Real and complex supersolvable line arrangements in the projective plane (Q2052826)

In the paper under review the authors study some combinatorial problems related to the geometry of supersolvable arrangements defined over the complex and real numbers. Let \(\mathcal{L} \subset \mathbb{P}^{2}_{\mathbb{C}}\) be an arrangement of \(s\) lines in the projective plane. A point \(p\) is called a modular point for \(\mathcal{L}\) if it is an intersection point with the additional property that whenever \(q\) is another crossing point, then the line through \(p\) and \(q\) is a configurational line of \(\mathcal{L}\). We say that \(\mathcal{L}\) is supersolvable if it has a modular point. The main aim of the paper under review is to provide a classification of complex supersolvable line arrangements with respect to the number of modular points. The first result of the paper tells us that if \(\mathcal{L}\) is a complex arrangement of \(s\) lines in the projective which is non-trivial (i.e. is not a pencil of lines), then it cannot have more than \(4\) modular points. Based on that observation, the authors provide a complete description of complex supersolvable line arrangements with \(3\) and \(4\) modular points. The ultimate goal of the paper under review was to verify a conjecture saying that if \(\mathcal{C}\) is a non-trivial complex supersolvable arrangement of \(s\) lines, then the number of double points \(t_{2}(\mathcal{L})\) is bounded from below by \(s/2\). It turned out that this conjecture is true, as it has been recently shown by \textit{T. Abe} in [Int. Math. Res. Notices 145 (2020; Zbl 1483.14096)]. Here we sum up the above loose discussion by a concrete theorem. Given a supersolvable line arrangement \(\mathcal{L}\), if it has two or more modular points and they do no all have the same multiplicity, we say that \(\mathcal{L}\) is not homogeneous, but if all modular points of \(\mathcal{L}\) have the same multiplicity, we say that \(\mathcal{L}\) is homogeneous, and \(m\)-homogeneous if the common multiplicity is \(m\). Theorem. Let \(\mathcal{L}\) be a line arrangement having \(\mu_{\mathcal{L}} > 0\) modular points over any field \(k\). a) If \(\mathcal{L}\) is not homogeneous, then either \(\mathcal{L}\) is a near pencil or \(\mu_{\mathcal{L}} = 2\); if \(\mu_{\mathcal{L}} = 2\), then \(\mathcal{L}\) consists of \(a\geq 2\) lines through one modular point, \(b>a\) lines through the other modular point, and we have \(s = a+b-1\) lines and \((a-1)(b-1)\) double intersection points. b) If \(\mathcal{L}\) has a modular point of multiplicity \(2\), then \(\mathcal{L}\) is a pencil of lines. c) If \(\mathcal{L}\) is complex and homogeneous with \(m > 2\), then \(1 \leq \mu_{\mathcal{L}} \leq 4\). If \(\mu_{\mathcal{L}} \in \{3,4\}\), we have the following possibilities: i) If \(\mu_{\mathcal{L}} = 4\), then \(s=6\), \(m=3\), the number of double points is \(3\), the number of triple points is \(4\). ii) If \(\mu_{\mathcal{L}} = 3\), then \(m>3\) and, up to change of coordinates, \(\mathcal{L}\) consists of the lines defined by the linear factors of \[xyz(x^{m-2} - y^{m-2})(x^{m-2} - z^{m-2})(y^{m-2}-z^{m-2}),\] and hence \(s = 3m-3\), the number of double points is \(3m-6\), the number of triple points is \((m-2)^2\), and the number of \(m\)-fold points is equal to \(3\).