1659: \hat \jo_0= \[\begin{array}{ccc} 1\\0\end{array}\].
1660: $$
1661: Therefore, $\jo(x,Q)$ can be viewed as a BA function for the singular curve $\Gin$.
1662:
1663:
1664: We also introduce the matrix BA function
1665: $$
1666: H_+(\l)=\[ \jo_-^{(1)}(\l),\jo_+^{(2)}(\l)\] \quad {\rm and}\quad\quad H_-(\l)=\[ \jo_+^{(1)}(\l),\jo_-^{(2)} (\l) \]
1667: $$ analytic in the upper/lower half-plane respectively. They are connected by the gluing condition
1668: \bay\label{RH}
1669: H_-(x, \l)= H_+(x,\l) S(\l),\qquad\qquad\qquad\text{where}\qquad\l \in \RB
1670: \ey
1671: and the scattering matrix $S(\l)$
1672: $$
1673: S(\l) = \frac{1}{ a} \[\begin{array}{ccccc} \;\; {1} & {\overline{b}}\\
1674: -{b}& {1}
1675: \end{array}\].
1676: $$
1677:
1678:
1679:
1680:
1681:
1682: The adjoint (dual) Jost solution $\jo^+$ at the point $Q$ is defined by the formula
1683: $$
1684: \jo^+(x,Q)\equiv \jo(x,Q)^T.
1685: $$
1686: Any Jost solution satisfies $\[J\partial_x- J V\]\jo=0$.
1687: By analogy with the periodic case one can prove that $\jo^+$ satisfies
1688: $\jo^+\[J\partial_x- J V\]=0.$
1689:
1690:
1691:
1692:
1693: The matrices $H_+^+$ and $H_-^+$ are defined as
1694: $$
1695: H_+^+(\l)=\sigma_1 H_+^T(\l)= \[\begin{array}{ccccccc} \jo_+^{(2)\,T}\\
1696: \jo_-^{(1)\,T}
1697: \end{array}\], \qquad\qquad
1698: H_-^+(\l)=\sigma_1 H_-^T(\l)= \[\begin{array}{ccccccc} \jo_-^{(2)\,T}\\
1699: \jo_+^{(1)\,T}
1700: \end{array}\].
1701: $$
1702: Extending $a$ into the lower half-plane by the formula $a^*(\l)= \overline{a(\bar \l)}$, we define
1703: $$
1704: H_+^*(\l)= -\frac{\sigma_3}{ a(\l)} H_+^+(\l)\qquad \qquad \qquad\qquad \text{for}\qquad \Im \l >0;
1705: $$
1706: and
1707: $$
1708: \qquad H_-^*(\l)= -\frac{\sigma_3}{ a^*(\l)} H_-^+(\l)\qquad \qquad \qquad \text{for}\qquad \Im \l < 0.
1709: $$
1710: It is easy to check that the dual gluing condition holds
1711: \bay\label{ARH}
1712: H_-^*(x,\l) =S^{-1}(\l) H_+^*(x,\l),\qquad\qquad\qquad{\rm where} \qquad \l\in \RB
1713: \ey
1714: and
1715: $$ S^{-1}(\l)= \frac{1}{ a^*}
1716: \[\begin{array}{ccccc} \;\; {1} & -{\overline{b}}\\
1717: \;\;{b}& {1}
1718: \end{array}\].
1719: $$
1720:
1721:
1722:
1723:
1724: Next two lemmas state asymptotic properties of Jost solutions which will be used in computations with symplectic forms.
1725: \begin{lem} (i) For fixed $x$ the following formulas hold
1726: $$
1727: \jo_+^{(2)}(x,\l)=e^{+i \lt x} \sum\limits_{s=0}^{\infty} \jo_s(x) \l^{-s} =
1728: e^{+i \lt x} \sum\limits_{s=0}^{\infty} \[\begin{array}{cccc}
1729: g_s \\
1730: k_s
1731: \end{array} \] \l^{-s},
1732: $$
1733: where $g_0=0, \quad k_0=1$, and\footnote{Operation $\hat{\phantom
1734: 0}$ applied to a scalar signifies complex
1735: conjugation and reversal of infinities, see formulas in the part (ii) below.}
1736: $$
1737: \jo_-^{(1)}(x,\l)=e^{-i \lt x} \sum\limits_{s=0}^{\infty} \hat \jo_s(x) \l^{-s}
1738: = e^{-i \lt x} \sum\limits_{s=0}^{\infty} \[\begin{array}{cccc}
1739: \hat k_s \\
1740: \hat g_s
1741: \end{array} \] \l^{-s} ,
1742: $$
1743: where $\hat g_0=0,\quad \hat k_0=1$. The expansion has an asymptotic character for
1744: $\l\;:\quad \d \leq \arg \l \leq \pi -\d,\;\d >0$.
1745:
1746: (ii) The coefficients $g_1, \; k_1$ are given by the formulas
1747: $$
1748: g_1= -i \psib,\quad \quad \quad \quad
1749: k_1= i \int_{x}^{+ \infty} |\psi(x')|^2 dx'
1750: $$
1751: and
1752: $$
1753: \hat g_1= i \psi, \quad \quad \quad \quad
1754: \hat k_1 = i \int_{-\infty}^{x} |\psi(x')|^2 dx'.
1755: $$
1756: \end{lem}
1757:
1758:
1759: {\it Proof} {\it (i).} Using\footnote{$\Gamma=\[\Gamma^{(1)},
1760: \Gamma^{(2)}\]$.} $\partial_\xi^n \Gamma_+^{(2)}(x,\xi)|_{\xi=\infty}= 0$, for
1761: $n=0,1,...$ and integrating $n$ times by parts,
1762: \bay
1763: \jo^{(2)}_+(x,\l)& = & e^{\frac{i\l x}{ 2}} \[\begin{array}{ccccc} 0\\ 1 \end{array} \] +
1764: \int_{x}^{\infty} \Gamma_+^{(2)} (x,\xi) e^{\frac{i\l \xi}{ 2}} d\xi \nonumber \\
1765: \dots & =& e^\frac{i\l x}{ 2} \[\begin{array}{ccccc} 0\\ 1 \end{array} \] -
1766: \frac{ e^\frac{i\l x}{ 2} }{ \(\frac{i\l}{ 2}\)} \Gamma^{(2)}_+ (x,x) +
1767: \frac{ e^\frac{i\l x}{ 2} }{ \(\frac{i\l}{ 2}\)^2} \partial_\xi \Gamma^{(2)}_+ (x,x)
1768: - \dots \nonumber \\
1769: & & \quad +(-1)^n \frac{ e^\frac{i\l x}{ 2} }{\(\frac{i\l}{ 2}\)^n} \partial_\xi^{n-1} \Gamma^{(2)}_+ (x,x)
1770: %& & \quad +(-1)^n { e^\frac{i\l x}{ 2} }
1771: %\(\frac{i\l}{ 2}\)^n} \partial_\xi^{n-1} \Gamma^{(2)} (x,x)\\
1772: + (-1)^n \frac{1}{ \(\frac{i\l}{ 2}\)^n} \int_{x}^{\infty} \partial_\xi^{n}
1773: \Gamma_+^{(2)} (x,\xi) e^\frac{i\l \xi}{ 2} d\xi. \nonumber
1774: \ey
1775: This implies the existence and asymptotic character of the expansion in the parameter
1776: $\l$. The other infinity can be treated similarly.
1777:
1778: {\it (ii).} Consider $\jo^{(2)}_+(x,\l)$ first. The differential equation $ \jo'=V\jo$ implies
1779: $$
1780: -\[\begin{array}{ccccc}
1781: g_s'\\
1782: k_s'
1783: \end{array}
1784: \] + Y_{0}
1785: \[\begin{array}{ccccc}
1786: g_s\\
1787: k_s
1788: \end{array}
1789: \]= \frac{i}{ 2} \(I + \sigma_3 \)
1790: \[\begin{array}{ccccc}
1791: g_{s+1}\\
1792: k_{s+1}
1793: \end{array}
1794: \], \quad\quad s=0,1,\ldots;
1795: $$
1796: and $g_0=0,\quad k_0=1$.
1797:
1798:
1799: This recurrent relation leads to the identities
1800: \bay\nonumber
1801: - g_s'+ \psib k_s &=& ig_{s+1}, \\
1802: - k_s' +\psi g_s &=& 0. \nonumber
1803: \ey
1804: For $s\geq 1$ we have the boundary condition
1805: $$
1806: g_s(x)|_{x=+ \infty}=k_s(x)|_{x=+ \infty}=0.
1807: $$
1808: These imply the stated formulas for $g_1,\, k_1$.
1809:
1810: For $\jo^{(1)}_-(x,\l)$ the differential equation implies
1811: $$
1812: -\[\begin{array}{ccccc}
1813: k_s'\\
1814: g_s'
1815: \end{array} \] + Y_{0}
1816: \[\begin{array}{ccccc}
1817: k_s\\
1818: g_s
1819: \end{array}
1820: \]= \frac{i}{ 2} \(- I + \sigma_3 \)
1821: \[\begin{array}{ccccc}
1822: k_{s+1}\\
1823: g_{s+1}
1824: \end{array}
1825: \], \quad\quad s=0,1,\ldots;
1826: $$
1827: and $k_0=1,\quad g_0=0$.
1828:
1829: The recurrent relation produces the identities
1830: \bay
1831: - g_s'+ \psi k_s&= &- ig_{s+1}, \nonumber\\
1832: - k_s' +\psib g_s&=&0. \nonumber
1833: \ey
1834: For $s\geq 1$ we have the boundary condition
1835: $$
1836: g_s(x)|_{x=- \infty}=k_s(x)|_{x=- \infty}=0.
1837: $$
1838: These imply the stated formulas for $\hat g_1,\, \hat k_1$. We are done. \qed
1839:
1840:
1841: \noi
1842: {\it Remark.} It is interesting to compare asymptotic expansions for $\jo_-^{(1)}/ \jo_+^{(2)}$ and
1843: $\e(x,Q)$. For the Jost solution $\jo_{-}^{(1)}(x,\l)$ normalized at the left
1844: $$
1845: \jo_{-}^{(1)}(x,\l)= e^{ -\frac{i\l x}{ 2} } \( \[\begin{array}{ccccc} 1\\0 \end{array} \] + \frac{1}{ \l}
1846: \[ \begin{array}{ccccc} i\int_{-\infty}^x |\psi|^2\\ i\psi \end{array}
1847: \] + \dots \)
1848: $$
1849: and
1850: $$
1851: e^{i\lt l} \e(x,Q)= e^{- \frac{i\l x}{ 2}} \( \[\begin{array}{ccccc} 1\\0 \end{array}
1852: \] + \frac{1}{ \l}
1853: \[ \begin{array}{ccccc} -i \psi_0 + i\int_{-l}^x |\psi|^2\\ i\psi \end{array}
1854: \] + \dots \), \quad \quad Q\in(P_-).
1855: $$
1856: If $\psi$ is compactly supported and $l$ becomes sufficiently large, then
1857: $e^{i\lt l} \e(x,Q)=\jo(x,\l)$.
1858: For the Jost solution normalized at the right the situation is slightly different. If one defines the {\it new}
1859: $\e(x,Q)\equiv \e(x,Q) w^{-1} (Q)$ which is the Floquet solution
1860: normalized at the right end
1861: $x=l$ of the interval, then for compactly supported $\psi$ and sufficiently large $l$
1862: the {\it new} $e^{-i\lt l} \e(x,Q)= \jo_+^{(2)}(x,\l)$.
1863:
1864:
1865: A result similar to Lemma 4.1 holds for Jost solutions analytic in the lower half-plane.
1866: \begin{lem} (i) For fixed $x$ the following formulas hold
1867: $$
1868: \jo_+^{(1)}(x,\l)=e^{-i \lt x} \sum\limits_{s=0}^{\infty} \jo_s(x) \l^{-s} =
1869: e^{-i \lt x} \sum\limits_{s=0}^{\infty} \[\begin{array}{cccc}
1870: h_s \\
1871: f_s
1872: \end{array} \] \l^{-s},
1873: $$
1874: where $h_0=1, \quad f_0=0$, and
1875: $$
1876: \jo_-^{(2)}(x,\l)=e^{+i \lt x} \sum\limits_{s=0}^{\infty} \hat \jo_s(x) \l^{-s}
1877: = e^{+i \lt x} \sum\limits_{s=0}^{\infty} \[\begin{array}{cccc}
1878: \hat f_s \\
1879: \hat h_s
1880: \end{array} \] \l^{-s} ,
1881: $$
1882: where $\hat h_0=1,\quad \hat f_0=0$. The expansion has an asymptotic character for
1883: $ \l\;:\quad- \d \geq \arg \l \geq -\pi + \d,\; \d >0$.
1884:
1885: (ii) The coefficients $h_1, \; f_1$ are given by the formulas
1886: $$
1887: f_1= i \psi,\quad \quad \quad \quad
1888: h_1= - i \int_{x}^{+ \infty} |\psi(x')|^2 dx'
1889: $$
1890: and
1891: $$
1892: \hat f_1= - i \psib, \quad \quad \quad \quad
1893: \hat h_1 =- i \int_{-\infty}^{x} |\psi(x')|^2 dx'.
1894: $$
1895: \end{lem}
1896:
1897:
1898:
1899:
1900: Similar to the periodic case we will need time dependent BA functions. They are obtained by an
1901: elementary construction.
1902:
1903:
1904:
1905: \begin{lem} \cite{I}.
1906: There exists the Jost solution $\jo(\tau,x,t, Q)$ on the curve $\Gin$ with three time parameters $\tau,\, x $ and $t$
1907: which satisfies the differential equations:
1908: $$
1909: [\partial_{\tau}-V_1(\tau,x,t)] \jo(\tau,x,t, Q) = 0,
1910: $$
1911: $$
1912: [\partial_{x}-V_2(\tau,x,t)] \jo(\tau,x,t, Q) = 0,
1913: $$
1914: $$
1915: [\partial_{t}-V_3(\tau,x,t)] \jo(\tau,x,t, Q) =0.
1916: $$
1917: \end{lem}
1918:
1919:
1920: \noi
1921: {\it Proof.}
1922: We consider "+" sheet and the upper half-plane where $\jo(x,Q)=\jo_+^{(2)}(x,\l)$.
1923: First, we construct $\jo(\tau,x,t, Q)$ such that
1924: $
1925: [\partial_{x}-V_2(\tau,x,t)] \; \jo(\tau,x,t, Q) =0
1926: $
1927: normalized for all $\tau$ and $t$ as
1928: $$
1929: \jo(\tau,x,t, Q) \sim e^{+i \lt x} \( \[\begin{array}{cccc}
1930: 0 \\ 1
1931: \end{array} \] + o(1) \), \quad \quad \quad {\rm when} \quad x\rightarrow \infty.
1932: $$
1933:
1934: Then, we construct $\jo(\tau,x,t, \epsilon_\pm Q)$ on the lower sheet as a solution
1935: $$
1936: [\partial_{x}-V_2(\tau,x,t)] \; \jo(\tau,x,t, \epsilon_\pm Q) =0
1937: $$
1938: for the same value of the spectral parameter $\l=\l(Q)$ normalized for all $\tau$ and $t$ as
1939: $$
1940: \jo(\tau,x,t, \epsilon_\pm Q) \sim e^{-i \lt x} \( \[\begin{array}{cccc}
1941: 1 \\ 0
1942: \end{array} \] + o(1) \), \quad \quad \quad {\rm when} \quad x\rightarrow -\infty.
1943: $$
1944: The solutions $\jo(\tau,x,t, Q)$ and $\jo(\tau,x,t, \epsilon_\pm Q)$ span the kernel of the operator $[\partial_{x}-V_2(\tau,x,t)]$.
1945: Now we introduce
1946: $$
1947: \jo_{new}(\tau,x,t, Q)\equiv e^{-i\frac{1}{ 2} \tau - i \frac{\l^2}{ 2} t} \jo(\tau,x,t, Q),
1948: $$
1949: which is the desired solution.
1950: Evidently,
1951: $$
1952: \[ \partial_{x} -V_2(\tau,x,t)\] \jo_{new}(\tau,x,t,Q)= 0.
1953: $$
1954:
1955:
1956:
1957: To prove the first identity of the statement we note, that
1958: commutativity of the operators $\partial_{\tau} - V_1$ and $\partial_{x} - V_2$ implies
1959: $$
1960: \[ \partial_{\tau} -V_1(\tau,x,t)\] \jo_{new}(\tau,x,t,Q)= c_1(\tau, t) \jo(\tau,x,t,Q) +
1961: c_2(\tau, t) \jo(\tau,x,t,\epsilon_{\pm}Q).
1962: $$
1963: From another side as $x\rightarrow + \infty$,
1964: $$
1965: \[ \partial_{\tau} -V_1(\tau,x,t)\] \jo_{new}(\tau,x,t,Q)= \[\partial_{\tau}- \frac{i}{2}\sigma_3\] e^{-i\frac{1}{ 2} \tau +
1966: i \frac{\l}{ 2}x - i \frac{\l^2}{ 2} t} \( \[\begin{array}{cccc}
1967: 0 \\
1968: 1
1969: \end{array} \] + o(1) \) =o(1).
1970: $$
1971: Due to the linear independence of the solutions $ \jo(\tau,x,t, Q)$ and
1972: $\jo(\tau,x,t,\epsilon_{\pm}Q)$ we have
1973: $c_1(\tau, t) =c_2(\tau, t) =0$.
1974:
1975: Similarly it can be proved that
1976: $$
1977: \[ \partial_{t} -V_3(\tau,x,t)\] \jo_{new}(\tau,x,t,Q)=0.
1978: $$
1979: Another sheet of $\Gin$ can be treated the same way. We are done. \qed
1980:
1981: \noi
1982: {\it Remark.}
1983: It is easy to see that,
1984: $$
1985: \jo_{new}(\tau,x,t, Q) = e^{\pm i \( -\frac{1}{ 2} \tau + \frac{\l}{ 2} x - \frac{\l^2}{ 2}t\)}
1986: \[ \jo_0/ \hat \jo_0 + o(1)\]\qquad\qquad\qquad Q\in (P_{\pm}).
1987: $$
1988: Thus the standard Jost solution with asymtotics \ref{J} can be obtained from the BA functions if
1989: one puts $\tau$ and $t$ equal to 0.
1990:
1991:
1992:
1993:
1994:
1995: \subsection{ The symplectic structures } We are ready to introduce the scattering version of the Krichever--Phong formula.
1996: The everaging is defined now as an integral over the entire line $<\bullet>=\int_{-\infty}^{+\infty}dx$.
1997:
1998: \begin{thm} The formula
1999: \bay\label{SY}
2000: \omega_0=
2001: \tr\, \R\; \frac{1}{ 2} \[<H_+^* J\delta V \wedge \delta H_+> +
2002: <H_-^* J\delta V \wedge \delta H_->\]
2003: d \lambda,
2004: \ey
2005: defines a closed 2--form $\omega_0$ on the space of operators
2006: $\partial_x - V_2$ with potential from the Schwartz class $S(\RB)$.
2007: The flows $e^{tX_m},\; m=1,2,\ldots$ on the space of operators defined by the
2008: formula
2009: $$
2010: [\partial_{\tau_m} - V_m, \partial_x -V_2]=0,
2011: $$
2012: are Hamiltonian with respect to the 2--form $\omega_0$ with Hamiltonian
2013: function $H_m$ (up to a non-essential constant factor).
2014: \end{thm}
2015:
2016:
2017: \noi
2018: {\it Remark 1.} The symbol
2019: $$
2020: \R\; <H_+^* J\delta V \wedge \delta H_+>
2021: $$
2022: means the coefficient corresponding to the term $\frac{1}{\l}$ in the power series expansion near infinity in the
2023: upper half--plane. The second term
2024: $$
2025: \R\; <H_-^* J\delta V \wedge \delta H_->
2026: $$
2027: is defined in the same way, only the upper half--plane plane is replaced with the lower half--plane.
2028:
2029: \noi
2030: {\it Remark 2.} The formula
2031: $$
2032: \omega_n=\tr\; \R\; \frac{\l^n}{ 2}\[ <H_+^* J\delta V \wedge \delta H_+> +
2033: <H_-^* J\delta V \wedge \delta H_->\]
2034: d \lambda,
2035: $$
2036: where $n=0, 1 \ldots,$ defines a closed 2--forms $\omega_n$ on the space of operators
2037: $\partial_x - V_2$ with potential from the Schwartz class $S(\RB)$ that satisfy the constraints
2038: $H_k=const,\; k=1,\ldots,n$.
2039: It is instructive to compute explicitly the symplectic forms for small $n$.
2040: The first few are given by the formulas
2041: $$
2042: \omega_0 = 2i< \d\overline{\psi}\wedge \d \psi>,
2043: $$
2044: and
2045: $$
2046: \omega_1 = <\d \psi \wedge \d \psib' + \d \psib \wedge \d \psi' +
2047: \delta \[\int_{-\infty}^x |\psi|^2 - \int^{\infty}_x |\psi|^2\]
2048: \wedge \d |\psi|^2>.
2049: $$
2050: subject to the constrain $H_1=const$. The derivation employs Lemmas 3.1-3.2 and similar to the
2051: periodic case.
2052:
2053:
2054:
2055:
2056:
2057:
2058:
2059: \noi
2060: {\it Proof. } Closeness of the form $\omega_0$ follows either from the explicit formula or from the result of the next theorem.
2061: The proof of the second statement we present for the first $e^{tX_1}$ flow. The time dependent Jost solutions entering into \ref{TT} are constructed in Lemma 3.4. Let $i_{\partial_t}$ be the construction operator produced by the vector field $X_1$. We will prove
2062: $i_{\partial_t}\omega_0=-\d 2H_1$. For the first term in \ref{SY}
2063: \begin{eqnarray}
2064: \tr & &\R\, < H_+^* J\delta V \wedge \delta H_+> \, d\l \nonumber \\
2065: & & = \R \frac{1}{ a} < \jo_-^{(1)T} J \d V \wedge \d \jo_+^{(2)}> d\l
2066: - \R \frac{1}{ a} < \jo_+^{(2)T} J \d V \wedge \d \jo_-^{(1)}> d\l \label{TT}
2067: \end{eqnarray}
2068: Applying the contraction operator to the first term in \ref{TT}
2069:
2070:
2071:
2072: \begin{eqnarray}
2073: i_{\partial_t}\, \R \;\frac{1}{ a}& < &\jo_-^{(1)T} J \d V \wedge \d \jo_+^{(2)}> d\l \nonumber \\
2074: & & = \R\; \frac{1}{ a} < \jo_-^{(1)T} J V^{\bullet} \d \jo_+^{(2)}> d\l
2075: - \R \;\frac{1}{ a} < \jo_-^{(1)T} J \d V \jo_+^{(2)\bullet}> d\l. \nonumber
2076: \end{eqnarray}
2077: From \ref{V} we have
2078: $$
2079: \frac{1}{ a}= a_0+\frac{a_1}{ \l}+\ldots, \qquad\qquad{\rm where} \qquad a_0=1, \qquad a_1=-i \int_{-\infty}^{+\infty} |\psi|^2.
2080: $$
2081: Using $V^{\bullet}=\[ \frac{i}{2} \sigma_3, V\]=i \sigma_3 V$, and Lemma 3.1, we have
2082: $$
2083: \R\; \frac{1}{ a} < \jo_-^{(1)T} J V^{\bullet} \d \jo_+^{(2)}> d\l
2084: =a_0< \hat \jo_0^T J i \sigma_3 V \d \jo_1>= -<\psi \d \psib>.
2085: $$
2086: Similarly, using $\jo^{\bullet} = \frac{i}{ 2} \sigma_3 \jo$, we have
2087: \begin{eqnarray*}
2088: \R\; \frac{1}{ a} < \jo_-^{(1)T} J \d V \jo_+^{(2)\bullet}> d\l
2089: & = & a_1<\hat \jo_0^T J \d V \frac{i}{ 2}\sigma_3 \jo_0> \\
2090: & + & a_0\[ <\hat \jo_0^T J \d V \frac{i}{ 2}\sigma_3 \jo_1> +
2091: <\hat \jo_1^T J \d V \frac{i}{ 2}\sigma_3 \jo_0> \].
2092: \end{eqnarray*}
2093: The first term vanishes, the second produces
2094: $$
2095: = \frac{1}{ 2} <\psib \d \psi - \psi \d \psib >.
2096: $$
2097: Finally,
2098: $$
2099: i_{\partial_t} \R \frac{1}{ a} < \jo_-^{(1)T} J \d V \wedge \d \jo_+^{(2)}> d\l \nonumber = - \d H_1.
2100: $$
2101: The second term in formula \ref{TT} can be treated similarly. Therefore,
2102: $$
2103: i_{\partial_t} \tr\, \R \; <H_+^* J\delta V \wedge \delta H_+> = -2 \d H_1.
2104: $$
2105: The second term in formula \ref{SY} produces the same result.
2106: The proof is finished. \qed
2107:
2108:
2109:
2110:
2111:
2112:
2113:
2114:
2115:
2116:
2117:
2118: \subsection{Action--angle variables}
2119:
2120:
2121: In response to infinitesimal deformations of the matrix
2122: $\tilde{V}= V + \d V $ the matrix $T(\l)$
2123: changes according to the rule: $\tilde{T}(\l)=T(\l) + \d T(\l) + \hdots$.
2124: The next result is similar to Lemma 2.3 of the periodic case.
2125: \begin{lem} The following formula holds
2126: $$
2127: < H_+^+ J\d V H_+>=\[\begin{array}{ccccccc}
2128: -\d a & \overline{b}\d a - a \d \overline{b} \\
2129: a\d b - b \d a & - \d a \end{array}\],
2130: $$
2131: with averaging defined as
2132: $$
2133: < H_+^+ J\d V H_+> =\int_{-\infty}^{+\infty}H_+^+(x,\l) J\d V(x) H_+(x,\l)\; dx.
2134: $$
2135: \end{lem}
2136:
2137: \noi
2138: {\it Proof.} Let us assume, first, that $\psi$ has compact support.
2139: We denote by $\tilde V,\; \tilde T$ and $\tilde \jo$ deformed matrices
2140: $V, T$ and the Jost solution $\jo$. We will derive the expression for $<\jo_+^{(2)\,T} J\d V \jo_{+}^{(1)}>$
2141: in the left-upper corner. First, we obtain the formula
2142: \bay\label{FF}
2143: < \jo^+ J\delta V \jo> + \;\text{lower order terms} = \left.\jo^+ J \tilde \jo \right|_{-L}^{+L}.
2144: \ey
2145: Indeed,
2146: \begin{eqnarray*}
2147: \jo^+ \([J\partial_x -J \tilde V]\tilde \jo\)& =& 0,\\
2148: \(\jo^+ [J\partial_x -J V]\)\tilde \jo& =& 0.
2149: \end{eqnarray*}
2150: Subtracting one identity from another we have
2151: $$
2152: \jo^+ J\delta V \jo + \;\text{lower order terms}= \jo^+\(J\partial \tilde \jo\) -\( \jo^+ J \partial\) \tilde \jo .
2153: $$
2154: Integrating the RHS in x variable we obtain
2155: $$
2156: \int_{-L}^{+L} \[ \jo^+ J \tilde \jo' + \jo^{+\prime} J \tilde \jo \] \, dx =\left.\jo^+ J \tilde \jo
2157: \right|_{-L}^{+L}.
2158: $$
2159: This implies \ref{FF}.
2160: Now using formulas for the asymptotics of $\jo^{(2)}_+$ and $\jo^{(1)}_-$ for the RHS of \ref{FF}, we have
2161: \begin{eqnarray*}
2162: \jo^{(2)T}_+ J \tilde \jo^{(1)}_-\left. \right|_{-L}^{+L}&=& \fr^TJ\[ \tilde a \fl +
2163: \tilde b \fr\]\left. \right|^{+L} -
2164: \[a \fr^T - \overline{b} \fl^T \]J \fl \left. \right|_{-L}\\
2165: &=& \tilde{a} \fr^T J \fl - a \fr^T J \fl= a-\tilde{a}=- \d a +\text{lower order terms}.
2166: \end{eqnarray*}
2167: Collecting terms of the same order, we obtain the
2168: result. The case of a potential with non-compact support
2169: can be considered using approximation arguments. For other entries the arguments are the same.
2170: Lemma is proved. \qed
2171:
2172:
2173:
2174:
2175:
2176:
2177: \begin{thm} The following formulas hold
2178: $$
2179: \omega_0= \frac{1}{ \pi i} \int\limits_{-\infty}^{+\infty}
2180: \frac{\d \bar b(\l)\wedge \d b(\l)}{ |a(\l)|^2} \, d\l.
2181: $$
2182: \end{thm}
2183: \noi
2184: {\it Proof.} By the Cauchy integral formula
2185: $$
2186: \frac{1}{ 2} \tr\; \R <H_+^* J\delta V \wedge \delta H_+>= -\frac{1}{ 2\pi i}
2187: \int\limits_{-\infty}^{+\infty} \tr <H_+^* J\delta V \wedge \delta H_+> \, d\l,
2188: $$
2189: and
2190: $$
2191: \frac{1}{ 2} \tr\; \R <H_-^* J\delta V \wedge \delta H_->= \frac{1}{ 2\pi i}
2192: \int\limits_{-\infty}^{+\infty} \tr <H_-^* J\delta V \wedge \delta H_-> \, d\l.
2193: $$
2194: Taking sum
2195: $$
2196: \omega_0= \frac{1}{ 2\pi i}
2197: \int\limits_{-\infty}^{+\infty} \tr <H_-^* J\delta V \wedge \delta H_-> \, d\l -
2198: \frac{1}{ 2\pi i}
2199: \int\limits_{-\infty}^{+\infty} \tr <H_+^* J\delta V \wedge \delta H_+> \, d\l.
2200: $$
2201: Using \ref{RH}, \ref{ARH}
2202: $$
2203: \d H_-=\d H_+ S +H_+\d S,
2204: $$
2205: we obtain
2206: \begin{eqnarray*}
2207: \omega_0& =& \frac{1}{ 2\pi i}
2208: \int\limits_{-\infty}^{+\infty} \tr\; S^{-1} <H_+^* J\delta V \wedge \delta H_+> S \; \, d\l
2209: +\int\limits_{-\infty}^{+\infty} \tr\; S^{-1} <H_+^* J\delta V H_+> \wedge \d S \, \, d\l\\
2210: & &-\frac{1}{ 2\pi i}\int\limits_{-\infty}^{+\infty}\tr <H_+^* J\delta V\wedge \delta H_+> \,d\l\\
2211: &=& \int\limits_{-\infty}^{+\infty} \tr <H_+^* J\delta V H_+> \wedge \d S S^{-1} \, d\l.
2212: \end{eqnarray*}
2213: Now, applying the result of Lemma 3.5, we have
2214: $$
2215: \omega_0= \frac{1}{ 2\pi i} \int\limits_{-\infty}^{+\infty} \tr\;- \frac{ \sigma_3}{ a^*}
2216: \[\begin{array}{ccccccc}
2217: -\d a & \overline{b}\d a - a \d \overline{b} \\
2218: a\d b - b \d a & - \d a \end{array}\]
2219: \wedge \d S\, S^{-1} \, d\l.
2220: $$
2221: After simple algebra we arrive at the stated identity. Theorem is proved. \qed
2222:
2223: \noi
2224: {\it Remark 1.} The formula of Theorem can be put easily into more familiar form using the identities
2225: $$
2226: |a|^2-|b|^2=1,\qquad\qquad\qquad \d \log |b|^2=\frac{2|a|\d|a|}{|a|^2-1}.
2227: $$
2228: Indeed,
2229: \bey
2230: \frac{1}{ i} \frac{\d \bar b(\l)\wedge \d b(\l)}{ |a(\l)|^2}
2231: &=& \frac{|b(\l)|^2 \d \log | b(\l)|^2 \wedge \d {\rm ph}\, b(\l)}{ |a(\l)|^2} \\
2232: &=& 2 \d \log|a(\l)|\wedge \d\, {\rm ph}\, b(\l).
2233: \eey
2234: Therefore,
2235: \bey
2236: \omega_0= \frac{1}{ \pi i} \int\limits_{-\infty}^{+\infty}
2237: \frac{\d \bar b(\l)\wedge \d b(\l)}{ |a(\l)|^2}= \frac{2}{ \pi} \int\limits_{-\infty}^{+\infty} \d \log|a(\l)|\wedge \d\, {\rm ph}\, b(\l)
2238: \; d\l.
2239: \eey
2240:
2241:
2242: \noi
2243: {\it Remark 2.} The formula
2244: $$
2245: \omega_n= \frac{1}{ \pi i} \int\limits_{-\infty}^{+\infty}
2246: \frac{\d \bar b(\l)\wedge \d b(\l)}{ |a(\l)|^2} \l^n\, d\l, \qquad\qquad\qquad n=1,2,\ldots;
2247: $$
2248: subject to the constrains $H_k=const,\; k=1,\ldots,n$; gives Darboux coordinates for higher symplectic forms.
2249:
2250:
2251:
2252:
2253:
2254:
2255:
2256: \begin{thebibliography}{***}
2257:
2258: \bibitem{A} V.I. Arnold {\it Mathematical methods of classical mechanics}.
2259: Springer-Verlag, 1978.
2260:
2261:
2262: \bibitem{BK} R.F. Bikbaev and S.B. Kuksin
2263: {\it On the parametrization of finite--gap solutions by frequency vector and wavenumber vector and a theorem of I. Krichever}
2264: Lett. Math. Phys. 166, 1994, pp. 115-122.
2265:
2266: \bibitem{DKP} Eric D'Hooker, I.M. Krichever and D.H Phong
2267: {\it Seiberg-Witten Theory, Symplectic Forms, and Hamiltonian Theory of Solitons},
2268: Beijing and Hangzhou 2002; preprint hep-th 0212313
2269:
2270: \bibitem{FT}
2271: L.D. Faddeev and L.A. Takhtadzian
2272: {\it Hamiltonian Methods in the Theory of Solitons.} Springer-Verlag, 1987.
2273:
2274: \bibitem{FZ} L.D. Faddeev and V.E. Zakharov {\it Korteweg-de Vries equation: A completely
2275: integrable Hamiltonian system.} Funk. Analiz and its Appl. Vol 5, 1971, 18--27.
2276:
2277:
2278: \bibitem{FM} H. Flashka and D. McLaughlin
2279: {\it Canonically conjugate variables for the Korteweg--de Vries equation and
2280: the Toda lattice with the periodic boundary conditions} Progr. Theoret. Phys.
2281: 55 (1976) no 2, 438--456.
2282:
2283: \bibitem{I} A.R. Its {\it The Liouville theorem and the inverse scattering method.}
2284: Journ. of Soviet Math. 31 (1985), 6, 299--334.
2285:
2286: \bibitem{K} M.G. Krein {\it On the theory of accelerant and $S$--matrices of canonical
2287: differential equations.} Dokl. Akad. Nauk. SSSR 111 1956, 1167--1170.
2288:
2289:
2290:
2291:
2292: \bibitem{KMA} M.G. Krein and P.E. Melik-Adamian {\it A contribution to the theory of
2293: $S$--matrices of canonical differential equations with summable potential.} Dokl. Akad. Nauk.
2294: Armjan. SSR 46 1968, 150--155.
2295:
2296:
2297:
2298:
2299: \bibitem{K1} I. M. Krichever
2300: {\it Methods of algebraic geometry in the theory of nonlinear equations}
2301: Uspehi Mat. Nauk 32, (1977), no 6, 183--208.
2302:
2303:
2304: \bibitem{K2} I. M. Krichever {\it Elliptic solutions to the difference
2305: nonlinear equations and nested Bethe ansatz equations},
2306: Calogero--Moser--Sutherland models (Montreal, QC,1997) CRM Ser. Math. Phys.
2307: Spinger, New York, 2000.
2308:
2309: \bibitem{K3} I. M. Krichever {\it Perturbation Theory in Periodic Problems for two Dimensional Integrable Systems}
2310: Sov. Sci. Rev. C. Math. Phys. Vol. 9, 1992, pp 1--103.
2311:
2312:
2313:
2314:
2315:
2316: \bibitem{KP} I.M. Krichever and D.H Phong
2317: {\it Symplectic forms in the theory of solitons},
2318: Surveys in Differential Geometry: Integrable Systems, v.4 (1999),
2319: International Press, Boston, 1999; preprint hep-th 9708170
2320:
2321:
2322:
2323: \bibitem{KR} P.P. Kulish and A.G. Reiman {\it Hierarchy of Symplectic Forms for the Schrodinger and Dirac
2324: Equations on a line.} Zap. LOMI, vol 77, 134-147, 1978.
2325:
2326: \bibitem{SNK} V.B. Kuznetzov, F.W. Nijhoff, E.K. Sklyanin
2327: {\it Separation of variables for the Ruijsenaars system} CMP, 189 (1997), no 3, 855-877.
2328:
2329: \bibitem{MA} P.E. Melik-Adamian {\it On the properties of $S$--matrices of canonical differential
2330: equations on the entire line.} Dokl. Akad. Nauk. Armjan. SSR 58 1974, 199-205
2331:
2332: \bibitem{MC1} H.P. McKean {\it Trace Formulas and the Canonical 1--form}. Algebraic Aspects of Integrable Systems.
2333: Progr. Nonlinear Differential Equations, 26; Birkhauser Boston, MA, 1997, 217-235.
2334:
2335:
2336:
2337: \bibitem{MC2} H.P. McKean {\it A Novel Aspect of Action--Angle Variables.}
2338: Proceedings of Symp in Pure Mathematics. vol 65, 1999, 155-160.
2339:
2340: \bibitem{MV} H.P. McKean and K.L. Vaninsky {\it Action-angle variables for the Cubic Schrodinger Equation.}
2341: CPAM, vol L, 1997, 489--562.
2342:
2343: \bibitem{MO} J. Moser {\it Integrable Hamiltonian systems and spectral theory}
2344: Lezioni Fermiane. Scuola Normale Superiore, Piza, 1983;
2345:
2346: \bibitem{NMP} S.P. Novikov, S.V. Manakov, L.P. Pitaevsky
2347: {\it Theory of Solitons. The inverse scattering method.} Trans from Russian,
2348: Consultants Bureau [Plenum], New York, 1984, 276 pp.
2349:
2350:
2351:
2352:
2353:
2354: \bibitem{NV} S.P. Novikov and A.P. Veselov
2355: {\it Poisson brackets and complex tori.} Trudy Math, Inst Steklov, 165 (1984),
2356: 49--61.
2357:
2358:
2359: \bibitem{SW1} N. Seiberg and E. Witten
2360: {\it Electro-magnetic duality, monopole condensation,
2361: and confinement in $N=2$ supersymmetric Yang-Mills theory},
2362: Nucl. Phys. B 426 (1994) 19-53, hep-th/9407087;
2363:
2364:
2365:
2366: \bibitem{SW2}
2367: N. Seiberg and E. Witten,
2368: {\it Monopoles, duality, and chiral symmetry breaking
2369: in $N=2$ supersymmetric QCD},
2370: Nucl. Phys. B 431 (1994) 494, hep-th/9410167.
2371:
2372:
2373:
2374: \bibitem{S} E.K. Sklyanin
2375: {\it Separation of variables -- new trends.} Quantum field theory,
2376: integrable models and beyond (Kyoto, 1994). Progr Theor Phys. Suppl.
2377: n 118 (1995) 35--60.
2378:
2379: \bibitem{V1} K.L. Vaninsky {\it A convexity theorem in the Scattering theory
2380: for the Dirac operator.} TAMS, 350, 1998, 1895-1911.
2381:
2382: \bibitem{V2} K.L. Vaninsky {\it The scattering divisor for the cubic Schr\"{o}dinger equation.}
2383: paper in preparation.
2384:
2385:
2386: \bibitem{ZM}
2387: V.E. Zakharov and S.V. Manakov {\it On the complete integrability of a nonlinear
2388: Schrodinger Equation.} Teor. and Math Phys, vol 19, 332-343, 1974.
2389:
2390:
2391:
2392: \end{thebibliography}
2393:
2394:
2395:
2396:
2397:
2398:
2399:
2400:
2401:
2402:
2403:
2404:
2405:
2406:
2407:
2408:
2409:
2410:
2411:
2412:
2413:
2414:
2415:
2416: \
2417:
2418:
2419:
2420:
2421:
2422:
2423:
2424:
2425:
2426:
2427:
2428:
2429:
2430:
2431:
2432:
2433:
2434:
2435:
2436:
2437:
2438:
2439:
2440: \end{document}
2441:
2442:
2443:
2444:
2445:
2446:
2447:
2448:
2449:
2450:
2451:
2452:
2453:
2454:
2455:
2456:
2457:
2458:
2459:
2460:
2461:
2462:
2463:
2464:
2465:
2466:
2467:
2468:
2469:
2470:
2471:
2472:
2473:
2474:
2475:
2476:
2477:
2478:
2479:
2480:
2481:
2482:
2483:
2484:
2485:
2486:
2487:
2488:
2489:
2490:
2491:
2492:
2493:
2494:
2495:
2496:
2497:
2498:
2499:
2500:
2501:
2502:
2503:
2504:
2505:
2506:
2507:
2508:
2509:
2510:
2511:
2512:
2513:
2514:
2515:
2516:
2517:
2518:
2519:
2520:
2521:
2522:
2523:
2524:
2525:
2526:
2527:
2528:
2529:
2530:
2531:
2532:
2533:
2534:
2535:
2536:
2537:
2538:
2539:
2540:
2541:
2542:
2543:
2544:
2545:
2546:
2547:
2548:
2549:
2550:
2551:
2552:
2553:
2554:
2555:
2556:
2557:
2558:
2559:
2560:
2561:
2562:
2563:
2564:
2565:
2566:
2567:
2568:
2569:
2570:
2571:
2572:
2573:
2574:
2575:
2576:
2577:
2578:
2579:
2580:
2581:
2582:
2583:
2584:
2585:
2586:
2587:
2588:
2589:
2590:
2591:
2592:
2593:
2594:
2595:
2596:
2597:
2598:
2599:
2600:
2601:
2602:
2603:
2604:
2605:
2606:
2607:
2608:
2609:
2610:
2611:
2612:
2613:
2614:
2615:
2616:
2617:
2618:
2619:
2620:
2621:
2622:
2623:
2624:
2625:
2626:
2627:
2628:
2629:
2630:
2631:
2632:
2633:
2634:
2635:
2636:
2637:
2638:
2639:
2640:
2641:
2642:
2643:
2644:
2645:
2646:
2647:
2648:
2649:
2650:
2651:
2652:
2653:
2654:
2655:
2656:
2657:
2658:
2659:
2660:
2661:
2662:
2663:
2664:
2665:
2666:
2667:
2668:
2669:
2670:
2671:
2672:
2673:
2674:
2675:
2676:
2677:
2678:
2679:
2680:
2681:
2682:
2683:
2684:
2685:
2686:
2687:
2688:
2689:
2690:
2691:
2692:
2693:
2694:
2695:
2696:
2697:
2698:
2699:
2700:
2701:
2702:
2703:
2704:
2705:
2706:
2707:
2708:
2709:
2710:
2711:
2712:
2713:
2714:
2715:
2716:
2717:
2718:
2719:
2720:
2721:
2722:
2723:
2724:
2725:
2726:
2727:
2728:
2729:
2730:
2731:
2732:
2733:
2734:
2735:
2736:
2737:
2738:
2739:
2740:
2741:
2742:
2743:
2744:
2745:
2746:
2747:
2748:
2749:
2750:
2751:
2752:
2753:
2754:
2755:
2756:
2757:
2758:
2759:
2760:
2761:
2762:
2763:
2764:
2765:
2766:
2767:
2768:
2769:
2770:
2771:
2772:
2773:
2774:
2775:
2776:
2777:
2778:
2779:
2780:
2781:
2782:
2783:
2784:
2785:
2786:
2787:
2788:
2789:
2790:
2791:
2792:
2793:
2794:
2795:
2796:
2797:
2798:
2799:
2800:
2801:
2802:
2803:
2804:
2805:
2806:
2807:
2808:
2809:
2810:
2811:
2812:
2813:
2814:
2815:
2816:
2817:
2818:
2819:
2820:
2821:
2822:
2823:
2824:
2825:
2826:
2827:
2828:
2829:
2830:
2831:
2832:
2833:
2834:
2835:
2836:
2837:
2838:
2839:
2840:
2841:
2842:
2843:
2844:
2845:
2846:
2847:
2848:
2849:
2850:
2851:
2852:
2853:
2854:
2855:
2856:
2857:
2858:
2859:
2860:
2861:
2862:
2863:
2864:
2865:
2866:
2867:
2868:
2869:
2870:
2871:
2872:
2873:
2874:
2875:
2876:
2877:
2878:
2879:
2880:
2881:
2882:
2883:
2884:
2885:
2886:
2887:
2888:
2889:
2890:
2891:
2892:
2893:
2894:
2895:
2896:
2897:
2898:
2899:
2900:
2901:
2902:
2903:
2904:
2905:
2906:
2907:
2908:
2909:
2910:
2911:
2912:
2913:
2914:
2915:
2916:
2917:
2918:
2919:
2920:
2921:
2922:
2923:
2924:
2925:
2926:
2927:
2928:
2929:
2930:
2931:
2932:
2933:
2934:
2935:
2936:
2937:
2938:
2939:
2940:
2941:
2942:
2943:
2944:
2945:
2946:
2947:
2948:
2949:
2950:
2951:
2952:
2953:
2954:
2955:
2956:
2957:
2958:
2959:
2960:
2961:
2962:
2963:
2964:
2965:
2966:
2967:
2968:
2969:
2970:
2971:
2972:
2973:
2974:
2975:
2976:
2977:
2978:
2979:
2980:
2981:
2982:
2983:
2984:
2985:
2986:
2987:
2988:
2989:
2990:
2991:
2992:
2993:
2994:
2995:
2996:
2997:
2998:
2999:
3000:
3001:
3002:
3003:
3004:
3005:
3006:
3007:
3008:
3009:
3010:
3011:
3012:
3013:
3014:
3015:
3016:
3017:
3018:
3019:
3020:
3021:
3022:
3023:
3024:
3025:
3026:
3027:
3028:
3029:
3030:
3031:
3032:
3033:
3034:
3035:
3036:
3037:
3038:
3039:
3040:
3041:
3042:
3043:
3044:
3045:
3046:
3047:
3048:
3049:
3050:
3051:
3052:
3053:
3054:
3055:
3056:
3057:
3058:
3059:
3060:
3061:
3062:
3063:
3064:
3065:
3066:
3067:
3068:
3069:
3070:
3071:
3072:
3073:
3074:
3075:
3076:
3077:
3078:
3079:
3080:
3081:
3082:
3083:
3084:
3085:
3086:
3087:
3088:
3089:
3090:
3091:
3092:
3093:
3094:
3095:
3096:
3097:
3098:
3099:
3100:
3101:
3102:
3103:
3104:
3105:
3106:
3107:
3108:
3109:
3110:
3111:
3112:
3113:
3114:
3115:
3116:
3117:
3118:
3119:
3120:
3121:
3122:
3123:
3124:
3125:
3126:
3127:
3128:
3129:
3130:
3131:
3132:
3133:
3134:
3135:
3136:
3137:
3138:
3139:
3140:
3141:
3142:
3143:
3144:
3145:
3146:
3147:
3148:
3149:
3150:
3151:
3152:
3153:
3154:
3155:
3156:
3157:
3158:
3159:
3160:
3161:
3162:
3163:
3164:
3165:
3166:
3167:
3168:
3169:
3170:
3171:
3172:
3173:
3174:
3175:
3176:
3177:
3178:
3179:
3180:
3181:
3182:
3183:
3184:
3185:
3186:
3187:
3188:
3189:
3190:
3191:
3192:
3193:
3194:
3195:
3196:
3197:
3198:
3199:
3200:
3201:
3202:
3203:
3204:
3205:
3206:
3207:
3208:
3209:
3210:
3211:
3212:
3213:
3214:
3215:
3216:
3217:
3218:
3219:
3220:
3221:
3222:
3223:
3224:
3225:
3226:
3227:
3228:
3229:
3230:
3231:
3232:
3233:
3234:
3235:
3236:
3237:
3238:
3239:
3240:
3241:
3242:
3243:
3244:
3245:
3246:
3247:
3248:
3249:
3250:
3251:
3252:
3253:
3254:
3255:
3256:
3257:
3258:
3259:
3260:
3261:
3262:
3263:
3264:
3265:
3266:
3267:
3268:
3269:
3270:
3271:
3272:
3273:
3274:
3275:
3276:
3277:
3278:
3279:
3280:
3281:
3282:
3283:
3284:
3285:
3286:
3287:
3288:
3289:
3290:
3291:
3292:
3293:
3294:
3295:
3296:
3297:
3298:
3299:
3300:
3301:
3302:
3303:
3304:
3305:
3306:
3307:
3308:
3309:
3310:
3311:
3312:
3313:
3314:
3315:
3316:
3317:
3318:
3319:
3320:
3321:
3322:
3323:
3324:
3325:
3326:
3327:
3328:
3329:
3330:
3331:
3332:
3333:
3334:
3335:
3336:
3337:
3338:
3339:
3340:
3341:
3342:
3343:
3344:
3345:
3346:
3347:
3348:
3349:
3350:
3351:
3352:
3353:
3354:
3355:
3356:
3357:
3358:
3359:
3360:
3361:
3362:
3363:
3364:
3365:
3366:
3367:
3368:
3369:
3370:
3371:
3372:
3373:
3374:
3375:
3376:
3377:
3378:
3379:
3380:
3381:
3382:
3383:
3384:
3385:
3386:
3387:
3388:
3389:
3390:
3391:
3392:
3393:
3394:
3395:
3396:
3397:
3398:
3399:
3400:
3401:
3402:
3403:
3404:
3405:
3406:
3407:
3408:
3409:
3410:
3411:
3412:
3413:
3414:
3415:
3416:
3417:
3418:
3419:
3420:
3421:
3422:
3423:
3424:
3425:
3426:
3427:
3428:
3429:
3430:
3431:
3432:
3433:
3434:
3435:
3436:
3437:
3438:
3439:
3440:
3441:
3442:
3443:
3444:
3445:
3446:
3447:
3448:
3449:
3450:
3451:
3452:
3453:
3454:
3455:
3456:
3457:
3458:
3459:
3460:
3461:
3462:
3463:
3464:
3465:
3466:
3467:
3468:
3469:
3470:
3471:
3472:
3473:
3474:
3475:
3476:
3477:
3478:
3479:
3480:
3481:
3482:
3483:
3484:
3485:
3486:
3487:
3488:
3489:
3490:
3491:
3492:
3493:
3494:
3495:
3496:
3497:
3498:
3499:
3500:
3501:
3502:
3503:
3504:
3505:
3506:
3507:
3508:
3509:
3510:
3511:
3512:
3513:
3514:
3515:
3516:
3517:
3518:
3519:
3520:
3521:
3522:
3523:
3524:
3525:
3526:
3527:
3528:
3529:
3530:
3531:
3532:
3533:
3534:
3535:
3536:
3537:
3538:
3539:
3540:
3541:
3542:
3543:
3544:
3545:
3546:
3547:
3548:
3549:
3550:
3551:
3552:
3553:
3554:
3555:
3556:
3557:
3558:
3559:
3560:
3561:
3562:
3563:
3564:
3565:
3566:
3567:
3568:
3569:
3570:
3571:
3572:
3573:
3574:
3575:
3576:
3577:
3578:
3579:
3580:
3581:
3582:
3583:
3584:
3585:
3586:
3587:
3588:
3589:
3590:
3591:
3592:
3593:
3594:
3595:
3596:
3597:
3598:
3599:
3600:
3601:
3602:
3603:
3604:
3605:
3606:
3607:
3608:
3609:
3610:
3611:
3612:
3613:
3614:
3615:
3616:
3617:
3618:
3619:
3620:
3621:
3622:
3623:
3624:
3625:
3626:
3627:
3628:
3629:
3630:
3631:
3632:
3633:
3634:
3635:
3636:
3637:
3638:
3639:
3640:
3641:
3642:
3643:
3644:
3645:
3646:
3647:
3648:
3649:
3650:
3651:
3652:
3653:
3654:
3655:
3656:
3657:
3658:
3659:
3660:
3661:
3662:
3663:
3664:
3665:
3666:
3667:
3668:
3669:
3670:
3671:
3672:
3673:
3674:
3675:
3676:
3677:
3678:
3679:
3680:
3681:
3682:
3683:
3684:
3685:
3686:
3687:
3688:
3689:
3690:
3691:
3692:
3693:
3694:
3695:
3696:
3697:
3698:
3699:
3700:
3701:
3702:
3703:
3704:
3705:
3706:
3707:
3708:
3709:
3710:
3711:
3712:
3713:
3714:
3715:
3716:
3717:
3718:
3719:
3720:
3721:
3722:
3723:
3724:
3725:
3726:
3727:
3728:
3729:
3730:
3731:
3732:
3733:
3734:
3735:
3736:
3737:
3738:
3739:
3740:
3741:
3742:
3743:
3744:
3745:
3746:
3747:
3748:
3749:
3750:
3751:
3752:
3753:
3754:
3755:
3756:
3757:
3758:
3759:
3760:
3761:
3762:
3763:
3764:
3765:
3766:
3767:
3768:
3769:
3770:
3771:
3772:
3773:
3774:
3775:
3776:
3777:
3778:
3779:
3780:
3781:
3782:
3783:
3784:
3785:
3786:
3787:
3788:
3789:
3790:
3791:
3792:
3793:
3794:
3795:
3796:
3797:
3798:
3799:
3800:
3801:
3802:
3803:
3804:
3805:
3806:
3807:
3808:
3809:
3810:
3811:
3812:
3813:
3814:
3815:
3816:
3817:
3818:
3819:
3820:
3821:
3822:
3823:
3824:
3825:
3826:
3827:
3828:
3829:
3830:
3831:
3832:
3833:
3834:
3835:
3836:
3837:
3838:
3839:
3840:
3841:
3842:
3843:
3844:
3845:
3846:
3847:
3848:
3849:
3850:
3851:
3852:
3853:
3854:
3855:
3856:
3857:
3858:
3859:
3860:
3861:
3862:
3863:
3864:
3865:
3866:
3867:
3868:
3869:
3870:
3871:
3872:
3873:
3874:
3875:
3876:
3877:
3878:
3879:
3880:
3881:
3882:
3883:
3884:
3885:
3886:
3887:
3888:
3889:
3890:
3891:
3892:
3893:
3894:
3895:
3896:
3897:
3898:
3899:
3900:
3901:
3902:
3903:
3904:
3905:
3906:
3907:
3908:
3909:
3910:
3911:
3912:
3913:
3914:
3915:
3916:
3917:
3918:
3919:
3920:
3921:
3922:
3923:
3924:
3925:
3926:
3927:
3928:
3929:
3930:
3931:
3932:
3933:
3934:
3935:
3936:
3937:
3938:
3939:
3940:
3941:
3942:
3943:
3944:
3945:
3946:
3947:
3948:
3949:
3950:
3951:
3952:
3953:
3954:
3955:
3956:
3957:
3958:
3959:
3960:
3961:
3962:
3963:
3964:
3965:
3966:
3967:
3968:
3969:
3970:
3971:
3972:
3973:
3974:
3975:
3976:
3977:
3978:
3979:
3980:
3981:
3982:
3983:
3984:
3985:
3986:
3987:
3988:
3989:
3990:
3991:
3992:
3993:
3994:
3995:
3996:
3997:
3998:
3999:
4000:
4001:
4002:
4003:
4004:
4005:
4006:
4007:
4008:
4009:
4010:
4011:
4012:
4013:
4014:
4015:
4016:
4017:
4018:
4019:
4020:
4021:
4022:
4023:
4024:
4025:
4026:
4027:
4028:
4029:
4030:
4031:
4032:
4033:
4034:
4035:
4036:
4037:
4038:
4039:
4040:
4041:
4042:
4043:
4044:
4045:
4046:
4047:
4048:
4049:
4050:
4051:
4052:
4053:
4054:
4055:
4056:
4057:
4058:
4059:
4060:
4061:
4062:
4063:
4064:
4065:
4066:
4067:
4068:
4069:
4070:
4071:
4072:
4073:
4074:
4075:
4076:
4077:
4078:
4079:
4080:
4081:
4082:
4083:
4084:
4085:
4086:
4087:
4088:
4089:
4090:
4091:
4092:
4093:
4094:
4095:
4096:
4097:
4098:
4099:
4100:
4101:
4102:
4103:
4104:
4105:
4106:
4107:
4108:
4109:
4110:
4111:
4112:
4113:
4114:
4115:
4116:
4117:
4118:
4119:
4120:
4121:
4122:
4123:
4124:
4125:
4126:
4127:
4128:
4129:
4130:
4131:
4132:
4133:
4134:
4135:
4136:
4137:
4138:
4139:
4140:
4141:
4142:
4143:
4144:
4145:
4146:
4147:
4148:
4149:
4150:
4151:
4152:
4153:
4154:
4155:
4156:
4157:
4158:
4159:
4160:
4161:
4162:
4163:
4164:
4165:
4166:
4167:
4168:
4169:
4170:
4171:
4172:
4173:
4174:
4175:
4176:
4177:
4178:
4179:
4180:
4181:
4182:
4183:
4184:
4185:
4186:
4187:
4188:
4189:
4190:
4191:
4192:
4193:
4194:
4195:
4196:
4197:
4198:
4199:
4200:
4201:
4202:
4203:
4204:
4205:
4206:
4207:
4208:
4209:
4210:
4211:
4212:
4213:
4214:
4215:
4216:
4217:
4218:
4219:
4220:
4221:
4222:
4223:
4224:
4225:
4226:
4227:
4228:
4229:
4230:
4231:
4232:
4233:
4234:
4235:
4236:
4237:
4238:
4239:
4240:
4241:
4242:
4243:
4244:
4245:
4246:
4247:
4248:
4249:
4250:
4251:
4252:
4253:
4254:
4255:
4256:
4257:
4258:
4259:
4260:
4261:
4262:
4263:
4264:
4265:
4266:
4267:
4268:
4269:
4270:
4271:
4272:
4273:
4274:
4275:
4276:
4277:
4278:
4279:
4280:
4281:
4282:
4283:
4284:
4285:
4286:
4287:
4288:
4289:
4290:
4291:
4292:
4293:
4294:
4295:
4296:
4297:
4298:
4299:
4300:
4301:
4302:
4303:
4304:
4305:
4306:
4307:
4308:
4309:
4310:
4311:
4312:
4313:
4314:
4315:
4316:
4317:
4318:
4319:
4320:
4321:
4322:
4323:
4324:
4325:
4326:
4327:
4328:
4329:
4330:
4331:
4332:
4333:
4334:
4335:
4336:
4337:
4338:
4339:
4340:
4341:
4342:
4343:
4344:
4345:
4346:
4347:
4348:
4349:
4350:
4351:
4352:
4353:
4354:
4355:
4356:
4357:
4358:
4359:
4360:
4361:
4362:
4363:
4364:
4365:
4366:
4367:
4368:
4369:
4370:
4371:
4372:
4373:
4374:
4375:
4376:
4377:
4378:
4379:
4380:
4381:
4382:
4383:
4384:
4385:
4386:
4387:
4388:
4389:
4390:
4391:
4392:
4393:
4394:
4395:
4396:
4397:
4398:
4399:
4400:
4401:
4402:
4403:
4404:
4405:
4406:
4407:
4408:
4409:
4410:
4411:
4412:
4413:
4414:
4415:
4416:
4417:
4418:
4419:
4420:
4421:
4422:
4423:
4424:
4425:
4426:
4427:
4428:
4429:
4430:
4431:
4432:
4433:
4434:
4435:
4436:
4437:
4438:
4439:
4440:
4441:
4442:
4443:
4444:
4445:
4446:
4447:
4448:
4449:
4450:
4451:
4452:
4453:
4454:
4455:
4456:
4457:
4458:
4459:
4460:
4461:
4462:
4463:
4464:
4465:
4466:
4467:
4468:
4469:
4470:
4471:
4472:
4473:
4474:
4475:
4476:
4477:
4478:
4479:
4480:
4481:
4482:
4483:
4484:
4485:
4486:
4487:
4488:
4489:
4490:
4491:
4492:
4493:
4494:
4495:
4496:
4497:
4498:
4499:
4500:
4501:
4502:
4503:
4504:
4505:
4506:
4507:
4508:
4509:
4510:
4511:
4512:
4513:
4514:
4515:
4516:
4517:
4518:
4519:
4520:
4521:
4522:
4523:
4524:
4525:
4526:
4527:
4528:
4529:
4530:
4531:
4532:
4533:
4534:
4535:
4536:
4537:
4538:
4539:
4540:
4541:
4542:
4543:
4544:
4545:
4546:
4547:
4548:
4549:
4550:
4551:
4552:
4553:
4554:
4555:
4556:
4557:
4558:
4559:
4560:
4561:
4562:
4563:
4564:
4565:
4566:
4567:
4568:
4569:
4570:
4571:
4572:
4573:
4574:
4575:
4576:
4577:
4578:
4579:
4580:
4581:
4582:
4583:
4584:
4585:
4586:
4587:
4588:
4589:
4590:
4591:
4592:
4593:
4594:
4595:
4596:
4597:
4598:
4599:
4600:
4601:
4602:
4603:
4604:
4605:
4606:
4607:
4608:
4609:
4610:
4611:
4612:
4613:
4614:
4615:
4616:
4617:
4618:
4619:
4620:
4621:
4622:
4623:
4624:
4625:
4626:
4627:
4628:
4629:
4630:
4631:
4632:
4633:
4634:
4635:
4636:
4637:
4638:
4639:
4640:
4641:
4642:
4643:
4644:
4645:
4646:
4647:
4648:
4649:
4650:
4651:
4652:
4653:
4654:
4655:
4656:
4657:
4658:
4659:
4660:
4661:
4662:
4663:
4664:
4665:
4666:
4667:
4668:
4669:
4670:
4671:
4672:
4673:
4674:
4675:
4676:
4677:
4678:
4679:
4680:
4681:
4682:
4683:
4684:
4685:
4686:
4687:
4688:
4689:
4690:
4691:
4692:
4693:
4694:
4695:
4696:
4697:
4698:
4699:
4700:
4701:
4702:
4703:
4704:
4705:
4706:
4707:
4708:
4709:
4710:
4711:
4712:
4713:
4714:
4715:
4716:
4717:
4718:
4719:
4720:
4721:
4722:
4723:
4724:
4725:
4726:
4727:
4728:
4729:
4730:
4731:
4732:
4733:
4734:
4735:
4736:
4737:
4738:
4739:
4740:
4741:
4742:
4743:
4744:
4745:
4746:
4747:
4748:
4749:
4750:
4751:
4752:
4753:
4754:
4755:
4756:
4757:
4758:
4759:
4760:
4761:
4762:
4763:
4764:
4765:
4766:
4767:
4768:
4769:
4770:
4771:
4772:
4773:
4774:
4775:
4776:
4777:
4778:
4779:
4780:
4781:
4782:
4783:
4784:
4785:
4786:
4787:
4788:
4789:
4790:
4791:
4792:
4793:
4794:
4795:
4796:
4797:
4798:
4799:
4800:
4801:
4802:
4803:
4804:
4805:
4806:
4807:
4808:
4809:
4810:
4811:
4812:
4813:
4814:
4815:
4816:
4817:
4818:
4819:
4820:
4821:
4822:
4823:
4824:
4825:
4826:
4827:
4828:
4829:
4830:
4831:
4832:
4833:
4834:
4835:
4836:
4837:
4838:
4839:
4840:
4841:
4842:
4843:
4844:
4845:
4846:
4847:
4848:
4849:
4850:
4851:
4852:
4853:
4854:
4855:
4856:
4857:
4858:
4859:
4860:
4861:
4862:
4863:
4864:
4865:
4866:
4867:
4868:
4869:
4870:
4871:
4872:
4873:
4874:
4875:
4876:
4877:
4878:
4879:
4880:
4881:
4882:
4883:
4884:
4885:
4886:
4887:
4888:
4889:
4890:
4891:
4892:
4893:
4894:
4895:
4896:
4897:
4898:
4899:
4900:
4901:
4902:
4903:
4904:
4905:
4906:
4907:
4908:
4909:
4910:
4911:
4912:
4913:
4914:
4915:
4916:
4917:
4918:
4919:
4920:
4921:
4922:
4923:
4924:
4925:
4926:
4927:
4928:
4929:
4930:
4931:
4932:
4933:
4934:
4935:
4936:
4937:
4938:
4939:
4940:
4941:
4942:
4943:
4944:
4945:
4946:
4947:
4948:
4949:
4950:
4951:
4952:
4953:
4954:
4955:
4956:
4957:
4958:
4959:
4960:
4961:
4962:
4963:
4964:
4965:
4966:
4967:
4968:
4969:
4970:
4971:
4972:
4973:
4974:
4975:
4976:
4977:
4978:
4979:
4980:
4981:
4982:
4983:
4984:
4985:
4986:
4987:
4988:
4989:
4990:
4991:
4992:
4993:
4994:
4995:
4996:
4997:
4998:
4999:
5000:
5001:
5002:
5003:
5004:
5005:
5006:
5007:
5008:
5009:
5010:
5011:
5012:
5013:
5014:
5015:
5016:
5017:
5018:
5019:
5020:
5021:
5022:
5023:
5024:
5025:
5026:
5027:
5028:
5029:
5030:
5031:
5032:
5033:
5034:
5035:
5036:
5037:
5038:
5039:
5040:
5041:
5042:
5043:
5044:
5045:
5046:
5047:
5048:
5049:
5050:
5051:
5052:
5053:
5054:
5055:
5056:
5057:
5058:
5059:
5060:
5061:
5062:
5063:
5064:
5065:
5066:
5067:
5068:
5069:
5070:
5071:
5072:
5073:
5074:
5075:
5076:
5077:
5078:
5079:
5080:
5081:
5082:
5083:
5084:
5085:
5086:
5087: