On (\(k,p^{e}\))-arcs in Desarguesian planes (Q596587)

A \((k,n)\)-arc in \(\text{ PG}(2,q)\), \(q=p^h\) with \(p\) prime, is a set of \(k\) points intersecting each line in at most \(n\) points. One can easily show that \(k \leq qn-q+n\). A \((k,n)\)-arc is called complete if it is not contained in a \((k+1,n)\)-arc. A \((k,n)\)-arc is called maximal if \(k=qn-q+n\). In [\textit{T. Szönyi}, Des. Codes Cryptography 18, No. 1--3, 235--246 (1999; Zbl 0964.51010)], it was proved that every \((qp-q+p-\varepsilon,p)\)-arc with \(\varepsilon \leq \frac{1}{2} \sqrt{q}\) can be embedded in a maximal arc. In the present paper, the author improves this result. She shows that every \((qp^e-q+p^e-\varepsilon,p^e)\)-arc can be embedded in a maximal arc if \(\varepsilon \leq \frac{1}{4} \sqrt{q}\) and \(p^e \leq \frac{1}{2} \sqrt{q}\). From the nonexistence of complete arcs in \(\text{PG}(2,p^h)\), \(p \not= 2\), see [\textit{S. Ball}, \textit{A. Blokhuis} and \textit{F. Mazzocca}, Combinatorica 17, No. 1, 31--41 (1997; Zbl 0880.51003)], an upper bound for the size of a \((k,p^e)\)-arc then easily follows. The author also gives a new proof of a result of Segre which says that any \((k,2)\)-arc (or shortly \(k\)-arc) in \(\text{PG}(2,2^h)\) can be extended to a hyperoval if \(k > q+1-\sqrt{q}\).