The distribution of zeros of an irreducible curve over a finite field (Q1920900)

If \(f \in F_p [X,Y]\) is an absolutely irreducible polynomial of total degree larger than 1, one expects that the set of zeros of \(f\) in \(F_p\times F_p\) should be ``randomly distributed'' in some sense. The results of the paper under review make this vague expectation a little more precise. If \(x\) and \(y\) are elements of \(F_p\), define \(d(x,y)\) to be \(|a-b |\), where \(a\) and \(b\) are integers in the interval \([0,p)\) that reduce to \(x\) and \(y\), respectively. For every \(\delta \in [0,1]\) let \(S_\delta\) be the set \(\{(x,y) \in F_p \times F_p: d(x,y) < \delta p\}\), and let \(V\) be the set of zeros of \(f\) in \(F_p \times F_p\). The main theorem of the paper states that \[ |\# (S_\delta \cap V) - \delta (2- \delta) p\mid \leq C_p^{1/2} \log p, \] where \(C\) is a number that depends only on the degree of \(f\) (and not on \(p)\). Since \(\# S_\delta\) is roughly \(\delta (2- \delta)\) times the cardinality of \(F_p \times F_p\), and since the Weil conjectures imply that \(\# V\) is roughly \(p\), the theorem can be interpreted as saying that the number of zeros of \(f\) in the region \(S_\delta\) is about as many as one would expect by chance alone. The definition of the regions \(S_\delta\) seems a little awkward to the reviewer, but the methods of the paper (see in particular Lemma 3 and its proof) give rise to analogous theorems that apply to more general subsets of \(F_p \times F_p\). The results of the paper depend on estimates for exponential sums over the set \(V\) given by \textit{E. Bombieri} [see Am. J. Math. 88, 71-105 (1966; Zbl 0171.41504)] and further investigated by \textit{J. H. H. Chalk} and \textit{R. A. Smith} [Acta Arith. 18, 191-212 (1971; Zbl 0219.12021)].