Numerical invariants and moduli spaces for line arrangements (Q2215989)

The authors study moduli spaces of line arrangements in the plane. Let \(S = \mathbb{C}[x,y,z]\) be the graded polynomial ring and \(S_{m}\) the vector space of degree-\(m\) homogeneous polynomials in \(S\). For an integer \(d\geq 1\) we consider \(\mathcal{C}(d) = \mathbb{P}(S_{d})\) the parameter space of degree-\(d\) curves in \(\mathbb{P}^{2}_{\mathbb{C}}\). We denote by \(\mathcal{C}(d)_{0}\) the subset of \(\mathcal{C}(d)\) corresponding to curves with only isolated singularities, and by \(A(d)\) the subset of \(\mathcal{C}(d)_{0}\) corresponding to line arrangements consisting of \(d\) distinct lines. One can show that \(\mathcal{C}(d)_{0}\) is Zariski open and dense in \(\mathcal{C}(d)\), and \(A(d)\) is a Zariski closed subset in \(\mathcal{C}(d)_{0}\). Let us define \[\psi : \mathbb{P}(S_{1})^{d} \rightarrow \mathbb{P}(S_{d})\] by \(\psi(\ell_{1}, ..., \ell_{d}) = \ell_{1} \cdots \ell_{d}\) and consider \[X(d) = \psi^{-1}(A(d)).\] One can observe that \(A(d)\) and \(X(d)\) are smooth irreducible algebraic varietie of dimension \(d\). Moreover, \(X(d)\) is simply connected and the fundamental group of \(A(d)\) is \(\Sigma_{d}\), the symmetric group on \(d\) elements. The set \(X(d)\) is exactly the set of linear forms \(\mathbb{P}(S_{1})^{d}\) such that \(\ell_{i} \neq \ell_{j}\) for \(i\neq j\). For \(f \in S_{d}\) we denote by \(J_{f}\) the Jacobian ideal of \(f\) generated by the partials. Consider \(M(f) =S / J_{f}\) the Milnor algebra and define \[m_{k}(f) = \dim \, M(f)_{k}.\] It turns out that the sets \[\{f \in S_{d} \, : \, m_{k}(f) \leq m\} \subset \mathcal{C}(d),\] \[\{f \in \mathcal{C}(d)_{0} \, : \, \tau(f) \leq m\} \subset \mathcal{C}(d)_{0},\] \[\{f \in A(d) \, : \, \tau(f) \leq m \} \subset A(d),\] where \(\tau(\cdot)\) denotes the total Tjurina number of a plane curve defined by \(f = 0\), are Zariski open and dense for any \(k\geq 0\) and \(m\geq 0\). The above observation presents stratifications of the spaces \(\mathcal{C}(d)_{0}\) and \(A(d)\) with respect to the Tjurina numbers. Furthermore, the authors study free and nearly free curves in the plane providing a new interesting homological characterization. Next, the authors study group actions on parameter spaces showing that some introduced partitions of \(A(d)\) and \(X(d)\) are \(G\)-equivarant, where \(G =\mathrm{Aut}(\mathbb{P}^{2}_{\mathbb{C}})\). In Section 4, for any integer \(d\geq 4\) the authors study the set of all possible intersection lattices of arrangements with \(d\) lines, denoted by \(\mathcal{L}(d)\). For each lattice \(L \in \mathcal{L}(d)\) (in fact the isomorphism class) one denotes by \(X(L)\) the subset of \(X(d)\) consisting of line arrangements having an intersection lattice isomorphic to \(L\), so the subsets \(X(L)\) for \(L \in \mathcal{L}(d)\) form the strata of a partition of \(X(d)\). One can show, for instance, that \(X(L)\) is constructible. In Section 5, the authors study the complexity of the stratification \(\{X(L)\}_{L \in \mathcal{L}(d)}\). For example, the authors describe all the strata when \(d = 4,5,6\). In the last section, the authors discuss Terao's conjecture, and they show, in particular, a certain generalization of the fact that generic arrangements of lines are not free for \(d\geq 4\).