Topological and arithmetic characteristics about products of projective lines with complex tori 0 0 footnotetext: E-mail address: J.-L. Mo: mojiali0722@126.com; M. Amram: meiravt@sce.ac.il; C. Gong: cgong@suda.edu.cn Jia-Li Mo, Meirav Amram, Cheng Gong Abstract In this paper, we study non-planar degeneracies with cylindrical configurations. They could be constructed by the product ℂ ​ ℙ 1 × T \mathbb{CP}^{1}\times T of the projective plane and a complex torus with embedding ( m , n ) (m,n) . We prove that their fundamental groups of Galois covers have an abelian subgroup of rank m ⁡ ( 2 ​ n − 1 ) m(2n-1) respectively, and the irregularity of these surfaces are at least 2 ​ m ​ n − 1 2mn-1 . Furthermore, we also use Chern numbers to compute the index of such surfaces and classify them. Keywords Chern number, complex torus, fundamental groups, Galois cover. MR(2020) Subject Classification 05E16, 14J10, 14J80, 20F55. 1 Scientific background There has been extensive research about the topological and arithmetic properties of non-simply connected surfaces. In this paper, we investigate some of these properties of the surfaces. An effective approach to studying the topological properties of algebraic surfaces is the application of degeneration techniques, particularly degeneration of the surface into a union of planes. The earliest application of this method was shown in the study of Hirzebruch surfaces in [ 18 ] and [ 22 ] . As demonstrated in [ 8 ] , [ 9 ] , and [ 10 ] , this method has been successfully applied to surfaces of degrees 5, 6, and 8, revealing their topological properties. Moreover, in references [ 5 ] , [ 6 ] , and [ 7 ] , we further investigated the topological properties of the Galois cover of Zappatic surfaces. Our results are related to Chisini’s conjecture about algebraic surfaces and can be applied to the classification of algebraic surfaces. For details, please refer to [ 14 ] and [ 15 ] . A promising direction for further work is the exploration of non-planar degenerations. Indeed, we have initiated preliminary investigations into such degenerations in [ 1 ] . This paper focuses on cylindrical non-planar degenerations, illustrated in Figure . Figure 1: Cylindrical non-planar degeneration Specifically, we degenerate the smooth surface (Figure -(1)) into a cylindrical configuration: a union of multiple planes. In Figure -(2), each plane is represented by a triangle. The unfolding of the degenerate surface is shown in Figure . Figure 2: The degeneration of the surface We use the symbol X m , n X^{m,n} to represent a smooth surface that has degenerated to have a length of n n and a width of m m (i.e., comprising an m × n m\times n array of squares). For detailed information, please refer to Section . The Galois cover X g ​ a ​ l X_{gal} of a surface X X serves as a geometric manifestation of Galois theory and reflects the topological properties of the surface. In [ 20 ] , the author elucidates the arithmetic properties of Galois covers and provides computational formulas for certain topological invariants. Furthermore, Galois covers can be applied to the study of moduli spaces of surfaces; for instance, in [ 24 ] , the authors compute decompositions of these moduli spaces. Here, we aim to study the structure of fundamental group π 1 ​ ( X g ​ a ​ l ) \pi_{1}(X_{gal}) of the Galois cover of a surface. In this paper, as a first step towards computing π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) for surfaces X m , n X^{m,n} , here m ≥ 1 , n ≥ 2 m\geq 1,\ n\geq 2 , we provide a complete presentation of these fundamental groups for families X 2 , 4 X^{2,4} , X 2 , n X^{2,n} , and X 3 , n X^{3,n} . By applying techniques from [ 25 ] , it follows directly that π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) is a subgroup of a certain quotient of Coxeter groups. Furthermore, we employ Coxeter group theory to derive certain properties of π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) , with a particular focus on its subgroups and quotient structures. By analyzing the quotient group of π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) , we are able to estimate the irregularity of the corresponding Galois cover in this setting. The following theorem constitutes our main result. Theorem 1 . Group π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) contains a subgroup that is isomorphic to ℤ m ⁡ ( 2 ​ n − 1 ) \mathbb{Z}^{m(2n-1)} . Since a simply connected space has trivial fundamental group, Theorem immediately implies that Galois covers X g ​ a ​ l m , n X^{m,n}_{gal} cannot be simply connected. Therefore, we obtain the following corollary. Corollary 2 . The Galois covers of the surfaces X m , n X^{m,n} are non-simply connected. Building on the findings presented in Theorem , we can derive significant implications for the structure of the first homology group and the irregularity of the algebraic surface. This subgroup structure of the fundamental group indicates a rich topological complexity, facilitating a deeper understanding of the associated properties of algebraic surface. Corollary 3 . ( ( 1 ) ) The rank of the first homology group H 1 ​ ( X g ​ a ​ l m , n , ℤ ) H_{1}(X^{m,n}_{gal},\mathbb{Z}) is at least 2 ​ ( 2 ​ m ​ n − 1 ) 2(2mn-1) . ( ( 2 ) ) The irregularity of the algebraic surface X g ​ a ​ l m , n X^{m,n}_{gal} is at least 2 ​ m ​ n − 1 2mn-1 . In algebraic geometry, when classifying a compact complex surface X X , in addition to its own topological invariants (such as Betti numbers), the Chern numbers of its tangent bundle, c 1 2 ​ ( X ) c^{2}_{1}(X) and c 2 ​ ( X ) c_{2}(X) , are two of the most fundamental numerical invariants. The positivity and negativity of Chern numbers is closely related to the Kodaira dimension of the surface. In [ 12 ] , the authors discuss the geography problem of surfaces with given Chern numbers. The method of Galois covers can also be employed to provide many examples in the geography of surfaces. We apply the formulas from [ 22 ] to determine the Chern numbers of the Galois covers. Another important topological invariant of algebraic surfaces X X is the index of the intersection form on H 2 ​ ( X , ℚ ) H^{2}(X,\mathbb{Q}) , denoted by τ ⁡ ( X ) \tau(X) . Hirzebruch’s formula (see [ 17 ] ) says that τ (X)= 1 3 ( c_1^2(X)-2c_2(X) ). By analyzing the index, we can draw connections between the topological characteristics of these surfaces and their geometrical properties. We have the following theorem: Theorem 4 . ( ( 1 ) ) τ ⁡ ( X g ​ a ​ l m , n ) = 1 3 ​ ( 4 ​ m ) ! ⋅ ( 8 ​ m − 9 ) \tau(X^{m,n}_{gal})=\frac{1}{3}(4m)!\cdot(8m-9) , if n = 2 n=2 . In particular, when m = 1 m=1 , X g ​ a ​ l m , n X^{m,n}_{gal} are surfaces with a negative index. When m ≥ 2 m\geq 2 , X g ​ a ​ l m , n X^{m,n}_{gal} are surfaces with a positive index. ( ( 2 ) ) τ ⁡ ( X g ​ a ​ l m , n ) = 1 3 ​ ( 2 ​ m ​ n ) ! ⋅ ( m ​ n − 3 ​ n + 3 ) \tau(X^{m,n}_{gal})=\frac{1}{3}(2mn)!\cdot(mn-3n+3) , if n ≥ 3 n\geq 3 . In particular, when m = 1 m=1 , X g ​ a ​ l m , n X^{m,n}_{gal} are surfaces with a negative index. When m = 2 , n = 3 m=2,n=3 , X g ​ a ​ l m , n X^{m,n}_{gal} is an index 0 surface. Moreover, when n > 3 n>3 , such surfaces have a negative index. When m ≥ 3 m\geq 3 , X g ​ a ​ l m , n X^{m,n}_{gal} are surfaces with a positive index. The index is said to be positive, zero, or negative according to whether the signature of the intersection form on H 2 ​ ( X g ​ a ​ l m , n , ℚ ) H^{2}(X^{m,n}_{gal},\mathbb{Q}) is positive, zero, or negative. This classification is particularly helpful when constructing divisors with prescribed properties on the surface. The paper is organized as follows. In Section , we give the constructions of the surfaces. In Section , we explain the methods we use and the terminology related to the paper. In Section , we use the Coxeter group theory to prove Theorem and give some open questions. In Section , we focus on the irregularity of the surfaces. In Section , we compute the Chern numbers and topological indices of the surfaces. 2 Construction of X m , n X^{m,n} We take surface X = ℂ ​ ℙ 1 × T X=\mathbb{CP}^{1}\times T , where T T is an elliptic curve in ℂ ​ ℙ 2 \mathbb{CP}^{2} . We now describe the construction of surface X m , n X^{m,n} . Note that ℂ ​ ℙ 1 \mathbb{CP}^{1} can be embedded in ℂ ​ ℙ m ​ ( m ≥ 2 ) \mathbb{CP}^{m}\penalty\ (m\geq 2) and T T can be embedded in ℂ ​ ℙ n − 1 ​ ( n ≥ 3 ) \mathbb{CP}^{n-1}\penalty\ (n\geq 3) . Consequently, via Segre embedding ℂ ​ ℙ m × ℂ ​ ℙ n − 1 ↪ ℂ ​ ℙ m ⁡ ( n − 1 ) + m + ( n − 1 ) \mathbb{CP}^{m}\times\mathbb{CP}^{n-1}\hookrightarrow\mathbb{CP}^{m(n-1)+m+(n-1)} , the surface ℂ ​ ℙ 1 × T \mathbb{CP}^{1}\times T can be embedded into a larger projective space. We call such an embedding the ( m , n ) (m,n) -embedding of ℂ ​ ℙ 1 × T \mathbb{CP}^{1}\times T and denote the image of the embedding by X m , n X^{m,n} . To study fundamental group π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) , we employ a degeneration approach following [ 2 ] , combined with braid monodromy techniques [ 19 , 23 ] . This method allows us to degenerate the surface X m , n X^{m,n} into a union of 2 ​ m ​ n 2mn planes, as illustrated in Figure . We now explain the degeneration process in detail. Let T T be a normal elliptic curve in ℂ ​ ℙ n − 1 \mathbb{CP}^{n-1} . It can degenerate into an n n -degree stick curve that forms a closed loop, as shown in Figure . Figure 3: An elliptic normal curve On the other hand, the rational normal curve ℂ ​ ℙ 1 \mathbb{CP}^{1} in ℂ ​ ℙ m \mathbb{CP}^{m} can degenerate into an m m -degree stick curve that forms a path; see Figure . Figure 4: A rational normal curve The product surface ℂ ​ ℙ 1 × T \mathbb{CP}^{1}\times T consequently degenerates into m ​ n mn quadric surfaces. These are represented by the small squares in Figure . Figure 5: The degeneration of the surfaces X m , n X^{m,n} Each quadric surface further degenerates into two planes, resulting in a total of 2 ​ m ​ n 2mn planes. These are denoted by small triangles in Figure . This degenerate configuration corresponds to the surface X m , n X^{m,n} introduced in Section . 3 Preliminaries and terminology In this paper, we study the structure of the fundamental groups of the Galois covers of the surfaces X m , n X^{m,n} , building upon our previous studies of X 1 , 3 X^{1,3} and X 1 , n X^{1,n} in [ 3 ] and [ 4 ] , respectively. We denote the degenerations of X m , n X^{m,n} by X 0 m , n X_{0}^{m,n} , and these degenerate surfaces with as shown in Figure will be used throughout the paper for certain values of m m and n n . Figure 6: The degeneration of the surfaces X m , n X^{m,n} To label edges in Figure , we use x=mn; y=n(m-1); z=n(m-2). In this section, we now begin to present all relevant information and lemmas concerning the general surfaces, whose notation will be X m , n X^{m,n} for m ≥ 1 , n ≥ 2 m\geq 1,\ n\geq 2 , and whose degeneration appears above in Figure . We aim to determine the structure of fundamental group π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) . First, we consider a generic projection from the degenerated surface X 0 m , n X^{m,n}_{0} to ℂ ​ ℙ 2 \mathbb{CP}^{2} , obtaining branch curve C 0 m , n C^{m,n}_{0} , which is a union of 3 ​ y + 2 ​ n 3y+2n lines. These lines correspond to the images of the 3 ​ y + 2 ​ n 3y+2n edges in Figure , where the edges are numbered from left to right and from bottom to top, taking into account the identifications resulting from the degeneration of the torus. Through the regeneration process, we then obtain branch curve C m , n C^{m,n} , which arises from a generic projection of the surface X m , n X^{m,n} to ℂ ​ ℙ 2 \mathbb{CP}^{2} . These processes can be intuitively understood from Figure . X m , n ⊆ ℂ ​ ℙ N → degeneration X 0 m , n ⊆ ℂ ​ ℙ N generic projection ↓ ↓ generic projection C m , n ⊂ ℂ ​ ℙ 2 ← regeneration C 0 m , n ⊂ ℂ ​ ℙ 2 \begin{CD}X^{m,n}\subseteq\mathbb{CP}^{N}@>{\text{degeneration}}>{}>X^{m,n}_{0}\subseteq\mathbb{CP}^{N}\\ @V{\text{generic\penalty\ projection}}V{}V@V{}V{\text{generic\penalty\ projection}}V\\ C^{m,n}\subset\mathbb{CP}^{2}@<{\text{regeneration}}<{} > . From the relations computed for group G 1 m , n {G^{m,n}_{1}} , it is evident that it contains all the E 6 E_{6} basic cycles, but none of the cycles of m m types, as follows (see Figures and ): • The cycle ( H 1 m , n ) (H^{m,n}_{1}) generated by generators 1 , 2 , … , 2 ​ n 1,2,...,2n ; • The cycle ( H 2 m , n ) (H^{m,n}_{2}) generated by generators 3 ​ n + 1 , 3 ​ n + 2 , … , 5 ​ n 3n+1,3n+2,...,5n ; • ⋯ ⋯ ⋯ \cdots\cdots\cdots • The cycle ( H m − 1 m , n ) (H^{m,n}_{m-1}) generated by generators 3 ​ z + 1 , 3 ​ z + 2 , … , 3 ​ z + 2 ​ n 3z+1,3z+2,...,3z+2n ; • The cycle ( H m m , n ) (H^{m,n}_{m}) generated by generators 3 ​ y + 1 , 3 ​ y + 2 , … , 3 ​ y + 2 ​ n 3y+1,3y+2,...,3y+2n . Also, here the following exact sequence holds: 0 {\lx@inpgf@ignorespaces 0} π G m , n {\lx@inpgf@ignorespaces G^{m,n}} 0 {\lx@inpgf@ignorespaces 0} 0 {\lx@inpgf@ignorespaces 0} G 1 m , n {\lx@inpgf@ignorespaces G^{m,n}_{1}} 0 {\lx@inpgf@ignorespaces 0} Proposition 4.1 in [ 25 ] states that the kernel K 1 ​ ( T ) K_{1}(T) in the second exact sequence is generated as a subgroup by the elements γ j ​ γ i − 1 \gamma_{j}\gamma^{-1}_{i} . For instance, let us consider the cycle H 1 m , n H^{m,n}_{1} generated by generators 1 , 2 , … , 2 ​ n 1,2,\ldots,2n in the dual graph. Define γ _i = (i+2) ⋯2n ⋅1 ⋯i, for i=1, …, 2n-2, and set γ _2n-1= 1 ⋅2⋯(2n-1), γ _2n=2⋅3⋯(2n). Then γ j ​ γ 1 − 1 ∈ K 1 ​ ( T ) \gamma_{j}\gamma^{-1}_{1}\in K_{1}(T) for j = 2 , … , 2 ​ n \penalty\ j=2,\ldots,2n . Hence, we obtain 2 ​ n − 1 2n-1 distinct elements from cycle H 1 m , n H^{m,n}_{1} in K 1 ​ ( T ) K_{1}(T) . And that happens for each of cycles H i m , n H^{m,n}_{i} , i = 1 , … , m i=1,\ldots,m . Via the lift map from K 1 ​ ( T ) K_{1}(T) to π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) , it follows directly that π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) contains the elements: γ _j γ ^-1_1, j=2,…,2n. Since cycle H 1 m , n H^{m,n}_{1} yields ( 2 ​ n − 1 ) (2n-1) elements ( γ j ​ γ 1 − 1 \gamma_{j}\gamma^{-1}_{1} for j = 2 , … , 2 ​ n j=2,\ldots,2n ), we apply a similar method to show that, via the lift map, π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) contains ( 2 ​ n − 1 ) (2n-1) such elements for each cycle H i m , n H^{m,n}_{i} with i = 1 , 2 , … , m i=1,2,\ldots,m . In total, this gives m ⁡ ( 2 ​ n − 1 ) m(2n-1) elements. Now, we prove the above elements are of infinite order and commute with each other. The proof is divided into two steps. In the first step, we prove that the elements are torsion-free. As established previously, the group C Y ​ ( T ) C_{Y}(T) does not contain any of the E 6 E_{6} basic cycle relations that are present in G 1 m , n G^{m,n}_{1} . According to Theorem 6.1 in [ 25 ] , we have the following short exact sequence: 0 →K(T) →C_Y(T) →S_2mn →0, where K ⁡ ( T ) K(T) is isomorphic to A t , n A_{t,n} , with t t denoting the number of cycles in dual graph T T . The generators of K ⁡ ( T ) K(T) are elements of the form γ j ​ γ i − 1 \gamma_{j}\gamma^{-1}_{i} , within each cycle. Now consider group G 1 m , n G^{m,n}_{1} . According to Lemma , for any E 6 E_{6} basic cycle, if one relation of the form γ j 0 ​ γ i 0 − 1 \gamma_{j_{0}}\gamma^{-1}_{i_{0}} is imposed, then all other relations of the form γ j ​ γ i − 1 \gamma_{j}\gamma^{-1}_{i} within the same cycle are automatically satisfied. Consequently, G 1 m , n G^{m,n}_{1} is isomorphic to the quotient of C Y ​ ( T ) C_{Y}(T) by the normal subgroup generated by the elements γ j ​ γ i − 1 \gamma_{j}\gamma^{-1}_{i} from every E 6 E_{6} basic cycle. Denoting this normal subgroup by B B , we obtain the following short exact sequence: 0 →K(T)/(B ∩K(T)) →G^m,n_1 →S_2mn →0. The kernel K ⁡ ( T ) / ( B ∩ K ⁡ ( T ) ) K(T)/(B\cap K(T)) is generated by all elements γ j ​ γ i − 1 \gamma_{j}\gamma^{-1}_{i} corresponding to the cycle ( H i m , n ) , i = 1 , 2 , … , m − 1 , m (H^{m,n}_{i}),\ i=1,2,\ldots,m-1,m (the elements arising from the E 6 E_{6} basic cycle relations have been quotiented out). Suppose that there exists a torsion element in one of these generators. Consider the preimage of the above torsion element in K ⁡ ( T ) K(T) . We denote it by h h . This would imply that a power of this generator can be written as a product of other generators corresponding to the preimage of γ j ​ γ i − 1 ​ ( i = 1 , 2 , … , m − 1 , m ) \gamma_{j}\gamma^{-1}_{i}\ (i=1,2,\ldots,m-1,m) from the E 6 E_{6} basic cycle. Using [ 25 , Theorem 6.1] , we can write such relations as follows: (h)^n=∑_i,j a_ij γ _j γ ^-1_i. (*) However, since K ⁡ ( T ) ≅ A t , n K(T)\cong A_{t,n} , and by the definition of A t , n A_{t,n} , its generators satisfy only commutation relations of the form [ x i ​ j , y k ​ l ] = 1 [x_{ij},y_{kl}]=1 and no others, it follows that A t , n A_{t,n} admits no relations of form ( ∗ ) (*) . This leads to a contradiction. Therefore, the elements γ i ​ γ 1 − 1 ∈ K ⁡ ( T ) \gamma_{i}\gamma_{1}^{-1}\in K(T) for i = 2 , … , 2 ​ n i=2,\ldots,2n (studied in Section ) are torsion-free. In the second step, we prove that these elements commute pairwise. The commutativity of elements from different cycles ( H i m , n ) (H^{m,n}_{i}) can be verified directly by computation. The key part is to prove that within the same cycle ( H i m , n ) (H^{m,n}_{i}) , elements of form γ i ​ γ 1 − 1 \gamma_{i}\gamma^{-1}_{1} and γ j ​ γ 1 − 1 \gamma_{j}\gamma^{-1}_{1} commute. To show that γ i ​ γ 1 − 1 ​ γ j ​ γ 1 − 1 = γ j ​ γ 1 − 1 ​ γ i ​ γ 1 − 1 \gamma_{i}\gamma^{-1}_{1}\gamma_{j}\gamma^{-1}_{1}=\gamma_{j}\gamma^{-1}_{1}\gamma_{i}\gamma^{-1}_{1} , it suffices to prove the equivalent relation: γ i ​ γ 1 − 1 ​ γ j = γ j ​ γ 1 − 1 ​ γ i \gamma_{i}\gamma^{-1}_{1}\gamma_{j}=\gamma_{j}\gamma^{-1}_{1}\gamma_{i} , where γ i = ( i + 2 ) ⋅ ( i + 3 ) ⋯ ( 2 n − 1 ) ⋅ ( 2 n ) ⋅ 1 ⋅ 2 ⋯ i \gamma_{i}=(i+2)\cdot(i+3)\cdots(2n-1)\cdot(2n)\cdot 1\cdot 2\cdots i , here i = 1 , 2 , … , 2 ​ n − 2 i=1,2,\dots,2n-2 , and γ 2 ​ n − 1 = 1 ⋅ 2 ⋯ ( 2 n − 2 ) ⋅ ( 2 n − 1 ) \gamma_{2n-1}=1\cdot 2\cdots(2n-2)\cdot(2n-1) , γ 2 ​ n = 2 ⋅ 3 ⋯ ( 2 n − 1 ) ⋅ ( 2 n ) \gamma_{2n}=2\cdot 3\cdots(2n-1)\cdot(2n) . Here we only prove that the case γ 2 ​ n − 1 ​ γ 1 − 1 ​ γ 2 ​ n = γ 2 ​ n ​ γ 1 − 1 ​ γ 2 ​ n − 1 \gamma_{2n-1}\gamma^{-1}_{1}\gamma_{2n}=\gamma_{2n}\gamma^{-1}_{1}\gamma_{2n-1} (37) holds for the cycle ( H 1 m , n ) (H^{m,n}_{1}) (the other cases have similar proofs). Since i = i − 1 i=i^{-1} for any generator i i in the group G 1 m , n G^{m,n}_{1} , we have γ 1 − 1 = 1 − 1 ⋅ ( 2 n ) − 1 ⋅ ( 2 n − 1 ) − 1 ⋯ 4 − 1 ⋅ 3 − 1 = 1 ⋅ ( 2 n ) ⋅ ( 2 n − 1 ) ⋯ 4 ⋅ 3 \gamma^{-1}_{1}=1^{-1}\cdot(2n)^{-1}\cdot(2n-1)^{-1}\cdots 4^{-1}\cdot 3^{-1}=1\cdot(2n)\cdot(2n-1)\cdots 4\cdot 3 . Now we expand both sides of relation ( ) by substituting the above expressions. LHS (resp. RHS) is the left (resp. right) side of ( ). We get the following equality: LHS \displaystyle\text{LHS} = 1 ⋅ 2 ⋯ ( 2 n − 2 ) ⋅ ( 2 n − 1 ) ⋅ 1 ⋅ ( 2 n ) ⋅ ( 2 n − 1 ) ⋯ 4 ⋅ 3 ⋅ 2 ⋅ 3 ⋯ ( 2 n − 1 ) ⋅ ( 2 n ) \displaystyle=1\cdot 2\cdots(2n-2)\cdot(2n-1)\cdot 1\cdot(2n)\cdot(2n-1)\cdots 4\cdot 3\cdot 2\cdot 3\cdots(2n-1)\cdot(2n) (38) = 2 ⋅ 3 ⋯ ( 2 n − 1 ) ⋅ ( 2 n ) ⋅ 1 ⋅ ( 2 n ) ⋅ ( 2 n − 1 ) ⋯ 4 ⋅ 3 ⋅ 1 ⋅ 2 ⋯ ( 2 n − 2 ) ⋅ ( 2 n − 1 ) = RHS . \displaystyle=2\cdot 3\cdots(2n-1)\cdot(2n)\cdot 1\cdot(2n)\cdot(2n-1)\cdots 4\cdot 3\cdot 1\cdot 2\cdots(2n-2)\cdot(2n-1)=\text{RHS}. We simplify ( ) by using relations [ 1 , i ] = e , i = 3 , 4 , … , ( 2 n − 1 ) [1,i]=e,\penalty\ i=3,\penalty\ 4,\penalty\ \dots,\penalty\ (2n-1) and ⟨ 1 , 2 ⟩ = e = ⟨ 1 , 2 ​ n ⟩ \langle 1,\penalty\ 2\rangle=e=\langle 1,\penalty\ 2n\rangle . We get the following equality: LHS \displaystyle\text{LHS} = 2 ⋅ 3 ⋯ ( 2 n − 2 ) ⋅ ( 2 n − 1 ) ⋅ ( 2 n ) ⋅ ( 2 n − 1 ) ⋯ 4 ⋅ 3 ⋅ 2 ⋅ 3 ⋯ ( 2 n − 1 ) ⋅ ( 2 n ) \displaystyle=2\cdot 3\cdots(2n-2)\cdot(2n-1)\cdot(2n)\cdot(2n-1)\cdots 4\cdot 3\cdot 2\cdot 3\cdots(2n-1)\cdot(2n) (39) = 3 ⋅ 4 ⋯ ( 2 n − 2 ) ⋅ ( 2 n − 1 ) ⋅ ( 2 n ) ⋅ ( 2 n − 1 ) ⋯ 4 ⋅ 3 ⋅ 2 ⋅ 3 ⋯ ( 2 n − 2 ) ⋅ ( 2 n − 1 ) = RHS . \displaystyle=3\cdot 4\cdots(2n-2)\cdot(2n-1)\cdot(2n)\cdot(2n-1)\cdots 4\cdot 3\cdot 2\cdot 3\cdots(2n-2)\cdot(2n-1)=\text{RHS}. Then we use relations [ 2 n , i ] = e , i = 2 , 3 , … , ( 2 n − 2 ) [2n,i]=e,\penalty\ i=2,\penalty\ 3,\penalty\ \dots,\penalty\ (2n-2) and ⟨ 2 ​ n − 1 , 2 ​ n ⟩ = e \langle 2n-1,\penalty\ 2n\rangle=e , and get LHS \displaystyle\text{LHS} = 2 ⋅ 3 ⋯ ( 2 n − 2 ) ⋅ ( 2 n − 1 ) ⋅ ( 2 n − 2 ) ⋅ ( 2 n ) ⋅ ( 2 n − 3 ) ⋯ 4 ⋅ 3 ⋅ 2 ⋅ 3 ⋯ ( 2 n − 1 ) ⋅ ( 2 n ) \displaystyle=2\cdot 3\cdots(2n-2)\cdot(2n-1)\cdot(2n-2)\cdot(2n)\cdot(2n-3)\cdots 4\cdot 3\cdot 2\cdot 3\cdots(2n-1)\cdot(2n) (40) = 3 ⋅ 4 ⋯ ( 2 n − 2 ) ⋅ ( 2 n − 1 ) ⋅ ( 2 n − 2 ) ⋅ ( 2 n ) ⋅ ( 2 n − 3 ) ⋯ 4 ⋅ 3 ⋅ 2 ⋅ 3 ⋯ ( 2 n − 2 ) ⋅ ( 2 n − 1 ) \displaystyle=3\cdot 4\cdots(2n-2)\cdot(2n-1)\cdot(2n-2)\cdot(2n)\cdot(2n-3)\cdots 4\cdot 3\cdot 2\cdot 3\cdots(2n-2)\cdot(2n-1) = RHS . \displaystyle=\text{RHS}. Now we use relations [ 2 n − 1 , i ] = e , i = 2 , 3 , … , ( 2 n − 3 ) [2n-1,i]=e,\penalty\ i=2,\penalty\ 3,\penalty\ \dots,\penalty\ (2n-3) and ⟨ 2 ​ n − 2 , 2 ​ n − 1 ⟩ = e \langle 2n-2,\penalty\ 2n-1\rangle=e to get LHS \displaystyle\text{LHS} = 2 ⋅ 3 ⋯ ( 2 n − 3 ) ⋅ ( 2 n − 2 ) ⋅ ( 2 n − 3 ) ⋅ ( 2 n − 1 ) ⋅ ( 2 n ) ⋅ ( 2 n − 4 ) ⋯ 4 ⋅ 3 ⋅ 2 ⋅ 3 ⋯ ( 2 n − 1 ) \displaystyle=2\cdot 3\cdots(2n-3)\cdot(2n-2)\cdot(2n-3)\cdot(2n-1)\cdot(2n)\cdot(2n-4)\cdots 4\cdot 3\cdot 2\cdot 3\cdots(2n-1) (41) ⋅ ( 2 n ) = 3 ⋅ 4 ⋯ ( 2 n − 3 ) ⋅ ( 2 n − 2 ) ⋅ ( 2 n − 3 ) ⋅ ( 2 n − 1 ) ⋅ ( 2 n ) ⋅ ( 2 n − 4 ) ⋯ 4 ⋅ 3 ⋅ 2 ⋅ 3 \displaystyle\cdot(2n)=3\cdot 4\cdots(2n-3)\cdot(2n-2)\cdot(2n-3)\cdot(2n-1)\cdot(2n)\cdot(2n-4)\cdots 4\cdot 3\cdot 2\cdot 3 ⋯ ( 2 n − 2 ) ⋅ ( 2 n − 1 ) = RHS . \displaystyle\cdots(2n-2)\cdot(2n-1)=\text{RHS}. The simplification proceeds by using [ 2 ​ n − 2 , i ] = e [2n-2,i]=e , i = 2 , 3 , … , ( 2 ​ n − 4 ) i=2,3,\dots,(2n-4) , and ⟨ 2 ​ n − 3 , 2 ​ n − 2 ⟩ = e \langle 2n-3,\penalty\ 2n-2\rangle=e . Ultimately, we obtain the relation: ⟨2n, 2n-1⟩= e. It is clear that this relation already exists, hence the equivalent relation ( ) holds. A similar argument applies to any other pair of commuting elements. Therefore, the m ⁡ ( 2 ​ n − 1 ) m(2n-1) elements generate a subgroup isomorphic to ℤ m ⁡ ( 2 ​ n − 1 ) \mathbb{Z}^{m(2n-1)} , which completes the proof of the theorem. ∎ 4.4 Open questions Computing the fundamental group π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) for a general surface X m , n X^{m,n} is considerably more complex than for the special cases previously studied. We conclude by posing the following interesting open questions: Question 1 . Is π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) an abelian group? Question 2 . Can one explicitly express π 1 ​ ( X g ​ a ​ l m , n ) \pi_{1}(X^{m,n}_{gal}) as a direct or semidirect product of simpler groups? Answers to these questions will significantly deepen our understanding of the moduli space of algebraic surfaces. 5 Estimation of the irregularity of the Galois covers In this section, we estimate the irregularity of the Galois covers using the method of Liedtke [ 20 ] . Let f : X ⟶ ℂ ​ ℙ 2 f:X\longrightarrow\mathbb{CP}^{2} be a generic projection of degree l l , with branch curve S ⊂ ℂ ​ ℙ 2 . S\subset\mathbb{CP}^{2}. Consider an affine chart ℂ ​ 𝔸 2 \mathbb{CA}^{2} of ℂ ​ ℙ 2 \mathbb{CP}^{2} , and let D := S ∩ ℂ ​ 𝔸 2 D:=S\cap\mathbb{CA}^{2} be the affine portion of the branch curve. From Lemma 3.2 of [ 19 ] there is a short exact sequence 1 → π _1(X^aff_gal) → π _1( CA ^2 - D, x_0)/⟨Γ_i^2 ⟩ ψ → S_l →1. Here Γ i {\Gamma_{i}} are the generators of π 1 ​ ( ℂ ​ 𝔸 2 − D , x 0 ) \pi_{1}(\mathbb{CA}^{2}-D,x_{0}) , and X gal aff X^{\mathrm{aff}}_{\mathrm{gal}} denotes an affine chart of the Galois cover X gal X_{\mathrm{gal}} for a general projection f f . We define now C p ​ r ​ o ​ j ¯ \overline{C_{proj}} . Definition 15 . Let C aff C_{\text{aff}} be the subgroup normally generated by the following elements inside π 1 ​ ( ℂ ​ 𝔸 2 − D , x 0 ) \pi_{1}(\mathbb{CA}^{2}-D,x_{0}) : • [ γ ​ Γ i ​ γ , Γ j − 1 ] [\gamma\Gamma_{i}\gamma,\Gamma^{-1}_{j}] if ψ ⁡ ( γ ​ Γ j ​ γ − 1 ) \psi(\gamma\Gamma_{j}\gamma^{-1}) and ψ ⁡ ( Γ i ) \psi(\Gamma_{i}) are disjoint transpositions, • < γ ​ Γ i ​ γ , Γ j − 1 > <\gamma\Gamma_{i}\gamma,{\Gamma^{-1}_{j}}> if ψ ⁡ ( γ ​ Γ i ​ γ − 1 ) \psi(\gamma\Gamma_{i}\gamma^{-1}) and ψ ⁡ ( Γ j ) \psi(\Gamma_{j}) have precisely one letter in common, where γ \gamma runs through π 1 ​ ( ℂ ​ 𝔸 2 − D , x 0 ) \pi_{1}(\mathbb{CA}^{2}-D,x_{0}) . We define the image of C aff C_{\text{aff}} in π 1 ​ ( X g ​ a ​ l ) \pi_{1}(X_{gal}) by C p ​ r ​ o ​ j C_{proj} . Let C p ​ r ​ o ​ j ¯ \overline{C_{proj}} be the image of C p ​ r ​ o ​ j C_{proj} in the abelianization of π 1 ​ ( X g ​ a ​ l ) \pi_{1}(X_{gal}) . Our main tool is the following result (see Corollary 4.9 of [ 19 ] ): Corollary 16 . The rank of H 1 ​ ( X g ​ a ​ l , ℤ ) / C p ​ r ​ o ​ j ¯ H_{1}(X_{gal},\mathbb{Z})/\overline{C_{proj}} as an abelian group is equal to ( l − 1 ) (l-1) times the rank of H 1 ​ ( X , ℤ ) H_{1}(X,\mathbb{Z}) , where H 1 ​ ( X g ​ a ​ l , ℤ ) H_{1}(X_{gal},\mathbb{Z}) and H 1 ​ ( X , ℤ ) H_{1}(X,\mathbb{Z}) are the first homology groups of X g ​ a ​ l X_{gal} and X X , respectively, and C p ​ r ​ o ​ j ¯ \overline{C_{proj}} is a subgroup of H 1 ​ ( X g ​ a ​ l , ℤ ) H_{1}(X_{gal},\mathbb{Z}) . In the context of this paper, it can be calculated that the degree of the general cover is 2 ​ m ​ n 2mn . Since X m , n X^{m,n} is an elliptic surface, the rank of H 1 ​ ( X m , n , ℤ ) H_{1}(X^{m,n},\mathbb{Z}) is 2 2 . Therefore, by Corollary , the rank of H 1 ​ ( X g ​ a ​ l m , n , ℤ ) / C p ​ r ​ o ​ j ¯ H_{1}(X^{m,n}_{gal},\mathbb{Z})/\overline{C_{proj}} is ( 2 ​ m ​ n − 1 ) × 2 = 4 ​ m ​ n − 2 (2mn-1)\times 2=4mn-2 . Consequently, the rank of H 1 ​ ( X g ​ a ​ l m , n , ℤ ) H_{1}(X^{m,n}_{gal},\mathbb{Z}) itself is at least 4 ​ m ​ n − 2 4mn-2 . According to the Hodge decomposition (see [ 16 ] ), we have rank ⁡ ( H 1 ​ ( X g ​ a ​ l m , n , ℤ ) ) = 2 ​ q ​ ( X g ​ a ​ l m , n ) \mathrm{rank}\penalty\ (H_{1}(X^{m,n}_{gal},\mathbb{Z}))=2q(X^{m,n}_{gal}) , where q ⁡ ( X g ​ a ​ l m , n ) q(X^{m,n}_{gal}) is the regularity of the Galois cover, and this implies that the irregularity satisfies q ⁡ ( X g ​ a ​ l m , n ) ≥ 2 ​ m ​ n − 1 q(X^{m,n}_{gal})\geq 2mn-1 . This lower bound is in fact sharp. To see this, consider the surfaces X 1 , n X^{1,n} studied in [ 3 ] . These surfaces are constructed as Galois covers of generic projections from the product surface E × ℙ 1 E\times\mathbb{P}^{1} , where E E is an elliptic curve. Let F F be a very ample line bundle of degree n n on E E . The line bundle L := F ⊠ O ℙ 1 ​ ( d ) L:=F\boxtimes O_{\mathbb{P}^{1}}(d) on X 1 , n X^{1,n} is very ample. If we choose n n sufficiently large and d = 1 d=1 , then L L gives rise to a generic projection f : X 1 , n → ℙ 2 f:X^{1,n}\rightarrow\mathbb{P}^{2} . The degree n = deg ​ f n=\mathrm{deg}f equals the self-intersection of L L , which is equal to 2 ​ n 2n . For this surface, it is shown that π 1 ​ ( X g ​ a ​ l 1 , n ) ≅ ℤ 4 ​ n − 2 \pi_{1}(X^{1,n}_{gal})\cong\mathbb{Z}^{4n-2} . Since the first homology group is the abelianization of the fundamental group, it follows that H 1 ​ ( X g ​ a ​ l 1 , n , ℤ ) ≅ ℤ 4 ​ n − 2 H_{1}(X^{1,n}_{gal},\mathbb{Z})\cong\mathbb{Z}^{4n-2} , which indicates that q ⁡ ( X g ​ a ​ l 1 , n ) = 4 ​ n − 2 2 = 2 ​ n − 1 q(X^{1,n}_{gal})=\frac{4n-2}{2}=2n-1 . 6 Chern number and index We study the surface X m , n X^{m,n} , which has a degeneration shown in Figure . In this section, we compute the Chern numbers of the Galois cover of this surface using the following theorem: Theorem 17 . [ 22 , Proposition 0.2] Let S S be the branch curve of an algebraic surface X X . Denote the degree of the generic projection by b b , and f : X → ℂ ​ ℙ 2 f:X\rightarrow\mathbb{CP}^{2} is the generic projection. The Chern numbers of the Galois covers of the surfaces are given in the following formulas: ( ( 1 ) ) c 1 2 ​ ( X g ​ a ​ l ) = b ! 4 ​ ( h − 6 ) 2 , c_{1}^{2}(X_{gal})=\frac{b!}{4}(h-6)^{2}, ( ( 2 ) ) c 2 ​ ( X g ​ a ​ l ) = b ! ​ ( h 2 2 − 3 ​ h 2 + 3 − 3 ​ d 4 − 4 ​ ρ 3 ) , c_{2}(X_{gal})=b!(\frac{h^{2}}{2}-\frac{3h}{2}+3-\frac{3d}{4}-\frac{4\rho}{3}), where h = deg ⁡ S h=\deg S , d = d= number of nodes in S S , ρ = \rho= number of cusps in S S . Figure , above, shows the degeneration of the surfaces X m , n X^{m,n} . From this configuration, we observe that all relations from nodes and cusps originate from vertices of two types (intersections of either two or six lines), and relations of type . The number of vertices where two lines intersect is 2 ​ n 2n . Each of the vertices V 2 , … , V n V_{2},\dots,V_{n} and V n + 1 , V n + 2 , … , V 2 ​ n V_{n+1},V_{n+2},\dots,V_{2n} contributes three cusps. Therefore, the total number of cusps arising from these vertices is 6 ​ n 6n . The number of vertices where six lines intersect is n ⁡ ( m − 1 ) n(m-1) . According to [ 11 ] , each such vertex contributes 24 nodes and 24 cusps. Therefore, the total number of nodes (respectively, cusps) arising from these vertices is 24 ​ n ​ ( m − 1 ) 24n(m-1) . Relations of type , introduced in the preliminaries, also contribute commutation relations. According to Figure , the number of such commutation relations is given by the formulas • 4 ⋅ ( C 3 ​ m ​ n − n 2 − 2 ​ n − 15 ​ ( m − 1 ) ​ n + ( m − 1 ) ) 4\cdot(C^{2}_{3mn-n}-2n-15(m-1)n+(m-1)) , if n = 2 n=2 , • 4 ⋅ ( C 3 ​ m ​ n − n 2 − 2 ​ n − 15 ​ ( m − 1 ) ​ n ) 4\cdot(C^{2}_{3mn-n}-2n-15(m-1)n) , if n ≥ 3 n\geq 3 , where C n r = n ! r ! ​ ( n − r ) ! C^{r}_{n}=\frac{n!}{r!\ (n-r)!} . In conclusion, we have b = 2 ​ m ​ n , h = 6 ​ m ​ n − 2 ​ n , ρ = 6 ​ n + 24 ​ n ⋅ ( m − 1 ) = 24 ​ m ​ n − 18 ​ n b=2mn,\ h=6mn-2n,\ \rho=6n+24\penalty\ n\cdot(m-1)=24mn-18n , and the following formulas: • when n = 2 n=2 , we have d = 24 ​ n ⋅ ( m − 1 ) + 4 ⋅ ( C 3 ​ m ​ n − n 2 − 2 ​ n − 15 ​ ( m − 1 ) ​ n + ( m − 1 ) ) = 72 ​ m 2 − 128 ​ m + 64 , d=24\penalty\ n\cdot(m-1)+4\cdot(C^{2}_{3mn-n}-2n-15(m-1)n+(m-1))=72m^{2}-128m+64, • and when n ≥ 3 n\geq 3 , we have d = 24 ​ n ⋅ ( m − 1 ) + 4 ⋅ ( C 3 ​ m ​ n − n 2 − 2 ​ n − 15 ​ ( m − 1 ) ​ n ) = 18 ​ m 2 ​ n 2 − 12 ​ m ​ n 2 − 42 ​ m ​ n + 2 ​ n 2 + 30 ​ n . d=24\penalty\ n\cdot(m-1)+4\cdot(C^{2}_{3mn-n}-2n-15(m-1)n)=18m^{2}n^{2}-12mn^{2}-42mn+2n^{2}+30n. Furthermore, in the degeneration of X m , n X^{m,n} , we have deg ⁡ S = 6 ​ m ​ n − 2 ​ n \deg S=6mn-2n . The computation of Chern numbers is a crucial step in studying the moduli spaces of algebraic surfaces, see [ 13 ] . By the formula in Theorem , we get the following Chern numbers: • when n = 2 n=2 , we have c 1 2 ​ ( X g ​ a ​ l m , n ) = ( 4 ​ m ) ! ​ ( 36 ​ m 2 − 60 ​ m + 25 ) c_{1}^{2}(X^{m,n}_{gal})=(4m)!\penalty\ (36m^{2}-60m+25) c 2 ​ ( X g ​ a ​ l m , n ) = ( 4 ​ m ) ! ​ ( 18 ​ m 2 − 34 ​ m + 17 ) , c_{2}(X^{m,n}_{gal})=(4m)!\penalty\ (18m^{2}-34m+17), • and when n ≥ 3 n\geq 3 , we have c 1 2 ​ ( X g ​ a ​ l m , n ) = ( 2 ​ m ​ n ) ! ​ ( 9 ​ m 2 ⋅ n 2 − 6 ​ m ​ n 2 − 18 ​ m ​ n + n 2 + 6 ​ n + 9 ) c_{1}^{2}(X^{m,n}_{gal})=(2mn)!\penalty\ (9m^{2}\cdot n^{2}-6mn^{2}-18mn+n^{2}+6n+9) c 2 ​ ( X g ​ a ​ l m , n ) = ( 2 ​ m ​ n ) ! ​ ( 9 / 2 ⋅ m 2 ​ n 2 − 3 ​ m ​ n 2 − 19 / 2 ⋅ m ​ n + 1 / 2 ⋅ n 2 + 9 / 2 ⋅ n + 3 ) . c_{2}(X^{m,n}_{gal})=(2mn)!\penalty\ (9/2\cdot m^{2}n^{2}-3mn^{2}-19/2\cdot mn+1/2\cdot n^{2}+9/2\cdot n+3). In the geography problem of surfaces, the index τ ⁡ ( X g ​ a ​ l ) = 1 3 ​ ( c 1 2 ​ ( X g ​ a ​ l ) − 2 ​ c 2 ​ ( X g ​ a ​ l ) ) \tau(X_{gal})=\frac{1}{3}\left(c_{1}^{2}(X_{gal})-2c_{2}(X_{gal})\right) can concisely reflect the topological properties of the surface. Moreover, the index of surfaces, in conjunction with the Seiberg–Witten theory and the Dirac operator index, constrains the existence of metrics with positive scalar curvature (see [ 12 ] ). Based on the results regarding the above calculations of the Chern numbers, we have: • τ ⁡ ( X g ​ a ​ l m , n ) = ( 4 ​ m ) ! / 3 ⋅ ( 8 ​ m − 9 ) \tau(X^{m,n}_{gal})=(4m)!/3\penalty\ \cdot(8m-9) , if n = 2 n=2 . • τ ⁡ ( X g ​ a ​ l m , n ) = ( 2 ​ m ​ n ) ! / 3 ⋅ ( m ​ n − 3 ​ n + 3 ) \tau(X^{m,n}_{gal})=(2mn)!/3\penalty\ \cdot(mn-3n+3) , if n ≥ 3 n\geq 3 . Then we get Theorem , which provides an exact expression for the τ \tau -invariant of X g ​ a ​ l m , n X^{m,n}_{gal} for different parameters, along with a detailed analysis of the sign of its index. The result unifies both the cases n = 2 n=2 and n ≥ 3 n\geq 3 , and reveals the significant influence of the parameters m m and n n on the geometric properties of the surface: as the parameters vary, the index of the surface is shown to transition from negative through zero to positive values, reflecting a profound relationship between its geometric structure and topological complexity. The proof of the theorem thus provides a pivotal computational foundation and theoretical reference for further study of the invariants and classification of such algebraic varieties. Acknowledgments: We express our gratitude to Prof. Xianhua Li for discussing the construction of the subgroups. This research was partly supported by NSFC (No.12331001, 12271384), Natural Science Research Project of Anhui Educational Committee (No.2025AHGXZK40337), Anhui Science and Technology University 2025 (Stable) Talent Recruitment Program (No.2025qhxm28), DFG Grant (No.537912683), and the Key Discipline Construction Project of Anhui Science and Technology University (No.XK-XJGY002).