Resolvent Trace Formula and Determinants of Laplacians on Orbifold Riemann Surfaces

For a nonnegative integer, we consider the -Laplacian Δₙ acting on the space of -differentials on a confinite Riemann surface X which has ramification points. The trace formula for the resolvent kernel is developed along the line à la Selberg. Using the trace formula, we compute the regularized...

Full description

Saved in:
Bibliographic Details
Published in:Symmetry, Integrability and Geometry: Methods and Applications
Date:2021
Main Author: Teo, Lee-Peng
Format: Article
Language:English
Published: Інститут математики НАН України 2021
Online Access:https://nasplib.isofts.kiev.ua/handle/123456789/211444
Tags: Add Tag
No Tags, Be the first to tag this record!
Journal Title:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Cite this:Resolvent Trace Formula and Determinants of Laplacians on Orbifold Riemann Surfaces. Lee-Peng Teo. SIGMA 17 (2021), 083, 40 pages

Institution

Digital Library of Periodicals of National Academy of Sciences of Ukraine
_version_ 1859723033553403904
author Teo, Lee-Peng
author_facet Teo, Lee-Peng
citation_txt Resolvent Trace Formula and Determinants of Laplacians on Orbifold Riemann Surfaces. Lee-Peng Teo. SIGMA 17 (2021), 083, 40 pages
collection DSpace DC
container_title Symmetry, Integrability and Geometry: Methods and Applications
description For a nonnegative integer, we consider the -Laplacian Δₙ acting on the space of -differentials on a confinite Riemann surface X which has ramification points. The trace formula for the resolvent kernel is developed along the line à la Selberg. Using the trace formula, we compute the regularized determinant of Δₙ + ( + 2 − 1), from which we deduce the regularized determinant of Δₙ, denoted by det′Δₙ. Taking into account the contribution from the absolutely continuous spectrum, det′Δₙ is equal to a constant Cₙ times ( ) when ≥ 2. Here ( ) is the Selberg zeta function of . When = 0 or = 1, ( ) is replaced by the leading coefficient of the Taylor expansion of ( ) around = 0 and = 1, respectively. The constants Cn are calculated explicitly. They depend on the genus, the number of cusps, as well as the ramification indices, but are independent of the moduli parameters.
first_indexed 2026-03-15T10:21:32Z
format Article
fulltext Symmetry, Integrability and Geometry: Methods and Applications SIGMA 17 (2021), 083, 40 pages Resolvent Trace Formula and Determinants of n Laplacians on Orbifold Riemann Surfaces Lee-Peng TEO Department of Mathematics, Xiamen University Malaysia, Jalan Sunsuria, Bandar Sunsuria, 43900, Sepang, Selangor, Malaysia E-mail: lpteo@xmu.edu.my Received April 07, 2021, in final form September 05, 2021; Published online September 13, 2021 https://doi.org/10.3842/SIGMA.2021.083 Abstract. For n a nonnegative integer, we consider the n-Laplacian ∆n acting on the space of n-differentials on a confinite Riemann surface X which has ramification points. The trace formula for the resolvent kernel is developed along the line à la Selberg. Using the trace formula, we compute the regularized determinant of ∆n + s(s + 2n − 1), from which we deduce the regularized determinant of ∆n, denoted by det′∆n. Taking into account the contribution from the absolutely continuous spectrum, det′∆n is equal to a constant Cn times Z(n) when n ≥ 2. Here Z(s) is the Selberg zeta function of X. When n = 0 or n = 1, Z(n) is replaced by the leading coefficient of the Taylor expansion of Z(s) around s = 0 and s = 1 respectively. The constants Cn are calculated explicitly. They depend on the genus, the number of cusps, as well as the ramification indices, but is independent of the moduli parameters. Key words: determinant of Laplacian; n-differentials; cocompact Riemann surfaces; Selberg trace formula 2020 Mathematics Subject Classification: 14H15; 11F72; 11M36 Dedicated to Professor Leon Takhtajan on the occasion of his 70th birthday 1 Introduction Let H be the upper half plane, and let Γ be a Fuchsian group, a discrete subgroup of PSL(2,R). The group Γ acts discontinuously on H and the quotient X = Γ\H is a Riemann surface. In this work, we consider the case where X is a cofinite Riemann surface. In other words, X has finite hyperbolic volume. This includes compact Riemann surfaces, as well as surfaces with finitely many cusps and ramification points. Let ∆0 = −y2 ( ∂2 ∂x2 + ∂2 ∂y2 ) be the Laplacian operator acting on functions. It is an invariant operator under the action of PSL(2,R). In the seminal paper [18], Selberg developed a theory that can be used to study the spectrum of ∆0 on Riemann surfaces, making full use of the invariant property of the Laplacian operator. The theory was subsequently developed and elaborated in [7, 11, 12, 19, 23, 24, 25]. One of the extensions of Selberg’s theory is to consider the spectrum of Dn for n ≥ 1, where Dn = −Kn−1Ln−n(n− 1), and Kn and Ln are the Maass operators1 (see Section 2 for more details). This paper is a contribution to the Special Issue on Mathematics of Integrable Systems: Classical and Quan- tum in honor of Leon Takhtajan. The full collection is available at https://www.emis.de/journals/SIGMA/Takhtajan.html 1Our Dn has a negative sign compared to that used in [6], while Fischer [7] denote it by −∆n. mailto:lpteo@xmu.edu.my https://doi.org/10.3842/SIGMA.2021.083 https://www.emis.de/journals/SIGMA/Takhtajan.html 2 L.-P. Teo The significance of this theory is an identity that relates the spectral trace to the geometric trace. In mathematical physics, the regularized determinant of the Laplacian is of special interest. This has been considered in [4, 17] for compact Riemann surfaces. Subsequently, Efrat [5] exten- ded the result to smooth Riemann surfaces with cusps. In the special case of arithmetic surfaces, this has also been considered in [14, 15]. In [9], Gong derived the regularized determinant in full generality using the trace formula developed in [7] directly, which can be applied to cofinite Riemann surfaces and for the operators Dn with unitary twists. The goal of this paper is to study the explicit expression for the resolvent trace formula and the regularized determinant of n-Laplacians ∆n on Riemann surfaces that have finite volumes. ∆n has the same spectrum as Dn + n(n − 1). In principle, one can extract the expression for the trace formula and the determinant of ∆n from the formulas obtained in [7, 9, 12, 23]. However, we find that a direct approach from a theory for ∆n would be more appealing for further application. This is the tasks undertaken in this paper. 2 Laplacians of n-differentials Let H be the upper half of the complex plane equipped with the hyperbolic metric ds2 = dx2 + dy2 y2 . The corresponding metric density is ρ(z) = (Im z)−2 and the area form is dµ(z) = dx dy y2 = ρ(z) d2z. The group PSL(2,R) acts on H transitively by Möbius transformations γ = ( a b c d ) : z 7→ γ(z) = az + b cz + d . The hyperbolic metric is an invariant metric under this group action. The distance between two points z and w on H, denoted by d(z, w), is given by cosh d(z, w) = 1 + 2u(z, w), where u(z, w) = |z − w|2 4 Im z Imw is a point-pair invariant, i.e., u(γz, γw) = u(z, w) for all γ ∈ PSL(2,R). Let Γ be a cofinite Fuchsian group, namely, it is a discrete subgroup of PSL(2,R) such that X = Γ\H is a Riemann surface with finite hyperbolic volume. Such Γ is finitely generated. More precisely, if X is a genus g Riemann surface with q-punctures and v ramification points, then Γ is generated by 2g hyperbolic elements α1, β1, . . . , αg, βg, q parabolic elements κ1, . . . , κq, as well as v elliptic elements τ1, . . . , τv of orders m1, . . . ,mv respectively, with 2 ≤ m1 ≤ m2 ≤ · · · ≤ mv. Resolvent Trace Formula and Determinants of n Laplacians 3 For each 1 ≤ j ≤ v, τ mj j = I. The generators of Γ satisfy the additional relation α1β1α −1 1 β−1 1 · · ·αgβgα −1 g β−1 g κ1 · · ·κqτ1 · · · τv = I, where I is the identity element. We say that the Riemann surface X and the group Γ are of type (g; q;m1,m2, . . . ,mv). Let K be the canonical bundle of X. For a nonnegative integer n, let S(n) be the space of sections of Kn/2 ⊗ K̄−n/2. A function in S(n) can be realized as a function f : H → C satisfying f(γz) ( cz̄ + d cz + d )n = f(z), for all γ = ( a b c d ) ∈ Γ. In the context of analytic number theory, it is natural to consider the Maass operators Kn: S(n) → S(n+ 1), Ln : S(n) → S(n− 1) and Dn : S(n) → S(n) given by [6] Kn = (z − z̄) ∂ ∂z + n, Ln = − (z − z̄) ∂ ∂z̄ − n, Dn = −y2 ( ∂2 ∂x2 + ∂2 ∂y2 ) + 2iny ∂ ∂x . These operators are invariant with respect to the PSL(2,R) action and Dn+1Kn = KnDn, Dn−1Ln = LnDn, Dn = −Ln+1Kn − n(n+ 1) = −Kn−1Ln − n(n− 1). For applications of the theory of Riemann surfaces in physics, it is more natural to consider the space of tensors f(z)(dz)n, or called n-differentials, on X. These are sections of Kn, which are functions f : H → C satisfying f(γz)γ′(z)n = f(z), for all γ ∈ Γ. We denote this space by H2 n(Γ). It is a Hilbert space with the inner product ⟨f, g⟩ = ∫∫ X f(z)g(z)ρ(z)−ndµ(z). The n-Laplacian operator ∆n = 4∂̄∗n∂̄n is an operator on H2 n(Γ) with explicit formula given by ∆n = −4y2−2n ∂ ∂z y2n ∂ ∂z̄ . This is a positive operator. We put a factor 4 in front of ∂̄∗n∂̄n so that when n = 0, we get the usual Laplacian on functions ∆0. Throughout this paper, the nonnegative integer n is fixed. Our goal is to study the trace formula for the resolvent of ∆n and the regularized determinant of ∆n. The isometry I: H2 n(Γ) → S(n) defined by f(z) 7→ ynf(z) conjugates ∆n with Dn + n(n− 1). Hence, ∆n and Dn + n(n− 1) have the same spectrum. For a Riemann surface with cusps, the spectrum of ∆n consists of a discrete part and a con- tinuous part. To study the spectrum of the continuous part, we need to consider Eisenstein series, which we discuss in next section. 4 L.-P. Teo A powerful tool to study the spectrum of ∆n is the Selberg trace formula. The Selberg trace formula for the operator Dn has been developed extensively in [7, 12, 23]. It can be adapted to ∆n. To find the determinants of Laplacian, we will follow Fischer [7] and derive the trace formula for the resolvent kernel of the Laplacian operator first. Let Ψ: R → C be a function that vanishes at infinity and satisfies other regularity conditions to be specified when needed. Define k : H×H → C by k(z, w) = Ψ(u(z, w)) (−4)n (z − w̄)2n = Ψ(u(z, w))Hn(z, w), (2.1) where Hn(z, w) = (−4)n (z − w̄)2n . Notice that Hn(z, z) = ρ(z)n. Since Hn(γz, γw)γ ′(z)nγ′(w) n = Hn(z, w), we find that k(γz, γw)γ′(z)nγ′(w) n = k(z, w). In other words, k(z, w) is a point-pair invariant kernel on H. Given s ∈ C, the function f(z) = (Im(z))s−n is an “eigenfunction” of ∆n with eigenvalue −(s− n)(s+ n− 1). The following proposition shows that f(z) is also an eigenfunction for the operator defined by the point-pair invariant kernel k(z, w). Proposition 2.1. Let Ψ: R → C be a continuous function with compact support and let k(z, w) = Ψ(u(z, w))Hn(z, w) be the corresponding point-pair invariant kernel. If s ∈ C, then∫∫ H k(z, w)(Imw)s−n(Imw)2ndµ(w) = Λs(Im z)s−n, where Λs = ∫∫ H k(i, w)(Imw)s−n(Imw)2ndµ(w). Proof. Given z ∈ H, notice that γ = ( a b 0 d ) , with 1 d = a = √ Im z and ab = Re z maps i to z. For such γ, γ′(z) = a2 = Im z. Let g(z) = ∫∫ H k(z, w)(Imw)s−n(Imw)2ndµ(w). Using k(γi, γw)γ′(i)nγ′(w) n = k(i, w) and Im(γw) = Im(w)|γ′(w)|, a change of variables w 7→ γw gives g(z) = ∫∫ H k(γi, γw)(Im γw)s+ndµ(w) = a2s−2n ∫∫ H k(i, w)(Imw)s+ndµ(w) = Λs(Im z)s−n, where Λs = ∫∫ H k(i, w)(Imw)s+ndµ(w). ■ Resolvent Trace Formula and Determinants of n Laplacians 5 Given the point-pair invariant kernel k(z, w) on H, we can define a point-pair invariant ker- nel K(z, w) for the Riemann surface X by K(z, w) = ∑ γ∈Γ k(γz, w)γ′(z)n, z, w ∈ H. (2.2) Notice that if f ∈ H2 n(Γ), then∫∫ X K(z, w)f(w)ρ(w)−ndµ(w) = ∫∫ H k(z, w)f(w)ρ(w)−ndµ(w). We want to find a function Ψn,s(u) = Ψ(u) so that the corresponding K(z, w) is the resolvent kernel of the operator ∆n+s(s+2n−1). The reason to use s(s+2n−1) instead of (s−n)(s+n−1) is so that when s = 0, we get ∆n. This function Ψ(u) must satisfy the differential equation ∆n(Ψ(u(z, w))Hn(z, w)) = −s(s+ 2n− 1)Ψ(u(z, w))Hn(z, w). Using the fact that ∆n(Ψ(u(z, w))Hn(z, w)) = ( −u(u+ 1) ∂2Ψ ∂u2 + [(2n− 2)u− 1] ∂Ψ ∂u ) Hn(z, w), we find that u(u+ 1) ∂2Ψ ∂u2 − [(2n− 2)u− 1] ∂Ψ ∂u − s(s+ 2n− 1)Ψ = 0. A solution is given by Ψn,s(u) = (u+ 1)−s 4π Γ(s)Γ(s+ 2n) Γ(2s+ 2n) 2F1 ( s, s+ 2n 2s+ 2n ; 1 u+ 1 ) , (2.3) where 2F1 ( a, b c ; z ) = Γ(c) Γ(a)Γ(b) ∞∑ k=0 Γ(a+ k)Γ(b+ k) Γ(c+ k) zk k! is the hypergeometric function. The normalization constant 1/4π is chosen so that when u→ 0+, Ψ(u) ∼ 1 4π log 1 u . More explicitly, one can show that as u→ 0+, Ψn,s(u) = 1 4π { log 1 u + 2ψ(1)− ψ(s+ 2n)− ψ(s) } +O(u). (2.4) Here ψ(s) = Γ′(s) Γ(s) is the logarithmic derivative of the gamma function Γ(s), and ψ(1) = −γ, where γ is the Euler constant. 6 L.-P. Teo 3 The Eisenstein series For Riemann surfaces that are not compact, it is well-known that the spectrum of the Laplacians contain a continuous part, which are related to the Eisenstein series. The Riemann surface X has q cusps corresponding to the q parabolic elements κ1, . . . , κq. For 1 ≤ i ≤ q, let xi ∈ R∪ {∞} be the fixed point of κi. Then xi is a representative of the cusp associated to κi. Let σi ∈ PSL(2,R) be an element that conjugates κi to ( 1 ±1 0 1 ) , namely, σ−1 i κiσi = ( 1 ±1 0 1 ) . Then σi(∞) = xi. If B is the parabolic subgroup generated by ( 1 1 0 1 ) , then Γi = σiBσ −1 i is the stabilizer of the cusp xi in Γ. Define the Eisenstein series associated to the cusp xi by Ei(z, s;n) = ∑ γ∈Γi\Γ [ Im ( σ−1 i γz )]s−n[( σ−1 i γ )′ (z) ]n , when Re s > 1. Here to simplify notation, we write σ−1 ◦ γ as σ−1γ, and we write ( σ−1 ◦ γ ) (z) as σ−1γz. One can check that the definition of the Eisenstein series here differs from the one used in [7, 12, 23] by the factor yn. This makes good sense in view of the isometry between H2 n(Γ) and S(n) that we discussed earlier. Using the fact that ∆ny s−n = −(s− n)(s+ n− 1)ys−n, and ∆n is invariant with respect to the action of PSL(2,R), we find that ∆nEi(z, s;n) = −(s− n)(s+ n− 1)Ei(z, s;n). The theory of Eisenstein series has been developed extensively in the books [7, 12, 13]. In the following, we follow [13] to give a brief exposition of the facts needed in this work. Given 1 ≤ i, j ≤ q, there is a double coset decomposition of the group σ−1 i Γσj into disjoint double cosets given by σ−1 i Γσj = δijΩ∞ ∪ ⋃ c>0 ⋃ dmod c Ωd/c. Here Ω∞ contains all the upper triangular matrices in σ−1 i Γσj , and Ωd/c is the double coset Bωd/cB, where ωd/c is an element of σ−1 i Γσj of the form ωd/c = ( ∗ ∗ c d ) . Using this, one can show that when y → ∞, Ei(σjz, s;n)σ ′ j(z) n = δijy s−n + φij(s;n)y 1−s−n + exponentially decaying terms, where φij(s;n) = √ π Γ(s)Γ ( s− 1 2 ) Γ(s+ n)Γ(s− n) ∑ c>0 ∑ dmod c 1 c2s is a Dirichlet series. From this explicit expression, one finds that φij(s;n) = Γ(s)2 Γ(s+ n)Γ(s− n) φij(s; 0). (3.1) Namely, there is a simple relation between φij(s;n) and φij(s; 0). Resolvent Trace Formula and Determinants of n Laplacians 7 The q × q matrix Φ(s) = [φij(s)] is called the scattering matrix and it plays an important role in the spectral theory. It has the following properties: Φ(s) = Φ(s̄), Φ(s)T = Φ(s). Moreover, if we denote by Ξ(z, s;n) the column matrix with components Ei(z, s;n), then Ξ(z, s;n) = Φ(s;n) Ξ(z, 1− s;n). It follows that Φ(s)Φ(1− s) = Iq. Denote by φ(s) the determinant of Φ(s), namely, φ(s) = detΦ(s). Then we find that φ(s) = φ(s̄), φ(s)φ(1− s) = 1. 4 The trace of the resolvent kernel The spectral theory for the Riemann surface X = Γ\H states that there is a countable orthonor- mal system {uk}k≥0 of eigenfunctions of ∆n with eigenvalues 0 = λ0 ≤ λ1 ≤ λ2 ≤ · · · , and eigenpackets given by the Eisenstein series so that for any f ∈ H2 n(Γ), f(z) = ∞∑ k=0 ⟨f, uk⟩uk(z) + 1 4π q∑ j=1 ∫ ∞ −∞ 〈 f,Ej ( ·, 1 2 + ir;n )〉 Ej ( z, 1 2 + ir;n ) dr. We want to use this formula to find the trace of the resolvent (∆n + s(s+2n− 1))−1. However, the kernel function K(z, w) for this resolvent defined by Ψn,s(u) in (2.3) has singularity along z = w. To circumvent this problem, we consider the kernel function K(z, w) with Ψ(u) = Ψn,s(u)−Ψn,a(u), for some fixed a. For this kernel function, ⟨K(·, w), uk⟩ = Λ(λk)uk(w),〈 K(·, w), Ej ( ·, 1 2 + ir;n )〉 = Λ̃(r)Ej ( w, 1 2 + ir;n ) , where Λ(λk) = 1 λk + s(s+ 2n− 1) − 1 λk + a(a+ 2n− 1) , Λ̃(r) = 1( s+ n− 1 2 )2 + r2 − 1( a+ n− 1 2 )2 + r2 . Therefore, the spectral decomposition of K(z, w) is K(z, w) = ∞∑ k=0 Λ(λk)uk(z)uk(w) + 1 4π q∑ j=1 ∫ ∞ −∞ Λ̃(r)Ej ( z, 1 2 + ir;n ) Ej ( w, 1 2 + ir;n ) dr. Setting z = w, we have K(z, z) = ∞∑ k=0 Λ(λk)|uk(z)|2 + 1 4π q∑ j=1 ∫ ∞ −∞ Λ̃(r) ∣∣∣∣Ej ( z, 1 2 + ir;n )∣∣∣∣2dr. 8 L.-P. Teo Integrating over X, one would have ∞∑ k=0 Λ(λk) = ∫∫ X ( K(z, z)− 1 4π q∑ j=1 ∫ ∞ −∞ Λ̃(r) ∣∣∣∣Ej ( z, 1 2 + ir;n )∣∣∣∣2dr ) y2ndµ(z). As in [13], some regularizations are needed to make this integral well-defined. This will be explained in the following. By definition (2.1) and (2.2), K(z, w) can be written as a series K(z, w) = ∑ γ∈Γ k(γz, w)γ′(z)n, where k(z, w) = (Ψn,s(u(z, w))−Ψn,a(u(z, w))) (−4)n (z − w̄)2n , with Φn,s(u) defined in (2.3). Decompose the elements of the group Γ into the set that con- tains only the identity, and the three sets that contain respectively hyperbolic elements, elliptic elements and parabolic elements, we have K(z, w) = k(z, w) + ∑ γ∈Γ γ is hyperbolic k(γz, w)γ′(z)n + ∑ γ∈Γ γ is elliptic k(γz, w)γ′(z)n + ∑ γ∈Γ γ is parabolic k(γz, w)γ′(z)n. Therefore, ∞∑ k=0 Λ(λk) = Ξ0 + ΞH + ΞE + ΞP , (4.1) where Ξ0 = ∫∫ X k(z, z)y2ndµ(z), ΞH = ∫∫ X ∑ γ∈Γ γ is hyperbolic k(γz, z)γ′(z)ny2ndµ(z), ΞE = ∫∫ X ∑ γ∈Γ γ is elliptic k(γz, z)γ′(z)ny2ndµ(z). The term ΞP contains the parabolic contribution as well as the absolutely continuous spectrum. We need to do some regularization to make it finite. Let F be a fundamental domain of X on H. Given Y > 0, let F Y = F \ q⋃ j=1 F Y j , where F Y j = F ∩ σj ( {x+ iy | y > Y } ) . Resolvent Trace Formula and Determinants of n Laplacians 9 Then we define ΞP = lim Y→∞ {∫∫ FY ∑ γ∈Γ γ is parabolic k(γz, w)γ′(z)ny2ndµ(z) − 1 4π q∑ j=1 ∫ ∞ −∞ Λ̃(r) ∫∫ FY ∣∣∣∣Ej ( z, 1 2 + ir;n )∣∣∣∣2 y2ndµ(z) dr } . In Appendix D, we show that this limit indeed exist. By (2.4), we have k(z, z)y2n = − 1 4π { ψ(s+ 2n) + ψ(s)− ψ(a+ 2n)− ψ(a) } . Therefore, Ξ0 = −|X| 4π { ψ(s+ 2n) + ψ(s)− ψ(a+ 2n)− ψ(a) } , where |X| is the hyperbolic area of X given by |X| = 2π { 2g − 2 + q + v∑ j=1 ( 1− 1 mj )} . The computations of ΞH , ΞE and ΞP are much more complicated. We leave them to Appen- dices B, C and D, and quote the results here. For the hyperbolic contributions, let P be the primitive hyperbolic conjugacy classes in Γ. For a representative γ of a primitive hyperbolic class, let N(γ) > 1 be the multiplier of γ. Then ΞH = EH(s)− EH(a), where EH(s) = 1 2s+ 2n− 1 ∑ [γ]∈P ∞∑ k=0 logN(γ) N(γ)s+n+k − 1 . The elliptic contribution is given by ΞE = EE(s)− EE(a), where EE(s) = 1 2s+ 2n− 1 v∑ j=1 mj−1∑ r=0 [ 2αmj (r − n) + 1−mj 2m2 j ψ ( s+ r mj ) + 2αmj (r + n) + 1−mj 2m2 j ψ ( s+ 2n+ r mj )] . If m is a positive integer greater than 1, and k is an integer, αm(k) is defined to be the least positive residue modulo m. Namely, it is the smallest nonnegative integer congruent to k mo- dulo m. Finally, we have ΞP = EP (s)− EP (a), where EP (s) = 1 (2s+ 2n− 1)2 [ q − TrΦ ( 1 2 )] + 1 2 Σ(s) + q 2(2s+ 2n− 1) { ψ(s) + ψ(s+ 2n)− 2 log 2− 2ψ ( s+ n+ 1 2 ) − 2ψ(s+ n) } , where Σ(s) = 1 2π ∫ ∞ −∞ 1 r2 + ( s+ n− 1 2 )2 φ′ φ ( 1 2 + ir ) dr. Gathering all the results, we obtain the resolvent trace formula for ∆n. 10 L.-P. Teo Theorem 4.1 (the resolvent trace formula). The resolvent trace formula of the n-Laplacian ∆n on the Riemann surface X is given by TS(s)− TS(a) = TG(s)− TG(a), where TS(s) and TG(s) are respectively the spectral trace and the geometric trace given by TS(s) = ∞∑ k=0 1 λk + s(s+ 2n− 1) − 1 2 Σ(s), TG(s) =− |X| 4π [ ψ(s+ 2n) + ψ(s) ] + 1 2s+ 2n− 1 ∑ [γ]∈P ∞∑ k=0 logN(γ) N(γ)s+n+k − 1 + 1 2s+ 2n− 1 v∑ j=1 mj−1∑ r=0 [ 2αmj (r − n) + 1−mj 2m2 j ψ ( s+ r mj ) + 2αmj (r + n) + 1−mj 2m2 j ψ ( s+ 2n+ r mj )] + q 2(2s+ 2n− 1) [ ψ(s) + ψ(s+ 2n)− 2 log 2− 2ψ ( s+ n+ 1 2 ) − 2ψ(s+ n) ] + A (2s+ 2n− 1)2 , and A is the constant A = Tr [ I − Φ ( 1 2 )] . In [22], we have explained that A is an even nonnegative integer. In applying the resolvent trace formula, we always take a to be a positive constant that is large enough. As mentioned in the introduction, ∆n has the same spectrum as Dn + n(n − 1). Therefore, ∆n + s(s+ 2n− 1) has the same spectrum as Dn + s(s+ 2n− 1) + n(n− 1) = Dn + (s+ n)(s+ n− 1). Hence, if we replace the s in the resolvent trace formula for Dn that were derived in [7, 12] by s+ n, one should get the resolvent trace formula for ∆n obtained in Theorem 4.1. In principle, we can directly quote the formulas in [7, 12] and proceed directly to derive the regularized determinants of n-Laplacians. However, we have found some inconsistencies in the results in these two references. One of the goals of this work is to resolve this inconsistency, giving a concise reference to the computations that lead to the resolvent trace formula, supplementing the n = 0 case done in the book [13]. When doing so, we find that it is more natural to consi- der ∆n which is a matter of taste. The use of ∆n+s(s+2n−1) instead of ∆n+(s−n)(s+n−1) is so that setting s = 0 in ∆n + s(s+ 2n− 1) give ∆n. 5 Dimensions of spaces holomorphic differentials For n ≥ 1, the space of holomorphic n-differentials is the subspace of H2 n(Γ) which are holomor- phic. Hence, they are eigenvectors of ∆n with eigenvalue 0. We also call them zero modes or holomorphic cusp forms of weight 2n. Let dn be the dimension of the space of holomorphic n-differentials of X. Although the formula for dn is well-known, we would like to show how this can be derived from the resolvent trace formula, similar to what have been done in [12]. Resolvent Trace Formula and Determinants of n Laplacians 11 When n = 0, d0 is the dimension of space of holomorphic functions, which is equal to 1, corresponding to the constant functions. Notice that the hyperbolic contribution to the resolvent trace formula can be written as ΞH = 1 2s+ 2n− 1 d ds logZ(s+ n)− 1 2a+ 2n− 1 d ds logZ(a+ n), where Z(s) is the Selberg zeta function of X defined as Z(s) = ∏ [γ]∈P ∞∏ k=0 ( 1−N(γ)−s−k ) . It is well known that Z(s) has a zero of order 1 at s = 1 and it is absolutely convergent when Re s > 1. Therefore, when n = 1, d ds logZ(s + n) has residue 1 at s = 0. When n ≥ 2, d ds logZ(s+ n) has residue 0 at s = 0. As we explained above, the dimension of the space of holomorphic n-differentials dn is equal to the multiplicity of 0 as a discrete eigenvalue of ∆n. According to Theorem D.1, the residue of Σ(s) at s = 0 is 0. Hence, dn is equal to (2n−1) times the residue at s = 0 of the function on the left hand side of the resolvent trace formula. To determine the value of dn from the resolvent trace formula, we need to find the residue at s = 0 of the function at the right-hand side of the trace formula. It then amounts to understanding the zeros and poles of Z(s) at positive integers, as well as the residues of ψ(s) at rational numbers. Since ψ(s) = −γ + ∞∑ k=0 ( 1 k + 1 − 1 s+ k ) , we find that the residues of ψ(s) at s = 0,−1,−2, . . . are all equal to −1. ψ(s) does not have poles at other points. Using the fact that αmj (−1) = mj − 1, we find that d1 = 1 2 [ 2g − 2 + q + v∑ j=1 ( 1− 1 mj )] + 1− 1 2 v∑ j=1 ( 1− 1 mj ) − q 2 = g, which is a well-known result, since the space of holomorphic one-differentials is exactly the space of abelian differentials. When n ≥ 2, dn = 2n− 1 2 [ 2g − 2 + q + v∑ j=1 ( 1− 1 mj )] + 1 2 v∑ j=1 mj − 1− 2αmj (−n) mj − q 2 . Since αmj (−n) mj = − n mj − ⌊ − n mj ⌋ , we find that when n ≥ 2, dn = (2n− 1)(g − 1) + (n− 1)q + v∑ j=1 (⌊ n− n mj ⌋) . 12 L.-P. Teo 6 The determinant of n-Laplacian In this section, we want to derive our main result – the formula for the determinant of n- Laplacian ∆n for a cofinite Riemann surface X in terms of the Selberg zeta function for the Riemann surface. This extends our result in [22] to the case where n ≥ 1. The results are not entirely new. For compact hyperbolic surfaces, the relation between the determinant of ∆n and the Selberg zeta function have been obtained by D’Hoker and Phong [4] and Sarnak [17]. In [5], Efrat considered the determinant of ∆0+ s(s− 1) for cofinite hyperbolic surfaces without elliptic points. For congruence subgroups Γ0(N), Γ1(N) and Γ(N), Koyama has obtained the relation in his work [14, 15]. Gong has considered the more general case of Laplacian operators on automorphic forms of nonzero weights in [9], but this work has not attracted much attention. We would first derive the determinant of ∆n + s(s + 2n − 1). Since ∆n contains absolutely continuous spectrum, we need to be careful with defining the determinant. Following [5, 15, 24], let ζ(w, s) = ∞∑ k=0 1 (λk + s(s+ 2n− 1))w − 1 4π ∫ ∞ −∞ 1[( s+ n− 1 2 )2 + r2 ]w φ′ φ ( 1 2 + ir ) dr be the spectral zeta function of X correspond to ∆n+s(s+2n−1). This expression is well-defi- ned when Rew is large enough. It can be analytically continued to a neighbourhood of w = 0. The zeta regularized determinant det(∆n + s(s+ 2n− 1)) is defined as det((∆n + s(s+ 2n− 1)) = exp (−ζw(0, s)) . By uniqueness of analytic continuation, we find that d ds 1 2s+ 2n− 1 d ds log det(∆n + s(s+ 2n− 1)) = d ds 1 2s+ 2n− 1 d ds (−ζw(0, s)) = d ds { ∞∑ k=0 [ 1 λk + s(s+ 2n− 1) − 1 λk + a(a+ 2n− 1) ] − 1 4π ∫ ∞ −∞ [ 1( s+ n− 1 2 )2 + r2 − 1( a+ n− 1 2 )2 + r2 ] φ′ φ ( 1 2 + ir ) dr } . To compute this using the resolvent trace formula, we first recall that the Alekseevskii–Barnes double gamma function Γ2(s) is defined as [1, 3]: Γ2(s+ 1) = 1 (2π) s 2 e s 2 + γ+1 2 s2 ∞∏ k=1 ( 1 + s k )−k es− s2 2k . It satisfies the equation Γ2(s+ 1) = Γ2(s) Γ(s) . Using d ds log Γ2(s+ 1) = −1 2 log(2π) + 1 2 + (γ + 1)s− s ∞∑ k=1 ( 1 k − 1 s+ k ) , and ψ(s) = d ds log Γ(s) = −γ − 1 s − ∞∑ k=1 ( 1 s+ k − 1 k ) , Resolvent Trace Formula and Determinants of n Laplacians 13 we find that d ds log (2π)2s+2n−1Γ2(s+ 2n)2Γ2(s) 2Γ(s+ 2n)2n−1 Γ(s)2n+1 = −(2s+ 2n− 1)(ψ(s+ 2n) + ψ(s)− 2). Hence, the resolvent trace formula says that d ds 1 2s+ 2n− 1 d ds log det(∆n + s(s+ 2n− 1)) = d ds 1 2s+ 2n− 1 d ds log [ Z∞(s)Z(s+ n)Zell(s)Zpar(s) ] , where Z∞(s) = [ (2π)2s+2n−1Γ2(s+ 2n)2Γ2(s) 2Γ(s+ 2n)2n−1 Γ(s)2n+1 ] |X| 4π , Zell(s) = v∏ j=1 mj−1∏ r=0 Γ ( s+ r mj ) 2αmj (r−n)+1−mj 2mj Γ ( s+ 2n+ r mj ) 2αmj (r+n)+1−mj 2mj = v∏ j=1 ∏mj−1 r=0 Γ ( s+r mj )αmj (r−n) mj Γ ( s+2n+r mj )αmj (r+n) mj[ (2π)mj−1m −(2s+2n−1) j Γ(s+ 2n)Γ(s) ] 1 2 ( 1− 1 mj ) , Zpar(s) = [ Γ(s)Γ(s+ 2n) 22s+2n−1Γ(s+ n)2Γ ( s+ n+ 1 2 )2]q/2(s+ n− 1 2 )A 2 . (6.1) It follows that det(∆n + s(s+ 2n− 1)) = Z∞(s)Z(s+ n)Zell(s)Zpar(s)e B(s+n− 1 2) 2 +D for some constants B and D. To determine B and D, we need to study the behavior of both sides when s→ ∞. First we define the heat kernel θ(t) = ∞∑ k=0 e−tλk − 1 4π ∫ ∞ −∞ e −t ( r2+[n− 1 2 ] 2 ) φ′ φ ( 1 2 + ir ) dr for t > 0. Notice that ζ(w, s) = 1 Γ(w) ∫ ∞ 0 tw−1θ(t)e−t(s(s+2n−1))dt. Although we can derive the explicit formula for θ(t) using the general trace formula, but we do not need it. We only need the asymptotic expansion of θ(t) as t → 0+. By comparison to the n = 0 or the general theory about heat kernel, we know that the asymptotic expansion of θ(t) has the form θ(t) = ( b t + c log t√ t + d√ t + h ) e−t(n− 1 2) 2 +O (√ t ) for some constants b, c, d and h. Let u = s+ n− 1 2 . 14 L.-P. Teo Then as u→ ∞, log det(∆n + s(s+ 2n− 1)) ∼ − ∂ ∂w ∣∣∣∣ w=0 1 Γ(w) ∫ ∞ 0 tw−1 ( b t + c log t√ t + d√ t + h ) e−tu2 dt. Using ∂ ∂w ∣∣∣∣ w=0 1 Γ(w) ∫ ∞ 0 tw−1e−tu2 dt = −2 log u, ∂ ∂w ∣∣∣∣ w=0 1 Γ(w) ∫ ∞ 0 tw−2e−tu2 dt = −u2 + 2u2 log u, ∂ ∂w ∣∣∣∣ w=0 1 Γ(w) ∫ ∞ 0 tw− 3 2 e−tu2 dt = −2 √ πu, ∂ ∂w ∣∣∣∣ w=0 1 Γ(w) ∫ ∞ 0 tw− 3 2 log t e−tu2 dt = −2 √ πu(2− 2 log 2− γ − 2 log u), we find that as u→ ∞, log det(∆n + s(s+ 2n− 1)) = bu2 − 2bu2 log u+ 2c √ πu(2− 2 log 2− γ − 2 log u) + 2d √ πu+ 2h log u+ o(1). On the other hand, it is obvious from definition that logZ(s + n) is o(1) as s → ∞. From the asymptotic behaviors log Γ(s) = ( s− 1 2 ) log s− s+ 1 2 log 2π + o(1), log Γ2(s+ 1) = −1 2 s2 log s+ 3 4 s2 − s 2 log(2π) + 1 12 log s− ζ ′(−1) + o(1), (6.2) we can deduce the following. As s→ ∞, logZ∞(s) = |X| 4π {( −2u2 + 2n2 − 1 6 ) log u+ 3u2 − 4ζ ′(−1) } + o(1), logZpar(s) = q 2 {−(2u+ 1) log u+ 2u− log(2π)− 2u log 2}+ A 2 log u+ o(1). The asymptotic behavior of Zell(s) is the most complicated one. We leave the computation to Appendix F and quote the result here. We find that as s→ ∞, logZell(s) = B log u+ D + o(1), where B = v∑ j=1 ( m2 j − 1 6mj − αmj (n)(mj − αmj (n)) mj ) , D = − v∑ j=1 ( m2 j − 1 6mj − αmj (n)(mj − αmj (n)) mj ) logmj . Resolvent Trace Formula and Determinants of n Laplacians 15 Hence, we find that as s→ ∞, log det(∆n + s(2 + 2n− 1)) = |X| 4π {( −2u2 + 2n2 − 1 6 ) log u+ 3u2 − 4ζ ′(−1) } + B log u+ D + q 2 { −(2u+ 1) log u+ 2u− log(2π)− 2u log 2 } + A 2 log u+Bu2 +D + o(1) = bu2 − 2bu2 log u+ 2c √ πu(2− 2 log 2− γ − 2 log u) + 2d √ πu+ 2h log u+ o(1). Comparing the two asymptotic expansions give B = −|X| 2π , D = |X| π ζ ′(−1) + q 2 log(2π) + v∑ j=1 ( m2 j − 1 6mj − αmj (n)(mj − αmj (n)) mj ) logmj . (6.3) This gives the following result. Theorem 6.1. Let n be a nonnegative integer and let ∆n be the n-Laplacian of the Riemann surface X. Then det (∆n + s(s+ 2n− 1)) = Z∞(s)Z(s+ n)Zell(s)Zpar(s)e B(s+n− 1 2) 2 +D, where Z(s) is the Selberg zeta function, Z∞(s), Zell(s) and Zpar(s) are defined in (6.1), and the constants B and D are given by (6.3). Since ∆n has zero eigenvalues, we need to remove these zero modes when we define the regularized determininant of ∆n. When n ≥ 2, φ(s+ n) is regular when s = 0. Hence, a reaso- nable definition is det′∆n = lim s→0 det(∆n + s(s+ 2n− 1)) [s(s+ 2n− 1)]dn = 1 (2n− 1)dn lim s→0 det(∆n + s(s+ 2n− 1)) sdn , (6.4) where dn is the dimension of holomorphic n-differentials. When n = 1, Proposition D.1 shows that we can also use (6.4) to define det′∆1. The n = 0 case is more complicated. As discussed in [23], the possible zero of φ(s) at s = 0 would give some extra contribution. If n0 is the order of zero of φ(s) at s = 0, then we should define det′∆0 = lim s→0 det(∆0 + s(s− 1)) [s(s− 1)]1−n0 = (−1)n0−1 lim s→0 det(∆0 + s(s− 1)) s1−n0 . Now we want to derive the formula for det′∆n from Theorem 6.1. We discuss the case n = 0 and n ≥ 1 separately. When n = 0, we find that when s→ 0, Z∞(s) ∼ ( 1 2π ) |X| 4π s− |X| 2π , Zell(s) ∼ v∏ j=1 s mj−1 mj m 1−mj mj j mj−1∏ r=1 Γ ( r mj ) 2r+1−mj mj , Zpar(s) ∼ ( −1 2 )A 2 ( 2 π ) q 2 . 16 L.-P. Teo These imply that det′∆0 = C0 lim s→∞ Z(s) s2g−1+q−n0 , where C0 = (−1) A 2 +1−n02q− A 2 (2π)− q 2 − |X| 4π e B 4 +D v∏ j=1 m 1−mj mj j mj−1∏ r=1 Γ ( r mj ) 2r+1−mj mj . (6.5) When n ≥ 1, we find that as s→ 0, Z∞(s) ∼ s(2n−1) |X| 4π [ (2π)2n−1Γ2(2n) 2Γ(2n)2n−1 ] |X| 4π , Zell(s) ∼ v∏ j=1 { s mj−1−2αmj (−n) 2mj m 2αmj (−n)+1−mj 2mj j × mj−1∏ r=1 Γ ( r mj ) 2αmj (r−n)+1−mj 2mj mj−1∏ r=0 Γ ( 2n+ r mj ) 2αmj (r+n)+1−mj 2mj } , Zpar(s) ∼ s− q 2 [ Γ(2n) 22n−1Γ(n)2Γ ( n+ 1 2 )2]q/2(n− 1 2 )A 2 . Using also the fact that Z(s) has a zero of order 1 at s = 1, and Z(n) is nonzero, we conclude that when n = 1, det′∆1 = C1Z ′(1), while when n ≥ 2, det′∆n = CnZ(n). The constant Cn is given by Cn = [ (2π)2n−1Γ2(2n) 2Γ(2n)2n−1 ] |X| 4π × v∏ j=1 { m 2αmj (−n)+1−mj 2mj j mj−1∏ r=1 Γ ( r mj ) 2αmj (r−n)+1−mj 2mj mj−1∏ r=0 Γ ( 2n+ r mj ) 2αmj (r+n)+1−mj 2mj } × [ 22n−1 πΓ(2n) ]q/2 (2n− 1)−dn ( n− 1 2 )A 2 eB(n− 1 2) 2 +D. (6.6) Finally, we obtain the main result of our paper. Theorem 6.2. When n ≥ 0, the regularized determinant of the n-Laplacian ∆n of X is given by det′∆n =  C0Z0, n = 0, C1Z ′(1), n = 1, CnZ(n), n ≥ 2, where Z0 = lim s→0 Z(s) s2g−1+q−n0 , C0 is given by (6.5) and for n ≥ 1, Cn is given by (6.6). Resolvent Trace Formula and Determinants of n Laplacians 17 This establishes the exact relation between the regularized determinant det′∆n and the Sel- berg zeta function. Notice that the constant Cn only depends on the type of the Riemann surface and is independent of the moduli. In [20] and [21], Takhtajan and Zograf established the local in- dex theorem for Riemann surfaces with cusps and with ramification points, using Z ′(1) and Z(n), n ≥ 2, as defintions for det′∆1 and det′∆n, n ≥ 2 respectively. Since they only considered the sec- ond variation of log det′∆n on the moduli space, Theorem 6.2 justifies their approach. However, if one wants to consider the holomorphic factorization of the determinant of Laplacian, as con- sidered in [16] for compact Riemann surfaces, the precise value of Cn becomes important. They would also be important for the extension of the work [8] by Freixas i Montplet and von Pippich. We would like to point out that our definition of the regularized determinant of Laplacian has taken into account the contribution from the absolutely continuous spectrum. For a generic cofinite Riemann surface with cusps, it is not known whether it has an infinite discrete spectrum. In fact, it has been conjectured to be the opposite. The inclusion of the absolutely continuous spectrum is essential to render the determinant to be equal to a moduli independent constant times Z(n) (for n ≥ 2). In the resolvent trace formula, the contribution from the absolutely continuous spectrum depends on φ(s), the determinant of the scattering matrix, which is moduli dependent. It would be interesting if one can compute the contribution of this term in the local index theorem explicitly. A The inversion formulas In this appendix, we want to derive a useful formula that is needed in the computation of the elliptic contribution to the trace equation (4.1) in Appendix C. When s = 1 2 + ir, we find that −(s− n)(s+ n− 1) = r2 + ( n− 1 2 )2 . Given a function H (λ), define h(r) = H ( r2 + ( n− 1 2 )2) . In this work, we let H (λ) be a function such that the function h(r) satisfies the following conditions: � h(r) is holomorphic in the strip | Im(r)| ≤ 1 2 + δ for some δ > 0, � there exist positive constants M and ε such that |h(r)| ≤ M (1+|r|)2+ε in the strip. Using Proposition 2.1 with k(z, w) the kernel of the operator H (∆n), we find that H (−(s− n)(s+ n− 1)) = ∫ ∞ 0 ∫ ∞ −∞ Ψ ( x2 + (y − 1)2 4y ) (−4)n (x− i(y + 1))2n ys+ndx dy y2 = 2(−1)n ∫ ∞ 0 ∫ ∞ −∞ Ψ ( x2 + (y − 1)2 4y ) 1( x− i (y + 1) 2 √ y )2n y s− 3 2dx dy. 18 L.-P. Teo Let Q(v) = 2(−1)n ∫ ∞ −∞ Ψ(x2 + v) 1( x− i √ v + 1 )2ndx = 2(−1)n ∫ ∞ 0 Ψ(x2 + v) ( x+ i √ v + 1 )2n + ( x− i √ v + 1 )2n( x2 + 1 + v )2n dx. Then h(r) = ∫ ∞ 0 Q ( (y − 1)2 4y ) yir−1dy = ∫ ∞ −∞ Q ( sinh2 t 2 ) eitrdt. Define g(t) = Q ( sinh2 t 2 ) . Then h(r) = ∫ ∞ −∞ g(t)eitrdt. Namely, h(r) is the Fourier transform of g(t). The theory of Fourier transform implies that g(t) = 1 2π ∫ ∞ −∞ h(r)e−irtdr. When Ψ(u) = Ψn,s(u)−Ψn,a(u), where Ψn,s(u) is defined in (2.3), we have h(r) = 1 r2 + ( n− 1 2 )2 + s(s+ 2n− 1) − 1 r2 + ( n− 1 2 )2 + a(a+ 2n− 1) = 1 r2 + ( s+ n− 1 2 )2 − 1 r2 + ( a+ n− 1 2 )2 . It is well-known that h(r) is the Fourier transform of the function g(t) = 1 2s+ 2n− 1 e−t(s+n− 1 2) − 1 2a+ 2n− 1 e−t(a+n− 1 2). Next we want to express Ψ(u) in terms of Q(v) in the particular case where Ψ(u) = Ψn,s(u) − Ψn,a(u). Inspired by the formula in [11], we claim that Ψ(x) = − 1 2π ∫ ∞ −∞ Q′(x+ t2 )(√ x+ 1 + t2 − t )2n dt = − 1 2π ∫ ∞ 0 Q′(x+ t2 ) [(√ x+ 1 + t2 − t )2n + (√ x+ 1 + t2 + t )2n] dt. Since Ψn,s(u) is a linear combination of terms of the form 1 (u+ 1)α , it suffices to consider the case where Ψ(x) = 1 (x+ 1)α . Resolvent Trace Formula and Determinants of n Laplacians 19 In this case, Q(v) = 4(−1)n ∫ ∞ 0 1 (x2 + v + 1)α+2n n∑ r=0 ( 2n 2r ) (v + 1)n−r(−1)n−rx2rdx = 2 n∑ r=0 ( 2n 2r ) (v + 1)n−r(−1)r ∫ ∞ 0 xr− 1 2 (x+ v + 1)α+2n dx = 2 (v + 1)α+n− 1 2 n∑ r=0 ( 2n 2r ) (−1)r Γ ( r + 1 2 ) Γ ( α+ 2n− r − 1 2 ) Γ(α+ 2n) . Some manipulations give Q(v) = 2 (v + 1)α+n− 1 2 Γ ( α+ 2n− 1 2 ) Γ(α+ 2n) Γ ( 1 2 ) n∑ r=0 (−n)r ( −n+ 1 2 ) r( 1 2 ) r r! ( 1 2 ) r( −α− 2n+ 3 2 ) r = 2 √ π (v + 1)α+n− 1 2 Γ ( α+ 2n− 1 2 ) Γ(α+ 2n) 2F1 ( −n, −n+ 1 2 −α− 2n+ 3 2 ; 1 ) = 2 √ π (v + 1)α+n− 1 2 Γ ( α+ 2n− 1 2 ) Γ(α+ 2n) Γ ( −α− 2n+ 3 2 ) Γ (−α+ 1) Γ (−α− n+ 1)Γ ( −α− n+ 3 2 ) = 2 √ π (v + 1)α+n− 1 2 Γ(α+ n)Γ ( α+ n− 1 2 ) Γ(α)Γ(α+ 2n) . It follows that Q′(v) = − 2 √ π (v + 1)α+n+ 1 2 Γ(α+ n)Γ ( α+ n+ 1 2 ) Γ(α)Γ(α+ 2n) . Next, we compute J(x) = ∫ ∞ −∞ 1( x+ 1 + t2 )α+n+ 1 2 (√ x+ 1 + t2 − t )2n dt = 2 ∫ ∞ 0 1( x+ 1 + t2 )α+n+ 1 2 n∑ r=0 ( 2n 2r ) (x+ 1 + t2)n−rt2rdt = 2 n∑ r=0 ( 2n 2r )∫ ∞ 0 t2r( x+ 1 + t2 )α+r+ 1 2 dt = 1 (x+ 1)α n∑ r=0 (−n)r ( −n+ 1 2 ) r( 1 2 ) r r! Γ ( r + 1 2 ) Γ (α) Γ ( α+ r + 1 2 ) . It follows that J(x) = Γ(α) Γ ( α+ 1 2 ) √ π (x+ 1)α n∑ r=0 (−n)r ( −n+ 1 2 ) r r! 1( α+ 1 2 ) r = Γ(α) Γ ( α+ 1 2 ) √ π (x+ 1)α 2F1 ( −n, −n+ 1 2 α+ 1 2 ; 1 ) = Γ(α) Γ ( α+ 1 2 ) √ π (x+ 1)α Γ ( α+ 1 2 ) Γ(α+ 2n) Γ(α+ n)Γ ( α+ n+ 1 2 ) . 20 L.-P. Teo This proves that for Ψ(x) = 1 (x+ 1)α , the inversion formula holds. Namely, if Q(v) = 2(−1)n ∫ ∞ −∞ Ψ ( x2 + v ) 1( x− i √ v + 1 )2ndx, then Ψ(x) = − 1 2π ∫ ∞ −∞ Q′(x+ t2 )(√ x+ 1 + t2 − t )2n dt. Thus, this formula works for Ψ(u) = Ψn,s(u)−Ψn,a(u). B The hyperbolic contribution In this section, we want to compute the hyperbolic contribution ΞH = ∑ γ∈Γ γ is hyperbolic ∫∫ Γ\H k(γz, z)γ′(z)ny2ndµ(z) to the trace equation (4.1). For any γ ∈ Γ, let Γγ = { α ∈ Γ | αγα−1 = γ } . It is a cyclic subgroup of Γ generated by some γ0 ∈ Γ such that γ = γℓ0 for some nonzero integer ℓ. The elements in Γ are partitioned into conjugacy classes. The conjugacy class of γ consists of all elements in Γ of the form αγα−1, when α runs through all elements in Γ. Two elements α and β in Γ define the same element in the class containing γ, namely, αγα−1 = βγβ−1, if and only if α−1β ∈ Γγ . Let γ ∈ Γ be a hyperbolic element. There exists σ ∈ PSL(2,R) and a real number λγ > 1 such that σ−1γσ = Dγ = ( λ 1/2 γ 0 0 λ −1/2 γ ) . λγ is called the multiplier of γ. Each hyperbolic element γ in Γ belongs to a conjugacy class. All elements in the same conjugacy class has the same multiplier. There exists γ0 ∈ Γ such that γ = γℓ0 for some nonzero integer ℓ, and Γγ is generated by γ0. γ0 is called a primitive hyperbolic element. Let P be the set of conjugacy classes of primitive hyperbolic elements in Γ. Then the set of hyperbolic elements in Γ can be written as ⋃ [γ0]∈P ∞⋃ ℓ=1 { αγℓ0α −1 | α ∈ Γγ0\Γ } . Resolvent Trace Formula and Determinants of n Laplacians 21 Given γ0 a representative of a primmitive hyperbolic conjugacy class, and ℓ a positive integer, let γ = γℓ0. We first consider the integral IH = ∑ α∈Γγ\Γ ∫∫ F k ( αγα−1z, z ) [ (αγα)′(z) ]n y2ndµ(z). Using the invariant property of the kernel k(z, w), we find that IH = ∫∫ σ−1(Γγ\H) k(λγz, z)λ n γy 2ndµ(z). In the following, we abbreviate λγ as λ. Let λ = λℓ0. Then σ−1 (Γγ\H) = {x+ iy | 1 ≤ y ≤ λ0} . It follows that IH = λn ∫ λ0 1 ∫ ∞ −∞ k(λz, z)y2ndx dy y2 = λn ∫ λ0 1 ∫ ∞ −∞ Ψ ( (λ− 1)2 4λ ( x2 + 1 )) (−4)n [(λ− 1)x+ i(λ+ 1)]2n dx dy y = 2(−1)nλ 1 2 λ− 1 lnλ0 ∫ ∞ −∞ Ψ ( x2 + (λ− 1)2 4λ ) 1[ x+ iλ+1 2 √ λ ]2ndx = √ λ λ− 1 lnλ0 g(lnλ) = 1 2s+ 2n− 1 λ−s−n 1− λ−1 lnλ0 − 1 2a+ 2n− 1 λ−a−n 1− λ−1 lnλ0. Now we can sum up the contributions from γ = γℓ0, where ℓ ranges from 1 to ∞. To avoid confusion with the eigenvalues of Laplacian, we denote λγ by N(γ). Since ∞∑ ℓ=1 λ−ℓ(s+n) 1− λ−ℓ = ∞∑ k=0 ∞∑ ℓ=1 λ−ℓ(s+n+k) = ∞∑ k=0 1 λs+k+n − 1 , we find that ΞH = EH(s)− EH(a), where EH(s) = 1 2s+ 2n− 1 ∑ [γ]∈P ∞∑ k=0 N(γ)−(s+n+k) 1−N(γ)−(s+n+k) lnN(γ). C Elliptic contribution In this appendix, we compute the elliptic contribution ΞE = ∑ γ∈Γ γ is elliptic ∫∫ Γ\H k(γz, z)γ′(z)ny2ndµ(z) to the trace equation (4.1). For simplicity, let us just do the calculation with Ψ(u) = Ψn,s(u), 22 L.-P. Teo understanding that the singularity at u = 0 is eliminated by subtracting Ψn,a(u). At the end, we will restore the result with Ψ(u) = Ψn,s(u)−Ψn,a(u). Recall that the Riemann surface X has v ramification points corresponding to the v elliptic elements τ1, . . . , τv of orders m1, . . . ,mv respectively. For each 1 ≤ j ≤ v, there is a σj ∈ PSL(2,R) such that σ−1 j τjσj = ( cos θj − sin θj sin θj cos θj ) , where θj = π mj . If γ is an elliptic element of Γ, then there exists α ∈ Γ, 1 ≤ j ≤ v and an integer 1 ≤ ℓ ≤ mj − 1 so that α−1γα = τ ℓj . This implies that ΞE = v∑ j=1 mj−1∑ ℓ=1 ∑ α∈Γτj \Γ ∫∫ Γ\H k ( ατ ℓjα −1z, z )[( ατ ℓjα −1 )′ (z) ]n y2ndµ(z) = v∑ j=1 mj−1∑ ℓ=1 ∫∫ Γτj \H k ( τ ℓj z, z )( τ ℓj )′ (z)ny2ndµ(z) = v∑ j=1 mj−1∑ ℓ=1 IE,j(θj,ℓ), where θj,ℓ = πℓ mj , and IE(θ) = 1 mj ∫∫ H k(Rθz, z)R ′ θ(z) ny2ndµ(z), Rθ = ( cos θ − sin θ sin θ cos θ ) . In the following, we omit the subscript j in our calculation of IE,j . Now, u(Rθz, z) + 1 = sin2 θ ( 1 + x2 + y2 )2 + 4 cos2 θy2 4y2 , (−4)n R′ θ(z) n (Rθz − z̄)2n = (−4)n( sin θ ( 1 + x2 + y2 ) − 2iy cos θ )2n . With t = 1 + x2 + y2 2y , we have x = √ 2yt− 1− y2, dx = y dt√ 2yt− 1− y2 . Then IE(θ) = (−1)n 2 m ∫ ∞ 0 ∫ ∞ 1+y2 2y Ψ ( t2 sin2 θ − sin2 θ ) (t sin θ − i cos θ)2n dt√ 2yt− 1− y2 dy y Resolvent Trace Formula and Determinants of n Laplacians 23 = (−1)n 2 m ∫ ∞ 1 ∫ t+ √ t2−1 t− √ t2−1 dy y √ 2yt− 1− y2 Ψ ( t2 sin2 θ − sin2 θ ) (t sin θ − i cos θ)2n dt. It has been computed that∫ t+ √ t2−1 t− √ t2−1 dy y √ 2yt− 1− y2 = π. Therefore, IE(θ) = (−1)n 2π m ∫ ∞ 1 Ψ ( t2 sin2 θ − sin2 θ )( t sin θ − i cos θ )2n dt. This integral is difficult to compute. Fischer [7] used the power series expansion of Ψn,s(u) followed by a lot of tedious calculations. The generalization from n = 0 to general positive n is not easy, and Hejhal was unable to proceed in his first volume [11]; only succeeded to circumvent the problem in his second volume [12]. We will use the method of Hejhal [12]. First of all, using the substitution u = t2 sin2 θ − sin2 θ, we find that IE(θ) = (−1)n π m sin θ ∫ ∞ 0 Ψ(u)(√ u+ sin2 θ − i cos θ )2n du√ u+ sin2 θ . Using the inversion formula (see Appendix A) Ψ(u) = − 1 2π ∫ ∞ −∞ Q′(u+ t2 )(√ u+ 1 + t2 − t )2n dt, we find that IE(θ) = (−1)n+1 2m sin θ ∫ ∞ 0 ∫ ∞ −∞ Q′(u+ t2 ) (√ u+ 1 + t2 − t )2n(√ u+ sin2 θ − i cos θ )2ndt du√ u+ sin2 θ = (−1)n+1 2m sin θ ∫ ∞ −∞ ∫ ∞ t2 Q′(v) (√ v + 1− t )2n(√ v − t2 + sin2 θ − i cos θ )2n dv√ v − t2 + sin2 θ dt = (−1)n+1 2m sin θ ∫ ∞ 0 Q′(v)H (v) dv, where H (v) = ∫ √ v − √ v (√ v + 1− t )2n(√ v − t2 + sin2 θ − i cos θ )2n dt√ v − t2 + sin2 θ . Making a change of variables t = √ v + sin2 θ sinφ, we find that H (v) = ∫ φ0(v) −φ0(v) (√ v + 1− √ v + sin2 θ sinφ )2n(√ v + sin2 θ cosφ− i cos θ )2n dφ, where φ0(v) = sin−1 √ v√ v + sin2 θ . 24 L.-P. Teo By definition, H (0) = 0, and hence, IE(θ) = (−1)n 2m sin θ ∫ ∞ 0 Q(v)H ′(v) dv. Now, we calculate H ′(v). It is straightforward to find that H ′(v) = φ′ 0(v) [(√ v + 1− √ v + sin2 θ sinφ0 )2n(√ v + sin2 θ cosφ0 − i cos θ )2n + (√ v + 1 + √ v + sin2 θ sinφ0 )2n(√ v + sin2 θ cosφ0 − i cos θ )2n ] + ∫ φ0(v) −φ0(v) d dv [√ v + 1− √ v + sin2 θ sinφ√ v + sin2 θ cosφ− i cos θ ]2n dφ. If we define ω so that sinhω = cos θ√ v + sin2 θ , we find that √ v + 1√ v + sin2 θ = coshω. Observe that d dv √ v + 1− √ v + sin2 θ sinφ√ v + sin2 θ cosφ− i cos θ = d dω coshω − sinφ cosφ− i sinhω dω dv = −i d dφ coshω − sinφ cosφ− i sinhω dω dv = − i cos θ 2(v + sin2 θ) √ v + 1 d dφ coshω − sinφ cosφ− i sinhω . These imply that H ′(v) = sin θ 2 √ v(v + sin2 θ) [(√ v + 1− √ v )2n (sin θ − i cos θ)2n + (√ v + 1 + √ v )2n (sin θ − i cos θ)2n ] − i cos θ 2(v + sin2 θ) √ v + 1 ∫ φ0(v) −φ0(v) d dφ [√ v + 1− √ v + sin2 θ sinφ√ v + sin2 θ cosφ− i cos θ ]2n dφ = (−1)n sin θ e2inθ 2 √ v(v + sin2 θ) [(√ v + 1− √ v )2n + (√ v + 1 + √ v )2n] − (−1)ni cos θ e2inθ 2(v + sin2 θ) √ v + 1 [(√ v + 1− √ v )2n − (√ v + 1 + √ v )2n] . It follows that H ′ ( sinh2 t 2 ) d dt sinh2 t 2 = (−1)ne2inθ 2 ( sinh2 t 2 + sin2 θ ){ sin θ cosh t 2 (etn + e−tn) + i cos θ sinh t 2 (etn − e−tn) } = (−1)ne2inθ cosh t− cos 2θ { sin θ ( e t 2 + e− t 2 ) (etn + e−tn) 2 + i cos θ ( e t 2 − e− t 2 ) (etn − e−tn) 2 } = (−1)ne2inθ cosh t− cos 2θ { sin θ e(n+ 1 2)t + e−(n+ 1 2)t + e(n− 1 2)t + e−(n− 1 2)t 2 Resolvent Trace Formula and Determinants of n Laplacians 25 + i cos θ e(n+ 1 2)t + e−(n+ 1 2)t − e(n− 1 2)t − e−(n− 1 2)t 2 } . This implies that IE(θ) = (−1)n 2m sin θ ∫ ∞ 0 Q ( sinh2 t 2 ) H ′ ( sinh2 t 2 ) d dt sinh2 t 2 dt = e2inθ 2m sin θ ∫ ∞ 0 g(t) cosh t− cos 2θ { sin θ e(n+ 1 2)t + e−(n+ 1 2)t + e(n− 1 2)t + e−(n− 1 2)t 2 + i cos θ e(n+ 1 2)t + e−(n+ 1 2)t − e(n− 1 2)t − e−(n− 1 2)t 2 } . Using g(t) = 1 2s+ 2n− 1 e−t(s+n− 1 2), and ∫ ∞ 0 e−µt cosh t− cos 2θ dt = 2 sin 2θ ∞∑ k=1 sin 2kθ µ+ k , we find that IE(θ) = 1 (2s+ 2n− 1) e2inθ 2m sin θ 1 sin 2θ × { sin θ [ ∞∑ k=1 sin 2kθ s+ k − 1 + ∞∑ k=1 sin 2kθ s+ k + ∞∑ k=1 sin 2kθ s+ k + 2n + ∞∑ k=1 sin 2kθ s+ k + 2n− 1 ] + i cos θ [ ∞∑ k=1 sin 2kθ s+ k − 1 − ∞∑ k=1 sin 2kθ s+ k + ∞∑ k=1 sin 2kθ s+ k + 2n − ∞∑ k=1 sin 2kθ s+ k + 2n− 1 ]} = 1 (2s+ 2n− 1) e2inθ 2m sin θ 1 sin 2θ × { sin θ [ ∞∑ k=0 sin(2k + 2)θ s+ k + ∞∑ k=0 sin 2kθ s+ k + ∞∑ k=0 sin 2kθ s+ k + 2n + ∞∑ k=0 sin(2k + 2)θ s+ k + 2n ] + i cos θ [ ∞∑ k=0 sin(2k+2)θ s+ k − ∞∑ k=0 sin 2kθ s+ k + ∞∑ k=0 sin 2kθ s+ k + 2n − ∞∑ k=0 sin(2k+2)θ s+ k + 2n ]} = 1 2s+ 2n− 1 ie2inθ 2m sin θ ( ∞∑ k=0 e−i(2k+1)θ s+ k − ∞∑ k=0 ei(2k+1)θ s+ 2n+ k ) . Now we find the sum m−1∑ ℓ=1 IE ( πℓ m ) . Notice that is k is an integer, Sk = m−1∑ ℓ=1 i sin πℓ m exp ( −πiℓ(2k + 1) m ) 26 L.-P. Teo = 1 2 [ m−1∑ ℓ=1 i sin πℓ m exp ( −πiℓ(2k + 1) m ) + m−1∑ ℓ=1 i sin π(m−ℓ) m exp ( −πi(m− ℓ)(2k + 1) m )] = m−1∑ ℓ=1 exp ( πiℓ(2k+1) m ) − exp ( −πiℓ(2k+1) m ) exp ( πiℓ m ) − exp ( −πiℓ m ) = m−1∑ ℓ=1 x2k+1 ℓ − x−2k−1 ℓ xℓ − x−1 ℓ , where xℓ = exp ( πiℓ m ) . Using the fact that m−1∑ ℓ=1 x2hℓ = { m− 1 if m divides h, −1 if m does not divide h, we find that if k is a nonnegative integer Sk = m−1∑ ℓ=1 ( x2kℓ + x2k−2 ℓ + · · ·+ x−2k+2 ℓ + x−2k ℓ ) = −2k − 1 +m ( 2 ⌊ k m ⌋ + 1 ) = −2m { k m } +m− 1. Here {x} = x−⌊x⌋. The case when k is a negative integer is more complicated. First we notice that for any integer x, if q = ⌊ x m ⌋ , then x = qm+ r, where 0 ≤ r ≤ m− 1. It follows that −x− 1 = −qm− r − 1 = (−q − 1)m+m− r − 1, where 0 ≤ m− r − 1 ≤ m− 1. This shows that⌊ −x− 1 m ⌋ = −q − 1 = − ⌊ x m ⌋ − 1. If k is a negative integer, let k′ = −k. Then k′ is a positive integer and Sk = − m−1∑ ℓ=1 x2k ′−1 ℓ − x−2k′+1 ℓ xℓ − x−1 ℓ = − m−1∑ ℓ=1 ( x2k ′−2 ℓ + x2k ′−4 ℓ + · · ·+ x−2k′+4 ℓ + x−2k′+2 ℓ ) = − { −2k′ + 1 +m ( 2 ⌊ k′ − 1 m ⌋ + 1 )} = −2k − 1−m ( −2 ⌊ k m ⌋ − 1 ) = −2m { k m } +m− 1. This shows that for any integer k, Sk = −2αm(k) +m− 1, Resolvent Trace Formula and Determinants of n Laplacians 27 where αm(k) = m { k m } is the least positive residue of k modulo m. We have established that αm(−k) = m− 1− αm(k). Hence, S−k = −Sk. It follows that m−1∑ ℓ=1 IE ( πℓ m ) = 1 2m 1 2s+ 2n− 1 { ∞∑ k=0 1 s+ k (−2αm(k − n) +m− 1) + ∞∑ k=0 1 s+ 2n+ k (−2αm(k + n) +m− 1) } . Writing k = mq + r, where 0 ≤ r ≤ m− 1, we have m−1∑ ℓ=1 IE ( πℓ m ) = 1 2m 1 2s+ 2n− 1 { ∞∑ q=0 m−1∑ r=0 1 s+mq + r (−2αm(r − n) +m− 1) + ∞∑ q=0 m−1∑ r=0 1 s+ 2n+mq + r (−2αm(r + n) +m− 1) } . Notice that when r runs from 0 to m−1, r−n and r+n respectively runs through a complete residue system modulo m. Hence, m−1∑ r=0 αm(r − n) = m−1∑ r=0 [m− 1− αm(r − n)] and m−1∑ r=0 αm(r + n) = m−1∑ r=0 [m− 1− αm(r + n)]. Therefore, m−1∑ r=0 (−2αm(r − n) +m− 1) = m−1∑ r=0 (−2αm(r + n) +m− 1) = 0. (C.1) Thus m−1∑ ℓ=1 IE ( πℓ m ) = 1 2m 1 2s+ 2n− 1 { ∞∑ q=0 m−1∑ r=0 [ 1 s+mq+ r − 1 m(q + 1) ] (−2αm(r − n) +m− 1) + ∞∑ q=0 m−1∑ r=0 [ 1 s+ 2n+mq + r − 1 m(q + 1) ] (−2αm(r + n) +m− 1) } 28 L.-P. Teo = 1 2s+ 2n− 1 m−1∑ r=0 [ 2αm(r − n) + 1−m 2m2 ψ ( s+ r m ) + 2αm(r + n) + 1−m 2m2 ψ ( s+ 2n+ r m )] . Finally, we find that the elliptic contribution is given by ΞE = EE(s)− EE(a), where EE(s) = 1 2s+ 2n− 1 v∑ j=1 mj−1∑ r=0 [ 2αmj (r − n) + 1−mj 2m2 j ψ ( s+ r mj ) + 2αmj (r + n) + 1−mj 2m2 j ψ ( s+ 2n+ r mj )] . D Contribution from parabolic elements and absolutely continuous spectrum In this appendix, we compute the parabolic contribution ΞP to the trace equation (4.1). D.1 Contribution from parabolic elements In this section, we compute the term ΞP,1 = ∫∫ FY ∑ γ∈Γ γ is parabolic k(γz, z)γ′(z)ny2ndµ(z). Again we first consider the case where Ψ(u) = Ψn,s(u). Recall that the Riemann surface X has q cusps corresponding to the q parabolic elements κ1, . . . , κq. For each 1 ≤ i ≤ q, there is a σi ∈ PSL(2,R) such that σ−1 i κiσi = ( 1 ±1 0 1 ) . If γ is a parabolic element of Γ, then there exists α ∈ Γ, 1 ≤ j ≤ q and a nonzero integer ℓ so that α−1γα = κℓj . For an integer ℓ, define Tℓ = ( 1 ℓ 0 1 ) . Then ΞP,1 = ∫∫ FY q∑ j=1 ∑ ℓ ̸=0 ∑ α∈Γκj \Γ k ( ακℓjα −1z, z )[( ακℓjα −1 )′ (z) ]n y2ndµ(z) = q∑ j=1 ∑ ℓ ̸=0 ∑ α∈σ−1 j Γκiσj\σ−1 j Γσj ∫∫ σ−1 j (FY ) k ( αTℓα −1z, z )[( αTℓα −1 )′ (z) ]n y2ndµ(z) = q∑ j=1 ∑ ℓ ̸=0 ∫∫ HY j k(Tℓz, z)T ′ ℓ(z) ny2ndµ(z). Resolvent Trace Formula and Determinants of n Laplacians 29 Here HY j = HY \ ⋃ k ̸=j ⋃ α∈Γκj \Γ σ−1 j ( α ( F Y k )) , where HY = { x+ iy | 0 ≤ x ≤ 1, 0 < y ≤ Y } . Since the hyperbolic area of H \HY is O ( Y −1 ) , we find that ΞP,1 = qIP +O ( Y −1 ) , where IP = ∑ ℓ̸=0 ∫∫ HY k(Tℓz, z)T ′ ℓ(z) ny2ndµ(z). Now u(Tℓz, z) = ℓ2 4y2 , (−4)n T ′ ℓ(z) n (Tℓz − z̄)2n = (−4)n (ℓ+ 2iy)2n . Hence, we have IP = (−1)n ∞∑ ℓ=1 ∫ Y 0 Ψ ( ℓ2 4y2 )[ (2y)2n (l + 2iy)2n + (2y)2n (l − 2iy)2n ] dy y2 = (−1)n ∞∑ ℓ=1 2 ℓ ∫ 2Y/ℓ 0 Ψ ( 1 y2 )[ y2n (1 + iy)2n + y2n (1− iy)2n ] dy y2 . As in [13], if F (y) is a well-behaved function, standard techniques in analytic number theory give ∞∑ ℓ=1 2 ℓ ∫ 2Y ℓ 0 F (y) dy = ∫ ∞ 1− ∫ 2Y v 0 F (y) dy d (∑ ℓ≤v 2 ℓ ) = ∫ 2Y 0 F (y) [∫ 2Y/y 1− d (∑ ℓ≤v 2 ℓ )] dy = ∫ 2Y 0 F (y) [ 2 log 2Y y + 2γ +O ( yY −1 )] dy. Hence, IP = (−1)n ∫ 2Y 0 Ψ ( 1 y2 )[ y2n (1 + iy)2n + y2n (1− iy)2n ][ 2 log 2Y y + 2γ +O ( yY −1 )]dy y2 = (log(2Y ) + γ)IP,0 + IP,1 +O ( Y −1 log Y ) , where IP,0 = 2(−1)n ∫ ∞ 0 Ψ ( 1 y2 )[ y2n (1 + iy)2n + y2n (1− iy)2n ] dy y2 , IP,1 = 2(−1)n ∫ ∞ 0 Ψ ( 1 y2 )[ y2n (1 + iy)2n + y2n (1− iy)2n ] log 1 y dy y2 . 30 L.-P. Teo A straightforward computation gives IP,0 = 2(−1)n ∫ ∞ 0 Ψ(x2) [ 1 (x+ i)2n + 1 (x− i)2n ] dx = Q(0) = 1 2s+ 2n− 1 . The computation of IP,1 is more complicated: IP,1 = 2(−1)n ∫ ∞ 0 Ψ(u2) [ 1 (u+ i)2n + 1 (u− i)2n ] log udu = (−1)n π ∫ ∞ 0 ∞∑ k=0 Γ(s+ k)Γ(s+ 2n+ k) Γ(2s+ 2n+ k) 1 k! 1 (u2 + 1)k+s+2n × n∑ m=0 ( 2n 2m ) u2m(−1)n−m log udu = 1 4π n∑ m=0 ( 2n 2m ) (−1)m ∞∑ k=0 Γ(s+ k)Γ(s+ 2n+ k) Γ(2s+ 2n+ k) 1 k! ∫ ∞ 0 um (u+ 1)k+s+2n log u 1√ u du. This implies that IP,1 = G′(0), where G(α) = 1 4π n∑ m=0 ( 2n 2m ) (−1)m ∞∑ k=0 Γ(s+ k)Γ(s+ 2n+ k) Γ(2s+ 2n+ k) 1 k! ∫ ∞ 0 um+α− 1 2 (u+ 1)k+s+2n du. From [10], we find that∫ ∞ 0 um+α− 1 2 (u+ 1)k+s+2n du = Γ ( m+ α+ 1 2 ) Γ ( s+ 2n+ k −m− α− 1 2 ) Γ(s+ 2n+ k) . Hence, G(α) = 1 4π n∑ m=0 ( 2n 2m ) (−1)mΓ ( m+ α+ 1 2 ) ∞∑ k=0 Γ(s+ k)Γ ( s+ 2n+ k −m− α− 1 2 ) Γ(2s+ 2n+ k) 1 k! . Using the definition and identity of hypergeometric functions (see [2]), we have ∞∑ k=0 Γ(s+ k)Γ ( s+ 2n+ k −m− α− 1 2 ) Γ(2s+ 2n+ k) 1 k! = Γ(s)Γ ( s+ 2n−m− α− 1 2 ) Γ (2s+ 2n) 2F1 ( s, s+ 2n−m− α− 1 2 2s+ 2n ; 1 ) = Γ(s)Γ ( s+ 2n−m− α− 1 2 ) Γ ( m+ α+ 1 2 ) Γ (s+ 2n) Γ ( s+m+ α+ 1 2 ) . Now, we notice that( 2n 2m ) = (2n)(2n− 1) · · · (2n− 2m+ 1) 1 · 2 · · · (2m− 1)(2m) = (−n)m ( −n+ 1 2 ) m( 1 2 ) m m! , (−1)mΓ ( s+ 2n−m− α− 1 2 ) = (−1)m Γ ( s+ 2n− α− 1 2 )( s+ 2n− α− 1 2 − 1 ) · · · ( s+ 2n− α− 1 2 −m ) Resolvent Trace Formula and Determinants of n Laplacians 31 = Γ ( s+ 2n− α− 1 2 )( −s− 2n+ α+ 1 2 + 1 ) · · · ( −s− 2n+ α+ 1 2 +m ) = Γ ( s+ 2n− α− 1 2 )( −s− 2n+ α+ 3 2 ) m . This gives G(α) = 1 4π Γ ( α+ 1 2 )2 Γ(s)Γ ( s+ 2n− α− 1 2 ) Γ(s+ 2n)Γ ( s+ α+ 1 2 ) H(α), where H(α) = n∑ m=0 (−n)m ( −n+ 1 2 ) m( 1 2 ) m m! ( α+ 1 2 ) m ( α+ 1 2 ) m( −s− 2n+ α+ 3 2 ) m ( s+ α+ 1 2 ) m = 4F3 ( −n, −n+ 1 2 , α+ 1 2 , α+ 1 2 1 2 , −s− 2n+ α+ 3 2 , s+ α+ 1 2 ; 1 ) . This is a balanced hypergeometric series. According to [2, Theorem 3.3.3], we find that H(α) = Γ(α+ 1− s)Γ(α+ s+ 2n)Γ ( α− s− 2n+ 3 2 ) Γ ( α+ s+ 1 2 ) Γ(α+ 1− s− n)Γ(α+ s+ n)Γ ( α− s− n+ 3 2 ) Γ ( α+ s+ n+ 1 2 ) × 4F3 ( −n, −n+ 1 2 , −α, −α 1 2 , s− α, 1− s− 2n− α ; 1 ) . As a function of α, 4F3 ( −n, −n+ 1 2 , −α, −α 1 2 , s− α, 1− s− 2n− α ; 1 ) = 1 +O ( α2 ) as α→ 0. Hence, IP,1 = G′(0) = L′(0), where L(α) is given by L′(α) = 1 4π Γ ( α+ 1 2 )2 Γ(s)Γ ( s+ 2n− α− 1 2 ) Γ(α+ 1− s)Γ(α+ s+ 2n)Γ ( α− s− 2n+ 3 2 ) Γ(s+ 2n)Γ(α+ 1− s− n)Γ(α+ s+ n)Γ ( α− s− n+ 3 2 ) Γ ( α+ s+ n+ 1 2 ) . It is then straightforward to obtain IP,1 = 1 4π Γ ( 1 2 )2 Γ(s)Γ ( s+ 2n− 1 2 ) Γ(1− s)Γ(s+ 2n)Γ ( −s− 2n+ 3 2 ) Γ (s+ 2n) Γ(1− s− n)Γ(s+ n)Γ ( −s− n+ 3 2 ) Γ ( s+ n+ 1 2 ) × { 2ψ ( 1 2 ) − ψ ( s+ 2n− 1 2 ) + ψ(1− s) + ψ(s+ 2n) + ψ ( −s− 2n+ 3 2 ) − ψ(1− s− n)− ψ(s+ n)− ψ ( −s− n+ 3 2 ) − ψ ( s+ n+ 1 2 )} . Using the fact that Γ(z + 1) = zΓ(z), we find that Γ ( s+ n+ 1 2 ) = ( s+ n− 1 2 ) Γ ( s+ n− 1 2 ) , and ψ ( s+ n+ 1 2 ) = 1 s+ n− 1 2 + ψ ( s+ n− 1 2 ) . 32 L.-P. Teo On the other hand, using Γ(z)Γ(1− z) = π sinπz , we find that Γ(s)Γ(1− s) Γ(1− s− n)Γ(s+ n) = sinπ(s+ n) sinπs = (−1)n, Γ(s+ 2n− 1 2)Γ ( −s− 2n+ 3 2 ) Γ ( −s− n+ 3 2 ) Γ ( s+ n− 1 2 ) = sinπ ( s+ n− 1 2 ) sinπ(s+ 2n− 1 2) = (−1)n, ψ(s)− ψ(1− s) + ψ(1− s− n)− ψ(s+ n) = 0, ψ ( s+ n− 1 2 ) − ψ ( −s− 2n+ 3 2 ) + ψ ( −s− n+ 3 2 ) − ψ ( s+ n− 1 2 ) = 0. Together with Γ ( 1 2 ) = √ π, ψ ≤ ( 1 2 ) = −γ − 2 log 2, we have 1 4π Γ ( 1 2 )2 Γ(s)Γ ( s+ 2n− 1 2 ) Γ(1− s)Γ ( −s− 2n+ 3 2 ) Γ(1− s− n)Γ(s+ n)Γ ( −s− n+ 3 2 ) Γ ( s+ n+ 1 2 ) = 1 2(2s+ 2n− 1) , and IP,1 = 1 2(2s+ 2n− 1) { ψ(s) + ψ(s+ 2n)− 4 log 2− 2ψ ( s+ n+ 1 2 ) − 2ψ(s+ n) } − γ 2s+ 2n− 1 + 1 (2s+ 2n− 1)2 . Collecting the various contributions, we find that IP = 1 2s+ 2n− 1 log Y + 1 (2s+ 2n− 1)2 + 1 2(2s+ 2n− 1) { ψ(s) + ψ(s+ 2n)− 2 log 2− 2ψ ( s+ n+ 1 2 ) − 2ψ(s+ n) } . Therefore, ΞP,1 = EP,1(s)− EP,1(a), where EP,1(s) = q 2s+ 2n− 1 log Y + q (2s+ 2n− 1)2 + q 2(2s+ 2n− 1) × { ψ(s) + ψ(s+ 2n)− 2 log 2− 2ψ ( s+ n+ 1 2 ) − 2ψ(s+ n) } +O ( Y −1 ) . D.2 Contribution of the absolutely continuous spectrum In this section, we want to compute J = 1 4π q∑ j=1 ∫ ∞ −∞ h(r) ∫∫ FY ∣∣∣∣Ej ( z, 1 2 + ir;n )∣∣∣∣2y2ndµ(z) dr, Resolvent Trace Formula and Determinants of n Laplacians 33 where h(r) = Λ̃ ( 1 4 + r2 ) = 1 r2 + ( s+ n− 1 2 )2 − 1 r2 + ( a+ n− 1 2 )2 . From Appendix E, we find that q∑ j=1 ∫∫ FY ∣∣∣∣Ej ( z, 1 2 + ir;n )∣∣∣∣2y2ndµ(z) = 1 2ir Tr [ Φ ( 1 2 − ir ) Y 2ir − Φ ( 1 2 + ir ) Y −2ir ] + 2q log Y − φ′ φ ( 1 2 + ir ) . Therefore, J = J1 + J2 + J3, where J1 = q log Y 2π ∫ ∞ −∞ h(r) dr, J2 = 1 8πi ∫ ∞ −∞ h(r) r Tr [ Φ ( 1 2 − ir ) Y 2ir − Φ ( 1 2 + ir ) Y −2ir ] dr, J3 = − 1 4π ∫ ∞ −∞ h(r) φ′ φ ( 1 2 + ir ) dr. Since 1 2π ∫ ∞ −∞ h(r) dr = g(0), we find that J1 = q log Y ( 1 2s+ 2n− 1 − 1 2a+ 2n− 1 ) . As in [13], J2 = 1 8πi ∫ ∞ −∞ h(r) r Tr [ Φ ( 1 2 − ir ) Y 2ir − Φ ( 1 2 + ir ) Y −2ir ] dr = 1 4 TrΦ ( 1 2 ) h(0) +O ( Y −2ε ) = [ 1 (2s+ 2n− 1)2 − 1 (2a+ 2n− 1)2 ] TrΦ ( 1 2 ) +O ( Y −2ε ) , where ε is an arbitrary positive number. Collecting everything together, we find that J = q log Y ( 1 2s+ 2n− 1 − 1 2a+ 2n− 1 ) + [ 1 (2s+ 2n− 1)2 − 1 (2a+ 2n− 1)2 ] TrΦ ( 1 2 ) − 1 4π ∫ ∞ −∞ h(r) φ′ φ ( 1 2 + ir ) dr +O ( Y −2ε ) . When h(r) = 1 r2+α2 for some α with Reα > 0, we can follow the method of [23] to compute the integral J3 = − 1 4π ∫ ∞ −∞ h(r) φ′ φ ( 1 2 + ir ) dr. 34 L.-P. Teo From (3.1) and the results for the case n = 0, we can deduce that φ(s) is holomorphic on the half plane Re s ≥ 1 2 except for for a finite number of poles on ( 1 2 , 1 ] , with order not larger than q. From the relation φ(s)φ(1 − s) = 1, we then deduce that on the half plane Re s < 1 2 , φ(s) can only have zeros on [ 0, 12 ) . Using these and the residue theorem, we find that 1 2π ∫ ∞ −∞ 1 r2 + α2 φ′ φ ( 1 2 + ir ) dr = 1 2s+ 2n− 1 φ′(s+ n) φ(s+ n) − ∑ ρ is a pole of φ(s) Re ρ< 1 2 order (ρ) α2 − ( ρ− 1 2 )2 + ∑ σ is a pole of φ(s) σ∈( 1 2 ,1] order (σ) α2 − ( σ − 1 2 )2 + c. (D.1) Here order(z) is the order of pole of φ(s) at s = z, c is a constant independent of α. We are interested in the particular case where α = s+ n− 1 2 . The following result is needed when we want to discuss the dimension of the space of holomorphic n-differentials. Proposition D.1. Given a positive integer n, let Σ(s) be the function Σ(s) = 1 2π ∫ ∞ −∞ 1 r2 + ( s+ n− 1 2 )2 φ′ φ ( 1 2 + ir ) dr. Then the residue of Σ(s) at s = 0 is 0. Proof. According to (D.1), Σ(s) is a sum of three terms. We denote these three terms by Σ1(s), Σ2(s) and Σ3(s) respectively. We need to discuss the case where n = 1 and n ≥ 2 separately. When n = 1, the residue of Σ1(s) at s = 0 is equal to −n0, where n0 is the order of pole of φ(s) at s = 1, which is equal to the order of zero of φ(s) at s = 0. Since φ(s) does not have pole at s = 0, the residue of Σ2(s) at s = 0 is 0. Since φ(s) has a pole of order n0 at s = 1, the residue of Σ3(s) at s = 0 is n0. This shows that the residue of Σ(s) at s = 0 is 0. When n ≥ 2, the residue of Σ1(s) is equal to 1 2n−1 times the order of zero of φ(s) at s = n. Since φ(s) does not have pole when Re s > 1, the residue of Σ3(s) at s = 0 is 0. For Σ2(s), the order of zero of φ(s) at s = n is equal to the order of pole of φ(s) at 1−n. Therefore, the residue of Σ2(s) at s = 0 is the negative of the residue of Σ1(s) at s = 0. This proves the assertion that the residues of Σ(s) at s = 0 is 0. ■ D.3 The term ΞP Collecting the results from Appendices D.1 and D.2, we find that the parabolic contribution ΞP = lim Y→∞ {∫∫ FY ∑ γ∈Γ γis parabolic k(γz, z)γ′(z)ny2ndµ(z) − 1 4π q∑ j=1 ∫ ∞ −∞ Λ̃(r) ∫∫ FY ∣∣∣∣Ej ( z, 1 2 + ir;n )∣∣∣∣2y2ndµ(z) dr } is given by ΞP = EP (s)− EP (a), where EP (s) = 1 (2s+ 2n− 1)2 [ q − TrΦ ( 1 2 )] + 1 4π ∫ ∞ −∞ 1 r2 + ( s+ n− 1 2 )2 φ′ φ ( 1 2 + ir ) dr + q 2(2s+ 2n− 1) { ψ(s) + ψ(s+ 2n)− 2 log 2− 2ψ ( s+ n+ 1 2 ) − 2ψ(s+ n) } . Resolvent Trace Formula and Determinants of n Laplacians 35 E Maass–Selberg relation The Maass–Selberg relation is very important in the computation of the contribution from the absolutely continuous spectrum. The results in this part follows almost the same as the case n = 0 as presented in [13]. Given a function f in H2 n(Γ) such that ∆nf = λf , it has the following Fourier expansion around the cusp associated to κj . f(σjz)σ ′ j(z) n = ∞∑ k=−∞ f (j) k (y)e2πikx. We define fY (z) = { f(z), z ∈ F Y , f(z)− f (j) 0 ( σ−1 j z ) , z ∈ F Y j . Theorem E.1. If f and g are functions in H2 n(Γ) such that ∆nf = λ1f, ∆ng = λ2g. Then (λ1 − λ2) ∫∫ F fY gY y2n−2dx dy = −Y 2n q∑ j=1 ( f (j) 0 (Y )g (j)′ 0 (Y )− f (j)′ 0 (Y )g (j) 0 (Y ) ) . Proof. (λ1 − λ2) ∫∫ FY fḡy2n−2dx dy = ∫∫ FY ∆nfḡy 2n−2dx dy − ∫∫ FY f∆ngy 2n−2dx dy = − 2i ∫∫ FY {( ∂ ∂z y2n ∂ ∂z̄ f ) ḡ − f ( ∂ ∂z̄ y2n ∂ ∂z ḡ )} dz ∧ dz̄ = − 2i ∫ ∂FY ( y2n ∂f ∂z̄ ḡdz̄ + fy2n ∂ḡ ∂z dz ) = q∑ j=1 −2i ∫ σ−1 j ∂FY j ( y2n ∂f ∂z̄ (σjz)σ ′ j(z) nσ′j(z) g(σjz)σ ′(z)n dz̄ + f(σjz)σ ′ j(z) ny2n ∂g ∂z̄ (σjz)σ′j(z) nσ′j(z) dz ) = Y 2n q∑ j=1 2i ∫ 1 0 ( ∂f ∂z̄ (σjz)σ ′ j(z) nσ′j(z) g(σjz)σ ′(z)n dx + f(σjz)σ ′ j(z) n∂g ∂z̄ (σjz)σ′j(z) nσ′j(z) dx ) . Notice that ∂f ∂z̄ (σjz)σ ′ j(z) nσ′j(z) = ∞∑ k=−∞ πikf (j) k (y)e2πikx + i 2 ∞∑ k=−∞ ∂f (j) k ∂y (y)e2πikx. Hence, (λ1 − λ2) ∫∫ FY fḡy2n−2dx dy = −Y 2n q∑ j=1 ∞∑ k=−∞ ( ∂f (j) k ∂y (Y )g (j) k (Y )− f (j) k (Y ) ∂g (j) k ∂y (Y ) ) . 36 L.-P. Teo Applying ∆nf = λf , where λ = −(s− n)(s+ n− 1), to the Fourier expansion, we find that y2 ∞∑ k=−∞ (2πk)2f (j) k (y)e2πikx − y2 ∞∑ k=−∞ ∂2f (j) k ∂y2 (y)e2πikx − 2ny ∞∑ k=−∞ 2πkf (j) k (y)e2πikx − 2ny ∞∑ k=−∞ ∂f (j) k ∂y (y)e2πikx = −(s− n)(s+ n− 1) ∞∑ k=−∞ f (j) k (y)e2πikx. Hence, y2 ∂2f (j) k ∂y2 (y) + 2ny ∂f (j) k ∂y (y)− ( 4πk2y2 − 4πnky + (s− n)(s+ n− 1) ) f (j) k = 0. This implies that f (j) 0 (y) = α0y s−n + β0y 1−s−n. On the other hand, ∂ ∂y y2n ( ∂f (j) k ∂y g (j) k − f (j) k ∂g (j) k ∂y ) = y2n−2[(s1− n)(s1 + n− 1)− (s2− n)(s2 + n− 1)]f (j) k g (j) k . Hence, if k ̸= 0, −Y 2n q∑ j=1 ∑ k ̸=0 ( ∂f (j) k ∂y (Y )g (j) k (Y )− f (j) k (Y ) ∂g (j) k ∂y (Y ) ) = −(λ1 − λ2) q∑ j=1 ∑ k ̸=0 ∫ ∞ Y y2n−2f (j) k g (j) k dy = −(λ1 − λ2) q∑ j=1 ∫∫ FY j fY ḡY y2n−2dx dy. It follows that (λ1 − λ2) ∫∫ FY fḡy2n−2dx dy = − Y 2n q∑ j=1 ( ∂f (j) 0 ∂y (Y )g (j) 0 (Y )− f (j) 0 (Y ) ∂g (j) 0 ∂y (Y ) ) − (λ1 − λ2) q∑ j=1 ∫∫ FY j fY ḡY y2n−2dx dy. Hence, (λ1 − λ2) ∫∫ F fY gY y2n−2dx dy = −Y 2n q∑ j=1 ( f (j) 0 (Y )g (j)′ 0 (Y )− f (j)′ 0 (Y )g (j) 0 (Y ) ) . ■ Theorem E.2 (Maass–Selberg relation). If s1 and s2 are regular points of the Eisenstein series Ei(z, s1;n) and Ei(z, s2;n), s1 ̸= s̄2 and s1 + s̄2 ̸= 1, then∫∫ F EY i (z, s1;n)EY j (z, s2;n)y 2n−2dx dy = δij 1 s1 + s2 − 1 Y s1+s2−1 + 1 s2 − s1 φij(s1)Y s2−s1 + 1 s1 − s2 φji(s2)Y s1−s2 − 1 s1 + s2 − 1 q∑ k=1 φik(s1)φjk(s2)Y 1−s1−s2 . Resolvent Trace Formula and Determinants of n Laplacians 37 Proof. By Theorem E.1, we have∫∫ F EY i (z, s1;n)EY j (z, s2;n)y 2n−2dx dy = Y 2n −(s1 − n)(s1 + n− 1) + (s2 − n)(s2 + n− 1) × q∑ k=1 ( [δikY s1−n+φik(s1)Y 1−s1−n][δjk(s2− n)Y s2n−1 + (1− s2− n)φjk(s2)Y −s2−n] − [δik(s1 − n)Y s1−n−1 + (1− s1 − n)φik(s1)Y −s1−n][δjkY s2−n + φjk(s2)Y 1−s2−n] ) = δij 1 s1 + s2 − 1 Y s1+s2−1 + 1 s2 − s1 φij(s1)Y s2−s1 + 1 s1 − s2 φji(s2)Y s1−s2 − 1 s1 + s2 − 1 q∑ k=1 φik(s1)φjk(s2)Y 1−s1−s2 . ■ Theorem E.3. As Y → ∞, q∑ j=1 ∫∫ FY ∣∣∣∣Ej ( z, 1 2 + ir;n )∣∣∣∣2y2ndµ(z) = 1 2ir Tr [ Φ ( 1 2 − ir ) Y 2ir − Φ ( 1 2 + ir ) Y −2ir ] + 2q log Y − φ′ φ ( 1 2 + ir ) . Proof. From the proof of the previous theorem, we find that∫∫ FY ∣∣∣∣Ej ( z, 1 2 + ir;n )∣∣∣∣2y2ndµ(z) = ∫∫ F ∣∣∣∣EY j ( z, 1 2 + ir;n )∣∣∣∣2y2ndµ(z) + exponentially decaying terms. Setting s1 = s2 = σ + ir in the previous theorem, we find that∫∫ F ∣∣∣∣EY j ( z, 1 2 + ir;n )∣∣∣∣2y2ndµ(z) = 1 2σ − 1 Y 2σ−1 − 1 2ir φjj(σ + ir)Y −2ir + 1 2ir φjj(σ + ir)Y 2ir − 1 2σ − 1 q∑ k=1 φjk(σ + ir)φjk(σ + ir)Y 1−2σ. We want to take the limit when σ → 1/2. We find that as σ → 1/2, Y 2σ−1 = 1 + (2σ − 1) log Y +O ( (2σ − 1)2 ) , φjk(σ + ir) = φjk ( 1 2 + ir ) + ( σ − 1 2 ) φ′ jk ( 1 2 + ir ) +O ( (2σ − 1)2 ) . Hence, q∑ j=1 ∫∫ F ∣∣∣∣EY j ( z, 1 2 + ir;n )∣∣∣∣2y2ndµ(z) = 1 2ir Tr [ Φ ( 1 2 − ir ) Y 2ir − Φ ( 1 2 + ir ) Y −2ir ] + q 2σ − 1 ( 1 + (2σ − 1) log Y ) − 1 2σ − 1 q∑ j=1 q∑ k=1 [ φjk ( 1 2 + ir ) + ( σ − 1 2 ) φ′ jk ( 1 2 + ir )] 38 L.-P. Teo × [ φkj ( 1 2 − ir ) + ( σ − 1 2 ) φ′ kj ( 1 2 − ir )]( 1− (2σ − 1) log Y ) +O(2σ − 1). Now the matrix Φ(s) = [ φij(s) ] satisfies Φ(s)Φ(1− s) = I. Hence, for each 1 ≤ j ≤ q, q∑ k=1 φjk ( 1 2 + ir ) φkj ( 1 2 − ir ) = 1. Moreover, Φ(1− s) = Φ(s)−1. Since d ds log detΦ(s) = Tr [ Φ′(s)Φ−1(s) ] , and d ds log detΦ(s) = − d ds log detΦ(1− s) we find that Tr [ Φ′ ( 1 2 + ir ) Φ−1 ( 1 2 + ir )] = Tr [ Φ′ ( 1 2 − ir ) Φ−1 ( 1 2 − ir )] . This gives q∑ j=1 ∫∫ F ∣∣∣∣EY j ( z, 1 2 + ir;n )∣∣∣∣2y2ndµ(z) = 1 2ir Tr [ Φ ( 1 2 − ir ) Y 2ir − Φ ( 1 2 + ir ) Y −2ir ] + 2q log Y − Tr [ Φ′ ( 1 2 + ir ) Φ−1 ( 1 2 + ir )] . The assertion of the theorem follows. ■ F Asymptotic behavior of the zeta function Zell(s) In this section, we want to compute the asymptotic behavior of logZell(s), where Zell(s) = v∏ j=1 mj−1∏ r=0 Γ ( s+ r mj ) 2αmj (r−n)+1−mj 2mj Γ ( s+ 2n+ r mj ) 2αmj (r+n)+1−mj 2mj , when u = s+ n− 1 2 is large. From the asymptotic behavior of log Γ(s) (6.2), we have logZell(s) = A u log u+ B log u+ Cu+ D + o(1), where A = v∑ j=1 αj mj , B = v∑ j=1 βj , C = v∑ j=1 1 mj (−αj − αj logmj), Resolvent Trace Formula and Determinants of n Laplacians 39 D = − v∑ j=1 βj logmj + 1 2 log(2π) v∑ j=1 αj , and αj = mj−1∑ r=0 ( 2αmj (r − n) + 1−mj 2mj + 2αmj (r + n) + 1−mj 2mj ) , βj = mj−1∑ r=0 { 2αmj (r − n) + 1−mj 2mj (−n+ 1 2 + r mj − 1 2 ) + 2αmj (r + n) + 1−mj 2mj ( n+ 1 2 + r mj − 1 2 )} . From (C.1), we find that αj = 0 for 1 ≤ j ≤ 0. Hence, A = C = 0, and D = − v∑ j=1 βj logmj . It remains to calculate βj . First we notice that βj = 1 mj mj−1∑ r=0 { r 2αmj (r − n) + 1−mj 2mj + r 2αmj (r + n) + 1−mj 2mj } because of (C.1). If αmj (n) = ℓ, then αmj (r + n) = { ℓ+ r, 0 ≤ r ≤ mj − ℓ− 1, ℓ+ r −mj , mj − ℓ ≤ r ≤ mj − 1, αmj (r − n) = { mj + r − ℓ, 0 ≤ r ≤ ℓ− 1, r − ℓ, ℓ ≤ r ≤ mj − 1. With the help of a standard computer algebra, we find that βj = m2 j − 1 6mj − αmj (n)(mj − αmj (n)) mj . Hence, B = v∑ j=1 ( m2 j − 1 6mj − αmj (n)(mj − αmj (n)) mj ) , D = − v∑ j=1 ( m2 j − 1 6mj − αmj (n)(mj − αmj (n)) mj ) logmj . Acknowledgements This research is supported by the Ministry of Higher Education Malaysia through the Fundamen- tal Research Grant Scheme (FRGS) FRGS/1/2018/STG06/XMU/01/1. We would like to thank L. Takhtajan and J. Friedman who have given helpful comments and suggestions. We would also like to thank the referees for reading the paper carefully and giving valuable comments. 40 L.-P. Teo References [1] Alekseevskii V.P., On functions similar to the gamma function, Comm. Proc. Kharkov Math. Soc. 1 (1889), 169–238. [2] Andrews G.E., Askey R., Roy R., Special functions, Encyclopedia of Mathematics and its Applications, Vol. 71, Cambridge University Press, Cambridge, 1999. [3] Barnes E.W., The theory of the G-function, Q. J. Math. 31 (1900), 264–314. [4] D’Hoker E., Phong D.H., On determinants of Laplacians on Riemann surfaces, Comm. Math. Phys. 104 (1986), 537–545. [5] Efrat I., Determinants of Laplacians on surfaces of finite volume, Comm. Math. Phys. 119 (1988), 443–451. [6] Fay J.D., Fourier coefficients of the resolvent for a Fuchsian group, J. Reine Angew. Math. 293–294 (1977), 143–203. [7] Fischer J., An approach to the Selberg trace formula via the Selberg zeta-function, Lecture Notes in Math., Vol. 1253, Springer-Verlag, Berlin, 1987. [8] Freixas i Montplet G., von Pippich A.-M., Riemann–Roch isometries in the non-compact orbifold setting, J. Eur. Math. Soc. (JEMS) 22 (2020), 3491–3564, arXiv:1604.00284. [9] Gong D.G., Zeta-determinant and torsion functions on Riemann surfaces of finite volume, Manuscripta Math. 86 (1995), 435–454. [10] Gradshteyn I.S., Ryzhik I.M., Table of integrals, series, and products, 6th ed., Academic Press Inc., San Diego, CA, 2000. [11] Hejhal D.A., The Selberg trace formula for PSL(2,R), Vol. 1, Lecture Notes in Math., Vol. 548, Springer- Verlag, Berlin, 1976. [12] Hejhal D.A., The Selberg trace formula for PSL(2,R), Vol. 2, Lecture Notes in Math., Vol. 1001, Springer- Verlag, Berlin, 1983. [13] Iwaniec H., Spectral methods of automorphic forms, 2nd ed., Graduate Studies in Mathematics, Vol. 53, Amer. Math. Soc., Providence, RI, 2002. [14] Koyama S., Determinant expression of Selberg zeta functions. I, Trans. Amer. Math. Soc. 324 (1991), 149–168. [15] Koyama S., Determinant expression of Selberg zeta functions. III, Proc. Amer. Math. Soc. 113 (1991), 303–311. [16] McIntyre A., Takhtajan L.A., Holomorphic factorization of determinants of Laplacians on Riemann surfaces and a higher genus generalization of Kronecker’s first limit formula, Geom. Funct. Anal. 16 (2006), 1291– 1323, arXiv:math.CV/0410294. [17] Sarnak P., Determinants of Laplacians, Comm. Math. Phys. 110 (1987), 113–120. [18] Selberg A., Harmonic analysis and discontinuous groups in weakly symmetric Riemannian spaces with applications to Dirichlet series, J. Indian Math. Soc. (N.S.) 20 (1956), 47–87. [19] Selberg A., Göttingen lectures on harmonic analysis, in Alte Selberg Collected Papers, Springer Collected Works in Mathematics, Springer-Verlag, Berlin, 1989, 626–675. [20] Takhtajan L.A., Zograf P.G., A local index theorem for families of ∂-operators on punctured Riemann surfaces and a new Kähler metric on their moduli spaces, Comm. Math. Phys. 137 (1991), 399–426. [21] Takhtajan L.A., Zograf P.G., Local index theorem for orbifold Riemann surfaces, Lett. Math. Phys. 109 (2019), 1119–1143, arXiv:1701.00771. [22] Teo L.-P., Ruelle zeta function for cofinite hyperbolic Riemann surfaces with ramification points, Lett. Math. Phys. 110 (2020), 61–82, arXiv:1901.07898. [23] Venkov A.B., Spectral theory of automorphic functions, Proc. Steklov Inst. Math. 153 (1982), 1–163. [24] Venkov A.B., Kalinin V.L., Faddeev L.D., A nonarithmetic derivation of the Selberg trace formula, J. Soviet Math. 8 (1977), 177–199. [25] Voros A., Spectral functions, special functions and the Selberg zeta function, Comm. Math. Phys. 110 (1987), 439–465. https://doi.org/10.1017/CBO9781107325937 https://doi.org/10.1007/BF01211063 https://doi.org/10.1007/BF01218082 https://doi.org/10.1515/crll.1977.293-294.143 https://doi.org/10.1007/BFb0077696 https://doi.org/10.4171/jems/992 https://arxiv.org/abs/1604.00284 https://doi.org/10.1007/BF02568004 https://doi.org/10.1007/BF02568004 https://doi.org/10.1007/BFb0079608 https://doi.org/10.1007/BFb0079608 https://doi.org/10.1007/BFb0061302 https://doi.org/10.1007/BFb0061302 https://doi.org/10.1090/gsm/053 https://doi.org/10.2307/2001500 https://doi.org/10.2307/2048513 https://doi.org/10.1007/s00039-006-0582-7 https://arxiv.org/abs/math.CV/0410294 https://doi.org/10.1007/BF01209019 https://doi.org/10.1007/BF02431886 https://doi.org/10.1007/s11005-018-01144-w https://arxiv.org/abs/1701.00771 https://doi.org/10.1007/s11005-019-01222-7 https://doi.org/10.1007/s11005-019-01222-7 https://arxiv.org/abs/1901.07898 https://doi.org/10.1007/BF01212422 1 Introduction 2 Laplacians of n-differential 3 The Eisenstein series 4 The trace of the resolvent kernel 5 Dimensions of spaces holomorphic differentials 6 The determinant of n-Laplacian A The inversion formulas B The hyperbolic contribution C Elliptic contribution D Contribution from parabolic elements and absolutely continuous spectrum D.1 Contribution from parabolic elements D.2 Contribution of the absolutely continuous spectrum D.3 The term Xi P E Maass–Selberg relation F Asymptotic behavior of the zeta function Z ell(s) References
id nasplib_isofts_kiev_ua-123456789-211444
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
issn 1815-0659
language English
last_indexed 2026-03-15T10:21:32Z
publishDate 2021
publisher Інститут математики НАН України
record_format dspace
spelling Teo, Lee-Peng
2026-01-02T08:35:30Z
2021
Resolvent Trace Formula and Determinants of Laplacians on Orbifold Riemann Surfaces. Lee-Peng Teo. SIGMA 17 (2021), 083, 40 pages
1815-0659
2020 Mathematics Subject Classification: 14H15; 11F72; 11M36
arXiv:2104.00895
https://nasplib.isofts.kiev.ua/handle/123456789/211444
https://doi.org/10.3842/SIGMA.2021.083
For a nonnegative integer, we consider the -Laplacian Δₙ acting on the space of -differentials on a confinite Riemann surface X which has ramification points. The trace formula for the resolvent kernel is developed along the line à la Selberg. Using the trace formula, we compute the regularized determinant of Δₙ + ( + 2 − 1), from which we deduce the regularized determinant of Δₙ, denoted by det′Δₙ. Taking into account the contribution from the absolutely continuous spectrum, det′Δₙ is equal to a constant Cₙ times ( ) when ≥ 2. Here ( ) is the Selberg zeta function of . When = 0 or = 1, ( ) is replaced by the leading coefficient of the Taylor expansion of ( ) around = 0 and = 1, respectively. The constants Cn are calculated explicitly. They depend on the genus, the number of cusps, as well as the ramification indices, but are independent of the moduli parameters.
This research is supported by the Ministry of Higher Education Malaysia through the Fundamental Research Grant Scheme (FRGS) FRGS/1/2018/STG06/XMU/01/1. We would like to thank L. Takhtajan and J. Friedman, who have given helpful comments and suggestions. We would also like to thank the referees for carefully reviewing the paper and providing valuable comments.
en
Інститут математики НАН України
Symmetry, Integrability and Geometry: Methods and Applications
Resolvent Trace Formula and Determinants of Laplacians on Orbifold Riemann Surfaces
Article
published earlier
spellingShingle Resolvent Trace Formula and Determinants of Laplacians on Orbifold Riemann Surfaces
Teo, Lee-Peng
title Resolvent Trace Formula and Determinants of Laplacians on Orbifold Riemann Surfaces
title_full Resolvent Trace Formula and Determinants of Laplacians on Orbifold Riemann Surfaces
title_fullStr Resolvent Trace Formula and Determinants of Laplacians on Orbifold Riemann Surfaces
title_full_unstemmed Resolvent Trace Formula and Determinants of Laplacians on Orbifold Riemann Surfaces
title_short Resolvent Trace Formula and Determinants of Laplacians on Orbifold Riemann Surfaces
title_sort resolvent trace formula and determinants of laplacians on orbifold riemann surfaces
url https://nasplib.isofts.kiev.ua/handle/123456789/211444
work_keys_str_mv AT teoleepeng resolventtraceformulaanddeterminantsoflaplaciansonorbifoldriemannsurfaces