Abstract. Let F be a p-adic field and E/F be a quadratic extension. In this paper, we prove the stability of Rankin-Selberg gamma factors for Sp 2n (F ), ‹ Sp 2n (F ) and U E/F (n, n) when the characteristic of the residue field of F is not 2.
IntroductionLet G n be Sp 2n , Sp 2n and U E/F (n, n), where E/F is a quadratic extension of local or global field. The global Rankin-Selberg zeta integrals for the generic irreducible cuspidal automorphic representations of G n twisted by generic irreducible cuspidal representations of GL m has been developed by Gelbart, Piatetski-Shapiro, Ginzburg, Rallis and Soudry, [GePS2,GiRS1,GiRS2]. Recently, the standard properties of such local γ-factors were established by Kaplan [Ka]. As a complimentary result of their work, in this paper, we prove the stability of the local gamma factor for a generic representation of G n (F ) when twisted by a sufficiently highly ramified character of GL 1 for a p-adic field F , when the residue field of F is not 2. More precisely, the main result of this paper is the following Theorem 0.1. Let F be a p-adic field such that the characteristic of its residue field is odd, E/F be a quadratic extension. Let ψ U be a generic character of a maximal unipotent subgroup of G n (F ) defined by a given nontrivial additive character ψ of F . Let π 1 , π 2 be two ψ U -generic irreducible smooth representations of G n (F ) with the same central character. If η is a highly ramified quasicharacter of F × , then γ(s, π 1 , η, ψ) = γ(s, π 2 , η, ψ).Here the γ-factors are the Rankin-Selberg gamma factors, see §1 for more details. We also notice that the main theorem also holds for U E/F if the residue characteristic of F is 2 and E/F is unramified.Here we remark that in the Sp 2n case, this result can be deduced from previous work. Cogdell, Kim, Piatetski-Shapiro and Shahidi proved the stability of gamma factors for classical groups (which at least includes Sp 2n and SO n ) in [CKPSS], where the gamma factors are defined using Langlands-Shahidi method. In [Ka], Kaplan proved that Rankin-Selberg gamma factors agree with the Langlands-Shahidi gamma factor. Thus our result in the Sp 2n case follows from the stability result in [CKPSS] and Kaplan's result on the agreement of the two type gamma factors. In the U E/F (n, n)-case, the stability of the Langlands-Shahidi gamma factors is proved in [KK]. Thus in principle, our result in the U E/F (n, n) case should follow from an agreement result of the two type γ-factors, which is unfortunately not included in [Ka].In this paper, we prove the stability of gamma factors for G n in the Rankin-Selberg context. Although one can deduce this by pulling back the Langlands-Shahidi gamma factors via [Ka], it is still important to have a proof of stability that remains within the context of integral representations, since there are L-functions that we have integral representations for that are not covered by the Langlands-Shahidi method. So developing methods that work in the integral representation context have an intrinsic valu...