Spectral Shift Functions of Lamé Operators

Ubong Sam IDIONG, Unanaowo Nyong BASSEY · 2025

Read Spectral Shift Functions of Lamé Operators by Ubong Sam IDIONG, Unanaowo Nyong BASSEY (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-19734.txt.

Introduction

Ubong Sam Idiong and Unanaowo Nyong Bassey wrote a 2025 math physics paper computing spectral shift functions for two classic operators. A spectral shift function measures how a perturbation moves the spectrum of an operator, with deep uses in scattering and quantum theory, but closed forms are rare. The authors tackle the Lamé operator in Weierstrass form and the Brioschi-Halphen operator, working through Green functions built from Fourier transforms of distributions. The hosted text is a brief abstract, so the resulting formulas are stated at that level.

Catalog record

CREATORUbong Sam IDIONG, Unanaowo Nyong BASSEY
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE50 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-19734

Browse the full library · About the Tower of Babel

Full text

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

Spectral Shift Functions of Lamé Operators Ubong Sam IDIONG, Unanaowo Nyong BASSEY Abstract The search for spectral shift functions of operators remains an open area of research. In this paper, the Kreĭn’s spectral shift functions are computed for the Lamé operator in the Weierstrass form and the Brioschi-Halphen operator through Green functions obtained by applying the technique of Fourier transform of distributions. Journal: Nuclear Physics B 2020 MSC 46C20, 34M45 1 Introduction The expression gives the Lamé operator in the Weierstrass form L_{s}:=-\;\frac{\mathrm{d}^{2}}{\mathrm{d}u^{2}}+s(s+1)\wp(u|g_{2},g_{3}),\qquad s\in\mathbb{R}, (1) where \wp is the Weierstrass elliptic \wp -function that satisfies the polynomial equation (\wp^{\prime}(u))^{2}=4\wp(u)^{3}-g_{2}\wp(u)-g_{3} , with g_{2},g_{3} being invariant constants. It is known by the works of Ince Ince (1940) that the Lamé operator (considered on the real line shifted by the imaginary half-periods \omega_{k}(k=1,2,3) with \sum_{k=1}^{3}\omega_{j}=0 for integer s has the property that its spectrum has exactly s\in\mathbb{N} gaps which are such that L_{s}:\mathscr{H}\rightarrow\mathscr{H} such that \mathscr{H}:=L^{2}((0,\dfrac{\pi}{r}),\mathrm{d}u)\subset\mathscr{C}_{c}^{\infty}(\Omega),\Omega=\mathbb{CP}^{1}\setminus\{e_{i}=\wp(\omega_{i})|i=1,2,3\} (see Ruijsenaars (1999) , Eq.(20), p.426). The ends of the spectrum E_{j} are given by the zeros of certain polynomials R_{2s+1}(E)=\prod_{i=0}^{2s}(E-E_{i}(s))=E^{2s+1}+b_{1}E^{2s}+b_{2}E^{2s-1}+\ldots+b_{2s+1}, (2) with b_{s} being functions of s . Here and hereafter, E denotes the energy variable corresponding to the operator L_{s} in ( 1 ). The polynomials given in ( 2 ) are called Lamé spectral polynomials and were first studied by Hermite and Halphen (see Grosset and Veselov (2005) , pp. 635-636 ). The Brioschi-Halphen operator obtained by a two-step transformation of the Lamé operator ( 1 ) by setting \psi=[\wp^{\prime}(\frac{1}{2}u)]^{-s}\varphi and w=\wp(u) is given as H_{s}=4\prod_{i=1}^{3}(w-e_{i})D^{2}-(2s-1)(6w^{2}-\frac{1}{2}g_{2})D+4s(2s-1)w-4B,\;\;(D=\frac{\mathrm{d}}{\mathrm{d}w}), (3) where B is a constant, s is restricted to be positive integer and the invariant constants g_{2} and g_{3} are given by g_{2}=4(e_{1}e_{2}+e_{2}e_{3}+e_{1}e_{3})\hskip 14.22636pt\textrm{and}\hskip 14.22636ptg_{3}=4e_{1}e_{2}e_{3} (4) where e_{j}=\wp(\omega_{j}),j=1,2,3 and each \omega_{j} is the half-period of the Weierstrass elliptic \wp -function. In what follows, \mathfrak{s}\ell(2,\mathbb{C}) denotes the Lie algebra of complex two-by-two traceless matrices. It is also realised as a Lie algebra of differential operators generated by \mathcal{J}_{+}:=w^{2}\frac{\mathrm{d}}{\mathrm{d}w}-2jw,\;\;\;\;\mathcal{J}_{0}:=w\frac{\mathrm{d}}{\mathrm{d}w}-j,\;\;\;\;\mathcal{J}_{-}:=\frac{\mathrm{d}}{\mathrm{d}w} (5) which obey the commutation relations [\mathcal{J}_{0},\mathcal{J}_{+}]=\mathcal{J}_{+},\;\;\;[\mathcal{J}_{-},\mathcal{J}_{+}]=2\mathcal{J}_{0},\;\;\;[\mathcal{J}_{-},\mathcal{J}_{0}]=\mathcal{J}_{-}\;\;. (cf: Turbiner (1988) , p.469). The quantum Euler top operator obtained by \mathfrak{s}\ell(2,\mathbb{C}) -algebraisation of Brioschi-Halphen operator is \mathbf{H}_{s}:=4\mathcal{J}_{+}^{2}-g_{2}\mathcal{J}_{0}^{2}-\frac{g_{3}}{2}\mathcal{J}_{-}\mathcal{J}_{0}-\frac{3j-1}{4}g_{2}\mathcal{J}_{0}-\frac{3}{16}(3j-1)^{2}g_{2}\mathcal{J}_{-} (6) with j=\frac{3s}{2} being the spin. The operator \mathbf{H}_{s} is found to be a Casimir operator with eigenvalues E_{j} given as E_{j}=\frac{j(j+1)g_{2}}{4}+\frac{3j(3j-1)}{4}. Now, the characteristic polynomial of \mathbf{H}_{s} coincides here with the Lamé spectral polynomial \det(E-\mathbf{H}_{s})=R_{2s+1}(E)=\prod_{j=0}^{2s}\left(E-\frac{j(j+1)g_{2}}{4}-\frac{3j(3j-1)}{4}\right). (7) It is well known that the spectral shift function (SSF) is a function that calculates spectral flow and is associated with the spectra of operators. It is a key object in scattering theory and has applications in spectral theory, quantum mechanics, and statistical mechanics. The relationship between the determinant and the trace of the operator under consideration will be required in the development of the SSFs presented below. For an operator X defined on a finite-dimensional space, the trace Tr(X) and the determinant \det(\exp(X) are related by the formula Tr(X)=\ln(\det(\exp(X))). (8) Similarly, the trace of the operator ( 7 ) can be written as \displaystyle Tr(\ln(E-\mathbf{H}_{s})) \displaystyle= \displaystyle\ln\det(\exp(\ln(E-\mathbf{H}_{s})) \displaystyle= \displaystyle\sum_{j=0}^{2s}\ln\left(E-\frac{j(j+1)g_{2}}{4}-\frac{3j(3j-1)}{4}\right). We note that when s=1 then E=\wp(\varepsilon) . In this case, we have R_{3}(\wp(\varepsilon))=Tr(\ln(\wp(\varepsilon)-\mathbf{H}_{2}))=\sum_{j=0}^{2}\ln\left(\wp(\varepsilon)-\frac{j(j+1)g_{2}}{4}-\frac{3j(3j-1)}{4}\right). Here, \varepsilon is the pullback of the Weierstrass elliptic function given by the expression \varepsilon=\wp^{-1}(E):=\int_{E}^{\infty}\frac{\mathrm{d}w}{\sqrt{4w^{3}-g_{2}w-g_{3}}}. The outline of the paper includes: Section 2 contains essential preliminaries, Section 3 contains the main results of this paper and Section 4 contains the conclusion and summary of the paper. 2 Preliminaries Let \mathscr{B}(\mathscr{H}) denote the space of all bounded linear operators defined on a Hilbert space \mathscr{H} , let T\in\mathscr{B}(\mathscr{H}) and let R_{z}(T):=(z-T)^{-1},z\in\mathbb{C} be the resolvent of the operator T . Following Kato ( Kato (1980) , p. 44), assume \varphi(\lambda) is an analytic ( = holomorphic) function in a domain \mathfrak{D}\subset\mathbb{C} containing all eigenvalues \lambda_{n} of T and let \Gamma\subset\mathfrak{D} be a simple closed smooth curve with positive direction enclosing all the \lambda_{n} in its interior. Then \varphi(T) is defined by the Dunford-Taylor integral (Dunford Dunford (1954) ) \varphi(T)=-\frac{1}{2\pi i}\int_{\Gamma}\varphi(\lambda)R_{\lambda}(T)\mathrm{d}\lambda=\frac{1}{2\pi i}\int_{\Gamma}\varphi(\lambda)(\lambda-T)^{-1}\mathrm{d}\lambda. (9) This is an analogue of the Cauchy integral formula in function theory [see Knopp Knopp (1945) , p.61]. More generally, \Gamma may consist of several simple closed curves \Gamma_{n} with interiors \mathfrak{D}_{n}^{\prime} such that the union of the \mathfrak{D}_{n}^{\prime} contains all the eigenvalues of T . The resolvent set of T is defined by \rho(T):=\{z|R_{z}(T)=(z-T)^{-1}\}. Let \sigma(T) denote the spectrum of T which is the defined as \sigma(T):=\mathbb{C}\setminus\rho(T) . An operator T\in\mathscr{B}(\mathscr{H}) is said to be compact if it takes a bounded sequence (x_{n}) to a sequence (Tx_{n}) with a convergent subsequence. The spaces \mathscr{B}_{1}\mathscr{(H)} and \mathscr{B}_{2}\mathscr{(H)} shall denote the set of trace-class and Hilbert-Schmidt operators respectively. We shall often have H and H_{0} as a pair of self-adjoint operators in with \sigma(H),\sigma(H_{0}) their spectra; \rho(H),\rho(H_{0}) their resolvent sets with \mathcal{R}(z) and \mathcal{R}_{0} their resolvents and E_{\lambda},E_{\lambda}^{0} the associated spectral families. The symbols \|\cdot\|,\|\cdot\|_{1}, and \|\cdot\|_{2} will denote operator norm, trace norm and Hilbert-Schmidt norm respectively, while \mathrm{Tr}(B) will stand for the trace of a trace-class operator B . The functions of operators as defined by Kato shall be stated in what follows. In what follows, the integration over the Riemann sphere which is the domain of the Lamé operator is discussed. This will assist us in the study of the compact nature of the Lamé integral operator. In this quest, we obtain a suitable measure for the integral relation. One of the important measures that exist for the Reimann sphere \mathbb{CP}^{1} is the area measure of 1-form. In local coordinates, \displaystyle\mathrm{d}z\wedge\mathrm{d}\overline{z} \displaystyle= \displaystyle(\mathrm{d}x+i\mathrm{d}y)\wedge(\mathrm{d}x-i\mathrm{d}y) \displaystyle= \displaystyle\left|\frac{\partial(\mathrm{d}z,\mathrm{d}\overline{z})}{\partial(\mathrm{d}x,\mathrm{d}y)}\right|\mathrm{d}x\mathrm{d}y \displaystyle= \displaystyle\left|\begin{array}[]{cc}1&i\\ 1&-i\\ \end{array}\right|\mathrm{d}x\mathrm{d}y \displaystyle= \displaystyle-2i\mathrm{d}x\mathrm{d}y. Following Royden ( Royden (1952) , §4, p.47), let f\in\mathscr{C}^{1}(\mathbb{CP}^{1}) be a piecewise smooth function over nets \{\Omega_{n}\}_{n\in\mathbb{N}} in \mathbb{CP}^{1} . Then if w=\mathrm{d}f , one obtains the norm 0\leqslant\|w\|^{2}=\frac{i}{2}\int_{\Omega_{n}}w\wedge\overline{w}=\int_{\Omega_{n}}|w|^{2}\mathrm{d}x\mathrm{d}y<\infty. It is also known ( McMullen (2025) , Theorem 7.1, p.72) that for any function g\in\mathscr{C}_{c}^{\infty}(\mathbb{C}), a solution to the equation \frac{\mathrm{d}f}{\mathrm{d}\overline{z}}=g is given by the convolution f(z)=g\ast\frac{1}{\pi z}=\frac{i}{2\pi}\int_{\mathbb{C}}\frac{g(w)}{z-w}\mathrm{d}w\wedge\mathrm{d}\overline{w}. Theorem 1 ( Kuttler (2008) , §9.10, p. 264) . Let f:X\times Y\longrightarrow[0,\infty] be measurable with respect to the \sigma -algebra \sigma(\mathcal{K}) and let \mu\times\nu the product measure so that (\mu\times\nu)(E)=\int_{Y}\int_{X}\chi_{E}\mathrm{d}\mu\mathrm{d}\nu=\int_{X}\int_{Y}\chi_{E}\mathrm{d}\nu\mathrm{d}\mu, (11) where \mu and \nu are the finite measure on the the measurable spaces (X,\mathcal{F}) and (Y,\mathcal{S}) respectively. Then \int_{X\times Y}f\mathrm{d}(\mu\times\nu)=\int_{Y}\int_{X}f\mathrm{d}\mu\mathrm{d}\nu=\int_{X}\int_{Y}f\mathrm{d}\nu\mathrm{d}\mu, (12) where \mathcal{K} is the family of compact sets. Proof. Since the measures are \sigma -finite there exists a pair of sequences \{X_{n}\} and \{Y_{n}\} such that \nu(X_{n})<\infty and \mu(Y_{n})<\infty then \int_{Y_{n}}\int_{X_{n}}f\mathrm{d}\mu\mathrm{d}\nu=\int_{X_{n}}\int_{Y_{n}}f\mathrm{d}\nu\mathrm{d}\mu, (13) so that as n\rightarrow\infty equation ( 13 ) gives \int_{X\times Y}f\mathrm{d}(\mu\times\nu)\equiv\int_{Y}\int_{X}f\mathrm{d}\mu\mathrm{d}\nu=\int_{X}\int_{Y}f\mathrm{d}\nu\mathrm{d}\mu, (14) In particular if we take f=\chi_{E} where E\in\mathcal{F}\times\mathcal{S} we obtain ( 11 ).∎ Lemma 2 . Let m_{k} be the Lebesgue measure on \mathbb{R}^{k} , if we define a measure \sigma_{k-1} on \mathbb{S}^{k-1} (a unit sphere in \mathbb{R}^{k} ). If further, A\subset\mathbb{S}^{k-1} is a Borel set, let \widetilde{A}:=\{r\mathbf{u}:0<r<1,\mathbf{u}\in A\} so that we define \sigma_{k-1}(A)=k\cdot m_{k}(\widetilde{A}). Then \displaystyle\int_{\mathbb{R}^{k}}f\mathrm{d}m_{k}=\int_{0}^{\infty}r^{k-1}\mathrm{d}r\int_{\mathbb{S}^{k-1}}f(r\mathbf{u})\mathrm{d}\sigma_{k-1}(\mathbf{u}). is valid for every Borel function f(\mathbf{x})\geq 0 on \mathbb{R}^{k} where, \mathbf{x}=r\mathbf{u}. Proof. Without loss of generality, the Borel set A is compact.We consider the sequence \left(T_{n}\right) of measurable sets defined by \displaystyle\displaystyle T_{1} \displaystyle= \displaystyle\left\{\mathbf{x}\in A\bigg|\frac{1}{2}\leq\varphi(\mathbf{x})<1\right\} \displaystyle T_{n} \displaystyle= \displaystyle\left\{\mathbf{x}\in A\bigg|\frac{1}{2^{n}}\leq\varphi(\mathbf{x})-\sum_{i=1}^{n-1}\frac{1}{2^{i}}\chi_{T_{i}}(\mathbf{x})<\frac{1}{2^{n-1}}\right\},\qquad\forall n\geq 2, where \displaystyle\chi_{T_{i}}(\mathbf{x}) is the characteristic function with respect to terms T_{i} , of the the sequence \left(T_{n}\right) and \varphi:\mathbb{R}^{k}\longrightarrow\mathbb{R}:\mathbf{x}\mapsto\varphi(\mathbf{x}) . . Also arguing by induction, it could be seen that for any \mathbf{x}\in A,i\in\mathbb{N},\chi_{T_{i}}(\mathbf{x})=a_{i},

Read the complete text