Startseite On large prime actions on Riemann surfaces
Artikel Öffentlich zugĂ€nglich

On large prime actions on Riemann surfaces

  • SebastiĂĄn Reyes-Carocca EMAIL logo und Anita M. Rojas
Veröffentlicht/Copyright: 31. MÀrz 2022

Abstract

In this article, we study compact Riemann surfaces of genus 𝑔 with an automorphism of prime order g + 1 . The main result provides a classification of such surfaces. In addition, we give a description of them as algebraic curves, determine and realise their full automorphism groups and compute their fields of moduli. We also study some aspects of their Jacobian varieties such as isogeny decompositions and complex multiplication. Finally, we determine the period matrix of the Accola–Maclachlan curve of genus four.

1 Introduction and statement of the results

Let M g denote the moduli space of compact Riemann surfaces (smooth irreducible complex algebraic curves) of genus g ⩟ 2 . It is classically known that M g is endowed with the structure of a complex analytic space of dimension 3 ⁹ g - 3 , and that, for g ⩟ 4 , its singular locus agrees with the branch locus of the canonical projection

T g → M g ,

where T g stands for the TeichmĂŒller space of genus 𝑔. In other words, if g ⩟ 4 , then

Sing ⁱ ( M g ) = { [ S ] ∈ M g : Aut ⁡ ( S ) ≠ 1 } ,

where Aut ⁡ ( S ) denotes the full automorphism group of 𝑆.

The classification of groups of automorphisms of compact Riemann surfaces is a classical problem which has attracted broad interest ever since it was proved that the full automorphism group of a compact Riemann surface 𝑆 of genus g ⩟ 2 is finite, and that

| Aut ⁥ ( S ) | ⩜ 84 ⁹ ( g - 1 ) .

It is well known that there are infinitely many values of 𝑔 for which there is no compact Riemann surfaces of genus 𝑔 possessing 84 ⁱ ( g - 1 ) automorphisms. Regarding this matter, Accola [2] and Maclachlan [46] proved that, for fixed 𝑔, the largest order n 0 ⁱ ( g ) of the full automorphism group of a compact Riemann surface of genus 𝑔 satisfies

(1.1) n 0 ⁹ ( g ) ⩟ 8 ⁹ ( g + 1 )

and that, for infinitely many values of 𝑔, the inequality (1.1) turns into an equality.

We denote by X 8 the so-called Accola–Maclachlan curve, namely the compact Riemann surface of genus 𝑔 with 8 ⁱ ( g + 1 ) automorphisms given by the algebraic curve

y 2 = x 2 ⁹ ( g + 1 ) - 1 .

The Accola–Maclachlan curve is a remarkable example of a compact Riemann surface determined by the order of its full automorphism group. More precisely, Kulkarni in [43] succeeded in proving that, up to finitely many values of the genus, if g ≱ 3 mod 4 , then X 8 is the unique compact Riemann surface of genus 𝑔 with exactly 8 ⁱ ( g + 1 ) automorphisms.

The analogous problem of finding n 0 ⁱ ( g ) but for uniparametric families of compact Riemann surfaces was studied in [22]. Concretely, the existence was proved of a closed equisymmetric complex one-dimensional family, henceforth denoted by C ¯ g , of hyperelliptic compact Riemann surfaces of genus 𝑔 with a group of automorphisms isomorphic to

D g + 1 × C 2   acting with signature ⁱ ( 0 ; 2 , 2 , 2 , g + 1 )

(we shall recall the precise definition of signature in § 2.1 and § 2.2). It was then shown that 4 ⁱ ( g + 1 ) is the largest order of the full automorphism group of complex one-dimensional families of compact Riemann surfaces of genus 𝑔 appearing for all 𝑔. These results were recently extended to the three and four-dimensional cases in [40] while the two-dimensional case is derived from the results of [56].

It is a well-known fact that if a compact Riemann surface of genus g ⩟ 2 has an automorphism of prime order 𝑞 such that q > g , then either q = 2 ⁹ g + 1 or q = g + 1 . The former case corresponds to the so-called Lefschetz surfaces. This paper deals with the latter case.

Let q ⩟ 5 be a prime number. Consider the singular sublocus

M q - 1 q ⊂ Sing ⁱ ( M q - 1 )

consisting of all those compact Riemann surfaces of genus q - 1 endowed with an automorphism of order 𝑞. This sublocus was studied by UrzĂșa in [68] from a hyperbolic geometry point of view, and later by Costa and Izquierdo in [21] when the existence of complex one-dimensional isolated strata of the singular locus of the moduli space was proved.

This paper is devoted to classifying and describing the surfaces lying in M q - 1 q and to study some aspects of the corresponding Jacobians in the singular locus of the moduli space of principally polarised abelian varieties of dimension q - 1 . In other words, we shall consider all those compact Riemann surfaces of genus g ⩟ 4 (and their Jacobian varieties) with a group of automorphisms of order

λ ⁹ ( g + 1 ) , where ⁹ λ ⩟ 1 ⁹ is an integer ,

under the assumption that q := g + 1 is a prime number.

The classification

The first result of the paper provides a classification of these surfaces.

Theorem 1

Let q ⩟ 7 be a prime number. If 𝑆 is a compact Riemann surface of genus g = q - 1 endowed with a group of automorphisms of order λ ⁹ q for some integer λ ⩟ 1 , then λ ∈ { 1 , 2 , 3 , 4 , 8 } .

Assume λ = 8 . Then 𝑆 is isomorphic to the Accola–Maclachlan curve X 8 .

Assume λ = 4 .

  1. If q ≡ 3 mod 4 , then 𝑆 belongs to the closed family C ¯ g .

  2. If q ≡ 1 mod 4 , then 𝑆 belongs to the closed family C ¯ g or 𝑆 is isomorphic to the unique compact Riemann surface X 4 with full automorphism group isomorphic to

    C q ⋊ 4 C 4   acting with signature ⁱ ( 0 ; 4 , 4 , q ) .

Moreover, if C g stands for the interior of C ÂŻ g , then

C ÂŻ g - C g = { X 8 } .

Assume λ = 3 . Then 𝑆 is isomorphic to the unique compact Riemann surface X 3 with full automorphism group isomorphic to

C q × C 3   acting with signature ⁱ ( 0 ; 3 , q , 3 ⁱ q ) .

Assume λ = 2 . Then one of the following statements holds.

  1. 𝑆 is isomorphic to one of the q - 3 2 pairwise non-isomorphic compact Riemann surfaces X 2 , k for k ∈ { 1 , 
 , q - 3 2 } with full automorphism group isomorphic to

    C q × C 2   acting with signature ⁱ ( 0 ; q , 2 ⁱ q , 2 ⁱ q ) .

  2. 𝑆 belongs to the closed family K ¯ g of compact Riemann surfaces with a group of automorphisms isomorphic to

    D q   acting with signature ⁱ ( 0 ; 2 , 2 , q , q ) .

Moreover, the closed family K ÂŻ g consists of at most

{ q + 3 4 if ⁱ q ≡ 1 mod 4 , q + 1 4 if ⁱ q ≡ 3 mod 4 ,

equisymmetric strata; one of them being C g . Furthermore, if K g stands for the interior of K ¯ g , then the full automorphism group of S ∈ K g - C g is isomorphic to D q and

K ¯ g - K g = { { X 4 , X 8 } if ⁱ q ≡ 1 mod 4 , { X 8 } if ⁱ q ≡ 3 mod 4 .

Remark 1

We point out some observations concerning Theorem 1.

  1. If S ∈ M q - 1 q , then either Aut ⁡ ( S ) ≅ C q or 𝑆 lies in one of the cases described in the theorem. These two possible situations were considered in [21] where the focus was put on finding isolated equisymmetric strata of Sing ⁱ ( M g ) . We shall discuss later the results of [21] in terms of our terminology (see Remark 2 in § 3).

  2. The case q = 5 is slightly different. As a matter of fact, if 𝑆 has genus 4 and is endowed with a group of automorphisms of order 5 ⁹ λ for some λ ⩟ 1 , then in addition to the case Aut ⁥ ( S ) ≅ C 5 and the possibilities given in the theorem, 𝜆 can equal 12 and 24. In the last two cases, 𝑆 is isomorphic to the classical Bring curve; see [19, 20, 44].

  3. We conjecture that the upper bound given in the theorem for the number of equisymmetric strata of the closed family K ÂŻ g is sharp. By means of computer routines developed in [7], one can see the sharpness of the bound for small primes ( q ⩜ 23 ).

  4. We emphasise that the equisymmetric family C ÂŻ g is contained in the family K ÂŻ g .

  5. An analogous classification as in the theorem but for the compact Riemann surfaces lying in M q + 1 q was obtained in a series of articles due to Belolipetsky, Izquierdo, Jones and the first author; see [8, 38, 39, 56].

Algebraic description

Although the literature still shows few general results in this direction, there is a great interest in providing descriptions of compact Riemann surfaces as algebraic curves in an explicit manner. The following result gives such a description for the surfaces appearing in Theorem 1, as well as a realisation of their full automorphism groups.

Theorem 2

Let q ⩟ 5 be a prime number, and let g = q - 1 . Set ω l = exp ⁥ ( 2 ⁹ π ⁹ i l ) .

If 𝑆 belongs to the closed family C ¯ g , then 𝑆 is isomorphic to the normalisation of the singular affine algebraic curve

X t : y 2 = ( x q - 1 ) ⁱ ( x q - t )   for some ⁱ t ∈ C - { 0 , 1 } .

In addition, if S ∈ C g , then the full automorphism group of S ≅ X t is generated by

( x , y ) ↩ ( ω q ⁹ x , - y )   and   ( x , y ) ↩ ( t q ⁹ 1 x , t ⁹ y x q ) .

Assume q ≡ 1 mod 4 , and choose ρ ∈ { 2 , 
 , q - 2 } such that ρ 4 ≡ 1 mod q . Then X 4 is isomorphic to the normalisation of the singular affine algebraic curve

y q = ( x - 1 ) ⁹ ( x - i ) ρ ⁹ ( x + 1 ) q - 1 ⁹ ( x + i ) q - ρ ,

where i 2 = - 1 . In the previous model, the full automorphism group of X 4 is generated by

( x , y ) ↩ ( x , ω q ⁹ y )   and   ( x , y ) ↩ ( i ⁹ x , φ ⁹ ( x ) ⁹ y ρ ) ,

where

φ ⁹ ( x ) = - ( x + i ) e - ρ ( x - i ) e - 1 ⁹ ( x + 1 ) ρ - 1   and   e = ρ 2 + 1 q .

The surface X 3 is isomorphic to the normalisation of the singular affine algebraic curve

y 3 = x q - 1 ,

and in this model, its full automorphism group is generated by

( x , y ) ↩ ( ω q ⁹ x , ω 3 ⁹ y ) .

For each k ∈ { 1 , 
 , q - 3 2 } , there exists n k ∈ { 1 , 
 , q - 1 } different from q - 2 such that X 2 , k is isomorphic to the normalisation of the singular affine algebraic curve

y q = x n k ⁹ ( x 2 - 1 )

and, in this model, its full automorphism group is generated by

( x , y ) ↩ ( x , ω q ⁹ y )   and   ( x , y ) ↩ ( - x , ( - 1 ) n k ⁹ y ) .

If 𝑆 belongs to the closed family K ¯ g , then 𝑆 is isomorphic to the normalisation of the singular affine algebraic curve

Z t : y q = ( x - 1 ) ⁹ ( x + 1 ) q - 1 ⁹ ( x - t ) ⁹ ( x + t ) q - 1   for some ⁹ t ∈ C - { 0 , ± 1 }

and, if S ≠ X 4 and S ∉ C ¯ g , then the full automorphism group of S ≅ Z t is generated by

( x , y ) ↩ ( x , ω q ⁹ y )   and   ( x , y ) ↩ ( - x , ϕ t ⁹ ( x ) ⁹ y - 1 ) ,

where ϕ t ⁹ ( x ) = ( x 2 - 1 ) ⁹ ( x 2 - t 2 ) .

The theorem above overlaps results obtained in [68, § 11].

Hyperelliptic surfaces

Arakelian and Speziali in [3] studied groups of automorphisms of large prime order of (non-necessarily smooth) projective absolutely irreducible algebraic curves over algebraically closed fields of any characteristic. In terms of our terminology, in [3, Theorem 4.7], they proved that if q ⩟ 7 is a prime number and 𝑆 is a compact Riemann surface of genus q - 1 with a group of automorphisms of order λ ⁹ q , then

S ⁹ is non-hyperelliptic implies ⁹ λ ∈ { 1 , 2 , 3 , 4 } .

The following result lengthens the implication above; it follows from Theorem 1.

Proposition 1

Let q ⩟ 7 be a prime number. The compact Riemann surfaces lying in M q - 1 q that are non-hyperelliptic are X 2 , k , X 3 , X 4 , the surfaces which belong to K g - C g and those for which Aut ⁥ ( S ) ≅ C q .

Jacobian variety

Let 𝑆 be a compact Riemann surface of genus g ⩟ 2 . We denote by J ⁹ S the Jacobian variety of 𝑆, that is, the quotient

J ⁹ S = H 1 ⁹ ( S , C ) * / H 1 ⁹ ( S , Z ) ,

where H 1 ⁱ ( S , C ) * stands for the dual of the 𝑔-dimensional complex vector space of holomorphic forms of 𝑆 and H 1 ⁱ ( S , Z ) stands for the first integral homology group of 𝑆.

We emphasise the following two classical facts (see, for example, [10]):

  1. J ⁱ S is an irreducible principally polarised abelian variety of dimension 𝑔, and

  2. up to isomorphism, the surface is determined by its Jacobian (Torelli’s theorem).

If đș is a group acting on 𝑆, then đș also acts on J ⁹ S and this action, in turn, induces the so-called group algebra decomposition of J ⁹ S . Concretely,

J ⁹ S ∌ A 1 × ⋯ × A r ∌ B 1 n 1 × ⋯ × B r n r ,

where the factors A j are pairwise non-đș-isogenous abelian subvarieties of J ⁹ S uniquely determined, and in correspondence with central idempotents generating the simple algebras decomposing the rational group algebra of đș. Each A j decomposes further as B j n j , where the abelian subvarieties B j are no longer unique and are related to the decomposition of each simple algebra as a product of minimal left ideals. The numbers 𝑟 and n j depend only on the algebraic structure of đș. See [16, 45].

The following result provides the group algebra decomposition of the Jacobian varieties of the surfaces of Theorem 1, with the exception of X 3 and X 2 , k . In fact, the group algebra decomposition of J ⁹ X 3 is trivial whilst that of J ⁹ X 2 , k agrees with the classical decomposition

J ⁹ X 2 , k ∌ J ⁹ ( X 2 , k / H ) × Prym ⁹ ( X 2 , k → X 2 , k / H ) ,

where Prym stands for the Prym variety and H ⩜ Aut ⁥ ( X 2 , k ) is isomorphic to C 2 .

Theorem 3

Let q ⩟ 5 be a prime number, and let g = q - 1 .

The Jacobian variety J ⁹ X 8 decomposes, up to isogeny, as the square power J ⁹ X 8 ∌ J ⁹ Y 8 2 , where Y 8 is quotient compact Riemann surface given by the action of ⟹ z ⟩ on X 8 , where

Aut ( X 8 ) ≅ ⟹ x , y , z : x 2 ⁹ q = y 2 = z 2 = 1 , [ x , y ] = [ z , y ] = 1 , z x z = x - 1 y ⟩ .

The Jacobian variety J ⁹ X 4 of X 4 decomposes, up to isogeny, as the fourth power J ⁹ X 4 ∌ J ⁹ Y 4 4 , where Y 4 is quotient compact Riemann surface given by the action of ⟹ B ⟩ on X 4 , where

Aut ( X 4 ) ≅ ⟹ A , B : A q = B 4 = 1 , B A B - 1 = A ρ ⟩

and 𝜌 is a primitive fourth root of unity in Z q .

The Jacobian variety J ⁹ S of S ∈ K g decomposes, up to isogeny, as the square power J ⁹ S ∌ J ⁹ X 2 , where 𝑋 is the quotient compact Riemann surface given by the action of ⟹ s ⟩ on 𝑆, where

Aut ⁡ ( S ) ≅ { D q if ⁱ S ∈ K g - C g , D q × C 2 if ⁱ S ∈ C g ,

and D q = ⟹ r , s : r q = s 2 = ( s r ) 2 = 1 ⟩ .

Field of moduli and fields of definition

Let Gal ⁱ ( C / Q ) denote the group of field automorphisms of ℂ. The correspondence

Gal ⁹ ( C / Q ) × M g → M g   given by ⁹ ( σ , [ S ] ) ↩ [ S σ ] ,

where S σ is the Galois 𝜎-transformed of 𝑆 (considered as algebraic curve), defines an action.

The field of moduli of a compact Riemann surface 𝑆 is the fixed field M ⁱ ( S ) of the isotropy group of 𝑆 under the aforementioned action, namely

M ⁹ ( S ) = fix ⁥ { σ ∈ Gal ⁹ ( C / Q ) : S σ ≅ S } .

The field of moduli of 𝑆 agrees with the intersection of all its fields of definition and, as proved by Koizumi in [41], 𝑆 can be defined over a finite degree extension of M ⁱ ( S ) .

Necessary and sufficient conditions under which 𝑆 can be defined over its field of moduli were provided by Weil in [69] (see also [34] for a constructive proof of Weil’s theorem); these conditions are trivially satisfied if 𝑆 has no non-trivial automorphisms. Besides, as proved by Wolfart in [71], if 𝑆 is quasiplatonic, then 𝑆 can be defined over its field of moduli.

The general question of deciding whether or not the field of moduli is a field of definition is a challenging problem; see, for example, [4, 27, 30, 33, 35, 42, 54]. In this direction, it is a known fact that if the genus of S / Aut ⁡ ( S ) is zero, then either 𝑆 can be defined over M ⁱ ( S ) or over a quadratic extension of it; see [23] and also [31] for recent results.

We now study the aforementioned problem for the compact Riemann surfaces of Theorem 1. First, note that, for the quasiplatonic ones, the problem is trivial. Indeed,

  1. as proved in Theorem 2, the surfaces X 3 , X 2 , k and X 8 are defined over ℚ, and therefore their fields of moduli are ℚ.

  2. As mentioned above, the fact that X 4 is quasiplatonic implies that it can be defined over its field of moduli. Moreover, the uniqueness of X 4 implies that its field of moduli is ℚ. In fact, we shall see later (Remark 3 in § 4) that X 4 is isomorphic to

    y q = x ⁹ ( x + 1 ) ρ ⁹ ( x - 1 ) q - ρ .

The remaining cases (that is, the surfaces lying in the family K g since it contains C g ) are given in the following proposition.

Proposition 2

Let q ⩟ 5 be prime, and let g = q - 1 . If 𝑆 belongs to the family K g and

S ≅ Z t = { ( x , y ) : y q = ( x - 1 ) ⁱ ( x + 1 ) q - 1 ⁱ ( x - t ) ⁱ ( x + t ) q - 1 }

for t ∈ C - { 0 , ± 1 } , then the field of moduli of 𝑆 is Q ⁹ ( t ) .

It it worth mentioning that a compact Riemann surface and its Jacobian variety can be defined over the same fields and that their fields of moduli agree; see [63] and also [47].

The following result is a direct consequence of the above.

Corollary 1

The compact Riemann surfaces of Theorem 1 and their Jacobian varieties can be defined over their fields of moduli.

The sublocus of A g with đș-action

It is well known that the moduli space A g of principally polarised abelian varieties of dimension 𝑔 is isomorphic to the quotient

π : H g → A g ≅ H g / Sp ⁥ ( 2 ⁹ g , Z )

of the Siegel upper half-space H g by the action of the symplectic group Sp ⁥ ( 2 ⁹ g , Z ) . If the isomorphism class of J ⁹ S is represented by Z S ∈ H g , then there is an isomorphism of groups

Aut ⁡ ( J ⁱ S ) ≅ Σ S := { R ∈ Sp ⁡ ( 2 ⁱ g , Z ) : R ⋅ Z S = Z S } ,

where Σ S is well-defined up to conjugation in Sp ⁥ ( 2 ⁹ g , Z ) . The subset of H g given by

S S := { Z ∈ H g : R ⋅ Z = Z ⁱ for all ⁱ R ∈ Σ S }

consists of those matrices representing principally polarised abelian varieties of dimension 𝑔 admitting an action which is equivalent to that of Aut ⁡ ( J ⁱ S ) . This subset is, indeed, an analytic submanifold of H g closely related with some special subvarieties and Shimura families of A g .

Observe that if U ¯ g is an equisymmetric family of compact Riemann surfaces of genus 𝑔 and if 𝑆 is any surface lying in the interior U g of U ¯ g , then

{ J ⁹ X : X ∈ U g } ⊆ π ⁹ ( S S ) .

In general, those loci of A g do not agree. Nonetheless, the uncommon cases in which these dimensions do agree have been useful in finding Jacobians with complex multiplication.

Although a satisfactory description of the matrices in S S seems to be a difficult problem, as we shall see in § 2.7, there is a simple representation-theoretic way to compute the dimension of the (component which contains J ⁱ S of) S S . We shall denote the aforementioned dimension by N S .

Theorem 4

Let q ⩟ 5 be prime, let g = q - 1 , and let S ∈ K g . Then

N X 8 = N X 3 = N X 2 , k = 0 , N X 4 = q - 1 4   and   N S = q - 1 2 .

According to results due to Streit in [67] (and later generalised in [26] for higher dimension), if N S is zero, then the full automorphism group of 𝑆 determines the period matrix for J ⁱ S and J ⁱ S admits complex multiplication. We refer to [49, 50] for recent applications of this result for quasiplatonic curves that are hyperelliptic and superelliptic.

As a direct consequence of the previous theorem, we recover the following known result.

Corollary 2

The Jacobian varieties of X 3 , X 2 , k and X 8 admit complex multiplication.

In spite of the fact that the problem of determining the period matrix of a given Jacobian variety is, in general, intractable, interesting results have been obtained for some famous Riemann surfaces. For instance, the period matrices of Macbeath’s curve of genus seven and of Bring’s curve were determined in [9] and [60] respectively. A method to find the period matrices of the Accola–Maclachlan and Kulkarni surfaces was given in [13]. In addition, in [13, Example 3.7], the authors went even further and employed their method to provide the period matrix of the Accola–Maclachlan curve of genus two in an explicit way.

At the end of the paper, we determine explicitly the period matrix of the Accola–Maclachlan curve of genus four.

This article is organised as follows. In § 2, we succinctly review the basic preliminaries: Fuchsian groups and group action on Riemann surfaces and abelian varieties. The proof of Theorem 1 is given in § 3, and the proofs of Theorem 2 and Proposition 2 are given in § 4. In § 5, we prove some basic algebraic lemmata needed to prove, in § 6, Theorems 3 and 4. Finally, we include an addendum in which the period matrix of the Accola–Maclachlan curve of genus four is computed.

2 Preliminaries

2.1 Fuchsian groups

A Fuchsian group is a discrete group of automorphisms of the upper half-plane ℍ. If Δ is a Fuchsian group and the orbit space H / Δ given by the action of Δ on ℍ is compact, then the algebraic structure of Δ is determined by its signature

(2.1) σ ⁹ ( Δ ) = ( Îł ; k 1 , 
 , k s ) ,

where đ›Ÿ is the genus of H / Δ and k 1 , 
 , k s are the branch indices in the universal canonical projection H → H / Δ . In this case, Δ has a canonical presentation in terms of canonical generators α 1 , 
 , α Îł , ÎČ 1 , 
 , ÎČ Îł , x 1 , 
 , x s and relations

(2.2) x 1 k 1 = ⋯ = x s k s = ∏ i = 1 Îł [ α i , ÎČ i ] ⁹ ∏ i = 1 s x i = 1 ,

where the brackets stand for the commutator. The TeichmĂŒller space of Δ is a complex analytic manifold homeomorphic to the complex ball of dimension 3 ⁹ Îł - 3 + s .

Let Δ â€Č be a group of automorphisms of ℍ such that Δ ⩜ Δ â€Č of finite index. Then Δ â€Č is also Fuchsian, and they are related by the so-called Riemann–Hurwitz formula

2 Îł - 2 + ∑ i = 1 s ( 1 - 1 k i ) = [ Δ â€Č : Δ ] ⋅ [ 2 Îł â€Č - 2 + ∑ i = 1 r ( 1 - 1 k i â€Č ) ] ,

where σ ⁹ ( Δ â€Č ) = ( Îł â€Č ; k 1 â€Č , 
 , k r â€Č ) .

2.2 Group action on Riemann surfaces

Let 𝑆 be a compact Riemann surface of genus g ⩟ 2 . A finite group đș acts on 𝑆 if there is a group monomorphism Ï” : G → Aut ⁥ ( S ) . The orbit space S / G given by the action of G ≅ Ï” ⁹ ( G ) on 𝑆 inherits naturally a Riemann surface structure such that the canonical projection S → S / G is holomorphic.

By the classical uniformisation theorem, there is a unique, up to conjugation, Fuchsian group Γ of signature ( g ; - ) such that S ≅ H / Γ . Moreover, đș acts on 𝑆 if and only if there is a Fuchsian group Δ containing Γ together with a group epimorphism

(2.3) ξ : Δ → G   such that   ker ⁡ ( ξ ) = Γ .

It is said that đș acts on 𝑆 with signature σ ⁹ ( Δ ) and that the action is represented by the surface-kernel epimorphism (2.3); henceforth, we write ske for short. Abusing notation, we shall also identify 𝜃 with the tuple of the images of the canonical generators of Δ.

2.3 Extending actions

Assume that G â€Č is a finite group such that G ⩜ G â€Č . The action of đș on 𝑆 represented by the ske (2.3) is said to extend to an action of G â€Č on 𝑆 if

  1. there is a Fuchsian group Δ â€Č containing Δ,

  2. the TeichmĂŒller spaces of Δ and Δ â€Č have the same dimension, and

  3. there exists an ske

    Θ : Δ â€Č → G â€Č   in such a way that   Θ | Δ = Ξ ⁹ and ⁹ ker ⁥ ( Ξ ) = ker ⁥ ( Θ ) .

An action is called maximal if it cannot be extended in the previous sense. Singerman in [66] determined the complete list of pairs of signatures of Fuchsian groups Δ and Δ â€Č for which it may be possible to have an extension as before. See also [62, 65].

2.4 Equivalence of actions

Two actions Ï” i : G → Aut ⁥ ( S ) are topologically equivalent if there exist ω ∈ Aut ⁥ ( G ) and an orientation preserving self-homeomorphism 𝑓 of 𝑆 such that

(2.4) Ï” 2 ⁹ ( g ) = f ⁹ Ï” 1 ⁹ ( ω ⁹ ( g ) ) ⁹ f - 1   for all ⁹ g ∈ G .

Each 𝑓 satisfying (2.4) yields an automorphism f * of Δ, where H / Δ ≅ S / G . If ℬ is the subgroup of Aut ⁥ ( Δ ) consisting of them, then Aut ⁥ ( G ) × B acts on the set of skes defining actions of đș on 𝑆 with signature σ ⁹ ( Δ ) by

( ( ω , f * ) , Ξ ) ↩ ω ∘ Ξ ∘ ( f * ) - 1 .

Two skes Ξ 1 , Ξ 2 : Δ → G define topologically equivalent actions if and only if they belong to the same ( Aut ⁥ ( G ) × B ) -orbit; see, for example, [12]. If the genus of S / G is zero, then ℬ is generated by the so-called braid transformations Ί i , for 1 ⩜ i < l , defined by

x i ↩ x i + 1 , x i + 1 ↩ x i + 1 - 1 ⁱ x i ⁱ x i + 1   and   x j ↩ x j   when ⁱ j ≠ i , i + 1 .

2.5 Equisymmetric stratification of M g

Following [11], the singular locus of M g admits an equisymmetric stratification where each equisymmetric stratum, if nonempty, corresponds to one topological class of maximal actions (see also [29]). More precisely,

Sing ⁱ ( M g ) = ⋃ G , ξ M ¯ g G , ξ ,

where the equisymmetric stratum M g G , Ξ consists of surfaces of genus 𝑔 with full automorphism group isomorphic to đș such that the action is topologically equivalent to 𝜃. In addition, the closure M ÂŻ g G , Ξ of M g G , Ξ is a closed irreducible algebraic subvariety of M g and consists of surfaces of genus 𝑔 with a group of automorphisms isomorphic to đș such that the action is topologically equivalent to 𝜃.

The subset F ÂŻ g ⁹ ( G , σ ) = F ÂŻ g of M g of all those compact Riemann 𝑆 surfaces of genus 𝑔 with a group of automorphisms isomorphic to a given group đș acting with a given signature 𝜎 will be called a closed family. Observe that if the signature of the action of đș on 𝑆 is (2.1), then

dim ⁥ ( F ¯ g ) = 3 ⁹ γ - 3 + s .

Assume that the action of đș is maximal. Then

  1. the interior F g of F ¯ g consists of those surfaces 𝑆 such that G = Aut ⁡ ( S ) ,

  2. F g is formed by finitely many equisymmetric strata that are in correspondence with the pairwise non-equivalent topological actions of đș, and

  3. the set F ¯ g - F g is formed by those surfaces 𝑆 such that G < Aut ⁡ ( S ) properly.

2.6 Abelian varieties

A complex abelian variety is a complex torus which is also a complex projective algebraic variety. Each abelian variety X = V / Λ admits a polarisation, that is, a non-degenerate real alternating form Θ on 𝑉 such that, for all v , w ∈ V ,

Θ ⁱ ( i ⁱ v , i ⁱ w ) = Θ ⁱ ( v , w )   and   Θ ⁱ ( Λ × Λ ) ⊂ Z .

If each elementary divisor of Θ | Λ × Λ is equal to 1, then Θ is called principal and 𝑋 is called a principally polarised abelian variety; from now on, we write ppav for short. In this case, there exists a basis for Λ such that the matrix for Θ Λ × Λ with respect to it is given by

(2.5) J = ( 0 I g - I g 0 ) , where ⁹ g = dim ⁥ ( X ) ;

such a basis is called symplectic. In addition, there exists a basis for 𝑉 with respect to which the period matrix for 𝑋 is

Π = ( I g ⁹ Z ) , where ⁹ Z ∈ H g = { Z ∈ M ⁹ ( g , C ) : Z = Z t , Im ⁥ ( Z ) > 0 } ,

with Z t denoting the transpose matrix of 𝑍. The space H g is called the Siegel upper half-space.

By an isomorphism of ppavs, we mean an isomorphism of the underlying complex tori preserving the involved polarisations. In other words, if ( I g ⁱ Z i ) is the period matrix of X i , then an isomorphism X 1 → X 2 is given by invertible matrices

(2.6) M ∈ GL ⁱ ( g , C ) ⁱ and ⁱ R ∈ GL ⁱ ( 2 ⁱ g , Z )   such that   M ⁱ ( I g ⁱ Z 1 ) = ( I g ⁱ Z 2 ) ⁱ R .

Since 𝑅 preserves the polarisation (2.5), it belongs to the symplectic group

Sp ⁹ ( 2 ⁹ g , Z ) = { R ∈ M ⁹ ( 2 ⁹ g , Z ) : R t ⁹ J ⁹ R = J } .

It follows from (2.6) that the correspondence Sp ⁱ ( 2 ⁱ g , Z ) × H g → H g given by

( R = ( A B C D ) , Z ) ↩ R ⋅ Z := ( A + Z ⁱ C ) - 1 ⁱ ( B + Z ⁱ D )

defines an action that identifies period matrices representing isomorphic ppavs. Hence

H g → A g := H g / Sp ⁱ ( 2 ⁱ g , Z )

is the moduli space of isomorphism classes of ppavs of dimension 𝑔. See [51].

2.7 Abelian varieties with đș-action

Let 𝑆 be a compact Riemann surface of genus g ⩟ 2 . Consider the Jacobian variety J ⁹ S and its full (polarisation-preserving) automorphism group Aut ⁥ ( J ⁹ S ) . Every automorphism of 𝑆 induces a unique automorphism of J ⁹ S . In fact,

[ Aut ( J S ) : Aut ( S ) ] ∈ { 1 , 2 }

according to whether or not 𝑆 is hyperelliptic; moreover, in the latter case,

Aut ⁥ ( J ⁹ S ) / Aut ⁥ ( S ) = { ± 1 } .

As mentioned in the introduction, once a symplectic basis of Λ = H 1 ⁱ ( S , Z ) is fixed, there is an isomorphism

Aut ⁡ ( J ⁱ S ) ≅ Σ S := { R ∈ Sp ⁡ ( 2 ⁱ g , Z ) : R ⋅ Z S = Z S } ,

where ( I g ⁹ Z S ) is the period matrix of J ⁹ S . A change of basis induces a different but equivalent choice of Z S and a conjugate subgroup Σ S . One obtains a well-defined analytic submanifold

S S := { Z ∈ H g : R ⋅ Z = Z ⁱ for all ⁱ R ∈ Σ S }

of H g whose points represent ppavs admitting an action equivalent to that of Aut ⁥ ( J ⁹ S ) in the symplectic group. Equivalently, as - 1 ∈ Σ S , the previous submanifold represents ppavs admitting an action equivalent to that of Aut ⁥ ( S ) . Clearly, Z S ∈ S S .

According to [67] (see also [26, Lemma 3.8]), the dimension N S of (the component which contains J ⁹ S of) S S agrees with

dim ⁥ ( Sym 2 ⁹ H 1 , 0 ⁹ ( S , C ) ) Aut ⁥ ( S ) ,

where Sym 2 ⁹ H 1 , 0 ⁹ ( S , C ) stands for the symmetric square of H 1 , 0 ⁹ ( S , C ) . It follows that

N S = ⟹ χ ρ a sym ∣ 1 ⟩ G , where ⁹ G = Aut ⁥ ( S )

and χ ρ a sym denotes the character of the symmetric square of the analytic representation ρ a of đș and the brackets denote the usual inner product of characters of đș.

It is worth mentioning that S S is related to some special subvarieties of A g . Indeed, as Aut ⁥ ( J ⁹ S ) can be considered as a subgroup of

L S := End Q ⁱ ( J ⁱ S ) = End ⁱ ( J ⁱ S ) ⊗ Z Q ,

one sees that S S contains a complex submanifold of H g of matrices representing ppavs containing L S in their endomorphism algebras. This submanifold is called a Shimura domain for 𝑆, and the corresponding ppavs form a so-called Shimura family for 𝑆; this a special subvariety of A g (see [48, § 3] for a precise definition). We refer to [70, § 3] for more details.

2.8 The group algebra decomposition

The action of a group đș on a compact Riemann surface 𝑆 induces a ℚ-algebra homomorphism from the rational group algebra of đș to L S ,

Ξ : Q ⁱ [ G ] → L S .

Let W 1 , 
 , W r be the rational irreducible representations of đș, and for each W l , let V l be a complex irreducible representation of đș associated to it. Following [45], the equality

(2.7) 1 = e 1 + ⋯ + e r   in ⁱ Q ⁱ [ G ] ,

where e l is a uniquely determined central idempotent associated to W l , yields an isogeny

J ⁹ S ∌ A 1 × ⋯ × A r , where ⁹ A l := Ξ ⁹ ( α l ⁹ e l ) ⁹ ( J ⁹ S )

which is đș-equivariant, with α l ⩟ 1 chosen to satisfy α l ⁹ e l ∈ Z ⁹ [ G ] . Additionally, there are idempotents f l ⁹ 1 , 
 , f l ⁹ n l such that

(2.8) e l = f l ⁱ 1 + ⋯ + f l ⁱ n l ,

where n l = d l / s l is the quotient of the degree d l and the Schur index s l of V l . These idempotents provide n l pairwise isogenous subvarieties of J ⁱ S . If we denote by B l one of them for each 𝑙, then (2.7) and (2.8) provide the isogeny

(2.9) J ⁹ S ∌ B 1 n 1 × ⋯ × B r n r

known as the group algebra decomposition of J ⁹ S with respect to đș. See [16].

Let đ» be a subgroup of đș. We denote by d l H the dimension of the vector subspace of V l of those elements which are fixed under đ». Following [16, Proposition 5.2], the group algebra decomposition (2.9) induces the following isogeny of the Jacobian J ⁹ ( S / H ) of the quotient S / H :

(2.10) J ⁹ ( S / H ) ∌ B 1 n 1 H × ⋯ × B r n r H , where ⁹ n l H = d l H / s l .

The previous isogeny has proved to be fruitful in finding Jacobians J ⁱ S isogenous to a product of Jacobians of quotients of 𝑆. See, for example, [58] and also [59].

Assume that ( Îł ; k 1 , 
 , k s ) is the signature of the action of đș on 𝑆 and that this action is represented by the ske Ξ : Δ → G , with Δ as in (2.2). Observe that if V 1 = W 1 denotes the trivial representation of đș, then B 1 ∌ J ⁹ ( S / G ) , and therefore dim ⁥ B 1 = Îł . If l ⩟ 2 , then according to [62, Theorem 5.12], we have that

(2.11) dim ⁥ B l = m l ⁹ [ d l ⁹ ( Îł - 1 ) + 1 2 ⁹ ∑ j = 1 s ( d l - d l ⟹ Ξ ⁹ ( x j ) ⟩ ) ] ,

where m l is the degree of Q ≀ L l with L l denoting a minimal field of definition for V l .

For decompositions of Jacobians and families of Jacobians with respect to special groups, we refer to the articles [5, 14, 15, 24, 25, 32, 52, 53, 57, 55, 61].

3 Proof of Theorem 1

The proof of Theorem 1 is presented as a consequence of a series of propositions proved in this section. Hereafter, we assume q ⩟ 7 to be prime and 𝑆 to be a Riemann surface of genus g := q - 1 with a group of automorphisms đș of order λ ⁹ q , where λ ⩟ 1 is an integer.

Proposition 3.1

If λ = 3 , then đș is cyclic and acts with signature ( 0 ; 3 , q , 3 ⁹ q ) . Moreover, 𝑆 is unique up to isomorphism, and đș is its full automorphism group.

Proof

Let ( Îł ; k 1 , 
 , k l ) be the signature of the action of đș on 𝑆. The Riemann–Hurwitz formula implies that

(3.1) 2 ⁹ ( q - 2 ) ⩟ 3 ⁹ q ⁹ ( 2 ⁹ Îł - 2 + 2 3 ⁹ l ) .

Observe that if Îł ⩟ 1 , then l = 0 , and therefore q = 2 , contradicting the assumption q ⩟ 7 . We then assume Îł = 0 , and therefore (3.1) shows that l = 3 . It follows that the signature of the action of đș is

( 0 ; k 1 , k 2 , k 3 ) , where ⁹ 1 k 1 + 1 k 2 + 1 k 3 = 1 3 + 4 3 ⁹ q ⁹ and ⁹ k j ∈ { 3 , q , 3 ⁹ q } .

After a routine computation, one sees that the unique solution of the previous equation is, up to permutation, k 1 = 3 , k 2 = q and k 3 = 3 ⁹ q . The last equality implies that

G ≅ C q × C 3 = ⟹ α , ÎČ : α q = ÎČ 3 = 1 , [ α , ÎČ ] = 1 ⟩ .

Consider the Fuchsian group Δ of signature ( 0 ; 3 , q , 3 ⁱ q ) canonically presented

Δ = ⟹ w 1 , w 2 , w 3 : w 1 3 = w 2 q = w 3 3 ⁹ q = w 1 w 2 w 3 = 1 ⟩ ,

and let Ξ : Δ → G be an ske representing an action of đș on 𝑆. It is not difficult to see that, up to an automorphism of đș, the ske 𝜃 is given by

Ξ ⁹ ( w 1 ) = ÎČ , Ξ ⁹ ( w 2 ) = α   and   Ξ ⁹ ( w 3 ) = α - 1 ⁹ ÎČ 2 ;

this proves the uniqueness of 𝑆. By the results of [66], if đș is strictly contained in the full automorphism group Aut ⁥ ( S ) of 𝑆, then Aut ⁥ ( S ) has order 12 ⁹ q , acts on 𝑆 with signature ( 0 ; 2 , 3 , 3 ⁹ q ) and đș is a non-normal subgroup of it. By the classical Sylow theorem, if a group of order 12 ⁹ q with q > 11 has a non-normal subgroup isomorphic to đș, then it is isomorphic to C q × A 4 , where A 4 stands for the alternating group of order 12. However, the product of an element of order two and an element of order three of C q × A 4 cannot have order 3 ⁹ q . The cases q ⩜ 11 are not realised either; see [18]. The proof is complete. ∎

Proposition 3.2

𝜆 is different from 5, 6 and 7.

Proof

If 𝜆 is equal to 5, 6 or 7, then đș is a large group of automorphisms (that is, | G | > 4 ⁹ ( g - 1 ) ), and therefore (see, for example, [43, § 2.3]), the signature of the action is either

  1. ( 0 ; k 1 , k 2 , k 3 ) for some 2 ⩜ k 1 ⩜ k 2 ⩜ k 3 ,

  2. ( 0 ; 2 , 2 , 2 , k ) for some k ⩟ 3 , or

  3. ( 0 ; 2 , 2 , 3 , k ) for some 3 ⩜ k ⩜ 5 .

If 𝜆 is equal to 5 or 7, then đș has no involutions; then the signature of the action đș is

( 0 ; k 1 , k 2 , k 3 )   for some ⁱ k j ∈ { 5 , q , 5 ⁱ q } ⁱ or ⁱ k j ∈ { 7 , q , 7 ⁱ q }

respectively. The Riemann–Hurwitz formula implies that

1 k 1 + 1 k 2 + 1 k 3 = 3 5 + 4 5 ⁱ q   and   1 k 1 + 1 k 2 + 1 k 3 = 5 7 + 4 7 ⁱ q

respectively, and this, in turn, implies that 𝑞 is negative, a contradiction.

We now assume that đș has order 6 ⁹ q . If the signature of the action of đș is ( 0 ; 2 , 2 , 2 , k ) , then by the Riemann–Hurwitz formula, q + 4 divides 6 ⁹ q , and therefore q = 2 , contradicting the assumption q ⩟ 7 . The signatures ( 0 ; 2 , 2 , 3 , 4 ) and ( 0 ; 2 , 2 , 3 , 5 ) cannot be realised either since a group of order 6 ⁹ q does not have elements of order 4 nor 5. Besides, a direct computation shows that the signature ( 0 ; 2 , 2 , 3 , 3 ) contradicts the Riemann–Hurwitz formula.

It follows that the signature of the action is

( 0 ; k 1 , k 2 , k 3 ) , where ⁹ k j ∈ { 2 , 3 , 6 , q , 2 ⁹ q , 3 ⁹ q , 6 ⁹ q }

satisfy, by the Riemann–Hurwitz formula, the equality

1 k 1 + 1 k 2 + 1 k 3 = 2 3 + 2 3 ⁹ q .

Set v = # ⁹ { k j : k j = 3 } . It is clear that v ≠ 2 , 3 . Assume v = 1 and say k 1 = 3 . If k 2 , k 3 ⩟ 6 , then q ⩜ 0 . Then we can assume k 2 = 2 , and therefore

1 k 3 = - 1 6 + 2 3 ⁹ q , showing that   q ⩜ 3 .

Thus v = 0 . Now, let u = # ⁱ { k j : k j = 2 } , and observe that u ≠ 2 , 3 . If u = 1 , then

1 k 2 + 1 k 3 = 2 3 ⁹ q - 1 6 , and therefore   q ⩜ 3 .

All the above ensures that each k j ⩟ 6 . It follows that

1 k 1 + 1 k 2 + 1 k 3 = 2 3 + 2 3 ⁹ q ⩜ 1 2 ,

and therefore q < 0 , contradicting the assumption q ⩟ 7 . ∎

Proposition 3.3

If λ ⩟ 8 , then λ = 8 and S ≅ X 8 .

Proof

If the order of đș is at least 8 ⁹ ( g + 1 ) , then following [2, p. 77], the signature of the action of đș is either

  1. ( 0 ; 2 , 2 , 2 , 3 ) ,

  2. ( 0 ; 2 , 3 , k ) , where k ⩟ 7 ,

  3. ( 0 ; 2 , 4 , k ) , where k ⩟ 5 ,

  4. ( 0 ; 2 , 5 , k ) , where 5 ⩜ k ⩜ 19 ,

  5. ( 0 ; 2 , 6 , k ) , where 6 ⩜ k ⩜ 11 ,

  6. ( 0 ; 2 , 7 , k ) , where 7 ⩜ k ⩜ 9 ,

  7. ( 0 ; 3 , 3 , k ) , where 4 ⩜ k ⩜ 11 , or

  8. ( 0 ; 3 , 4 , k ) , where 4 ⩜ k ⩜ 5 .

We observe that cases (1), (5) and (8) are not realised. Indeed, this fact follows from the contradiction between the fourth and fifth columns in the following table.

Case Signature | G | Condition Riemann–Hurwitz formula
(1) ( 0 ; 2 , 2 , 2 , 3 ) 6 ⁹ λ â€Č ⁹ q λ â€Č ⩟ 2 λ â€Č = 1 - 2 q
(5) ( 0 ; 2 , 6 , k ) 6 ⁹ λ â€Č ⁹ q λ â€Č ⩟ 2 λ â€Č = k k - 3 ⁹ ( 1 - 2 q )
(8.1) ( 0 ; 3 , 4 , 4 ) 12 ⁹ λ â€Č ⁹ q λ â€Č ⩟ 1 λ â€Č = ( 1 - 2 q )
(8.2) ( 0 ; 3 , 4 , 5 ) 60 ⁹ λ â€Č ⁹ q λ â€Č ⩟ 1 λ â€Č = 2 13 ⁹ ( 1 - 2 q )

We also note that cases (4), (6) and (7) are not realised.

Case Signature | G | Condition Riemann–Hurwitz formula
(4) ( 0 ; 2 , 5 , k ) 10 ⁹ λ â€Č ⁹ q λ â€Č ⩟ 1 λ â€Č = 2 ⁹ k 3 ⁹ k - 10 ⁹ ( 1 - 2 q )
(6) ( 0 ; 2 , 7 , k ) 2 ⁹ λ â€Č ⁹ q λ â€Č ⩟ 4 λ â€Č = 14 ⁹ k 5 ⁹ k - 14 ⁹ ( 1 - 2 q )
(7) ( 0 ; 3 , 3 , k ) 3 ⁹ λ â€Č ⁹ q λ â€Č ⩟ 3 λ â€Č = 2 ⁹ k k - 3 ⁹ ( 1 - 2 q )

Note that,

  1. in case (4), we have λ â€Č = 1 , and therefore q = 4 ⁹ k / ( 10 - k ) and 5 ⩜ k ⩜ 9 . However, for each 𝑘 as before, we obtain that 𝑞 is not prime;

  2. in case (6), we have λ â€Č = 4 , and therefore 𝑞 is equal to 14, 28 and 126 for k = 7 , 8 and 9 respectively;

  3. in case (7), the facts that λ ⩟ 3 and q ⩟ 7 imply that k ⩜ 8 . If k = 4 , then 𝑞 is not prime, if k = 5 , then λ â€Č = 3 or 4 and q = 5 or 10, and if k = 6 , 7 , 8 , then λ â€Č = 3 and 𝑞 is not prime.

We claim that case (2) is not realised either. Indeed, note that otherwise the order of đș is equal to 6 ⁹ λ â€Č ⁹ q for some λ â€Č ⩟ 2 and the Riemann–Hurwitz formula reads

k = 6 ⁹ q ⁹ λ â€Č q ⁹ ( λ â€Č - 2 ) + 4 , and therefore   k â€Č := 6 ⁹ λ â€Č q ⁹ ( λ â€Č - 2 ) + 4 ∈ Z + .

  1. If k â€Č = 1 , then λ â€Č = 2 + 8 / ( q - 6 ) , showing that q = 7 and λ â€Č = 10 . Consequently, the order of đș is 420 and acts on 𝑆 of genus six with signature ( 0 ; 2 , 3 , 7 ) . However, such a Riemann surface does not exist because the maximal number of automorphisms that a Riemann surface of genus six can admit is 150 (see, for instance, [18]).

  2. If k â€Č ⩟ 2 , then λ â€Č ⩜ 2 + 2 / ( q - 3 ) , and therefore λ â€Č = 2 . It follows that đș has order 12 ⁹ q and acts on 𝑆 with signature ( 0 ; 2 , 3 , 3 ⁹ q ) . Observe that the signature of the action shows, in particular, that 𝑆 has a cyclic subgroup H < G of automorphisms of order 3 ⁹ q . However, as proved in Proposition 3.1, if 𝑆 has a group of automorphisms of order 3 ⁹ q , then 𝑆 does not have more automorphisms, a contradiction.

This proves the claim.

All the above ensures that the signature of the action is ( 0 ; 2 , 4 , k ) for some k ⩟ 5 . Observe that the order of đș is 4 ⁹ λ â€Č ⁹ q for some λ â€Č ⩟ 2 and the Riemann–Hurwitz formula says

λ â€Č = 2 ⁹ k k - 4 ⁹ ( 1 - 2 q ) < 2 ⁹ k k - 4 .

It follows that one of the following statements holds.

  1. k = 5 and λ â€Č ∈ { 3 , 
 , 9 } , and therefore q = 20 / ( 10 - λ â€Č ) .

  2. k = 6 and λ â€Č ∈ { 3 , 4 , 5 } , and therefore q = 12 / ( 6 - λ â€Č ) .

  3. k ⩟ 5 and λ â€Č = 2 , and therefore q = k / 2 .

The first two cases must be disregarded because 𝑞 is not prime; then đș has order 8 ⁹ q and acts with signature ( 0 ; 2 , 4 , 2 ⁹ q ) . By [43, § 5], we obtain that S ≅ X 8 , as desired. ∎

We recall that, following [22], the closed family C ÂŻ g consists of all those compact Riemann surfaces of genus 𝑔 endowed with a group of automorphisms đș isomorphic to

D q × C 2 ≅ D 2 ⁱ q

acting with signature ( 0 ; 2 , 2 , 2 , q ) . Moreover, if 𝑆 belongs to the interior of C ÂŻ g , then đș agrees with the full automorphism group of 𝑆. It was also observed in [22] that X 8 ∈ C ÂŻ g - C g .

Proposition 3.4

C ÂŻ g - C g = { X 8 } .

Proof

Observe that if 𝑋 belongs to C ÂŻ g - C g , then its automorphism group has order 4 ⁹ t ⁹ q for some t ⩟ 2 . It follows from Proposition 3.3 that t = 2 and that X ≅ X 8 . ∎

For later and repeated use, we recall here that

(3.2) Aut ( X 8 ) ≅ ⟹ x , y , z : x 2 ⁹ q = y 2 = z 2 = 1 , [ x , y ] = [ z , y ] = 1 , z x z = x - 1 y ⟩ ,

and its action on X 8 is represented by the ske

(3.3) Θ : Δ 8 → Aut ⁡ ( X 8 )   given by ⁱ ( z 1 , z 2 , z 3 ) → ( z , z ⁱ x , x - 1 ) ,

where Δ 8 is a Fuchsian group of signature ( 0 ; 2 , 4 , 2 ⁱ q ) presented as

(3.4) Δ 8 = ⟹ z 1 , z 2 , z 3 : z 1 2 = z 2 4 = z 3 2 ⁹ q = z 1 z 2 z 3 = 1 ⟩ .

Proposition 3.5

If 𝑋 is a compact Riemann surface of genus 𝑔 with a group of automorphisms isomorphic to C q × C 2 2 acting with signature ( 0 ; 2 , 2 ⁱ q , 2 ⁱ q ) , then X ≅ X 8 .

Proof

Let Δ 2 be a Fuchsian group of signature ( 0 ; 2 , 2 ⁱ q , 2 ⁱ q ) presented as

(3.5) Δ 2 = ⟹ y 1 , y 2 , y 3 : y 1 2 = y 2 2 ⁹ q = y 3 2 ⁹ q = y 1 y 2 y 3 = 1 ⟩ ,

and consider the group G ≅ C q × C 2 2 presented as

⟹ A , B , C : A q = B 2 = C 2 = ( B ⁹ C ) 2 = [ A , B ] = [ A , C ] = 1 ⟩ .

Let Ξ : Δ 2 → G be an ske representing the action of đș on 𝑋. Up to an automorphism of đș, we can assume Ξ ⁹ ( y 1 ) = B . Moreover, after considering the automorphism of đș given by

A ↩ A , B ↩ B , C ↩ B ⁱ C ,

we can assume that ξ ⁱ ( y 2 ) is equal to either A i ⁱ B or A i ⁱ C for some i ∈ Z q * . Note that the former case is impossible since ξ ⁱ ( y 1 ⁱ y 2 ) would not have order 2 ⁱ q . Thus, after sending 𝐮 to an appropriate power of it, we obtain that 𝜃 is equivalent to

(3.6) Δ 2 → C q × C 2 2   given by ⁱ ( y 1 , y 2 , y 3 ) ↩ ( B , A ⁱ C , ( A ⁱ B ⁱ C ) - 1 ) .

Observe that, with the notation of (3.4), the elements

y ^ 1 := z 2 - 2 , y ^ 2 := z 3 - 1   and   y ^ 3 = z 2 - 1 ⁱ z 3 - 1 ⁱ z 2

generate a subgroup of Δ 8 isomorphic to Δ 2 , and the restriction of (3.3) to it

(3.7) Δ 2 → C q × C 2 2   is given by ⁱ ( y ^ 1 , y ^ 2 , y ^ 3 ) ↩ ( y , x , x - 1 ⁱ y ) .

By letting x = A ⁱ C and y = B , we see that (3.6) and (3.7) agree; consequently, X ≅ X 8 . ∎

Proposition 3.6

If 𝑌 is a compact Riemann surface of genus 𝑔 with a group of automorphisms isomorphic to C q ⋊ 2 C 4 acting with signature ( 0 ; 4 , 4 , q ) , then Y ≅ X 8 .

Proof

Let Δ 3 be a Fuchsian group of signature ( 0 ; 4 , 4 , q ) presented as

(3.8) Δ 3 = ⟹ y 1 , y 2 , y 3 : y 1 4 = y 2 4 = y 3 q = y 1 y 2 y 3 = 1 ⟩ ,

and consider the group G ≅ C q ⋊ 2 C 4 presented as

⟹ A , B : A q = B 4 = 1 , B A B - 1 = A - 1 ⟩ .

Let Ξ : Δ 3 → G be an ske representing the action of đș on 𝑌. Then, after sending đ” to B - 1 if necessary, 𝜃 is given by

( y 1 , y 2 , y 3 ) ↩ ( A i ⁱ B , A j ⁱ B - 1 , A k )   for some ⁱ i , j ∈ Z q ⁱ and ⁱ k ∈ Z q * .

Up to conjugation, we can assume i = 0 , and after sending 𝐮 to an appropriate power of it, we can assume k = 1 . It follows that j = 1 , and therefore 𝜃 is equivalent to

(3.9) Δ 3 → C q ⋊ 2 C 4   given by ⁱ ( y 1 , y 2 , y 3 ) ↩ ( B , A ⁱ B - 1 , A ) .

Now, as in the previous proposition, with the notation of (3.4), we see that

y ~ 1 := z 2 , y ~ 2 := z 3 ⁱ z 2 ⁱ z 3 - 1   and   y ~ 3 = z 3 2

generate a subgroup of Δ 8 isomorphic to Δ 3 , and the restriction of (3.3) to it

(3.10) Δ 3 → C q ⋊ 2 C 4   is given by   ( y ~ 1 , y ~ 2 , y ~ 3 ) ↩ ( z ⁱ x , x - 3 ⁱ z , x - 2 ) .

Write

A = x - 2   and   B = z ⁱ x

to see that (3.9) and (3.10) agree; consequently Y ≅ X 8 . ∎

Proposition 3.7

Assume q ≡ 1 mod 4 . There exists a unique, up to isomorphism, compact Riemann surface X 4 of genus 𝑔 with full automorphism group isomorphic to C q ⋊ 4 C 4 acting on it with signature ( 0 ; 4 , 4 , q ) .

Proof

Consider the Fuchsian group Δ 3 as in (3.8), and the group

(3.11) G ≅ C q ⋊ 4 C 4 = ⟹ A , B : A q = B 4 = 1 , B A B - 1 = A ρ ⟩ ,

where 𝜌 is a primitive fourth root of unity in Z q . If Ξ : Δ 3 → C q ⋊ 4 C 4 is an ske representing the action of đș on a compact Riemann surface 𝑍 of genus 𝑔, then by proceeding similarly to the previous proposition, one sees that 𝜃 is equivalent to

ξ 1 ⁱ ( y 1 , y 2 , y 3 ) = ( A - 1 ⁱ B , B - 1 , A )   or   ξ 2 ⁱ ( y 1 , y 2 , y 3 ) = ( A - 1 ⁱ B - 1 , B , A ) .

It follows that, up to isomorphism, there are at most two surfaces 𝑍 as before, namely

Z j := H / K j , where ⁱ K j = ker ⁡ ( ξ j )   for ⁱ j = 1 , 2 .

Observe that if the full automorphism group of Z j is different from đș, then by Proposition 3.3, necessarily Z j ≅ X 8 and, in particular, Aut ⁥ ( X 8 ) contains a subgroup isomorphic to C q ⋊ 4 C 4 . However, this is not possible. Indeed, with the notation of (3.2), if Îč ∈ Aut ⁥ ( X 8 ) has order 4, then Îč 2 is equal to the central element 𝑩. It follows that Aut ⁥ ( Z j ) ≅ G for j = 1 , 2 .

We record here that Z 1 and Z 2 are isomorphic if and only if K 1 and K 2 are conjugate in Aut ⁡ ( H ) . As the normaliser of each K j is Δ 3 , it can be seen that K 1 and K 2 are conjugate if and only if they are conjugate in the normaliser N ⁱ ( Δ 3 ) of Δ 3 . Now, the action by conjugation of N ⁱ ( Δ 3 ) on { K 1 , K 2 } has orbits of length [ N ⁱ ( Δ 3 ) : Δ 3 ] which is, by [66, Theorem 1], equal to 2. Hence K 1 and K 2 are conjugate, and thus X 4 := Z 1 ≅ Z 2 , as desired. ∎

Proposition 3.8

If λ = 4 and S ∉ C ÂŻ g , then q ≡ 1 mod 4 and S ≅ X 4 .

Proof

As in the proof of Proposition 3.2, the signature of the action of đș on 𝑆 is either

  1. ( 0 ; k 1 , k 2 , k 3 ) for some 2 ⩜ k 1 ⩜ k 2 ⩜ k 3 ,

  2. ( 0 ; 2 , 2 , 2 , k ) for some k ⩟ 3 , or

  3. ( 0 ; 2 , 2 , 3 , k ) for some 3 ⩜ k ⩜ 5 .

The third case must be disregarded since there is no group of order 4 ⁱ q with an element of order three. If the signature is as in the second case, then the Riemann–Hurwitz formula implies that k = q . We now assume the signature to be as in the first case. The Riemann–Hurwitz formula says

(3.12) 1 k 1 + 1 k 2 + 1 k 3 = 1 2 + 1 q .

Note that, among k 1 , k 2 , k 3 , not two or three of them can be equal to 2. It follows that, up to permutation, there are two cases to consider.

  1. Assume k 1 = 2 and k 2 , k 3 ⩟ 4 . Then (3.12) turns into 1 k 2 + 1 k 3 = 1 q . Note that if k 2 = 4 , then k 3 ⩜ 0 . It follows that

    k 2 , k 3 ⩟ q , and therefore   k 2 = k 3 = 2 ⁹ q .

  2. Assume k 1 , k 2 , k 3 ⩟ 4 . If the number of periods k j that are equal to 4 is 2, then (3.12) implies that the signature is ( 0 ; 4 , 4 , q ) ; otherwise q ⩜ 4 .

Thereby, the signature of the action of đș on 𝑆 is either

( 0 ; 2 , 2 , 2 , q ) , ( 0 ; 2 , 2 ⁱ q , 2 ⁱ q )   or   ( 0 ; 4 , 4 , q ) .

We recall that if q ≡ 3 mod 4 , then đș is isomorphic to either C 4 ⁹ q , C q × C 2 2 , D 2 ⁹ q or C q ⋊ 2 C 4 , and if q ≡ 1 mod 4 , then in addition, đș can be isomorphic to C q ⋊ 4 C 4 .

  1. If đș acts with signature ( 0 ; 2 , 2 , 2 , q ) , then đș is generated by three involutions, and therefore G ≅ D 2 ⁹ q , showing that S ∈ C ÂŻ g .

  2. If đș acts with signature ( 0 ; 2 , 2 ⁹ q , 2 ⁹ q ) , then đș is generated by two elements of order 2 ⁹ q whose product is an involution; thus G ≅ C q × C 2 2 . By Proposition 3.5, we see that S ≅ X 8 , and therefore S ∈ C ÂŻ g .

  3. If đș acts with signature ( 0 ; 4 , 4 , q ) , then đș is generated by two elements of order 4, and therefore đș is isomorphic to C q ⋊ 2 C 4 or C q ⋊ 4 C 4 . By Proposition 3.6, the former case implies S ≅ X 8 , whilst by Proposition 3.7, the latter case implies S ≅ X 4 .

This finishes the proof.∎

Proposition 3.9

If λ = 2 and đș is cyclic acting with signature ( 0 ; 2 , 2 , q , q ) , then S ∈ C ÂŻ g .

Proof

Let Δ 4 be a Fuchsian group of signature ( 0 ; 2 , 2 , q , q ) presented as

(3.13) Δ 4 = ⟹ x 1 , x 2 , x 3 , x 4 : x 1 2 = x 2 2 = x 3 q = x 4 q = x 1 x 2 x 3 x 4 = 1 ⟩ ,

and consider the cyclic group of order 2 ⁹ q generated by 𝑎 of order 𝑞 and 𝑏 of order two. As đș has only one involution, it is clear that, after sending 𝑎 to a suitable power of it, each ske representing an action of đș on 𝑆 is equivalent to

(3.14) ξ 0 : Δ 4 → C q × C 2   such that   ξ 0 ⁱ ( x 1 , x 2 , x 3 , x 4 ) = ( b , b , a , a - 1 ) .

Then such surfaces 𝑆 form an equisymmetric complex one-dimensional family. Let

(3.15) Δ 1 = ⟹ y 1 , y 2 , y 3 , y 4 : y 1 2 = y 2 2 = y 3 2 = y 4 q = y 1 y 2 y 3 y 4 = 1 ⟩

be a Fuchsian group of signature ( 0 ; 2 , 2 , 2 , q ) , and consider the group

(3.16) D 2 ⁹ q = ⟹ R , T : R 2 ⁹ q = T 2 = ( T R ) 2 = 1 ⟩ .

We recall that, following [22], the action of D 2 ⁹ q on S â€Č ∈ C ÂŻ g is represented by the ske

(3.17) ξ : Δ 1 → D 2 ⁱ q   given by   ( y 1 , y 2 , y 3 , y 4 ) ↩ ( R q , T , T ⁱ R , R q - 1 )

Now, the elements of Δ 1 ,

x ^ 1 := y 1 , x ^ 2 := y 2 ⁱ y 1 ⁱ y 2 , x ^ 3 := y 4   and   x ^ 4 := y 1 ⁱ y 2 ⁱ y 4 ⁱ y 2 ⁱ y 1 ,

generate a Fuchsian group isomorphic to Δ 4 . The restriction of (3.17) to it,

(3.18) Δ 4 → ⟹ R ⟩ ≅ G , is given by   ( x ^ 1 , x ^ 2 , x ^ 3 , x ^ 4 ) ↩ ( R q , R q , R q - 1 , R 1 - q ) .

Set a := R q - 1 and b := R q to see that (3.18) agrees with (3.14), and the result follows. ∎

Proposition 3.10

Let λ = 2 , and let đș be a cyclic group acting with signature ( 0 ; q , 2 ⁹ q , 2 ⁹ q ) . Then either S ≅ X 8 or the full automorphism group of 𝑆 agrees with đș. In the latter case, there are exactly q - 3 2 pairwise non-isomorphic compact Riemann surfaces.

Proof

Let Δ 5 be a Fuchsian group of signature ( 0 ; q , 2 ⁱ q , 2 ⁱ q ) presented as

Δ 5 = ⟹ x 1 , x 2 , x 3 : x 1 q = x 2 2 ⁹ q = x 3 2 ⁹ q = x 1 x 2 x 3 = 1 ⟩ ,

and consider the cyclic group of order 2 ⁱ q generated by 𝑎 of order 𝑞 and 𝑏 of order two.

If Ξ : Δ 5 → G is an ske representing the action of đș on 𝑆, then after sending 𝑎 to an appropriate power of it, we see that 𝜃 is equivalent to

ξ j = ( a , a j ⁱ b , a - j - 1 ⁱ b )   for some ⁱ j ≠ - 1 , 0 .

Let S j be the compact Riemann surface defined by ξ j , and write j * = q - 1 2 . We claim that S j * ≅ X 8 . To prove that, we note that, by Proposition 3.5, it suffices to verify that ξ j * extends to the action of C q × C 2 2 with signature ( 0 ; 2 , 2 ⁱ q , 2 ⁱ q ) . With the notation of the proof of Proposition 3.5, the elements

x ^ 1 := y 3 2 , x ^ 2 := y 1 ⁱ y 2 ⁱ y 1   and   x ^ 3 := y 2

generate a subgroup of (3.5) isomorphic to Δ 5 , and the restriction of (3.6) to it,

(3.19) Δ 5 → ⟹ A , C ⟩ ≅ G , is given by   ( x ^ 1 , x ^ 2 , x ^ 3 ) ↩ ( A - 2 , A ⁹ C , A ⁹ C ) .

By setting a := A - 2 , b := C , we see that (3.19) is equivalent to Ξ j * , as desired.

Let j ≠ j * . If Aut ⁥ ( S j ) ≠ G , then by [66] and Proposition 3.3, the action Ξ j must extend to the action (3.6) of C q × C 2 2 with signature ( 0 ; 2 , 2 ⁹ q , 2 ⁹ q ) . Observe that an element 𝑩 of (3.5) has order 2 ⁹ q if and only if it is conjugate to y 2 k or to y 3 k for some k ∈ { 1 , 
 , 2 ⁹ q - 1 } odd and different from 𝑞. As the target group is abelian, the image of 𝑩 under (3.6) is either A ⁹ C or ( A ⁹ B ⁹ C ) - 1 . Now, if Δ â€Č is a subgroup of (3.5) isomorphic to Δ 5 , then the restriction of (3.6) to the canonical generators of Δ â€Č must be

( A - 2 , A ⁱ C , A ⁱ C ) , ( B , A ⁱ C , ( A ⁱ B ⁱ C ) - 1 )   or   ( A 2 , ( A ⁱ B ⁱ C ) - 1 , ( A ⁱ B ⁱ C ) - 1 ) .

The second case is impossible since it does not have the required signature; the other two cases are equivalent to ξ j * . We conclude that Aut ⁡ ( S j ) = G if j ≠ j * , and in particular, S j is not isomorphic to S j * .

Write K j = ker ⁡ ( ξ j ) for each j ∈ Z q - { - 1 , 0 , j * } . As argued in the proof of Proposition 3.7, we have that S j 1 and S j 2 are isomorphic if and only if K j 1 and K j 2 are conjugate in the normaliser N ⁱ ( Δ 5 ) of Δ 5 . The action by conjugation of N ⁱ ( Δ 5 ) on

{ K j : j ∈ Z q - { - 1 , 0 , j * } }

has orbits of length [ N ⁱ ( Δ 5 ) : Δ 5 ] which is, by [66, Theorem 1], equal to 2. Hence

{ S j : j ∈ Z q - { - 1 , 0 , j * } }

splits into q - 3 2 isomorphism classes. Finally, observe that the elements x 2 and x 3 of Δ 5 are conjugate in N ⁹ ( Δ 5 ) ; thus S j and S - j - 1 are isomorphic, and therefore the isomorphism classes are represented by S j , where 1 ⩜ j ⩜ q - 3 2 .∎

Proposition 3.11

There exists a closed family K ¯ g of compact Riemann surfaces with a group of automorphisms isomorphic to the dihedral group of order 2 ⁹ q acting with signature ( 0 ; 2 , 2 , q , q ) . The number of equisymmetric strata of K ¯ g is at most

{ q + 3 4 if ⁱ q ≡ 1 mod 4 , q + 1 4 if ⁱ q ≡ 3 mod 4 ,

and, independently of 𝑞, one of them is equal to C g .

Proof

Let Δ 4 be a Fuchsian group of signature ( 0 ; 2 , 2 , q , q ) presented as in (3.13), and consider the dihedral group of order 2 ⁱ q ,

G ≅ D q = ⟹ r , s : r q = s 2 = ( s r ) 2 = 1 ⟩ .

The existence of the family K ÂŻ g follows after considering the ske

Ω : Δ 4 → G   given by   ( x 1 , x 2 , x 3 , x 4 ) ↩ ( s , s , r - 1 , r ) .

Let us now assume that Ξ : Δ 4 → G is an ske representing the action of đș on S ∈ K ÂŻ g . If Ξ ⁹ ( x 1 ) = Ξ ⁹ ( x 2 ) , then after a conjugation and after sending 𝑟 to an appropriate power of it, we see that 𝜃 is equivalent to Ί. On the other hand, if Ξ ⁹ ( x 1 ) ≠ Ξ ⁹ ( x 2 ) , then after considering a suitable automorphism of đș, we see that 𝜃 is equivalent to the ske

ξ i := ( s , s ⁱ r , r i , r - i - 1 )   for some ⁱ i ∈ Z q - { - 1 , 0 } .

The braid transformation Ω 3 (see § 2.4) shows that ξ i ≅ ξ - i - 1 . The rule

i ↩ - i - 1

has order two, restricts to a bijection of Z q - { - 1 , 0 } and has exactly one fixed point, namely i * = q - 1 2 . Observe that if φ u is the automorphism of đș given by r ↩ r u , then

Ί = Ί 2 2 ∘ φ ( i * ) - 1 ⁹ ( Ξ i * ) .

All the above ensures that 𝜃 is equivalent to either

Ω ⁱ or ⁱ ξ i   for some ⁱ i ∈ { 1 , 
 , q - 3 2 } .

Now, for each i ∈ { 1 , 
 , q - 3 2 } , the transformation φ i - 1 ∘ Ί 2 2 provides an equivalence

ξ i ≅ ξ - i ⁱ ( 2 ⁱ i + 1 ) - 1 .

The rule i ↩ - i ⁹ ( 2 ⁹ i + 1 ) - 1 has order two and (up to the identification i ∌ - i - 1 ) restricts to a bijection of { 1 , 
 , q - 3 2 } ; it has a fixed point if and only if

(3.20) - i - 1 = - i ⁱ ( 2 ⁱ i + 1 ) - 1 ⇔ 2 ⁱ i 2 + 2 ⁱ i + 1 = 0 ,

and the quadratic equation above has solution in Z q if and only if q ≡ 1 mod 4 . It follows that the number of pairwise non-equivalent skes 𝜃 is at most

1 + 1 2 ⁱ ( q - 3 2 ) = q + 1 4   and   2 + 1 2 ⁱ ( q - 3 2 - 1 ) = q + 3 4

if q ≡ 3 mod 4 and q ≡ 1 mod 4 respectively. Finally, with the notation of (3.15), define

x ^ 1 := y 1 ⁱ y 2 ⁱ y 1 , x ^ 2 := y 2 , x ^ 3 := y 2 ⁱ y 1 ⁱ y 4 ⁱ y 1 ⁱ y 2   and   x ^ 4 := y 4 ,

and note that they generate a Fuchsian group isomorphic to Δ 4 . The restriction of (3.17) to it is given by

(3.21) ( x ^ 1 , x ^ 2 , x ^ 3 , x ^ 4 ) ↩ ( T , T , R 1 - q , R q - 1 ) .

If we write s := T , r := R q - 1 , we see that (3.21) agrees with Ω. Hence the action of Ω extends to (3.17), and therefore the stratum defined by Ω agrees with C g . ∎

Proposition 3.12

If K g stands for the interior of the closed family K ¯ g , then the full automorphism group of S ∈ K g is isomorphic to either D q or D 2 ⁹ q . In addition,

K ¯ g - K g = { { X 8 , X 4 } if ⁱ q ≡ 1 mod 4 , { X 8 } if ⁱ q ≡ 3 mod 4 .

Proof

We keep the notation of the proof of Proposition 3.11. The first statement is clear since the full automorphism group of 𝑆 is isomorphic to D 2 ⁱ q or D q according to whether or not ξ i is equivalent to Ω.

Let K g i denote the equisymmetric stratum defined by Ξ i .

  1. If Ξ i is equivalent to Ί, then Proposition 3.11 and Proposition 3.4 imply that K ¯ g i - K g i = { X 8 } .

  2. If Ξ i is non-equivalent to Ί, then by Proposition 3.8, we see that,

    1. if q ≡ 3 mod 4 , then K ¯ g i - K g i is empty, and

    2. if q ≡ 1 mod 4 , then K ¯ g i - K g i is either empty or { X 4 } .

Assume q ≡ 1 mod 4 . We recall that the full automorphism group of X 4 is isomorphic to (3.11), and the corresponding action is given by the ske

(3.22) Δ 3 → Aut ⁡ ( X 4 )   such that   ( y 1 , y 2 , y 3 ) ↩ ( A - 1 ⁱ B , B - 1 , A ) ,

where Δ 3 is as in (3.8). The elements

x ^ 1 := y 1 ⁱ y 2 2 ⁱ y 1 - 1 , x ^ 2 := y 1 2 , x ^ 3 := y 1 - 1 ⁱ y 3 ⁱ y 1   and   x ^ 4 := y 3

generate a Fuchsian group isomorphic to Δ 4 . The restriction of (3.22) to it is equivalent to

(3.23) ( x ^ 1 , x ^ 2 , x ^ 3 , x ^ 4 ) → ( B 2 , B 2 ⁹ A , A e , A - e - 1 ) , where ⁹ e := - ρ ⁹ ( ρ - 1 ) - 1 .

By letting r := A and s := B 2 , we see that (3.23) agrees with ξ e . As 𝑒 solves equation (3.20), we conclude that K ¯ g i - K g i is equal to { X 4 } if 𝑖 solves (3.20) and is empty otherwise. ∎

Proposition 3.13

If λ = 2 , then 𝑆 belongs to K ÂŻ g or 𝑆 is isomorphic to one of the q - 3 2 pairwise non-isomorphic surfaces of Proposition 3.10.

Proof

Assume that the signature of the action of đș on 𝑆 is ( Îł ; k 1 , 
 , k l ) . Observe that Îł = 0 . Indeed, otherwise the Riemann–Hurwitz formula implies that

l ⩜ 2 ⁹ ( 1 - 2 q ) , showing that   l = 0 ⁹ or ⁹ l = 1 .

In both cases, we see that Îł = 1 , and therefore q = 2 and 4 respectively, a contradiction.

As each k j ⩟ 2 , we see that

l ⩜ 6 - 4 q , and therefore   l ∈ { 3 , 4 , 5 } .

Let v = # ⁹ { k j : k j = 2 } .

  1. If l = 5 , then

    ∑ j = 1 5 1 k j = 2 + 2 q ⩜ v 2 + 5 - v q ,

    showing that v = 5 . But, in this case, q = 4 , contradicting the fact that 𝑞 is prime.

  2. If l = 4 , then

    ∑ j = 1 4 1 k j = 1 + 2 q ⩜ v 2 + 4 - v q ,

    showing that v ∈ { 2 , 3 , 4 } . Note that if v = 4 , then q = 2 , and if v = 3 , then k 4 < 0 . It follows that v = 2 , and therefore k 3 = k 4 = q .

  3. If l = 3 , then clearly v ≠ 3 . If v = 2 , then k 3 < 0 , and if v = 1 , then q < 4 . Thus

    (3.24) 1 k 1 + 1 k 2 + 1 k 3 = 2 q , where ⁹ k j ∈ { q , 2 ⁹ q } .

    It is easy to verify that the unique solution of (3.24) is k 1 = q , k 2 = k 3 = 2 ⁹ q .

Thereby, the signature of the action of đș on 𝑆 is either

( 0 ; 2 , 2 , q , q )   or   ( 0 ; q , 2 ⁱ q , 2 ⁱ q ) .

We record here the simple fact that a group of order 2 ⁹ q is either cyclic or dihedral.

  1. If the signature is ( 0 ; q , 2 ⁹ q , 2 ⁹ q ) , then đș is cyclic, and therefore, by Proposition 3.10, we obtain that S ≅ X 8 ∈ K ÂŻ g or 𝑆 is isomorphic to one of the q - 3 2 pairwise non-isomorphic surfaces with full automorphism group isomorphic to C q × C 2 .

  2. If the signature is ( 0 ; 2 , 2 , q , q ) , then the following holds.

    1. If đș is cyclic, then by Proposition 3.9, we have that S ∈ C ÂŻ g ⊂ K ÂŻ g .

    2. If đș is dihedral, then by Proposition 3.11, we have that S ∈ K ÂŻ g .

This proves the proposition. ∎

Remark 2

Let S ∈ M q - 1 q , and denote by 𝑡 the automorphism of 𝑆 of order 𝑞. According to [21], the action of G = ⟹ t ⟩ on 𝑆 is equivalent to the action represented by one of the following skes:

Ξ 1 = ( t , t , t , t - 3 ) , Ξ 2 = ( t , t , t - 1 , t - 1 ) , Ξ 3 , i = ( t , t - 1 , t i , t - i ) , Ξ 4 , i = ( t , t , t i , t q - 2 - i ) , Ξ 5 , i ⁹ j = ( t , t i , t j , t q - 1 - i - j ) ,

where 2 ⩜ i ⩜ q - 1 2 , j ∉ { 1 , q - 1 , i , q - i } and q - 1 - i - j ∉ { 1 , i , j } .

In terms of our terminology, the results of [21] allow us to claim that the stratum determined by Ξ 1 contains X 3 , the stratum determined by Ξ 2 agrees with the family C g , the strata determined by Ξ 3 ⁹ i agree with the equisymmetric strata of the family K ¯ g that are different from C g , and the strata determined by Ξ 4 ⁹ i contain the surfaces X 2 , k . We also mention that the isolated strata of dimension one of Sing ⁹ ( M q - 1 ) are those determined by the skes Ξ 5 ⁹ i ⁹ j .

4 Proof of Theorem 2 and Proposition 2

Let q ⩟ 5 be a prime number, and set g = q - 1 . We write ω l = exp ⁥ ( 2 ⁹ π ⁹ i l ) .

The family C ÂŻ g

We recall that if S ∈ C ¯ g , then the action of

G = D 2 ⁹ q = ⟹ R , T : R 2 ⁹ q = T 2 = ( T R ) 2 = 1 ⟩

on 𝑆 has signature ( 0 ; 2 , 2 , 2 , q ) and is represented by the ske

Ξ = ( R q , T , T ⁹ R , R q - 1 ) .

Let H = ⟹ R q ⟩ ≅ C 2 . We consider the associated two-fold regular covering map

π : S → ÎŁ := S / H

and observe that Σ has genus zero and 𝜋 ramifies over 2 ⁱ q values. If we denote them by

(4.1) u 1 , 
 , u q   and   v 1 ⁱ 
 , v q ,

then it is classically known that 𝑆 is isomorphic to the normalisation of

y 2 = ∏ i = 1 q ( x - u i ) ⁱ ( x - v i )

Note that ÎŁ ≅ P 1 admits an action of K = G / H ≅ D q in such a way that ÎŁ / K ≅ S / G . It follows that the values in (4.1) form two orbits of length 𝑞 under the action of the cyclic subgroup of order 𝑞 of đŸ. Without loss of generality, we can assume that

u i = ω q i   and   v i = λ ⁹ ω q i , where ⁹ 1 ⩜ i ⩜ q

and 𝜆 is a nonzero complex number such that t := λ q ≠ 1 . Hence 𝑆 is isomorphic to the normalisation of the singular affine algebraic curve

X t := { ( x , y ) ∈ C 2 : y 2 = ( x q - 1 ) ⁹ ( x q - t ) }

for some t ≠ 0 , 1 . It is a direct computation to verify that the transformations

r ⁹ ( x , y ) = ( ω q ⁹ x , - y )   and   t ⁹ ( x , y ) = ( t q ⁹ 1 x , t ⁹ y x q )

restrict to automorphisms of X t and that ⟹ r , t ⟩ ≅ D 2 ⁹ q .

Note that r q ⁱ ( x , y ) = ( x , - y ) is the hyperelliptic involution and that X 8 ≅ X - 1 .

The surface X 4

Following the proof of Proposition 3.7, the action of

G = C q ⋊ 4 C 4 = ⟹ A , B : A q = B 4 = 1 , B A B - 1 = A ρ ⟩

on X 4 has signature ( 0 ; 4 , 4 , q ) and is represented by the ske Ξ = ( A - 1 ⁹ B , B - 1 , A ) . Let Q = ⟹ a ⟩ , and observe that the associated 𝑞-fold regular covering map

π : X 4 → ÎŁ := X 4 / Q ≅ P 1

ramifies over four values marked with 𝑞. As Σ admits the action of

K = G / Q ≅ C 4   with signature ⁱ ( 0 ; 4 , 4 )

and ÎŁ / K ≅ S / G , the branch values of 𝜋 form one orbit under the action of đŸ. Without loss of generality, we can assume these branch values to be 1, 𝑖, −1 and - i (where i 2 = - 1 ) and that their corresponding rotation numbers (modulo 𝑞) are 1, 𝜌, ρ 2 and ρ 3 respectively. Then, following [28] (see also [29, 72]), X 4 is isomorphic to the normalisation of

(4.2) y q = ( x - 1 ) ⁹ ( x - i ) ρ ⁹ ( x + 1 ) q - 1 ⁹ ( x + i ) q - ρ .

Since 𝜌 has order 4 in Z q , there exists e ∈ Z such that ρ 2 + 1 = e ⁹ q . Set

a ⁹ ( x , y ) = ( x , ω q ⁹ y )   and   b ⁹ ( x , y ) = ( i ⁹ x , - ( x + i ) e - ρ ( x - i ) e - 1 ⁹ ( x + 1 ) ρ - 1 ⁹ y ρ ) ,

and note that they restrict to automorphisms of (4.2). If 𝜑 is as in the statement of the theorem, then routine computations show that

φ ⁹ ( - i ⁹ x ) ⁹ φ ⁹ ( - x ) ρ ⁹ φ ⁹ ( i ⁹ x ) ρ 2 ⁹ φ ⁹ ( x ) ρ 3 = y 1 - ρ 4 .

This implies that 𝔟 has order four. Now, it is direct to see that ⟹ a , b ⟩ ≅ C q ⋊ 4 C 4 .

Remark 3

The Möbius transformation Îč : P 1 ≅ ÎŁ → P 1 given by

Îč ⁹ ( z ) = i ⁹ z - 1 z + 1   satisfies   Îč ⁹ ( 1 , i , - 1 , - i ) = ( 0 , - 1 , ∞ , 1 )

and lifts to obtain an isomorphism between X 4 and the Riemann surface given by

y q = x ⁹ ( x + 1 ) ρ ⁹ ( x - 1 ) q - ρ

This provides the explicit model for X 4 defined over its field of moduli ℚ.

The surface X 3

Similarly to before, the normalisation of y 3 = x q - 1 defines a Riemann surface of genus 𝑔 and ( x , y ) ↩ ( ω q ⁹ x , ω 3 ⁹ y ) restricts to an automorphism of it of order 3 ⁹ q .

The surfaces X 2 , k

Following the proof of Proposition 3.10, the action of

G = C q × C 2 = ⟹ a , b : a q = b 2 = [ a , b ] = 1 ⟩

on X 2 , k has signature ( 0 ; q , 2 ⁹ q , 2 ⁹ q ) and is determined by the ske

Ξ k = ( a , a k ⁹ b , a - k - 1 ⁹ b ) , where ⁹ 1 ⩜ k ⩜ q - 3 2 .

If Q = ⟹ a ⟩ , then the associated regular covering map

π k : X 2 , k → ÎŁ k := X 2 , k / Q ≅ P 1

ramifies over four values marked with 𝑞. As Σ k admits the action of

K = G / Q ≅ C 2   with signature ⁱ ( 0 ; 2 , 2 )

and ÎŁ / K ≅ S / G , two branch values of π k form an orbit, and the remaining ones are fixed under the action of đŸ. Thus we can assume that the branch values of π k are 1, −1, ∞ and 0, where the first two form an orbit. It follows that X 2 , k is isomorphic to

(4.3) y q = n n k ⁹ ( x - 1 ) ⁹ ( x + 1 )

for some 1 ⩜ n k ⩜ q - 1 such that n k ≠ q - 2 . It is easy to see that

a ⁹ ( x , y ) = ( x , ω q ⁹ y )   and   b k ⁹ ( x , y ) = ( - x , ( - 1 ) n k ⁹ y )

are automorphisms of (4.3) and that ⟹ a , b k ⟩ ≅ C q × C 2 .

The family K ÂŻ g

Following the proof of Proposition 3.11, the action of

(4.4) G = D q = ⟹ r , s : r q = s 2 = ( s r ) 2 = 1 ⟩

on S ∈ K ¯ g has signature ( 0 ; 2 , 2 , q , q ) and is determined by the ske

Ξ i = ( s , s ⁹ r , r i , r - i - 1 )   for some ⁹ 1 ⩜ i ⩜ q - 1 2 .

If Q = ⟹ r ⟩ , then the associated regular covering map

π : S → ÎŁ := S / Q ≅ P 1

ramifies over four values marked with 𝑞. As Σ admits the action of

K = G / Q ≅ C 2   with signature ⁱ ( 0 ; 2 , 2 )

and ÎŁ / K ≅ S / G , the branch values of 𝜋 form two orbits under the action of đŸ. We can assume these values to be 1 , - 1 and t , - t for some t ≠ 0 , ± 1 . In addition, the rotation numbers are 1, q - 1 , 1 and q - 1 respectively. Thus 𝑆 is isomorphic to

Z t := { ( x , y ) ∈ C 2 : y q = ( x - 1 ) ⁹ ( x + 1 ) q - 1 ⁹ ( x - t ) ⁹ ( x + t ) q - 1 }

It is straightforward to verify that the transformations

r ⁹ ( x , y ) = ( x , ω q ⁹ y )   and   s ⁹ ( x , y ) ↩ ( - x , ( x 2 - 1 ) ⁹ ( x 2 - t 2 ) ⁹ y - 1 )

restrict to automorphisms of Z t and ⟹ r , s ⟩ ≅ D q .

Proof of Proposition 2

Assume that S ∈ K q , and let đș be as in (4.4). The covering

Σ → Σ / K ≅ S / G   can be chosen as   z ↩ z 2 ,

and therefore, if S ≅ Z t , then the branch values of S → S / G are ∞ and 0 marked with 2, and 1 and t 2 marked with 𝑞.

Let σ ∈ Gal ⁹ ( C / Q ) , and assume S t and ( S t ) σ = S σ ⁹ ( t ) to be isomorphic. The facts that

  1. G ≅ Aut ⁡ ( S ) provided that S ∈ K g - C g , and

  2. đș is the unique group of automorphisms of 𝑆 isomorphic to D q provided that S ∈ C g

imply that there is a Möbius transformation φ : P 1 → P 1 such that

φ ⁹ ( { ∞ , 0 } ) = { ∞ , 0 }   and   φ ⁹ ( { 1 , t 2 } ) = { 1 , σ ⁹ ( t ) 2 } .

We have two possible cases.

  1. If φ ⁹ ( 0 ) = 0 and φ ⁹ ( ∞ ) = ∞ , then either φ ⁹ ( z ) = z or φ ⁹ ( z ) = σ ⁹ ( t ) 2 ⁹ z .

  2. If φ ⁹ ( 0 ) = ∞ and φ ⁹ ( ∞ ) = 0 , then either φ ⁹ ( z ) = 1 z or φ ⁹ ( z ) = σ ⁹ ( t ) 2 z .

Observe that the latter case in (1) and the former case in (2) imply t ⁹ σ ⁹ ( t ) = ± 1 , a contradiction. It follows that σ ⁹ ( t ) 2 = t 2 showing that σ ⁹ ( t ) = t . Conversely, if σ ⁹ ( t ) = t , then it is clear that S = ( S t ) σ . Hence the field of moduli of S ≅ Z t is

fix ⁥ { σ ∈ Gal ⁹ ( C / Q ) : σ ⁹ ( t ) = t } = Q ⁹ ( t ) ,

as desired.

5 Some algebraic lemmata

In this section, we collect some facts related to the representations of the groups appearing in Theorem 1; these results will be needed to prove Theorems 3 and 4. Set ω l = exp ⁥ ( 2 ⁹ π ⁹ i l ) .

Rational and complex irreducible representations

Lemma 1

Let q ⩟ 5 be a prime number. The group

G 8 = ⟹ x , y , z : x 2 ⁹ q = y 2 = z 2 = 1 , [ x , y ] = [ z , y ] = 1 , z x z = x - 1 y ⟩

has four complex irreducible representations of degree one given by

χ 0 1 + : { x ↩ 1 , y ↩ 1 , z ↩ 1 ,   χ 0 2 + : { x ↩ 1 , y ↩ 1 , z ↩ - 1 ,   χ q 1 + : { x ↩ - 1 , y ↩ 1 , z ↩ 1 ,   χ q 2 + : { x ↩ - 1 , y ↩ 1 , z ↩ - 1 ,

and 2 ⁹ q - 1 complex irreducible representations of degree two given by the following table.

Representation đ‘„ 𝑩 𝑧
χ j + , 1 ⩜ j ⩜ q - 1 diag ⁥ ( ω 2 ⁹ q j , - ω 2 ⁹ q q - j ) I 2 J 2
χ j 1 - , 0 ⩜ j ⩜ q - 1 2 diag ⁥ ( ω 2 ⁹ q j , ω 2 ⁹ q q - j ) - I 2 J 2
χ j 2 - , 1 ⩜ j ⩜ q - 1 2 diag ⁥ ( - ω 2 ⁹ q j , - ω 2 ⁹ q q - j ) - I 2 J 2

Here I 2 stands for the 2 × 2 identity matrix and J 2 = ( 0 1 1 0 ) .

Proof

The proof is an application of the method of Wigner and Mackey to build the complex irreducible representations of certain semidirect products. See, for example, [64, § 8.2]. ∎

Lemma 2

Let q ⩟ 5 be a prime number such that q ≡ 1 mod 4 . Let 𝜌 be a primitive fourth root of unity in Z q , and choose a maximal subset P ⊂ Z q * of representatives of the relation k ∌ k ⁹ ρ ∌ - k ∌ - k ⁹ ρ over Z q * . The group

G 4 = ⟹ A , B : A q = B 4 = 1 , B A B - 1 = A ρ ⟩

has four complex irreducible representations of degree one given by

χ 1 : { A ↩ 1 B ↩ 1 , χ i : { A ↩ 1 , B ↩ i ,   χ - 1 : { A ↩ 1 , B ↩ - 1 ,   χ - i : { A ↩ 1 , B ↩ - i ,

and q - 1 4 complex irreducible representations of degree four given by

ϕ j : A ↩ diag ⁥ ( ω q j , ω q ρ ⁹ j , ω q - j , ω q - ρ ⁹ j ) , B ↩ ( 0 0 0 1 1 0 0 0 0 1 0 0 0 0 1 0 )

for j ∈ P . In addition, the rational irreducible representations of G 4 are

χ 1 , χ - 1 , χ i ⊕ χ - i ⁹ and ⁹ ϱ = ⚁ j ∈ P ϕ j

Proof

The construction of the representations follows from [64, § 8.2], and we only need to prove the last statement. As G 4 has four conjugacy classes of cyclic subgroups, it has four pairwise non-equivalent rational irreducible representations: three of them are χ 1 , χ - 1 and χ i ⊕ χ - i . It follows that ϕ j are Galois conjugate and added up together produce the remaining rational irreducible representation. ∎

Symmetric square character formula

Let đ» be a finite group, and let ρ : H → GL ⁹ ( V ) be a complex representation of đ». Consider the associated representation of đ» on the symmetric square vector space of 𝑉,

Sym 2 ⁹ ( ρ ) : H → GL ⁹ ( Sym 2 ⁹ ( V ) ) .

According to [64, Proposition 2.3], the character χ ρ sym of Sym 2 ⁹ ( ρ ) is given by

(5.1) χ ρ sym ⁹ ( h ) = 1 2 ⁹ [ χ ρ ⁹ ( h ) 2 + χ ρ ⁹ ( h 2 ) ]   for ⁹ h ∈ H ,

where χ ρ denotes the character of 𝜌.

Lemma 3

Let 𝜒 be a character of đ», and let χ ÂŻ denote its complex-conjugate. Then

∑ h ∈ H χ sym = 1 2 ⁹ [ ∑ h ∈ H ( χ + χ ÂŻ ) sym ⁹ ( h ) - | H | ⁹ ⟹ χ ∣ χ ⟩ H ] .

Proof

According to [64, Exercise 2.1], for any pair of characters χ 1 and χ 2 , we have

( χ 1 + χ 2 ) sym = χ 1 sym + χ 2 sym + χ 1 ⁹ χ 2 .

If we write χ 1 = χ and χ 2 = χ ÂŻ , then the previous equality implies

∑ h ∈ H χ sym ⁹ ( h ) = ∑ h ∈ H [ ( χ + χ ÂŻ ) sym ⁹ ( h ) - ( χ ÂŻ ) sym ⁹ ( h ) - χ ⁹ ( h ) ⁹ χ ÂŻ ⁹ ( h ) ] .

Since

∑ h ∈ H ( χ ÂŻ ) sym ⁹ ( h ) = ∑ h ∈ H χ sym ⁹ ( h - 1 ) = ∑ h ∈ H χ sym ⁹ ( h ) ,

the conclusion follows after noting that ∑ h ∈ H χ ⁹ ( h ) ⁹ χ ÂŻ ⁹ ( h ) = | H | ⁹ ⟹ χ ∣ χ ⟩ H . ∎

The analytic representation

Let 𝑆 be a compact Riemann surface of genus 𝑔, and let đ» be a group of automorphisms of 𝑆. The action of đ» induces a complex representation

ρ a : H → GL ⁹ ( H 1 , 0 ⁹ ( S , C ) ) ≅ GL ⁹ ( g , C ) ,

called the analytic representation of đ». Let Irr ⁹ ( H ) denote the set of complex irreducible representations of đ», up to equivalence. If we write

ρ a ≅ ⚁ ρ ∈ Irr ⁹ ( H ) ÎŒ ρ ⁹ ρ , where ⁹ ÎŒ ρ ∈ N âˆȘ { 0 } ,

then ÎŒ ρ can be computed using the classically known Chevalley–Weil formula; see [17].

Lemma 4

Assume that the action of đ» on 𝑆 has signature σ = ( 0 ; k 1 , 
 , k s ) and is represented by the ske Ξ : Δ → H , where Δ is a Fuchsian group of signature 𝜎 canonically presented as in (2.2). Then ÎŒ ρ = 0 if 𝜌 is the trivial representation; otherwise

(5.2) ÎŒ ρ = - d ρ + ∑ l = 1 s ∑ j = 1 k l N l , j ρ ⁹ ( 1 - j k l ) ,

where d ρ is the degree of 𝜌 and N l , j ρ is the number of eigenvalues of ρ ⁹ ( Ξ ⁹ ( x l ) ) that are equal to ω k l j .

Computation of the dimension N S , G

Let 𝑆 be a compact Riemann surface of genus 𝑔. As mentioned in § 2.7, following [67] and [26, Lemma 3.8] together with formula (5.1), the dimension of (the component which contains J ⁱ S of) the submanifold S S of H g of matrices representing ppavs admitting an action equivalent to that of Aut ⁡ ( S ) is

(5.3) N S , G = ⟹ χ ρ a sym ∣ 1 ⟩ G = 1 | G | ⁹ ∑ g ∈ G χ ρ a sym ⁹ ( g ) = 1 2 ⁹ | G | ⁹ ∑ g ∈ G [ χ ρ a ⁹ ( g ) 2 + χ ρ a ⁹ ( g 2 ) ] ,

where ρ a is the analytic representation of G = Aut ⁥ ( S ) .

As a direct consequence of Lemma 3, the previous equality can be rewritten as follows.

Lemma 5

Assume that χ ρ a denotes the character of the analytic representation of G = Aut ⁥ ( S ) . Then

N S , G = 1 2 ⁹ | G | ⁹ [ ∑ g ∈ G ( χ ρ a + χ ÂŻ ρ a ) sym ⁹ ( g ) - | G | ⁹ ⟹ χ ρ a ∣ χ ρ a ⟩ G ] .

Remark 4

Note the following.

  1. The computation of N S , G depends both on the group đș and on its action on the Riemann surface 𝑆. It then makes sense to compute N S , H for a subgroup đ» of G = Aut ⁥ ( S ) . When considering G = Aut ⁥ ( S ) , we write N S instead of N S , G .

  2. Observe that if H 1 ⩜ H 2 are two groups of automorphisms of 𝑆, then we have N S , H 2 ⩜ N S , H 1 . In particular, if N S , H = 0 for some group of automorphisms đ» of 𝑆, then N S = 0 .

  3. According to (5.1), the summand ( χ ρ a + χ ÂŻ ρ a ) sym ⁹ ( g ) in Lemma 5 is equal to

    (5.4) ( χ ρ a + χ ÂŻ ρ a ) sym ⁹ ( g ) = 1 2 ⁹ [ ( χ ρ a + χ ÂŻ ρ a ) 2 ⁹ ( g ) + ( χ ρ a + χ ÂŻ ρ a ) ⁹ ( g 2 ) ]

    and is, indeed, the sum of two rational numbers.

6 Proof of Theorems 3 and 4

The surface X 8

With the notation of Lemma 1, the representations

χ j 1 - ⁹ and ⁹ χ j 2 -   for ⁹ j = 1 , 
 , q - 1 2

of G 8 are pairwise Galois conjugate, showing that their direct sum yields a rational irreducible representation of degree 2 ⁱ ( q - 1 ) . We denote this last representation by 𝜚, namely

ϱ ≅ ( ⚁ j = 1 q - 1 2 χ j 1 - ) ⊕ ( ⚁ j = 1 q - 1 2 χ j 2 - ) .

As explained in § 2.8, the group algebra decomposition of J ⁹ X 8 with respect to G 8 ≅ Aut ⁥ ( X 8 ) has the form J ⁹ X 8 ∌ B n × P , where đ” is the abelian subvariety of J ⁹ X 8 associated to 𝜚 and 𝑃 is the product of the factors associated to the remaining rational irreducible representations of G 8 . Following [36, Proposition (10.8)], the Schur index of χ 1 1 - is one, and thus n = 2 .

We recall that the action of G 8 on X 8 is represented by the ske Ξ = ( z , z ⁹ x , x - 1 ) . The dimension of the fixed subspaces of χ 1 1 - under the action of the subgroups ⟹ z ⟩ , ⟹ z ⁹ x ⟩ and ⟹ x - 1 ⟩ is 1, 0 and 0 respectively; this is clear by noting that χ 1 1 - ⁹ ( z ) = J 2 and that

χ 1 1 - ⁹ ( z ⁹ x ) = ( 0 ω 2 ⁹ q q - 1 ω 2 ⁹ q 0 )   and   χ 1 1 - ⁹ ( x - 1 ) = diag ⁥ ( ω 2 ⁹ q - 1 , ω 2 ⁹ q 1 - q )

do not have 1 as an eigenvalue. In addition, it is easy to see that the character field of χ 1 1 - has degree q - 1 over the rationals. We then apply equation (2.11) to conclude that

dim ⁥ B = ( q - 1 ) ⁹ [ - 2 + 1 2 ⁹ ( ( 2 - 1 ) + ( 2 - 0 ) + ( 2 - 0 ) ) ] = q - 1 2 .

Since the dimension of J ⁹ X 8 is q - 1 , it follows that P = 0 , and therefore

(6.1) J ⁹ X 8 ∌ B 2 .

Finally, we consider the subgroup ⟹ z ⟩ of G 8 and write Y 8 = X 8 / ⟹ z ⟩ . The induced isogeny (2.10) applied to (6.1) implies that

J ⁹ Y 8 = J ⁹ ( X 8 / ⟹ z ⟩ ) ∌ B , and therefore   J ⁹ X 8 ∌ J ⁹ Y 8 2 ,

as claimed in Theorem 3.

We now proceed to prove that N X 8 = 0 . Let H = ⟹ x ⟩ ≅ C 2 ⁹ q , and consider the maps

ρ k : H → C   given by   x ↩ ω 2 ⁹ q k   for ⁹ k = 0 , 
 , 2 ⁹ q - 1 .

We claim that the analytic representation ρ a of đ» decomposes as the direct sum

ρ a ≅ ⚁ j = q + 1 2 ⁹ q - 1 ρ j .

To prove that, observe that { ρ 0 , 
 , ρ 2 ⁹ q - 1 } is a full set of pairwise non-equivalent complex irreducible representations of đ». Besides, as noted in the proof of Proposition 3.10, the induced action of đ» on X 8 has signature ( 0 ; q , 2 ⁹ q , 2 ⁹ q ) and is represented by the ske ( x 2 , x - 1 , x - 1 ) (see also the algorithm in [7] based on [66]). If we write

ρ a ≅ ⚁ k = 0 2 ⁹ q - 1 ÎŒ k ⁹ ρ k   for some ⁹ ÎŒ k ∈ N âˆȘ { 0 } ,

then according to Lemma 4, we have Ό 0 = 0 and

ÎŒ k = - 1 + 2 ⁹ ∑ j = 1 2 ⁹ q - 1 N 1 , j k ⁹ ( 1 - j 2 ⁹ q ) + ∑ j = 1 q - 1 N 2 , j k ⁹ ( 1 - j q )

for each k ∈ { 1 , 
 , 2 ⁱ q - 1 } , where N 1 , j k = 1 if and only if j = 2 ⁱ q - k and

N 2 , j k = { 1 if and only if ⁹ j = k ⁹ for ⁹ k ⩜ q , 1 if and only if ⁹ j = k - q ⁹ for ⁹ k ⩟ q + 1 .

In this way, we obtain if k ⩜ q , then Ό k = 0 , and if k ⩟ q + 1 , then Ό k = 1 . The claim follows.

We then can construct the following table.

ℎ Order ( χ ρ a + χ ÂŻ ρ a ) 2 ⁹ ( h ) ( χ ρ a + χ ÂŻ ρ a ) ⁹ ( h 2 )
1 1 ( 2 ⁹ ( q - 1 ) ) 2 2 ⁹ ( q - 1 )
x q 2 0 2 ⁹ ( q - 1 )
x 2 ⁹ k , 1 ⩜ k ⩜ q - 1 𝑞 ( - 2 ) 2 −2
x 2 ⁹ k - 1 , 1 ⩜ k ⩜ q , k ≠ q + 1 2 2 ⁹ q 0 −2

Then we obtain, by (5.4), that ∑ h ∈ H ( χ ρ a + χ ÂŻ ρ a ) sym ⁹ ( h ) is equal to

1 2 ⁹ ( 4 ⁹ ( q - 1 ) 2 + 2 ⁹ ( q - 1 ) + 2 ⁹ ( q - 1 ) + 2 ⁹ ( q - 1 ) - 2 ⁹ ( q - 1 ) ) = 2 ⁹ ( q 2 - q ) .

Now, the fact that | H | ⁹ ⟹ χ ρ a ∣ χ ρ a ⟩ H = 2 ⁹ q ⁹ ( q - 1 ) together with Lemma 3 imply that

∑ h ∈ H χ ρ a sym ⁹ ( h ) = 1 2 ⁹ [ 2 ⁹ ( q 2 - q ) - 2 ⁹ q ⁹ ( q - 1 ) ] = 0 ,

and therefore we obtain N X 8 , H = 0 . Hence N X 8 = 0 , as claimed in Theorem 4.

The surface X 4

With the notation of Lemma 2, the group algebra decomposition of J ⁹ X 4 with respect to G 4 ≅ Aut ⁥ ( X 4 ) is J ⁹ X 4 ∌ D 1 × D i × D - 1 × D 4 , where the factor D k is associated to the representation χ k and đ· is associated to 𝜚. Observe that the character field of each ϕ j has degree q - 1 4 over the rationals.

We recall that the action of G 4 on X 4 is represented by the ske

Θ = ( A - 1 ⁹ B , B - 1 , A ) .

The dimension of the fixed subspace of ϕ j under the action of ⟹ A - 1 ⁹ B ⟩ , ⟹ B - 1 ⟩ and ⟹ A ⟩ is 1, 1 and 0 respectively. Consequently, equation (2.11) implies

dim ⁥ D = q - 1 4 ⁹ [ - 4 + 1 2 ⁹ ( ( 4 - 1 ) + ( 4 - 1 ) + ( 4 - 0 ) ) ] = q - 1 4 .

The previous equality shows, in addition, that

D 1 = D i = D - 1 = 0 , and therefore   J ⁹ X 4 ∌ D 4 .

Finally, we consider the subgroup ⟹ B ⟩ of G 4 and write Y 4 = X 4 / ⟹ B ⟩ . The induced isogeny (2.10) applied to the previous isogeny implies that

J ⁹ Y 4 = J ⁹ ( X 4 / ⟹ B ⟩ ) ∌ D , and therefore   J ⁹ X 4 ∌ J ⁹ Y 4 4 ,

as claimed in Theorem 3.

We now proceed to prove that N X 4 = q - 1 4 . A routine application of Lemma 4 shows that the analytic representation ρ a of G 4 is

ρ a ≅ ⚁ j ∈ P ϕ j , and therefore   | G 4 | ⁹ ⟹ χ ρ a ∣ χ ρ a ⟩ G 4 = q ⁹ ( q - 1 ) ,

where đ’« is as in Lemma 2. It is not difficult to see that

( χ ρ a + χ ÂŻ ρ a ) ⁹ ( g ) = { 2 ⁹ ( q - 1 ) if ⁹ g = 1 , 0 if ⁹ | g | = 2 , 4 , - 2 if ⁹ | g | = q ,

and therefore (5.4) allows us to write that ( χ ρ a + χ ÂŻ ρ a ) sym ⁹ ( 1 ) = ( q - 1 ) ⁹ ( 2 ⁹ q - 1 ) and that

∑ g ∈ G 4 , | g | = Ï” ( χ ρ a + χ ÂŻ ρ a ) sym ⁹ ( g ) = { q ⁹ ( q - 1 ) if ⁹ Ï” = 2 , 0 if ⁹ Ï” = 4 , q - 1 if ⁹ Ï” = q .

Thus ∑ g ∈ G 4 ( χ ρ a + χ ÂŻ ρ a ) sym ⁹ ( g ) = 3 ⁹ q ⁹ ( q - 1 ) . Now, it follows from Lemma 3 that

∑ g ∈ G 4 χ ρ a sym ⁹ ( g ) = 1 2 ⁹ ( 3 ⁹ q ⁹ ( q - 1 ) - q ⁹ ( q - 1 ) ) = q ⁹ ( q - 1 ) ,

and consequently, Lemma 5 says that N X 4 = q - 1 4 , as desired.

The surface X 3

The complex irreducible representations of G 3 = ⟹ x : x 3 ⁹ q = 1 ⟩ are

χ k : G 3 → C , x ↩ ω 3 ⁹ q k   for ⁹ k ∈ { 0 , 
 , 3 ⁹ q - 1 } .

We denote by ρ r the complexification of the rational representation corresponding to the action of G 3 on X 3 . We claim that

ρ r ≅ ⚁ σ ∈ G χ 1 σ , where ⁹ G = Gal ⁥ ( Q ⁹ ( ω 3 ⁹ q ) / Q )

In fact, according to [62, Theorem 5.10], the multiplicity of χ 1 in the decomposition of ρ r as a sum of irreducible representations is one. In addition, since ρ r is indeed defined over the rationals, we can deduce that all the orbit of χ 1 under 𝒱 appears in the decomposition of ρ r . The claim follows after noting that the aforementioned orbit has length 2 ⁹ ( q - 1 ) , and this number agrees with the degree of ρ r .

Since ρ r ≅ ρ a ⊕ ρ ÂŻ a , the previous claim says that ρ a decomposes into q - 1 pairwise non-equivalent complex irreducible representations of degree one of G 3 , and thereby

(6.2) ⟹ χ ρ a ∣ χ ρ a ⟩ G 3 = q - 1 .

In order to determine the character χ ρ r of ρ r , it is convenient to decompose ρ r in the following different but equivalent way. Let

Λ ⁱ ( q ) = { t ∈ { 1 , 
 , 3 ⁱ q - 1 } : gcd ⁡ ( t , 3 ⁱ q ) = 1 } ,

and consider the subset Λ â€Č ⁹ ( q ) of Λ ⁹ ( q ) of cardinality q - 1 obtained by removing the additive inverses modulo 3 ⁹ q (that is, k ∈ Λ â€Č ⁹ ( q ) , then - k mod 3 ⁹ q ∉ Λ â€Č ). It is not difficult to see that

ρ r ≅ ⚁ k ∈ Λ â€Č ⁹ ( q ) ( χ k ⊕ χ ÂŻ k ) , and therefore   χ ρ r = ∑ k ∈ Λ â€Č ⁹ ( q ) ( χ k + χ ÂŻ k ) .

  1. Clearly, χ ρ r ⁹ ( 1 ) = 2 ⁹ ( q - 1 ) , the degree of ρ r .

  2. For the elements of order 3 (that is, x q and x 2 ⁹ q ), we have that

    ( χ k + χ ÂŻ k ) ⁹ ( x q ) = ω 3 ⁹ q k ⁹ q + ω 3 ⁹ q - k ⁹ q = - 1 ⁹ for each ⁹ k ∈ Λ â€Č âŸč χ ρ r ⁹ ( x q ) = - ( q - 1 ) .

    Analogously, one sees that χ ρ r ⁹ ( x 2 ⁹ q ) = - ( q - 1 ) .

  3. For the elements of order 𝑞 (that is, x 3 ⁱ j with j ∈ { 1 , 
 , q - 1 } ), we have that

    ( χ k + χ ÂŻ k ) ⁹ ( x 3 ⁹ j ) = ω 3 ⁹ q 3 ⁹ k ⁹ j + ω 3 ⁹ q - 3 ⁹ k ⁹ j = ω q k ⁹ j + ω q - k ⁹ j

    for every k ∈ Λ â€Č . Then

    χ ρ r ⁹ ( x 3 ⁹ j ) = ∑ k ∈ Λ â€Č ⁹ ( q ) ( χ k + χ ÂŻ k ) ⁹ ( x 3 ⁹ j ) = ∑ k ∈ Λ â€Č ⁹ ( q ) ( ω q k ⁹ j + ω q - k ⁹ j ) ,

    showing that

    χ ρ r ⁹ ( x 3 ⁹ j ) = - 2   for all ⁹ j ∈ { 1 , 
 , q - 1 } .

  4. For the elements of order 3 ⁱ q (that is, x t with t ∈ Λ ⁱ ( q ) ), we have that

    χ ρ r ⁹ ( x t ) = ∑ k ∈ Λ â€Č ⁹ ( q ) ( χ k + χ ÂŻ k ) ⁹ ( x t ) = ∑ k ∈ Λ â€Č ⁹ ( q ) ( ω 3 ⁹ q k ⁹ t + ω 3 ⁹ q - k ⁹ t ) ,

    and this corresponds to the sum of all primitive 3 ⁹ q -th roots of unity. It is a known fact that this sum corresponds to the Möbius function ÎŒ ⁹ ( 3 ⁹ q ) , which is 1; thus

    χ ρ r ⁹ ( x t ) = 1   for every ⁹ t ∈ Λ ⁹ ( q ) .

We summarise all the above in the third column of the following table; the fourth column follows from all the above and (5.4).

Order 𝑔 χ ρ r ⁹ ( g ) = ( χ ρ a + χ ÂŻ ρ a ) ⁹ ( g ) 2 ⁹ ( χ ρ a + χ ÂŻ ρ a ) sym ⁹ ( g )
1 1 2 ⁹ ( q - 1 ) 2 ⁹ ( q - 1 ) + ( 2 ⁹ ( q - 1 ) ) 2
3 x q , x 2 ⁹ q - ( q - 1 ) - ( q - 1 ) + ( q - 1 ) 2
𝑞 x 3 ⁱ j , j = 1 , 
 , q - 1 −2 - 2 + 4 = 2
3 ⁱ q x t , t ∈ Λ ⁱ ( q ) 1 1 + 1 = 2

It follows that

∑ g ∈ G 3 ( χ ρ a + χ ÂŻ ρ a ) sym ⁹ ( g ) = 3 ⁹ q ⁹ ( q - 1 ) ,

and the desired result N X 3 = 0 follows from Lemma 3 together with equality (6.2).

The surfaces X 2 , k

Let 1 ⩜ k ⩜ q - 3 2 . Consider the complex irreducible representations of

Aut ( X 2 , k ) ≅ G 2 = ⟹ a , b : a q = b 2 = [ a , b ] = 1 ⟩

given by χ 1 : x ↩ ω 2 ⁹ q and χ 2 : x ↩ ω q , where x := a ⁹ b . If G 1 and G 2 are the Galois groups of the extensions of ℚ by Q ⁹ ( ω 2 ⁹ q ) and Q ⁹ ( ω q ) respectively, then

⚁ σ ∈ G 1 χ 1 σ   and   ⚁ σ ∈ G 2 χ 2 σ

are rational irreducible representations of G 2 of degree q - 1 .

Let ρ r denote the complexification of the rational representation corresponding to the action of G 2 on X 2 , k . By arguing as in the case of X 3 , one obtains that

ρ r ≅ ( ⚁ σ ∈ G 1 χ 1 σ ) ⊕ ( ⚁ σ ∈ G 2 χ 2 σ ) ,

and if ρ a is the analytic representation of the involved action, then

(6.3) ⟹ χ ρ a ∣ χ ρ a ⟩ G 2 = q - 1 ,

where χ ρ a is the character of ρ a . Moreover, the character χ ρ r of ρ r is given by

χ ρ r ⁹ ( g ) = { 2 ⁹ ( q - 1 ) if ⁹ g = 1 , 0 if ⁹ | g | = 2 , - 2 if ⁹ | g | = q , 0 if ⁹ | g | = 2 ⁹ q ,

and therefore we can construct the following table.

Order 𝑔 χ ρ r ⁹ ( g ) = ( χ ρ a + χ ÂŻ ρ a ) ⁹ ( g ) 2 ⁹ ( χ ρ a + χ ÂŻ ρ a ) sym ⁹ ( g )
1 1 2 ⁹ ( q - 1 ) 2 ⁹ ( q - 1 ) + ( 2 ⁹ ( q - 1 ) ) 2
2 𝑏 0 2 ⁱ ( q - 1 ) + 0 2
𝑞 a j , j = 1 , 
 , q - 1 −2 - 2 + ( - 2 ) 2 = 2
2 ⁹ q a j ⁹ b , gcd ⁥ ( j , 2 ⁹ q ) = 1 0 - 2 + 0 2

The desired conclusion N X 2 , k = 0 follows from the previous table, equation (6.3) and Lemma 5.

The family C g

The complex representation ψ : D 2 ⁹ q → GL ⁹ ( 2 , C ) (with the dihedral group presented as in (3.16)) given by

ψ ⁹ ( R ) = diag ⁥ ( ω 2 ⁹ q , ω 2 ⁹ q - 1 )   and   ψ ⁹ ( T ) = ( 0 1 1 0 )

has Schur index 1 and field of characters of degree q - 1 2 over the rationals. It follows that the group algebra decomposition of J ⁹ S for each S ∈ C g with respect to D 2 ⁹ q has the form

(6.4) J ⁹ S ∌ B 2 × P ,

where đ” is the abelian subvariety of J ⁹ S associated to 𝜓. The dimension of the fixed subspaces of 𝜓 under the action of ⟹ R q ⟩ , ⟹ T ⟩ , ⟹ T ⁹ R ⟩ and ⟹ R q - 1 ⟩ equal 0, 1, 1 and 0 respectively. Thereby, equation (2.11) together with the fact that the action is represented by the ske Ξ = ( R q , T , T ⁹ R , R q - 1 ) imply that

dim ⁥ B = q - 1 2 ⁹ ( - 2 + 1 2 ⁹ ( ( 2 - 0 ) + ( 2 - 1 ) + ( 2 - 1 ) + ( 2 - 0 ) ) ) = q - 1 2 ,

and therefore P = 0 . Now, we consider the subgroup ⟹ T ⟩ ≅ C 2 of D 2 ⁹ q and write X = S / ⟹ T ⟩ . The induced isogeny (2.10) applied to (6.4) implies that

J ⁹ X = J ⁹ ( S / ⟹ T ⟩ ) ∌ B , and therefore   J ⁹ S ∌ J ⁹ X 2 .

A routine application of (5.2) permits us to see that the analytic representation ρ a of the action of D 2 ⁹ q on 𝑆 is equivalent to the Galois orbit of 𝜓, namely

ρ a ≅ ⚁ σ ψ σ ,

where 𝜎 runs over the Galois group associated to character field of 𝜓.

The following table (taken from [37, Proposition 6.1]) collects the character of ρ a and of its symmetric square for representatives of the conjugacy classes of the group.

𝑔 χ ρ a χ ρ a sym
1 q - 1 ( q - 1 ) 2 + ( q - 1 )
R 2 ⁱ j - 1 , j = 1 , 
 , q , j ≠ q + 1 2 1 1 + ( - 1 )
R 2 ⁱ j , j = 1 , 
 , q - 1 −1 1 + ( - 1 )
R q - ( q - 1 ) ( q - 1 ) 2 + ( q - 1 )
𝑇 0 0 + ( q - 1 )
T ⁹ R 0 0 + ( q - 1 )

Thus, by equation (5.3), we obtain that

N S = 1 8 ⁹ q ⁹ ( ( q - 1 ) 2 + ( q - 1 ) + ( q - 1 ) 2 + ( q - 1 ) + 2 ⁹ q ⁹ ( q - 1 ) ) = q - 1 2 .

The family K g

It is well known that D q has two complex irreducible representations of degree one, and q - 1 2 of degree two given by

ψ j : r ↩ diag ⁥ ( ω q j , ω q - j )   and   s ↩ ( 0 1 1 0 ) , where ⁹ j ∈ { 1 , 
 , q - 1 2 } ;

all of them are Galois conjugate. Clearly, their direct sum

W ≅ ⚁ j = 1 q - 1 2 ψ j

is a rational irreducible representation of D q of degree q - 1 . It follows that the group algebra decomposition of J ⁹ S for each S ∈ K g - C g with respect to D q has the form

(6.5) J ⁹ S ∌ B 2 × P ,

where đ” is the abelian subvariety of J ⁹ S associated to 𝑊. Observe that

  1. the dimension of the fixed subspace of ψ 1 under ⟹ r i ⟩ is equal to 0 for each 1 ⩜ i ⩜ q - 1 ;

  2. the dimension of the fixed subspace of ψ 1 under ⟹ s ⁹ r i ⟩ is equal to 1 for each 0 ⩜ i ⩜ q - 1 .

It follows that, independently of the equisymmetric stratum to which 𝑆 belongs (see the ske ξ i given in the proof of Proposition 3.11), equation (2.11) implies

dim ⁥ B = q - 1 2 ⁹ [ - 2 + 1 2 ⁹ ( ( 2 - 1 ) + ( 2 - 1 ) + ( 2 - 0 ) + ( 2 - 0 ) ) ] = q - 1 2 ,

and therefore P = 0 . Now, if X = S / ⟹ s ⟩ , then (2.10) applied to (6.5) implies that J ⁹ S ∌ J ⁹ X 2 .

By Lemma 4, one sees that the analytic representation ρ a of the action of D q on 𝑆 is equivalent to 𝑊. The character of ρ a and of ρ a sym is summarised in the following table.

𝑔 χ ρ a χ ρ a sym
1 q - 1 ( q - 1 ) 2 + ( q - 1 )
s ⁱ r j , j = 0 , 
 , q - 1 0 q - 1
r j , j = 1 , 
 , q - 1 −1 - 1 + ( - 1 ) 2

Finally, by equation (5.3), we obtain that

N S = 1 4 ⁹ q ⁹ ( ( q - 1 ) 2 + ( q - 1 ) + q ⁹ ( q - 1 ) ) = q - 1 2 .

Award Identifier / Grant number: 1180073

Award Identifier / Grant number: 11180024

Award Identifier / Grant number: 1190991

Funding statement: Partially supported by Fondecyt Grants 1180073, 11180024, 1190991 and Redes Grant 2017-170071.

A Addendum

We recall here the fact that the full automorphism group of the Accola–Maclachlan curve X 8 determines its Jacobian variety J ⁹ X 8 (that is, N X 8 = 0 ), and therefore it allows us to determine its period matrix. In this addendum, we determine explicitly the period matrix ( I 4 ⁹ Z ) , where Z ∈ H 4 , of J ⁹ X 8 provided that the genus of X 8 is four (that is, for q = 5 ). To accomplish this task, we apply the results on adapted hyperbolic polygons and algorithms programed in [6] to realise the action of the full automorphism group of X 8 in the symplectic group. Explicitly, the rational representation ρ r : Aut ⁥ ( X 8 ) → Sp ⁹ ( 8 , Z ) is given by

ρ r ⁹ ( x - 1 ) = ( 0 1 1 1 - 1 0 0 0 1 0 0 0 0 - 1 0 0 0 0 0 0 0 1 - 1 0 0 0 0 0 0 0 1 - 1 1 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 1 0 0 0 0 ) , ρ r ⁹ ( z ⁹ x ) = ( 0 0 1 0 0 0 0 0 1 0 0 0 0 - 1 0 0 - 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 - 1 1 0 0 1 1 0 0 0 0 0 0 1 1 0 - 1 0 0 0 0 0 0 - 1 0 0 0 0 ) .

Now, if we write

Z = ( a b c d b e f g c f h j d g j k ) ,

then with the notation of § 2.7, the fact that

R ⋅ Z = Z   for each ⁹ R ∈ ⟹ ρ r ⁹ ( x - 1 ) , ρ r ⁹ ( z ⁹ x ) ⟩

implies that the coefficients of 𝑍 satisfy the relations

  1. a = - 100 / 11 ⁹ k 3 - 140 / 11 ⁹ k ,

  2. b = - 5 ⁹ k 2 - 31 / 4 ,

  3. c = d = 5 ⁹ k 2 + 29 / 4 ,

  4. e = 20 / 11 ⁹ k 3 + 39 / 11 ⁹ k ,

  5. f = - 10 / 11 ⁹ k 3 - 39 / 22 ⁹ k ,

  6. g = - 50 / 11 ⁹ k 3 - 151 / 22 ⁹ k ,

  7. h = k ,

  8. j = 60 / 11 ⁹ k 3 + 84 / 11 ⁹ k ,

where the parameter 𝑘 satisfies the following equation:

(A.1) k 4 + 5 / 2 ⁹ k 2 + 121 / 80 = 0 .

The solutions of (A.1) are

k 1 = - 1 2 ⁹ i ⁹ 2 5 ⁹ 5 + 5 , k 2 = 1 2 ⁹ i ⁹ 2 5 ⁹ 5 + 5 , k 3 = - 1 2 ⁹ i ⁹ - 2 5 ⁹ 5 + 5 , k 4 = 1 2 ⁹ i ⁹ - 2 5 ⁹ 5 + 5 .

The values k 1 and k 4 must be disregarded; indeed,

a ⁱ ( k 1 ) := - 100 / 11 ⁱ k 1 3 - 140 / 11 ⁱ k 1   and   a ⁱ ( k 4 ) := - 100 / 11 ⁱ k 4 3 - 140 / 11 ⁱ k 4

do not have imaginary part positive, and therefore the corresponding matrices do not belong to H 4 . Now, the fact that Z ∈ H 4 also implies that

Ύ := det ⁥ Im ⁥ ( a b b e ) ⁹ must be positive .

With the help of numerical approximations of [73] one sees that 𝛿 is positive only for k = k 2 .

We replace the value of k = k 2 in equalities ( 1 ) , 
 , ( 8 ) to finally obtain that the period matrix of J ⁱ X 8 is ( I 4 ⁱ Z ) , where 𝑍 is given below:

( 25 22 ⁹ i ⁹ z 3 2 - 70 11 ⁹ i ⁹ z 1 2 ⁹ 5 - 3 2 - 1 2 ⁹ 5 + 1 - 1 2 ⁹ 5 + 1 1 2 ⁹ 5 - 3 2 - 5 22 ⁹ i ⁹ z 3 2 + 39 22 ⁹ i ⁹ z 5 44 ⁹ i ⁹ z 3 2 - 39 44 ⁹ i ⁹ z 25 44 ⁹ i ⁹ z 3 2 - 151 44 ⁹ i ⁹ z - 1 2 ⁹ 5 + 1 5 44 ⁹ i ⁹ z 3 2 - 39 44 ⁹ i ⁹ z 1 2 ⁹ i ⁹ z - 15 22 ⁹ i ⁹ z 3 2 + 42 11 ⁹ i ⁹ z - 1 2 ⁹ 5 + 1 25 44 ⁹ i ⁹ z 3 2 - 151 44 ⁹ i ⁹ z - 15 22 ⁹ i ⁹ z 3 2 + 42 11 ⁹ i ⁹ z 1 2 ⁹ i ⁹ z )

with z = 2 5 ⁹ 5 + 5 .

Remark 5

The results of [6] extend those of [13]; thus, if we repeat this procedure for the Accola–Maclachlan curve of genus two, then we recover the period matrix given in [13].

Acknowledgements

The authors are very grateful to the referee for his/her valuable comments and suggestions.

  1. Communicated by: Robert Guralnick

References

[1] R. Accola, On the number of automorphisms of a closed Riemann surface, Trans. Amer. Math. Soc. 131 (1968), 398–408. 10.1090/S0002-9947-1968-0222281-6Suche in Google Scholar

[2] R. Accola, Topics in the Theory of Riemann Surfaces, Lecture Notes in Math. 1595, Springer, Berlin, 1994. 10.1007/BFb0073575Suche in Google Scholar

[3] N. Arakelian and P. Speziali, Algebraic curves with automorphism groups of large prime order, Math. Z. 299 (2021), no. 3–4, 2005–2028. 10.1007/s00209-021-02749-zSuche in Google Scholar

[4] E. Badr and F. Bars, Plane model-fields of definition, fields of definition, and the field of moduli for smooth plane curves, J. Number Theory 194 (2019), 278–283. 10.1016/j.jnt.2018.07.010Suche in Google Scholar

[5] P. Barraza and A. M. Rojas, The group algebra decomposition of Fermat curves of prime degree, Arch. Math. (Basel) 104 (2015), no. 2, 145–155. 10.1007/s00013-015-0724-ySuche in Google Scholar

[6] A. Behn, R. E. Rodríguez and A. M. Rojas, Adapted hyperbolic polygons and symplectic representations for group actions on Riemann surfaces, J. Pure Appl. Algebra 217 (2013), no. 3, 409–426. 10.1016/j.jpaa.2012.06.030Suche in Google Scholar

[7] A. Behn, A. M. Rojas and M. Tello-Carrera, A SAGE Package for 𝑛-gonal equisymmetric stratification of M g , Exp. Math. (2020), 10.1080/10586458.2020.1763872. 10.1080/10586458.2020.1763872Suche in Google Scholar

[8] M. Belolipetsky and G. A. Jones, Automorphism groups of Riemann surfaces of genus p + 1 , where 𝑝 is prime, Glasg. Math. J. 47 (2005), no. 2, 379–393. 10.1017/S0017089505002612Suche in Google Scholar

[9] K. Berry and M. Tretkoff, The period matrix of Macbeath’s curve of genus seven, Curves, Jacobians, and Abelian Varieties, Contemp. Math. 136, American Mathematical Society, Providence (1992), 31–40. 10.1090/conm/136/1188192Suche in Google Scholar

[10] C. Birkenhake and H. Lange, Complex Abelian Varieties, 2nd ed., Grundlehren Math. Wiss. 302, Springer, Berlin, 2004. 10.1007/978-3-662-06307-1Suche in Google Scholar

[11] S. A. Broughton, The equisymmetric stratification of the moduli space and the Krull dimension of mapping class groups, Topology Appl. 37 (1990), no. 2, 101–113. 10.1016/0166-8641(90)90055-7Suche in Google Scholar

[12] S. A. Broughton, Classifying finite group actions on surfaces of low genus, J. Pure Appl. Algebra 69 (1991), no. 3, 233–270. 10.1016/0022-4049(91)90021-SSuche in Google Scholar

[13] E. Bujalance, A. F. Costa, J. M. Gamboa and G. Riera, Period matrices of Accola–Maclachlan and Kulkarni surfaces, Ann. Acad. Sci. Fenn. Math. 25 (2000), no. 1, 161–177. Suche in Google Scholar

[14] A. Carocca, S. Recillas and R. E. Rodríguez, Dihedral groups acting on Jacobians, Complex Manifolds and Hyperbolic Geometry, Contemp. Math. 311, American Mathematical Society, Providence (2002), 41–77. 10.1090/conm/311/05446Suche in Google Scholar

[15] A. Carocca and S. Reyes-Carocca, Riemann surfaces of genus 1 + q 2 with 3 ⁱ q 2 automorphisms, J. Algebra 588 (2021), 440–470. 10.1016/j.jalgebra.2021.09.001Suche in Google Scholar

[16] A. Carocca and R. E. Rodríguez, Jacobians with group actions and rational idempotents, J. Algebra 306 (2006), no. 2, 322–343. 10.1016/j.jalgebra.2006.07.027Suche in Google Scholar

[17] C. Chevalley, A. Weil and E. Hecke, Über das Verhalten der Integrale erster Gattung bei Automorphismen des Funktionenkörpers, Abh. Math. Sem. Univ. Hambg. 10 (1934), no. 1, 358–361. 10.1007/978-1-4757-1705-1_10Suche in Google Scholar

[18] M. Conder, https://www.math.auckland.ac.nz/~conder/TriangleGroupQuotients101.txt. Suche in Google Scholar

[19] A. F. Costa and M. Izquierdo, Equisymmetric strata of the singular locus of the moduli space of Riemann surfaces of genus 4, Geometry of Riemann Surfaces, London Math. Soc. Lecture Note Ser. 368, Cambridge University, Cambridge (2010), 120–138. 10.1017/CBO9781139194266.007Suche in Google Scholar

[20] A. F. Costa and M. Izquierdo, On the connectedness of the branch locus of the moduli space of Riemann surfaces of genus 4, Glasg. Math. J. 52 (2010), no. 2, 401–408. 10.1017/S0017089510000091Suche in Google Scholar

[21] A. F. Costa and M. Izquierdo, On the existence of connected components of dimension one in the branch locus of moduli spaces of Riemann surfaces, Math. Scand. 111 (2012), no. 1, 53–64. 10.7146/math.scand.a-15213Suche in Google Scholar

[22] A. F. Costa and M. Izquierdo, One-dimensional families of Riemann surfaces of genus g with 4 ⁱ g + 4 automorphims, Rev. R. Acad. Cienc. Exactas Fís. Nat. Ser. A Mat. RACSAM 112 (2018), no. 3, 623–631. 10.1007/s13398-017-0429-0Suche in Google Scholar

[23] P. Dùbes and M. Emsalem, On fields of moduli of curves, J. Algebra 211 (1999), no. 1, 42–56. 10.1006/jabr.1998.7586Suche in Google Scholar

[24] R. Donagi and E. Markman, Spectral covers, algebraically completely integrable, Hamiltonian systems, and moduli of bundles, Integrable Systems and Quantum Groups (Montecatini Terme 1993), Lecture Notes in Math. 1620, Springer, Berlin (1996), 1–119. 10.1007/BFb0094792Suche in Google Scholar

[25] A. Fischer, M. Liu and J. Paulhus, Jacobian varieties of Hurwitz curves with automorphism group PSL ⁱ ( 2 , q ) , Involve 9 (2016), no. 4, 639–655. 10.2140/involve.2016.9.639Suche in Google Scholar

[26] P. Frediani, A. Ghigi and M. Penegini, Shimura varieties in the Torelli locus via Galois coverings, Int. Math. Res. Not. IMRN 2015 (2015), no. 20, 10595–10623. 10.1093/imrn/rnu272Suche in Google Scholar

[27] Y. Fuertes and G. González-Diez, Fields of moduli and definition of hyperelliptic covers, Arch. Math. (Basel) 86 (2006), no. 5, 398–408. 10.1007/s00013-005-1433-8Suche in Google Scholar

[28] G. González Díez, Loci of curves which are prime Galois coverings of P 1 , Proc. Lond. Math. Soc. (3) 62 (1991), no. 3, 469–489. 10.1112/plms/s3-62.3.469Suche in Google Scholar

[29] W. J. Harvey, On branch loci in TeichmĂŒller space, Trans. Amer. Math. Soc. 153 (1971), 387–399. 10.1090/S0002-9947-1971-0297994-0Suche in Google Scholar

[30] R. A. Hidalgo, Non-hyperelliptic Riemann surfaces with real field of moduli but not definable over the reals, Arch. Math. (Basel) 93 (2009), no. 3, 219–224. 10.1007/s00013-009-0025-4Suche in Google Scholar

[31] R. A. Hidalgo, A remark on the field of moduli of Riemann surfaces, Arch. Math. (Basel) 114 (2020), no. 5, 515–526. 10.1007/s00013-019-01411-9Suche in Google Scholar

[32] R. A. Hidalgo, L. JimĂ©nez, S. Quispe and S. Reyes-Carocca, Quasiplatonic curves with symmetry group Z 2 2 ⋊ Z m are definable over ℚ, Bull. Lond. Math. Soc. 49 (2017), no. 1, 165–183. 10.1112/blms.12014Suche in Google Scholar

[33] R. A. Hidalgo and S. Reyes, Fields of moduli of classical Humbert curves, Q. J. Math. 63 (2012), no. 4, 919–930. 10.1093/qmath/har017Suche in Google Scholar

[34] R. A. Hidalgo and S. Reyes-Carocca, Weil’s Galois descent theorem from a computational point of view, Geometry at the Frontier—Symmetries and Moduli Spaces of Algebraic Varieties, Contemp. Math. 766, American Mathematical Society, Providence (2021), 217–228. 10.1090/conm/766/15383Suche in Google Scholar

[35] B. Huggins, Fields of moduli of hyperelliptic curves, Math. Res. Lett. 14 (2007), no. 2, 249–262. 10.4310/MRL.2007.v14.n2.a8Suche in Google Scholar

[36] I. M. Isaacs, Character Theory of Finite Groups, Pure Appl. Math. 69, Academic Press, New York, 2006. 10.1090/chel/359Suche in Google Scholar

[37] M. Izquierdo, L. JimĂ©nez and A. Rojas, Decomposition of Jacobian varieties of curves with dihedral actions via equisymmetric stratification, Rev. Mat. Iberoam. 35 (2019), no. 4, 1259–1279. 10.4171/rmi/1084Suche in Google Scholar

[38] M. Izquierdo, G. A. Jones and S. Reyes-Carocca, Groups of automorphisms of Riemann surfaces and maps of genus p + 1 where 𝑝 is prime, Ann. Fenn. Math. 46 (2021), no. 2, 839–867. 10.5186/aasfm.2021.4649Suche in Google Scholar

[39] M. Izquierdo and S. Reyes-Carocca, A note on large automorphism groups of compact Riemann surfaces, J. Algebra 547 (2020), 1–21. 10.1016/j.jalgebra.2019.11.012Suche in Google Scholar

[40] M. Izquierdo, S. Reyes-Carocca and A. M. Rojas, On families of Riemann surfaces with automorphisms, J. Pure Appl. Algebra 225 (2021), no. 10, Paper No. 106704. 10.1016/j.jpaa.2021.106704Suche in Google Scholar

[41] S. Koizumi, The fields of moduli for polarized abelian varieties and for curves, Nagoya Math. J. 48 (1972), 37–55. 10.1017/S0027763000015063Suche in Google Scholar

[42] A. Kontogeorgis, Field of moduli versus field of definition for cyclic covers of the projective line, J. ThĂ©or. Nombres Bordeaux 21 (2009), no. 3, 679–692. 10.5802/jtnb.694Suche in Google Scholar

[43] R. S. Kulkarni, A note on Wiman and Accola–Maclachlan surfaces, Ann. Acad. Sci. Fenn. Ser. A I Math. 16 (1991), no. 1, 83–94. 10.5186/aasfm.1991.1615Suche in Google Scholar

[44] I. Kuribayashi and A. Kuribayashi, Automorphism groups of compact Riemann surfaces of genera three and four, J. Pure Appl. Algebra 65 (1990), no. 3, 277–292. 10.1016/0022-4049(90)90107-SSuche in Google Scholar

[45] H. Lange and S. Recillas, Abelian varieties with group action, J. Reine Angew. Math. 575 (2004), 135–155. 10.1515/crll.2004.076Suche in Google Scholar

[46] C. Maclachlan, A bound for the number of automorphisms of a compact Riemann surface, J. Lond. Math. Soc. 44 (1969), 265–272. 10.1112/jlms/s1-44.1.265Suche in Google Scholar

[47] J. S. Milne, Jacobian varieties, Arithmetic Geometry, Springer, New York (1986), 167–212. 10.1007/978-1-4613-8655-1_7Suche in Google Scholar

[48] B. Moonen and F. Oort, The Torelli locus and special subvarieties, Handbook of Moduli. Vol. II, Adv. Lect. Math. (ALM) 25, International Press, Somerville (2013), 549–594. Suche in Google Scholar

[49] N. MĂŒller and R. Pink, Hyperelliptic curves with many automorphisms, preprint (2017), https://arxiv.org/abs/1711.06599. 10.1142/S1793042122500488Suche in Google Scholar

[50] A. Obus and T. Shaska, Superelliptic curves with many automorphisms and CM Jacobians, Math. Comp. 90 (2021), no. 332, 2951–2975. 10.1090/mcom/3639Suche in Google Scholar

[51] F. Oort, Singularities of coarse moduli schemes, SĂ©minaire d’AlgĂšbre Paul Dubreil, 29Ăšme annĂ©e (Paris 1975–1976), Lecture Notes in Math. 586, Springer, Berlin (1977), 61–76. 10.1007/BFb0087122Suche in Google Scholar

[52] J. Paulhus and A. M. Rojas, Completely decomposable Jacobian varieties in new genera, Exp. Math. 26 (2017), no. 4, 430–445. 10.1080/10586458.2016.1199981Suche in Google Scholar

[53] S. Recillas and R. E. RodrĂ­guez, Jacobians and representations of S 3 , Workshop on Abelian Varieties and Theta Functions (Morelia 1996), Aportaciones Mat. Investig. 13, Sociedad MatemĂĄtica Mexicana, MĂ©xico (1998), 117–140. Suche in Google Scholar

[54] S. Reyes-Carocca, Field of moduli of generalized Fermat curves, Q. J. Math. 63 (2012), no. 2, 467–475. 10.1093/qmath/haq049Suche in Google Scholar

[55] S. Reyes-Carocca, On the one-dimensional family of Riemann surfaces of genus 𝑞 with 4 ⁱ q automorphisms, J. Pure Appl. Algebra 223 (2019), no. 5, 2123–2144. 10.1016/j.jpaa.2018.07.011Suche in Google Scholar

[56] S. Reyes-Carocca, On Riemann surfaces of genus 𝑔 with 4 ⁱ g - 4 automorphisms, Israel J. Math. 237 (2020), no. 1, 415–436. 10.1007/s11856-020-2010-3Suche in Google Scholar

[57] S. Reyes-Carocca, On p ⁹ q -fold regular covers of the projective line, Rev. R. Acad. Cienc. Exactas Fís. Nat. Ser. A Mat. RACSAM 115 (2021), no. 1, Paper No. 23. 10.1007/s13398-020-00965-6Suche in Google Scholar

[58] S. Reyes-Carocca and R. E. Rodríguez, A generalisation of Kani-Rosen decomposition theorem for Jacobian varieties, Ann. Sc. Norm. Super. Pisa Cl. Sci. (5) 19 (2019), no. 2, 705–722. 10.2422/2036-2145.201706_003Suche in Google Scholar

[59] S. Reyes-Carocca and R. E. Rodríguez, On Jacobians with group action and coverings, Math. Z. 294 (2020), no. 1–2, 209–227. 10.1007/s00209-019-02263-3Suche in Google Scholar

[60] G. Riera and R. E. Rodríguez, The period matrix of Bring’s curve, Pacific J. Math. 154 (1992), no. 1, 179–200. 10.2140/pjm.1992.154.179Suche in Google Scholar

[61] J. Ries, The Prym variety for a cyclic unramified cover of a hyperelliptic Riemann surface, J. Reine Angew. Math. 340 (1983), 59–69. 10.1515/crll.1983.340.59Suche in Google Scholar

[62] A. M. Rojas, Group actions on Jacobian varieties, Rev. Mat. Iberoam. 23 (2007), no. 2, 397–420. 10.4171/RMI/500Suche in Google Scholar

[63] T. Sekiguchi, On the fields of rationality for curves and for their Jacobian varieties, Nagoya Math. J. 88 (1982), 197–212. 10.1017/S0027763000020171Suche in Google Scholar

[64] J.-P. Serre, Linear Representations of Finite Groups, Grad. Texts in Math. 42, Springer, New York, 1977. 10.1007/978-1-4684-9458-7Suche in Google Scholar

[65] D. Singerman, Subgroups of Fuschian groups and finite permutation groups, Bull. Lond. Math. Soc. 2 (1970), 319–323. 10.1112/blms/2.3.319Suche in Google Scholar

[66] D. Singerman, Finitely maximal Fuchsian groups, J. Lond. Math. Soc. (2) 6 (1972), 29–38. 10.1112/jlms/s2-6.1.29Suche in Google Scholar

[67] M. Streit, Period matrices and representation theory, Abh. Math. Sem. Univ. Hambg. 71 (2001), 279–290. 10.1007/BF02941477Suche in Google Scholar

[68] G. UrzĂșa, Riemann surfaces of genus 𝑔 with an automorphism of order 𝑝 prime and p > g , Manuscripta Math. 121 (2006), no. 2, 169–189. 10.1007/s00229-006-0034-6Suche in Google Scholar

[69] A. Weil, The field of definition of a variety, Amer. J. Math. 78 (1956), 509–524. 10.1007/978-1-4757-1705-1_78Suche in Google Scholar

[70] J. Wolfart, Regular dessins, endomorphisms of Jacobians, and transcendence, A Panorama of Number Theory or the View from Baker’s Garden (ZĂŒrich 1999), Cambridge University, Cambridge (2002), 107–120. 10.1017/CBO9780511542961.008Suche in Google Scholar

[71] J. Wolfart, A ⁱ B ⁱ C for polynomials, dessins d’enfants and uniformization—a survey, Elementare und Analytische Zahlentheorie, Schr. Wiss. Ges. Johann Wolfgang Goethe Univ. Frankfurt am Main 20, Franz Steiner, Stuttgart (2006), 313–345. Suche in Google Scholar

[72] A. Wootton, Defining equations for cyclic prime covers of the Riemann sphere, Israel J. Math. 157 (2007), 103–122. 10.1007/s11856-006-0004-4Suche in Google Scholar

[73] SageMath: The Sage Mathematics Software System (Version 9.0), The Sage Developers, 2019, https://www.sagemath.org. Suche in Google Scholar

Received: 2020-09-09
Revised: 2022-01-18
Published Online: 2022-03-31
Published in Print: 2022-09-01

© 2022 Walter de Gruyter GmbH, Berlin/Boston

Heruntergeladen am 4.11.2025 von https://www.degruyterbrill.com/document/doi/10.1515/jgth-2020-0140/html
Button zum nach oben scrollen