Rigidity via modular properties of theta functions

Indraneel Tambe · 2025

Read Rigidity via modular properties of theta functions by Indraneel Tambe (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-2505-02792.txt.

Introduction

Indraneel Tambe proves a rigidity result in a 2025 paper in geometry and topology. Rigidity here means a certain number attached to a twisted operator cannot change under deformation. The operators were introduced by other authors on special bundles. The proof runs through the classic theta functions of Jacobi. It uses their shift and symmetry properties together with a fixed-point formula.

Catalog record

CREATORIndraneel Tambe
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE45 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-02792

Browse the full library · About the Tower of Babel

Full text

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

Rigidity via modular properties of theta functions Indraneel Tambe Abstract In this paper we use methods of Liu [ 14 , 15 ] to show that the twisted Dirac operators \mathscr{D} on certain bundles \Phi considered by Guan and Wang [ 22 ] are rigid. To do so, we use a Lefschetz formula and Atiyah-Bott localization to obtain formulas for the Lefschetz numbers \mathscr{L} of these operators \mathscr{D} in terms of Jacobi theta functions; then, using the translational and modular transformation properties of theta functions and the properties of their zeros, we prove \mathscr{L} is constant provided certain conditions on characteristic classes hold, thus showing the rigidity of \mathscr{D} on \Phi under these conditions. 1. Introduction Let G be a Lie group and let P be a G -equivariant Fredholm differential operator acting on the sections of a G -equivariant smooth vector bundle W\rightarrow M . The kernel and cokernel of P are finite-dimensional representations of G , and we obtain a well-defined function L on G given by L(g):=\mathrm{trace}(g|_{\mathrm{ker}P})-\mathrm{trace}(g|_{\mathrm{coker}P}) . This function L is called the Lefschetz number of P with respect to the G -action. We say the operator P is rigid for the G -action if its Lefschetz number L is constant as a function on G . One application of rigidity is as follows: if we can obtain an expression for the Lefschetz number L of P which does not depend on the choice of G -action on M and W , and if we have proved P is rigid in general for such G -actions, then showing the non-constancy of L for a particular pair (W\rightarrow M,P) implies that (W\rightarrow M,P) in fact does not admit an equivariant G -action. In [ 7 ] , Witten conjectured the rigidity of the twisted Dirac operator \mathscr{D} on a certain element of K(M)[[q^{1/2}]] for M a spin manifold. This \mathscr{D} is viewed as an analog of the classical Dirac operator on the (free) loop space of M , and the rigidity of this \mathscr{D} is the loop-space analog of the classically known rigidity of the Dirac operator on a finite-dimensional spin manifold. Witten’s rigidity conjectures were first proved in the work of Hirzebruch [ 8 ] , Taubes [ 10 ] , Bott and Taubes [ 11 ] , and Krichever [ 12 ] . Next, in [ 14 ] and [ 15 ] , Liu introduced new methods that simplified and generalized the proof. His approach uses the modularity properties of the Jacobi theta functions to circumvent the technical arguments used in [ 8 ] and [ 11 ] . Subsequently, Dessai [ 16 ] extended the work to the spin c case, Liu and Ma [ 17 , 18 ] explored extensions to families of operators. Further studies on the odd-dimensional case were conducted by Liu and Wang [ 19 ] , as well as Han and Yu [ 20 ] . Let M be a closed even-dimensional spin manifold with an S^{1} -action and V a real even-rank S^{1} -equivariant spin vector bundle on M . Let \Phi_{\lambda} (actually an element of K(M)[[q^{1/2}]] , where q^{1/2}=e^{\pi\mathbf{i}\tau} where \tau\in\mathbb{H} is a complex variable in the upper half-plane), \lambda=1,2,3 , be one of the bundles considered in the paper [ 22 ] of Guan and Wang. The bundle \Phi_{\lambda} is constructed from TM and V using standard operations such as direct sums, tensor products, and symmetric and exterior powers. Let \mathscr{D}_{\lambda} be the twisted Dirac operator on (the spinor vector bundle of) \Phi_{\lambda} . The goal of this paper is to prove the following theorem: Theorem 1.1 . The operators \mathscr{D}_{\lambda} , \lambda=1,2,3 , are rigid if the first S^{1} -equivariant Pontryagin class of V vanishes. The proof of Theorem 1.1 is given in Sections 5 and 6 , and has the following outline. Following the method of [ 15 ] , we find that if we assume the first S^{1} -equivariant Pontryagin class of V vanishes, under the modular action, the Lefschetz numbers L_{\lambda} for the \mathscr{D}_{\lambda} actually transform into each other, up to possible multiplication by nowhere-vanishing holomorphic functions. Following [ 15 ] , this “modularity” property of the L_{\lambda} ’s, together with an analysis of their periodicity properties and possible pole locations, is used to prove the constancy of the L_{\lambda} ’s, hence the rigidity of the \mathscr{D}_{\lambda} ’s. 2. Definitions and conventions Convention 2.1 . In this paper, unless otherwise stated, all manifolds are smooth, compact, connected, and without boundary, all vector bundles are smooth and of finite rank and are real except when explicitly declared to be complex, and all actions by Lie groups on manifolds are smooth. Notation 2.2 . • We use \mathbf{i} to denote the imaginary unit, as opposed to i which is used for just indices and other variables. We denote the i th Chern class as \mathbf{c}_{i} , as opposed to c_{i} which will be used for Chern roots. • We denote by \xi_{G} the “ G -equivariant version” of a given characteristic class \xi . • Let \mathbf{H}:=\{\tau\in\mathbb{C}\,|\,\mathrm{Im}(\tau)>0\} . We let \tau\in\mathbf{H} , and we define q:=e^{2\pi\mathbf{i}\tau} and q^{r}:=e^{2\pi r\mathbf{i}\tau} for any positive real number r . • A formal difference of vector bundles V and W will be denoted V-W . Definition 2.3 . Let \pi:V\rightarrow M be a vector bundle. A G -equivariant action on V\rightarrow M consists of: • a G -action \tilde{\rho}:G\times V\rightarrow V that linearly maps fibers to fibers, and • the quotient G -action \rho:G\times M\rightarrow M , which satisfies \pi\circ\tilde{\rho}=\rho\circ(\mathrm{id}_{G}\times\pi) . Let V\rightarrow M be a G -equivariant vector bundle. Let \Gamma(V) denote the space of smooth sections of V and let \mathscr{D}:\Gamma(V)\rightarrow\Gamma(V) be a linear differential operator which is G -equivariant. Then the kernel and cokernel of \mathscr{D} are representations of G . Suppose further that \mathscr{D} is also Fredholm (e.g. an elliptic operator), i.e., the kernel and cokernel of \mathscr{D} are finite-dimensional. Definition 2.4 . In the above situation, the Lefschetz number of \mathscr{D} is the function L=L_{\mathscr{D}}:G\rightarrow\mathbb{F} given by L_{\mathscr{D}}(g):=\mathrm{trace}(g|\mathrm{ker}(\mathscr{D}))-\mathrm{trace}(g|\mathrm{coker}(\mathscr{D})), where we use the convention that the codomain \mathbb{F} of L_{\mathscr{D}} is \mathbb{R} or \mathbb{C} according to whether V is real or complex. Remark 2.5 . Later, instead of a single vector bundle V on M with a differential operator \mathscr{D} , we will consider a formal power series \sum_{n=0}^{\infty}V_{(n)}q^{n} where q is a formal variable and each V_{(n)} is itself an S^{1} -equivariant vector bundle with an S^{1} -equivariant linear Fredholm differential operator \mathscr{D}_{(n)} . In this case, the “Lefschetz number” L of the collection (\mathscr{D}_{(n)})_{n=0,1,2,\ldots} is the function L:G\rightarrow\mathbb{F}[[q]] given by L(g)=\sum_{n=0}^{\infty}L_{n}(g)q^{n} where L_{n} is the Lefschetz number of \mathscr{D}_{(n)} for each n\geq 0 . Again \mathbb{F} is \mathbb{R} or \mathbb{C} according to whether the V_{n} ’s are real or complex. Definition 2.6 . The operator \mathscr{D} is G - rigid if the Lefschetz number L of \mathscr{D} is constant on G , which is the case if and only if for every g\in G we have L(g)=L(1_{G})=\mathrm{dim}\,\mathrm{ker}(\mathscr{D})-\mathrm{dim}\,\mathrm{coker}(\mathscr{D}). We note the RHS is just the ordinary Fredholm index of \mathscr{D} . Remark 2.7 (Lefschetz number is well-defined on K -group) . Given G -equivariant vector bundles A,B,A^{\prime},B^{\prime} on M with respective G -equivariant

Read the complete text