Intersection numbers of modular correspondences for genus zero modular curves (Q2288307)

From MaRDI portal
Revision as of 11:24, 21 July 2024 by ReferenceBot (talk | contribs) (‎Changed an Item)





scientific article
Language Label Description Also known as
English
Intersection numbers of modular correspondences for genus zero modular curves
scientific article

    Statements

    Intersection numbers of modular correspondences for genus zero modular curves (English)
    0 references
    0 references
    17 January 2020
    0 references
    Let \(j(E)\) denote the invariant of the elliptic curve \(E\) over \(\mathbb{C}\). For \(N\ge1\) define \(\Phi_N:\mathbb{C}\times\mathbb{C}\rightarrow\mathbb{C}\) by \[ \Phi_N(j(E),j(E'))=\prod_{(f,E_1')}(j(E)-j(E_1')), \] where the product ranges over the isomorphism classes of pairs \((f,E_1')\), with \(E_1'\) an elliptic curve over \(\mathbb{C}\) and \(f:E_1'\rightarrow E'\) an isogeny of degree \(N\). It is known that the function \(\Phi_N\) is given by a symmetric polynomial with coefficients in \(\mathbb{Z}\). Let \(T_N\) denote the affine plane curve defined by \(\Phi_N(X,Y)=0\). In [Math. Ann. 88, 26--52 (1922; JFM 48.1164.04)] \textit{A. Hurwitz} showed that \(T_{N_1}\) and \(T_{N_2}\) intersect properly if and only if \(N_1N_2\) is not a square. In this case he computed the intersection number \(T_{N_1}\cdot T_{N_2}\) in terms of equivalence classes of positive definite binary quadratic forms. The paper under review generalizes Hurwitz's formula by replacing the affine \(j\)-line \(Y_0(1)\) with a modular curve of the form \(Y_0(M)\) for any \(M\ge1\) for which the compactification \(X_0(M)\) has genus 0. Given such an \(M\) the author finds a parameter \(t\) on \(X_0(M)\) which is holomorphic on \(X_0(M)\smallsetminus\{\infty\}\), has a simple pole at \(\infty\) with residue 1, and whose Fourier expansion at the cusp \(\infty\) has coefficients in \(\mathbb{Z}\). For \(N\) relatively prime to \(M\) this allows him to define the modular polynomial \(\Phi_N^{\Gamma_0(M)}(X,Y)\) of level \(N\) for the congruence subgroup \(\Gamma_0(M)\), with \(t\) playing the role of \(j\) in the definition of \(\Phi_N(X,Y)\). Once again, \(\Phi_N^{\Gamma_0(M)}(X,Y)\) is a symmetric polynomial with coefficients in \(\mathbb{Z}\). Let \(T_N^{\Gamma_0(M)}\) denote the curve in \(Y_0(M)\times Y_0(M)\) defined by \(\Phi_N^{\Gamma_0(M)}(X,Y)=0\). A point on \(T_N^{\Gamma_0(M)}\) corresponds to two pairs \((E,C)\), \((E',C')\), each consisting of an elliptic curve over \(\mathbb{C}\) and a cyclic subgroup of order \(M\) such that there is an isogeny \(f:E\rightarrow E'\) of degree \(N\) with \(f(C)=C'\). Let \(N_1,N_2\) be positive integers which are relatively prime to \(M\). The author shows that \(T_{N_1}^{\Gamma_0(M)}\) and \(T_{N_2}^{\Gamma_0(M)}\) intersect properly if and only if \(N_1N_2\) is not a square. When this holds he gives a formula for the intersection number \(T_{N_1}^{\Gamma_0(M)}\cdot T_{N_2}^{\Gamma_0(M)}\) in terms of equivalence classes of positive definite binary quadratic forms. In the case where \(M\) is prime he gives a version of the intersection number formula which expresses \(T_{N_1}^{\Gamma_0(M)}\cdot T_{N_2}^{\Gamma_0(M)}\) in terms of coefficients of a Siegel Eisenstein series.
    0 references
    0 references
    modular correspondences
    0 references
    Kudla's problem
    0 references
    intersection number
    0 references
    modular curves
    0 references
    modular polynomials
    0 references
    quadratic forms
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references