Lattice 3 3 -polytopes of lattice width 2 2 and corresponding toric hypersurfaces

Martin Bohnert · 2024

Read Lattice 3 3 -polytopes of lattice width 2 2 and corresponding toric hypersurfaces by Martin Bohnert (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-17545.txt.

Introduction

Martin Bohnert classifies three-dimensional lattice polytopes of lattice width two and studies the toric surfaces they define. The paper from twenty twenty-four leans on recent work by Victor Batyrev. A hypersurface's Kodaira dimension can be read off its Newton polytope's Fine interior. That interior keeps points at integral distance at least one from all bounding planes.

When the Fine interior drops one dimension, the surface is of general type and the polytope has width two. Bohnert sorts all two-dimensional Fine interiors among polytopes with at most a set count of interior points. Each case yields a surface of general type with bounded genus. The census gives builders a stock of worked examples. It ties polytope shape, a discrete trait, to surface class, a deep geometric one.

Catalog record

CREATORMartin Bohnert
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE39 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-17545

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Lattice 3 3 -polytopes of lattice width 2 2 and corresponding toric hypersurfaces Martin Bohnert Abstract . The Kodaira dimension of a nondegenerate toric hypersurface can be computed from the dimension of the Fine interior of its Newton polytope according to recent work of Victor Batyrev, where the Fine interior of the Newton polytope is the subpolytope consisting of all points which have an integral distance of at least 1 1 to all integral supporting hyperplanes. In particular, if we have a Fine interior of codimension 1 1 , then the hypersurface is of general type and the Newton polytope has lattice width 2 2 . In this article we study this situation for lattice 3 3 -polytopes and the corresponding surfaces of general type. In particular, we classify all 2 2 -dimensional Fine interiors of those lattice 3 3 -polytopes which have at most 40 40 interior lattice points, thus obtaining many examples of surfaces of general type and genus at most 40. 1. Introduction Let M\cong\mathds{Z}^{d} be a lattice of rank d\in\mathds{Z}_{\geq 1} and N=\mathrm{Hom}(M\mathds{Z})\cong\mathds{Z}^{d} the dual lattice. We have the natural pairing \left<\cdot,\cdot\right>\colon M\times N\to\mathds{Z} , which extends to the real vector spaces M_{\mathds{R}}=M\otimes\mathds{R} and N_{\mathds{R}}=N\otimes\mathds{R} . For a rational d -polytope P in M_{\mathds{R}} , i.e. P\subseteq M_{\mathds{R}} is the convex hull of a finite number of points of M_{\mathds{Q}}=M\otimes\mathds{Q} and P is of full dimension, we have the Fine interior of P defined as \displaystyle F(P):=\{x\in M_{\mathds{R}}\mid\left<x,n\right>\geq\mathrm{Min}_{P}(n)+1\ \forall n\in N\setminus\{0\}\}\subseteq P with \displaystyle\mathrm{Min}_{P}(n):=\min\{\left<x,n\right>\mid x\in P\}. The Fine interior is itself a rational polytope, and if P is the Newton polytope of a nondegenerate toric hypersurface Z and has a non-empty Fine interior, then recent work by Victor Batyrev [ Bat23 ] shows that the Fine interior and additional combinatorial data can be used to construct a minimal model \hat{Z} of the hypersurface. Moreover, we can compute some important invariants of the minimal model directly from the Fine interior. The Kodaira dimension is given by [ Bat23 , 9.2] as \displaystyle\kappa(\hat{Z})=\min\{\dim F(P),d-1\} and, for example, for the top intersection number in the case \dim F(P)=d-1 we have the formula ( [ Gie22 , 5.1] , [ Bat23 , 9.4] ) \displaystyle(K_{\hat{Z}})^{d-1}=2\mathrm{Vol}_{d-1}(F(P)), where \mathrm{Vol}_{d-1} means that we are looking at the normalized volume in the affine subspace generated by F(P) , where normalized means that we have volume 1 for a simplex whose vertices form an affine lattice basis of the sublattice in the subspace. We focus in this article on the case d=3 with \dim F(P)=2 , so our nondegenerate toric hypersurface becomes a surface of general type. This implies that our Newton polytope P\subseteq M_{\mathds{R}} is a lattice 3 -polytope with lattice width 2 , where we should remember that the lattice width \mathrm{lw}(P) is given by \displaystyle\mathrm{lw}(P):=\min\{-\mathrm{Min}_{P}(-n)-\mathrm{Min}_{P}(n)\mid n\in N\setminus\{0\}. and we call n_{lw}\in N a lattice width direction if \mathrm{lw}(P)=-\mathrm{Min}_{P}(-n_{lw})-\mathrm{Min}_{P}(n_{lw}) . Why is the combinatorics in this situation easy enough to handle? As a first step, we will see in Theorem 2.5 in section 2 , that the Fine interior of a lattice d -polytope with lattice width 2 and lattice width direction n_{lw} is completely determined by its half-integral middle polytope , i.e. the rational polytope of dimension d-1 , which we get by the intersection of P with the hyperplane \displaystyle\{x\in M_{\mathds{R}}\mid\left<x,n_{lw}\right>=\mathrm{Min}_{P}(n_{lw})+1\}. In particular, we can work for lattice 3 -polytopes with a middlepolygon and only have to do combinatorics from there on only in dimension 2 . As a second step, we will see in section 3 that in this case not only the middle polygon but also the Fine interior is a half-integral polygon (Theorem 3.2 ). Moreover, we will see there that the Fine interiors can always be described by the union of a lattice polygon and some very special triangles that we can have on edges of the lattice polygons having two lattice points (see Theorem 3.9 ). It turns out that this description of the Fine interior is restrictive enough that we end up with an efficient classification algorithm for these Fine interiors in section 4 . We will use this algorithm there to obtain our main classification result in 4.1 : There are up to affine unimodular equivalence exactly 24324158 different 2 -dimensional Fine interiors those lattice 3 -polytopes which have at most 40 interior lattice points. In the last section we use this classification to classify as a corollary the Chern numbers for nondegenerated toric hypersurfaces of genus at most 40 which have as Newton polytope a lattice 3 -polytope with 2 -dimensional Fine interior. Having the Chern numbers of the surfaces, we can arrange them in the geography of the surfaces of general type. 2. The Fine interior of lattice polytopes of lattice width 2 In this section we will see that the Fine interior of a lattice d -polytope of lattice width 2 for arbitrary dimension d is determined by its half-integral middle polytope, which allows us to do our Fine interior computations in codimension 1 , or in other words, without loss of generality, we can restrict ourselves for Fine interior computations to the case of pyramids of heigth 2 which have the same middle polytope. We begin with the following lemma, which shows why the case of lattice width 2 is important for Fine interior computations. Lemma 2.1 . Let P\subseteq M_{\mathds{R}} be a rational d -polytope. If \mathrm{lw}(P)<2 , then F(P)=\emptyset . If \mathrm{lw}(P)=2 , then \dim F(P)<\dim P , and if \dim F(P)=\dim(P)-1 , then \mathrm{lw}(P)=2 . Proof. If n_{lw}\in N is a lattice width direction, then we have \displaystyle F(P)\subseteq\{x\in M_{\mathds{R}}\mid\left<x,n_{lw}\right>\geq\mathrm{Min}_{P}(n_{lw})+1,\left<x,-n_{lw}\right>\geq\mathrm{Min}_{P}(-n_{lw})+1\} and the set on the right side is empty for \mathrm{lw}(P)<2 and at of dimension d-1 for \mathrm{lw}(P)=2 . If we have \dim F(P)=\dim(P)-1 , then we must have a normal vector n\in N for F(P) such that -n is also a normal vector and thus -\mathrm{Min}_{P}(-n)-\mathrm{Min}_{P}(n)=2 . This implies \mathrm{lw}(P)\leq 2 and we have \mathrm{lw}(P)\geq 2 since this is always the case for non-empty Fine interior by the first part of the lemma. ∎ Remark 2.2 . There are many examples of lattice polytopes P\subseteq M_{\mathds{R}} with lattice width greater than 2 and \dim F(P)<\dim P-1 . Since \dim F((d+1)\Delta_{d})=0 for the standard simplex \Delta_{d} , i.e. the convex hull of an affine lattice basis, and \mathrm{lw}((d+1)\Delta_{d})=d+1 , we have for d,k,w\in\mathds{Z},d\geq 2,0\leq k<d-1,3\leq w\leq d-k+1 the lattice d -polytope (d-k+1)\Delta_{d-k}\times[0,w]^{k} with lattice width w>2 and \dim(F(P))=k . Remark 2.3 . We can see that \dim(F(P))=d-1 implies \mathrm{lw}(P)\leq 2 also from the general theory about lattice projections along the Fine interior. By [ Bat23 , 9.1] we have a lattice projection along the Fine interior on a lattice polytope with a Fine interior of dimension 0 . So from a Fine interior of dimension d-1 we get a projection onto a lattice polytope of dimension 1 with a Fine interior of dimension 0 . But up to lattice translation there is only [-1,1] with this property, and a lattice projection onto [-1,1] is obviously equivalent to \mathrm{lw}(P)\leq 2 . We now rearrange our coordinates for a good working setup with the middle polytope and split the dual lattice N with respect to the middle polytope. Definition

Read the complete text