K3 surfaces, Igusa cusp forms, and string theory

From MaRDI portal





Let \(\mathfrak H_g\) denote the Siegel upper half space and write \(\Gamma_g:=\text{Sp}(2g,{\mathbb Z})\) for the Siegel modular group acting properly discontinuously on \(\mathfrak H_g\). Also, for a finite index subgroup \(\Gamma\subset\Gamma_g\), and character \(\rho:\Gamma \rightarrow\mathbb C^{\times}\), write \(\mathfrak M^g_k(\Gamma,\rho)\) for the weight \(k\) \(\Gamma\)-modular forms, and \(\mathfrak M^g_k(\Gamma)\) for \(\mathfrak M^g_k(\Gamma,\text{id}_{\Gamma})\). One knows that (Igusa) \({\bigoplus_{k=0}^{\infty}\mathfrak M_k^2(\Gamma_2)\simeq\mathbb C[ \mathcal E_4,\mathcal E_6,\chi_{10},\chi_{12},\chi_{35}]}\), where \(\mathcal E_4\) and \(\mathcal E_6\) are the genus two Eisenstein series of weight \(4\) and \(6\), respectively, and where the \(\chi_{10}\), \(\chi_{12}\) and \(\chi_{35}\) are certain cusp forms of weights \(10\), \(12\) and \(35\), respectively. \(\chi_{35}^2\) can be expressed as a polynomial in the other generators. One defines the integer Heisenberg group \(\text{H}_{\mathbb Z}\) and a normal subgroup \(\Gamma^J\subset\text{H}_{\mathbb Z}\), the Jacobi group, such that \(\Gamma^J\simeq\Gamma_1\ltimes\text{H}_{\mathbb Z}\). For a character \(\rho^J:\Gamma^J\rightarrow\mathbb C^{\times}\), an integer \(k\) and \(m\in{1\over 2}\mathbb N\), one can define so-called Jacobi forms of weight \(k\) and index \(m\) with character \(\rho^J\). In particular, one has cusp Jacobi forms and weak Jacobi forms. Let \(M\) be a compact \(d\)-dimensional Kähler manifold with holomorphic tangent bundle \(T\rightarrow M\) and fixed rank \(\ell\) holomorphic vector bundle \(E\rightarrow M\). One may define the elliptic genus \(Z_E[M](\tau,z)\) as a function on \(\mathfrak H_1\times\mathbb C\) in terms of the Chern classes of \(T\) and \(E\). One has the following result: If \(M\) is a Calabi-Yau manifold, then the elliptic genus \(Z_{T^*}[M]\) is a weak Jacobi form of weight \(0\), index \(d/2\) and character \(\text{id}_{\Gamma_1}\times\rho^d_ {\text{H}}:\Gamma^J\rightarrow\mathbb C^{\times}\). In case \(M\) is even dimensional the character is trivial. -- In particular, for a \(K_3\)-surface \(M\) one obtains explicit expressions for the elliptic genus in terms of theta functions. In string theory one often encounters (regularized) integrals \(\mathcal I(\Omega)\), \(\Omega\in\mathfrak H_2\). In Phys. Lett. B 372, No. 1-2, 59-64 (1996), \textit{T. Kawai} treated a particular case and showed that an appropriate regularization of such a particular integral can be written as \(-\log(Y^{10}|\chi_{10}(\Omega) |^2)\), where \(Y=\det\text{Im}(\Omega)\). As an ingredient of the proof one uses infinite product representations of genus two theta functions whose square are just \(\chi_{10}(\Omega)\). This is due to work of Gritsenko and Nikulin who use a lifting procedure of Borcherds'. Here the Hecke operators for the Jacobi group enter the scene. Let \(M\) be a smooth projective surface, and write \(M^{(m)}=S^mM\) for the \(m\)-th symmetric product, and let \(M^{[m]}=\text{Hilb}^m (M)\) denote the Hilbert scheme of zero-dimensional subschemes of length \(m\). Göttsche proved the remarkable formula for the Euler characteristics of \(M^{(m)}\): \[ \sum_{m=0}^{\infty}p^m\chi(M^{[m]})=\prod_{k=1}^{\infty}{1\over(1-p^k) ^{\chi(M)}} \] In particular, for a \(K3\) surface \(M\) this leads to interesting expressions for \(Z[M^{(m)}](\tau,z)\). Several worked out examples are given: The first five \(Z[M^{(m)}](\tau,z)\)'s can be expressed as rational polynomials in the Eisenstein series \(E_4\), \(E_6\), the Weierstraß \(\wp \)-function, and the function \({K(\tau,z)=\sqrt{-1} {\theta_1(\tau,z)\over\eta(\tau)^3}}\) of a special form, e.g. \[ Z[M^{(5)}](\tau,z)= \biggl(176256\wp ^5+3720E_4\wp ^3+186E_6\wp ^2+ 17E_4^2\wp +{31\over 72}E_4E_6 \biggr)K^{10} \] In physical terms one gets that the elliptic genus \(Z[M^{(m)}]\) is equal to the sum of the (chiral) two-point functions of all the physical states having the mass squared \(2m-2\) on an elliptic curve, divided by the (chiral) tachyon two-point function on the same elliptic curve. For the entire collection see [Zbl 0905.00081].











This page was built for publication: K3 surfaces, Igusa cusp forms, and string theory

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4249111)