Endomorphisms of superelliptic Jacobians (Q1006820)

Let \(p\) be a prime and \({\zeta}_p\) be a primitive \(p\)-th root of unity, \(Q( {\zeta}_p) \) its \(p\)-th-cyclotomic field and \(Z[ {\zeta}_p]\) the ring of integers. Let \(C\) be a smooth projective model of the smooth affine curve \( y^p = f(x)\). The map \((x,y)\mapsto (x, {\zeta} y)\) gives rise to a \(K\)-automorphism \( \delta \) of \(C\) of period \(p\). The author's main result is the following theorem here stated as theorem 1.1. Suppose that \(\text{char}(K) =0, n \geq 4 \) and \(p\) is an odd prime that does not divide \(n\), that \(K\) contains a primitive \(p\)-th root of unity and \(\text{Gal}(f)\) contains a doubly transitive subgroup \(G\) such that if \(p |(n-1) \) with \(n \neq p +1 \) then \(G\) does not contain a proper subgroup, whose index divides \((n-1)/p\). Then the author shows that the algebra of complex multiplications of \(J(C)\) is a simple \(Q\)-algebra. If \(G\) does not contain a proper normal subgroup whose index divides \(n-1\) then the algebra of complex multiplications of \(J(C)\) is a central simple \( Q[\delta]\)-algebra. Moreover, if either \( n= p+1\) or \( p \) does not divide \(n-1\) then the algebra of complex multiplications is in fact \(Q[ \delta] \) and the algebra of endomorphisms of \(J(C)\) is \( Z[ \delta] \). Previous work for the case \(n=3\) is discussed in \textit{K. A. Ribet} [Ann. Math. (2) 101, 555--562 (1975; Zbl 0305.14016)] and [\textit{Y. G. Zarhin}, Diophantine geometry. Selected papers of a the workshop, Pisa, Italy, April 12--July 22, 2005. Pisa: Edizioni della Normale. Centro di Ricerca Matematica Ennio De Giorgi (CRM) Series (Nuova Serie) 4, 363--390 (2007; Zbl 1152.14029)]. The paper is organized as follows. In section two he recalls basic definitions and properties of endomorphisms of algebras of abelian varieties. In section three the author considers \(E\) a number field and a pair \( (X, \imath)\) consisting of an abelian variety and an embedding \( \imath \) of \(E\) into its algebra of complex multiplications. Let \(O\) be the ring of integers in \(E\), \(G\) the Galois group of the field extension of the field of definition of all points in \(X_{\lambda}\) over \(K\), \( d = 2 \text{dim}(X)/[E: Q]\), denote by \(\text{End}^0(X, \imath)\) the centralizer of \(\imath(E)\) in \(\text{End}^0(X)\), where \( n(X,\imath) = \text{dim}_{K}( \text{Lie}_{K}(X)_{\tau})\). The author's main result in this section is theorem 3.12 which is the following. Assume \(E\) is a number field that is normal over \(Q\). Suppose that \(Q\) is a field of \(\text{char}(K) = 0 \) that contains a subfield isomorphic to \(E\). If \(X\) is defined over \(Q\) and \( i(O) \subset \text{End}_K(X) \). Denote by \( n_{X,E} \) the greatest common divisor of all \( n(X, \imath)\) where it runs through the set of embeddings \(E \hookrightarrow K\). Suppose that \(\lambda\) is the only maximal ideal in \(O\) that lies above the prime \(l\), the group \(G\) does not contain a subgroup, whose index divides \(n_{X,E}\), except \(G\) and \(\text{End}_G( X_{\lambda}) = k(\lambda)\) then: [1] The algebra of complex multiplications of \(X\) is a simple \(Q\)-algebra. [2] Suppose that \(G\) does not contain a normal subgroup, whose index divides \(d\) except for \(G\). Then: [3] \(\text{End}^0(E, \imath)\) is a simple central \(E\)-algebra. [4] If \( n= 1\) then \( \text{End}^0(X, \imath) = \imath(E), \text{End}(X,\imath) = \imath(O)\). In section four the author proves theorem 1.1. An essential part of the proof is contained in theorem 3.12 where it is applied for the case of \(E = Q( \zeta_p), O = Z[ \zeta_p], l=p, \lambda = (1-\zeta_p) Z[ \zeta_p], k(\lambda) = F_p, X = J(C) \) and previous work form the author. In the last section, section five the author states and proves theorem 5.2 of similar type as theorem 1.1 which considers the case \( q= p^r\) for \( r >1\) a positive integer, assuming that \(\text{Gal}(f) \) contains a doubly transitive subgroup \(G\) such that if \( q|n\) and \( n \neq q + 1\) then \(G\) does not contain a proper subgroup, whose index divides \((n-1)/q\). In case \(p=2\) he assumes additionally that \(G\) does not contain a subgroup of index two.