Szemerédi-Trotter-type theorems in dimension 3 (Q482631)

The Szemerédi-Trotter theorem asserts that the number of incidences of \(m\) lines and \(n\) points in \(\mathbb{R}^d\) with \(d = 2\) is bounded \(\tfrac{5}{2} m^{2/3}n^{2/3}+m+n\). By a projection argument, the same bound holds true for point-line incidences in higher dimensions. This article seeks improved bounds in similar situations for \(d \geq 3\). Another projection argument allows to restrict to \(d = 3\). Due to the non-duality of points and lines in three dimensions, the bounds are no longer symmetric in \(m\) and \(n\). Over the complex numbers, the main term is \(mn^{1/3}\) and its constant coefficient depends on a bound for the number of coplanar lines. Over finite fields, the main term is \(mn^{2/5}\). In order to avoid trivial quadratic growth of the number of intersecting lines in \(\mathbb{C}^3\), it is necessary to assume bounds on the number of coplanar lines and the number of lines on quadrics. Here, the main term is \(m^{3/2}\). Most of these results holds true over finite fields of sufficiently large characteristic. The main idea of all proves is to interpret the union \(C\) of all lines as part of the complete intersection of low degree algebraic surfaces. Bounds for the surface degrees can be extracted from very old papers by \textit{G. Salmon} [A treatise on the analytic geometry of three dimensions. Revised by R. A. P. \textit{Rogers}. Fifth edition, in 2 vols. Vol. I. London: Longmans, Green \(\and\) Co (1911; JFM 42.0587.01)] and \textit{G.Monge} [Application de l'analyse à la géométrie (1809)]. Modern algebraic geometry techniques are then used to efficiently bound the algebraic genus of their intersection curve, the number of its singular points and hence also the number of intersection points in~\(C\).