The monodromy property for K3 surfaces allowing a triple-point-free model

Abstract

The aim of this thesis is to study under which conditions $K3$ surfaces allowing a triple-point-free model satisfy the monodromy property. This property is a quantitative relation between the geometry of the degeneration of a Calabi-Yau variety $X$ and the monodromy action on the cohomology of $X$: a Calabi-Yau variety $X$ satisfies the monodromy property if poles of the motivic zeta function $Z_{X,\omega}(T)$ induce monodromy eigenvalues on the cohomology of $X$. In this thesis, we focus on $K3$ surfaces allowing a triple-point-free model, i.e., $K3$ surfaces allowing a strict normal crossings model such that three irreducible components of the special fiber never meet simultaneously. Crauder and Morrison classified these models into two main classes: so-called flowerpot degenerations and chain degenerations. This classification is very precise, which allows to use a combination of geometrical and combinatorial techniques to check the monodromy property in practice. The first main result is an explicit computation of the poles of $Z_{X,\omega}(T)$ for a $K3$ surface $X$ allowing a triple-point-free model and a volume form $\omega$ on $X$. We show that the motivic zeta function can have more than one pole. This is in contrast with previous results: so far, all Calabi-Yau varieties known to satisfy the monodromy property have a unique pole. We prove that $K3$ surfaces allowing a flowerpot degeneration satisfy the monodromy property. We also show that the monodromy property holds for $K3$ surfaces with a certain chain degeneration. We don't know whether all $K3$ surfaces with a chain degeneration satisfy the monodromy property, and we investigate what characteristics a $K3$ surface $X$ not satisfying the monodromy property should have.