Abstract

We characterize the functional space of the planar mixed automorphic forms with respect to an equivariant pair and given lattice as the image of the Landau automorphic forms (involving special multiplier) by an appropriate isomorphic transform.

1. Introduction

Mixed automorphic forms of type arise naturally as holomorphic forms on elliptic varieties [1] and appear essentially in the context of number theory and algebraic geometry. Roughly speaking, they are a class of functions defined on a given (Hermitian symmetric) space and satisfying a functional equation of type for every and . Here, ; is an automorphic factor associated to an appropriate action of a group on , and is an equivariant pair for the data . Such notion was introduced by Stiller [2] and extensively studied by Lee in the case of being the upper half-plane. They include the classical ones as a special case. Nontrivial examples of them have been constructed in [3, 4]. We refer to [5] for an exhaustive list of references.

In this paper, we are interested in the space of mixed automorphic forms defined on the complex plane with respect to a given lattice in and an equivariant pair . We find that is isomorphic to the space of Landau automorphic forms [6], of “weight” with respect to a special pseudocharacter defined on and given explicitly through (5.3) below. The crucial point in the proof is to observe that the quantity is in fact a real constant independent of the complex variable .

The exact statement of our main result (Theorem 5.1) is given and proved in Section 5. In Sections 2 and 3, we establish some useful facts that we need to introduce the space of planar mixed automorphic forms . We have to give necessary and sufficient condition to ensure the nontriviality of such functional space. In Section 4, we introduce properly the function that serves to define the pseudocharacter .

2. Group Action

Let be the semidirect product group of the unitary group and the additive group . acts on the complex plane by the holomorphic mappings ; , , so that can be realized as Hermitian symmetric space .

By a -equivariant pair , we mean that is a -endomorphism and is a compatible mapping, that is, Now, for given real numbers , and an equivariant pair , we define to be the complex-valued mapping where ; is the “automorphic factor” given by Here and elsewhere, denotes the imaginary part of the complex number and the usual Hermitian scalar product on . Thus, one can check the following.

Proposition 2.1. The mapping satisfies the chain rule where is the real-valued function defined on by

Proof. For every and , we have Next, one can see that the automorphic factor satisfies for every and . This gives rise to Finally, (2.5) follows by making use of the equivariant condition .

Remark 2.2. According to Proposition 2.1 above, the unitary transformations for varying define then a projective representation of the group on the space of functions on .

3. The Space of Planar Mixed Automorphic Forms

Let be a uniform lattice of the additive group that can be seen as a discrete subgroup of by the identification so that the action of on is the one induced from this of , that is, Associated to such and given fixed data of and as above, we perform the vector space of smooth complex-valued functions on satisfying the functional equation for every and .

Definition 3.1. The space is called the space of planar mixed automorphic forms of biweight with respect to the equivariant pair and the lattice .

We assert the following.

Proposition 3.2. The functional space is nontrivial if and only if the real-valued function in (2.6) takes integral values on .

Proof. The proof can be handled in a similar way as in [6] making use of (2.5) combined with the equivariant condition (2.2). Indeed, assume that is nontrivial, and let be a nonzero function belonging to . According to (2.5), we get for every . On the other hand, we can write Now, by equating the right hand sides of (3.4) and (3.5), keeping in mind that is not identically zero, we get necessarily
Conversely, by classical analysis, we pick an arbitrary nonzero and compactly supported function with support contained in a fundamental domain of the lattice . Next, we consider the associated Poincaré series given by Then, it can be shown that the function is and a nonzero function on for being discrete and . Indeed, for every , we have Furthermore, under the condition that takes integral values on , we see that belongs to . In fact, for every and , we have Finally, since (by hypothesis on ), it follows that The last equality, that is, holds for every using the chain rule (2.5) and taking into account the assumption made on . This completes the proof.

Remark 3.3. The condition involved in Proposition 3.2 ensures that can be realized as the space of cross-sections on a line bundle over the complex torus .

4. On the Function

In order to prove the main result of this paper, we need to introduce the function .

Proposition 4.1. The first-order differential equation admits a solution such that is constant.

Proof. By writing the -endomorphism as , and differentiating the equivariant condition it follows that Hence, for and being in , we deduce that Therefore, is a real-valued constant function (since the only -invariant functions on are the constants). Now, by considering the differential 1-form one checks that , where . Therefore, there exists a function such that and satisfying the first-order partial differential equation This completes the proof.

Remark 4.2. The partial differential equation (4.6) satisfied by can be reduced further to the following: with .

5. Main Result

Let be the real part of , where is a complex-valued function on as in Proposition 4.1. Define to be the special transformation given by We have the following.

Theorem 5.1. The image of by the transform (5.1) is the space of Landau -automorphic functions. More exactly, one has with and is the pseudocharacter defined on by

For the proof, we begin with the following.

Lemma 5.2. The function defined on by is independent of the variable .

Proof. Differentiation of with respect to the variable gives On the other hand, using the equivariant condition and (4.6), one gets where we have set . Thus, from (5.5) and (5.6), we conclude that . Similarly, one gets also . This ends the proof of Lemma 5.2.

Proof of Theorem 5.1. We have to prove that belongs to whenever , where Indeed, we have Whence by Lemma 5.2, we see that , and therefore The proof is complete.

Corollary 5.3. The function satisfies the following pseudocharacter property: if and only if in (2.6) takes its values in on .

Acknowledgment

The author is indebted to Professor A. Intissar for valuable discussions and encouragement.