Calabi-Yau threefolds fibred by high rank lattice polarized \(K3\) surfaces (Q2288008)

This paper studies Calabi-Yau threefolds fibred by high rank lattice polarized \(K3\) surfaces. More precisely, specific \(K3\) surfaces considered here are the so-called \(M_n\)-polarized \(K3\) surfaces. Let \(M_n\) be the lattice of rank \(19\) defined by \(M_n:=H\oplus E_8\oplus E_8\oplus \langle -2n\rangle \) where \(H\) denotes the hyperbolic lattice of rank \(2\), \(E_8\) the rank \(8\) negative definite root lattice, and \(\langle -2n\rangle \) is the rank \(1\) lattice with self-intersection \(-2n\) for some natural number \(n\). Calabi-Yau threefolds fibered by such \(K3\) surfaces are interesting from mirror symmetric point of view. This article attempts to answer the following questions: (1) For which values of \(n\) do there exist Calabi-Yau threefolds fibered non-isotrivially by \(M_n\)-polarized \(K3\) surfaces? and (2) Is it possible to classify them? The first main result is the following theorem. Theorem 1: Let \(n\geq 2\), and let \(\pi^U:\mathcal{X}_U\to U\) denote a non-isotrivial \(M_n\)-polarized family of \(K3\) surfaces over a quasi-projective curve \(U\), such that the Néron-Severi group of a general fibre of \(\mathcal{X}_U\) is isomorphic to \(M_n\). Then \(\mathcal{X}_U\) is uniquely determined (up to isomorphism) by its generalized functional invariant map \(g: U\to\mathcal{M}_{M_n}\). Here \(\mathcal{M}_{M_n}\) denotes the compactified moduli space of \(M_n\)-polarized \(K3\) surfaces. This is a generalization of a theorem for \(M_2\)-polarized \(K3\) surfaces in [Adv. Math. 298, 369--392 (2016; Zbl 1339.14024)] by the same set of authors. This result is then used to give a complete explicit description of all Calabi-Yau threefolds fibred by \(M_n\)-polarized \(K3\) surfaces. Theorem 2: A Calabi-Yau threefold may only admits a non-isotrivial fibration by \(M_n\)-polarized \(K3\) surfaces if \[ n\in\{1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,19,23\} \] and moreover, the resulting Calabi-Yau threefolds are rigid (i.e., \(h^{1,1}=1\)) unless \[ n\in\{1,2,3,4,5,6,7,8,9,11\}. \] The paper then gives birational models for all such Calabi-Yau threefolds, explicit examples.