Theta lifting from elliptic cusp forms to automorphic forms on \(\mathrm{Sp}(1, q)\) (Q926246)

In this paper the author introduces (after Arakawa) the space of bounded automorphic forms \(A_0(\Gamma \setminus G,\omega_{\kappa})\) on the real symplectic group \(G=\text{Sp}(1,q)\) of signature \((1+,q-).\) On the other hand, there are unpublished works of Arakawa studying the theta lifts from elliptic cusp forms to automorphic forms on \(\text{Sp}(1,q.)\) The pair \((\text{SL}_2(\mathbb R),\text{Sp}(1,q)),\) unless \(q=1,\) is not a (standard) dual reductive pair in any symplectic group. So, there is no a classical theta lifting by means of Weil representation, but \(\text{Sp}(1,q)\) is viewed as a subgroup of the special orthogonal group \(\text{SO}(4,4q),\) so the restriction of the classical theta correspondence for the pair \( (\text{SL}_2(\mathbb R),\text{SO}(4,4q))\) (classical dual pair) to the pair \((\text{SL}_2(\mathbb R),\text{Sp}(1,q))\) is studied. More precisely, Arakawa constructed theta series on the product of the complex upper half plane and \(\text{SO}(4,4q)\) and studied it's restriction to the product of the upper half-plane and \(\text{Sp}(1,q).\) Arakawa proved that, under some choice of \(\Gamma,\) the images of his lifting belong to \(A_0(\Gamma \setminus G,\omega_{\kappa})\) when \(q=1.\) The main result of this paper is a generalization of Arakawa's results to the case \(q>1\) (when \(\kappa >4q+2\)). We describe this space of automorphic forms in more detail: Here \(\kappa\) is a natural number, and \(\sigma_{\kappa}\) is a pull back of the \(\kappa-\)th symmetric tensor representation of \(M_2(C)\) to \(H\) (Hamilton quaternions in the matrix realization). Then, to \(\sigma_{\kappa}\) is attached certain irreducible finite-dimensional representation \((\tau_{\kappa},V_{\kappa})\) of the maximal compact subgroup \(K\) of \(G.\) To \(\tau_{\kappa}\) is attached a discrete series representation \(\pi_{\kappa}\) of \(G\) with the minimal \(K\)-type \(\tau_{\kappa}\) (such \(\pi_{\kappa}\) exists when \(\kappa >2q-1\) by the regularity of the Harish-Chandra parameter of \(\pi_{\kappa}\)). To this \(\pi_{\kappa}\) certain \(K\)--spherical function \(\omega_{\kappa}:G\to \text{End}(V_{\kappa})\) is attached. For \(\kappa>4q\) and arithmetic subgroup \(\Gamma\) of \(G\) contained in \(G(Q)\), Arakawa defined a space of automorphic forms \(A_0(\Gamma\setminus G,\omega_{\kappa})\) as space of continuous, \(V_{\kappa}\)--valued functions on \(G\) satisfying some natural conditions (cf. Definition 2.1). In his previous work, the author showed that these forms generate the quaternionic discrete series \(\pi_{\kappa}.\) Now we briefly describe the content of the paper: in the first section the author recalls the definition of the group \(\text{Sp}(1,q)\) and introduces the relevant notation. In the second section, there is a review of the automorphic forms on \(\text{Sp}(1,q)\) generating quaternionic discrete series (for given \(\kappa\)). Also, some useful lemmas on the Fourier expansion of an automorphic form at a cusp are given. In the third section, the theta series on the product of the complex upper half plane and \(\text{SO}(p,q)\) is studied. The theta series is constructed in the following way: for \(p+q=n\) and \(V=\mathbb R^n,\) there is a Weil representation of \(\text{Sp}(V\times V)\) on the space \(L^2(V)\) of square-integrable functions on \(V.\) \(\text{SL}_2(\mathbb R)\) can be naturally considered to be a subgroup of \(\text{Sp}(V\times V),\) so that \(\text{SL}_2(\mathbb R)\) acts on \(L^2(V).\) The author then constructs a special test function in the space of \(L^2(V);\) this function will be used to construct theta series (it is defined using homogeneous polynomials on \(V\)). The rest of the first part of the third section is devoted to the explicit calculation of the action of \(\text{SL}_2(\mathbb R)\) on this test function, and to the construction of the theta function on the product of the upper half plane and \(\text{SO}(p,q)\) (using this test function). The second part of the third section is devoted to the study of the restriction of the theta series to the product of the upper half plane and the group \(\text{Sp}(1,q),\) which is realized as a subgroup of (conveniently chosen realization of) the real orthogonal group \(\text{SO}(4,4q).\) Now, the test function is a bit specialized version of the test function from the first part of the third section; now the definition of the test function includes a special choice of the homogeneous polynomial related to the representation \(\sigma_{\kappa}.\) Also, an appropriate arithmetic subgroup of \(G= \text{Sp}(1,q)\) is introduced. In the fourth section, a theta integral is defined; the integration is over the quotient of the upper half plane with the appropriate arithmetic subgroup. The integrand consists of the product of an elliptic cusp form of weight \(\kappa -2q+2\) (with respect to this subgroup) against the theta kernel (introduced in the previous section). The rest of the fourth section is devoted to the proof that this theta integral, applied to every vector from \(V_{\kappa},\) gives a theta lifting from the space of elliptic cusp forms of weight \(\kappa -2q+2\) to the space of the automorphic forms \(A_0(\Gamma \setminus G,\omega_{\kappa}),\) described above. In the appendix, the author gives two other constructions of the automorphic forms generating quaternionic discrete series \(\pi_{\kappa};\) these two constructions include Eisenstein series and Poincaré series. The author also shows that the images of these two constructions in the space of the automorphic forms, span this space.