Almost-regular dessins d'enfant on a torus and sphere (Q1637166)

The paper under review gives a contribution to the realization problem, which consists in classifying which ramification data over the 2-dimensional sphere are realisable. The authors consider ramification data for which -- when realisable -- the covering space is either the torus or the sphere. A family \(T=(T_i)_{i\in \mathbb N}\) of ramification data with \(r\) branch points is called \textit{almost regular of type \([k_1,\dots,k_r]\) and error at most \(\epsilon\),} if there exist a tuple \(A_j\) of positive integers different from \(k_j\) for each \(j=1,\dots,r,\) such that each \(T_i\) is of the form \([A_1,k_1^*],\dots,[A_r, k_r^*]\) with degree tending to infinity, and such that the sum \(\Sigma_{j=1}^{r}\Sigma_{a\in A_j}a\leq \epsilon\). They then show the following theorem: Theorem 1.1. Let \(T\) be a family of almost-regular genus 1 ramification data of type \([k_1,\dots,k_r]\), error at most \(\epsilon\), where \(T_i\) is not one of the exceptional types \((A)\)-\((D)\) (see below) for \(i \in \mathbb{N}\). Then all but finitely many members of \(T\) are realizable if \(\epsilon \leq 6\), or if \([k_1,\dots,k_r] \in \{[2,2,2,2][3,3,3]\}\) and \(\epsilon \leq 10\). The exceptional types \((A)\)-\((D)\) are \[ (A) \;[1,3,2^*] [2^*][2^*] [2^*]; (B) \;[2, 4, 3^*][3^*][3^*; (C) \;[2^*][3, 5, 4^*][4^*]; (D) \;[2^*][3^*][5, 7, 6^*]. \] Another result about ramification data of genus 1, in the same spirit, is proved, and for genus zero the authors show: Theorem 1.4. With the exception of \([2^*][1, 3^*][2, 2, 6^*]\), every family of almost-regular ramification data of genus \(0\) with \(\epsilon \leq 6\) is realisable in infinitely many degrees. Moreover, for families of type \([2, 2, 2, 2]\) or \([3, 3, 3]\), the same assertion holds with \(\epsilon\leq 10\). The authors use a mix of techniques with a focus on the study of a graph on a surface of genus \(g\) which comes from the ramification data, called a ``dessin d'enfant''. In the introduction the authors provide useful and detailed information about the state of art of this problem.