Motivic Milnor fiber of a quasi-ordinary hypersurface (Q2450168)

The motivic zeta function, and its specializations the topological zeta function and the motivic Milnor fibre, are important singularity invariants of hypersurface germs of analytic functions \(f:(\mathbb C^{d+1},0)\to (\mathbb C,0)\). They were introduced by Denef and Loeser in various papers. Note for instance that the motivic Milnor fibre specializes further to the well-known Hodge-Steenbrink spectrum of \(f\) at \(0\). There are formulas for all these invariants in terms of an embedded resolution of \(f=0\), but, for large \(d\) and arbitrary \(f\), embedded resolutions are very hard or impossible to compute in practice. In this paper the authors prove explicit formulas when \(f\) defines an irreducible \textsl{quasi-ordinary} hypersurface singularity germ \((S,0)\), meaning that there exists a finite map \((S,0)\to (\mathbb C^{d},0)\), which is unramified outside a normal crossings divisor of \((\mathbb C^{d},0)\). The embedded topology of \((S,0) \subset (\mathbb C^{d+1},0)\) is characterized by the so-called \textsl{characteristic monomials} of \((S,0)\), a concept that generalizes the characteristic pairs of plane branches. A crucial aspect of the formulas in the paper is that they are in terms of the characteristic monomials of \((S,0)\); hence all invariants above (and other specializations of the motivic zeta function) are determined by the embedded topology of \((S,0)\). The proofs involve toric geometry, Newton polyhedra, and the so-called semi-roots of \(f\).