Characterization of the Fermat curve as the most symmetric nonsingular algebraic plane curve (Q2509932)

Since XIX century, the problem of obtaining and classifying (complex) algebraic curves, or equivalently Riemann surfaces, with large automorphism groups has been a question studied for a long time. The first fundamental result was obtained by Hurwitz who proved that, if \(X\) is a complex projective irreducible algebraic curve of genus \(g \geq 2\), then the order of \(\mathrm{Aut}(X)\) is at most \(84(g-1)\). This bound is attained for infinitely many values of \(g\). The paper under review considers projective nonsingular plane algebraic curves defined over an algebraically closed field \(k\) of characteristic \(0\). If the degree of such a curve \(X\) is \(d \geq 4\), the Hurwitz bound is \(42d(d-3)\), which is attained only for small values of \(d\). So a natural question is to determine which are the curves with maximum order automorphism group for a given \(d\). This result was already known for \(d \leq 7\), as follows: the Klein quartic for \(d=4\), with \(168\) automorphisms; the Wiman sextic for \(d=6\), with \(360\) automorphisms; and the Fermat curve of equation \(x^d + y^d + z^d = 0\), for \(d=5,7\). The author completes here the classification, by proving that for any \(d \geq 8\), also the Fermat curve is just the curve attaining the maximum possible order for a given \(d\), which is proven to be \(6d^2\). The proof studies the subgroups of \(\mathrm{PGL}(3,k)\), because \(\mathrm{Aut}(X)\) is one of those subgroups for \(d \geq 4\), and splits into two possibilities. The general case is \(d \geq 11\), to which Section 4 is devoted. If \(G\) is the automorphism group of \(X\), the analysis is organized according to the possible \(G\)-invariant point-line configurations. Afterwards, Section 5 deals with \(d=8,9,10\) with a thorough and lengthy case-by-case analysis, and the global result is obtained. These results imply that the bound \(42d(d-3)\) is attained just for \(d=4\).