Maximal ideal cycles over normal surface singularities of Brieskorn type (Q411716): Difference between revisions

Let \(\pi : (\tilde X, E) \longrightarrow (X, o)\) be a normal surface singularity \((X, o)\). For an element \(h \in \mathfrak m_{X,o} \subset \mathcal O_{X,o}\), we consider a cycle \(\sum\limits^{r}_{j=1} v_{E_j} (h \circ \pi) E_j\) on \(E\), where \(E = \bigcup\limits^{r}_{j=1} E_j\) is the irreducible decomposition of \(E\). Among all such cycles, there is the minimal one \(M_E\) (S. S. T.Yau). It is called the maximal ideal cycle on \(E\). In generally, we have \(Z_{E} \leqq M_{E}\) for the fundamental cycle \(Z_E\). In the normal surface singularity theory, it is a natural problem to compare \(M_E\) and \(Z_E\). There are several results on it (see references of this paper). If \((X, o)\) is a hypersurface singularity defined by \(z^{a_1}_{1} +\cdots + z^{a_n}_{n} = 0 (a_j \geqq 2\) for any \(j\)), then it is called a hypersurface singularity of Brieskorn type. In this paper, for 2-dimensional hypersurface singularities of Brieskorn type (i.e., \(n = 3\)), the authors computed the maximal ideal cycle for the minimal good resolution of \((X, o)\) by considering a cyclic covering of \(\mathbb C^2\). Namely, they determined the \(M_E\) from \(a, b, c\), and compared \(M_E\) and \(Z_E\) . They obtain a necessary and sufficient condition to coincide them.