A Hodge theoretic criterion for finite Weil-Petersson degenerations over a higher dimensional base

From MaRDI portal
Publication:1990273



Abstract: We give a Hodge-theoretic criterion for a Calabi--Yau variety to have finite Weil--Petersson distance on higher dimensional bases up to a set of codimension geq2. The main tool is variation of Hodge structures and variation of mixed Hodge structures. We also give a description on the codimension 2 locus for the moduli space of Calabi--Yau threefolds. We prove that the points lying on exactly one finite and one infinite divisor have infinite Weil--Petersson distance along angular slices. Finally, by giving a classification of the dominant term of the candidates of the Weil--Petersson potential, we prove that the points on the intersection of exact two infinite divisors have infinite distance measured by the metric induced from the dominant terms of the candidates of the Weil--Petersson potential.


This paper addresses the validity of the following conjecture due to \textit{C.-L. Wang} [Math. Res. Lett. 4, No. 1, 157--171 (1997; Zbl 0881.32017)]. \par Conjecture. Let ${\mathfrak{X}}/S$ be an $n$-dimensional Calabi-Yau degeneration which is smooth outside a simple normal crossing divisor $\bigcup_i E_i$. Then $X:={\mathfrak{X}}_s$ with $s\in\bigcup_i E_i$, has finite Weil-Petersson distance if and only if $N_iF^n_{\infty}(s)=0$ for all $i$ with $s\in E_i$. Here $F^n_{\infty}$ us the $n$-th piece of Schmid's limiting Hodge filtration and $N_i$ is the nilpotent part of monodromy around $E_i$. \par The main results of this article are formulated as follows. \par Theorem 1: The conjecture holds for the degeneration of a Calabi-Yau $n$-folds, up to a set of codimension $\geq 2$ in the base $S$. \par Proof uses variation of mixed Hodge structures on codimension one boundary points. This imposes a strong constraint on the Weil-Petersson potential and hence the Weil-Petersson metric can be controlled. \par The Weil-Petersson metric on Calabi-Yau threefolds is computed on the covering ${\mathbb{H}}^2\times(\Delta)^{r-2}\to (\Delta^*)^2\times (\Delta)^{r-2}$, where $\Delta:=\{\, z\in{\mathbb{C}}\,|\, |z|<1\,\}$, $\Delta^*:=\Delta\setminus \{0\}$. \par Theorem 2: Let ${\mathfrak{X}}/S$ be a degeneration of Calabi-Yau threefolds as in the conjecture. Suppose that $s\in S$ lies on exactly two boundary divisors, say $s\in E_1\cap E_2$, with $E_1$ infinite and $E_2$ finite. Then $s$ has infinite Weil-Petersson distance along the angular slices $\{\,(z_1,z_2)\in {\mathbb{H}}^2\,|\, \text{Re}(z_j)=c_j\in{\mathbb{R}}\,\}\times(\Delta)^{r-2}.$ \par The next result handles the case when $s$ lies in the intersection of two infinite divisors. \par Theorem 3: In the case of Calabi-Yau threefolds, suppose $s\in S$ lies exactly on two boundary divisors, say $s\in E_1\cap E_2$ with both $E_i$ infinite. Then it has infinite distance measured by the dominant term of the candidates of the Weil-Petersson potentials.











This page was built for publication: A Hodge theoretic criterion for finite Weil-Petersson degenerations over a higher dimensional base

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1990273)