Hecke operator

In mathematics, in particular in the theory of modular forms, a Hecke operator, studied by Hecke (1937), is a certain kind of "averaging" operator that plays a significant role in the structure of vector spaces of modular forms and more general automorphic representations.

History

Mordell (1917) used Hecke operators on modular forms in a paper on the special cusp form of Ramanujan, ahead of the general theory given by Hecke (1937). Mordell proved that the Ramanujan tau function, expressing the coefficients of the Ramanujan form,

$\Delta(q)=q\left(\prod_{n=1}^{\infty}(1-q^n)\right)^{24}= \sum_{n=1}^{\infty} \tau(n)q^n, \quad q=e^{2\pi i\tau},$

is a multiplicative function:

$\tau(mn)=\tau(m)\tau(n) \quad \text{ for } (m,n)=1.$

The idea goes back to earlier work of Hurwitz, who treated algebraic correspondences between modular curves which realise some individual Hecke operators.

Mathematical description

Hecke operators can be realized in a number of contexts. The simplest meaning is combinatorial, namely as taking for a given integer n some function f(Λ) defined on the lattices of fixed rank to

$\sum f(\Lambda')$

with the sum taken over all the Λ′ that are subgroups of Λ of index n. For example, with n=2 and two dimensions, there are three such Λ′. Modular forms are particular kinds of functions of a lattice, subject to conditions making them analytic functions and homogeneous with respect to homotheties, as well as moderate growth at infinity; these conditions are preserved by the summation, and so Hecke operators preserve the space of modular forms of a given weight.

Another way to express Hecke operators is by means of double cosets in the modular group. In the contemporary adelic approach, this translates to double cosets with respect to some compact subgroups.

Explicit formula

Let M_m be the set of 2×2 integral matrices with determinant m and Γ = M₁ be the full modular group SL(2, Z). Given a modular form f(z) of weight k, the mth Hecke operator acts by the formula

$T_m f(z) = m^{k-1}\sum_{\left(\begin{smallmatrix}a & b\\ c & d\end{smallmatrix}\right)\in\Gamma\backslash M_m}(cz+d)^{-k}f\left(\frac{az+b}{cz+d}\right),$

where z is in the upper half-plane and the normalization constant m^k−1 assures that the image of a form with integer Fourier coefficients has integer Fourier coefficients. This can be rewritten in the form

$T_m f(z) = m^{k-1}\sum_{a,d>0, ad=m}\frac{1}{d^k}\sum_{b \pmod d} f\left(\frac{az+b}{d}\right),$

which leads to the formula for the Fourier coefficients of T_mf(z) = ∑ b_nqⁿ in terms of the Fourier coefficients of f(z) = ∑ a_nqⁿ:

$b_n = \sum_{r>0, r|(m,n)}r^{k-1}a_{mn/r^2}.$

One can see from this explicit formula that Hecke operators with different indices commute and that if a₀ = 0 then b₀ = 0, so the subspace S_k of cusp forms of weight k is preserved by the Hecke operators. If a (non-zero) cusp form f is a simultaneous eigenform of all Hecke operators T_m with eigenvalues λ_m then a_m = λ_ma₁ and a₁ ≠ 0. Hecke eigenforms are normalized so that a₁ = 1, then

$T_m f = a_m f, \quad a_m a_n = \sum_{r>0, r|(m,n)}r^{k-1}a_{mn/r^2},\ m,n\geq 1.$

Thus for normalized cuspidal Hecke eigenforms of integer weight, their Fourier coefficients coincide with their Hecke eigenvalues.

Hecke algebras

Algebras of Hecke operators are called Hecke algebras, and are commutative rings. Other, related, mathematical rings are called Hecke algebras, although the link to Hecke operators is not entirely obvious. These algebras include certain quotients of the group algebras of braid groups. The presence of this commutative operator algebra plays a significant role in the harmonic analysis of modular forms and generalisations. In the classical elliptic modular form theory, the Hecke operators T_n with n coprime to the level acting on the space of cusp forms of a given weight are self-adjoint with respect to the Petersson inner product. Therefore, the spectral theorem implies that there is a basis of modular forms that are eigenfunctions for these Hecke operators. Each of these basic forms possesses an Euler product. More precisely, its Mellin transform is the Dirichlet series that has Euler products with the local factor for each prime p is the inverse of the Hecke polynomial, a quadratic polynomial in p^−s. In the case treated by Mordell, the space of cusp forms of weight 12 with respect to the full modular group is one-dimensional. It follows that the Ramanujan form has an Euler product and establishes the multiplicativity of τ(n).

See also

• Eichler–Shimura congruence relation

References

• (See chapter 8.)

• Jean-Pierre Serre, A course in arithmetic.

• Don Zagier, Elliptic Modular Forms and Their Applications, in The 1-2-3 of Modular Forms, Universitext, Springer, 2008 ISBN 978-3-540-74117-6