The study of curves over finite fields is a classical area of mathematics with connections to coding theory and cryptography. A central question is: how many "rational points" (solutions with coordinates in a finite field) can a curve of a given complexity have? For smooth curves, this is answered by the Hasse-Weil-Serre bound. When curves are allowed to have singularities, meaning self-intersections or cusps, an analogous bound called the Aubry-Perret bound applies. A curve that actually achieves this maximum is called "maximal," and finding explicit examples of maximal singular curves has been an open challenge.
The authors develop a systematic method to build maximal singular curves starting from well-understood smooth curves, specifically elliptic curves (think of a donut-shaped surface) and hyperelliptic curves (a generalization of those). By applying geometric transformations called linear projections, they deliberately introduce controlled singularities, either nodes (where the curve crosses itself) or cusps (where the curve has a sharp pinch point). The key insight is that these operations can be done in a way that preserves or relates the count of rational points, allowing the new singular curves to still achieve the Aubry-Perret bound. They rely on established mathematical frameworks, particularly work by Stour on a certain class of singular curves and a theory developed by Rosa and Stour for another class.
As a concrete application, the authors apply their construction to specific families of smooth maximal curves previously identified by Tafazolian and by Tafazolian and Torres, producing new families of maximal singular curves explicitly. They also compute the "gonality" of all these curves, which is a number measuring how complex the curve is in terms of the simplest maps it admits to a line. The result is a rich new supply of explicit, well-characterized maximal singular curves, expanding a relatively sparse landscape and providing tools that may be useful in algebraic geometry, coding theory, and related fields.