Splitting Varieties for Cup Products with $\mathbb Z/3$-Coefficients

Brandon Boggess
DOI: https://doi.org/10.48550/arXiv.1601.06387
2016-01-24
Abstract:We connect Veronese embeddings to splitting varieties of cup products. We then give an algorithm for constructing splitting varieties for cup products with $\mathbb Z/n$ coefficients, with an explicit calculation for $n=3$. An application to the automatic realization of Galois groups is given.
Number Theory,Algebraic Geometry
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to construct splitting varieties in order to study the vanishing problem of specific cohomology classes. Specifically, for a given field \(F\) and a cohomology class \(\eta_F\in H^*(\text{Spec}F,\mathbb{Z}/n)\) on it, find a scheme \(X\) defined on the base field \(k\) such that \(X\) has an \(F\)-point if and only if \(\eta_F\) vanishes. More specifically, the paper focuses on constructing splitting varieties for studying cup products and has carried out detailed calculations especially for the case where the coefficient is \(\mathbb{Z}/3\). By introducing Veronese embeddings and the representation theory of the Heisenberg group, the author has developed an algorithm to construct these splitting varieties. The following are the key formulas and conclusions in the paper: 1. **Definition of the Heisenberg group**: Let \(H\) be the upper - triangular \(3\times3\) matrix modulo \(n\) Heisenberg group, whose diagonal elements are all \(1\) and the coefficients are in \(\mathbb{Z}/n\). The function \(a_{ij}:H\rightarrow\mathbb{Z}/n\) maps a matrix to its \((i, j)\) - th element. Denote \(E_{ij}\) as the matrix satisfying \(a_{ij}(E_{ij}) = 1\) and other non - diagonal elements are \(0\). 2. **Induced representation**: \(N\subset H\) is the subgroup \(\text{Ker}(a_{12}:H\rightarrow\mathbb{Z}/n)\). We have a one - dimensional representation \(\rho:N\rightarrow GL_1(k)\) defined as \(g\mapsto\zeta_n^{a_{13}(g)}\), where \(\zeta_n = e^{2\pi i/n}\). From this, we can obtain the induced representation \(\text{Ind}_N^H\rho\) of \(H\). 3. **Main results**: The main result of the paper is the construction of the splitting variety \(X(a, b)\) and the proof of the following theorems: - **Theorem 1.1**: The scheme \(X(a, b)\) has an \(F\)-point if and only if \(\kappa(a)\smile\kappa(b)=0\) in \(H^2(\text{Spec}F,\mathbb{Z}/3)\). - **Theorem 1.2**: Suppose \(a, b\in F^*\) such that \(F(\sqrt[3]{a},\sqrt[3]{b})/F\) is a \(\mathbb{Z}/3\times\mathbb{Z}/3\)-Galois extension, then the following conditions are equivalent: 1. There exists a \(\mathbb{Z}/3\)-Galois extension \(L/F(\sqrt[3]{a},\sqrt[3]{b})\) such that \(L/F\) is an \(H\)-Galois extension; 2. \(\kappa(a)\smile\kappa(b) = 0\) in \(H^2(\text{Spec}F,\mathbb{Z}/3)\); 3. The scheme \(X(a, b)\) has an \(F\)-point. These results not only provide a method for constructing splitting varieties, but also show how to use these varieties to automatically realize Galois groups.