On images of Mori dream spaces (Q261436): Difference between revisions

Mori dream spaces were introduced by \textit{Y. Hu} and \textit{S. Keel} [Mich. Math. J. 48, 331--348 (2000; Zbl 1077.14554)] as varieties that have the best possible Mori theoretic properties. Indeed, a Mori dream space \(X\) is a normal projective variety for which the \(D\)-minimal model program exists and terminates for every \(\mathbb Q\)-divisor \(D\) on \(X\). In addition, the birational contractions obtained by running \(D\)-minimal model programs \(X\dashrightarrow X_D\) can be understood in terms of variations of GIT (VGIT) on the Cox ring of \(X\). Toric varieties, Fano-type varieties and algebraic \(K3\) surfaces with finite automorphism groups are all examples of Mori dream spaces. While some aspects of the birational geometry of Mori dream spaces are well understood, it is known that this class of varieties is not stable under birational modifications or under small deformations and new examples are hard to construct. First, the author proposes a slight refinement of the definition of Mori dream spaces which ensures that the notion behaves as expected over any field of arbitrary characteristic. The main result of the paper is that if \(f: X \to Y\) is a surjective morphism between normal projective varieties and if \(X\) is a Mori dream space, then so is \(Y\). The author then goes on to compare the geometry of \(Y\) and of \(X\) when this is the case. More precisely, if \(X\) is a Mori dream space, there is a fan structure on the effective cone \(\mathrm{Eff }(X)\) which encodes Zariski decompositions of line bundles on \(X\) on the one hand, and VGIT on the Cox ring of \(X\) on the other hand. The author shows that when \(f: X \to Y\) is a surjective morphism between Mori dream spaces as above, then the fan structure on \(\mathrm{Eff}(Y)\) is the restriction of the fan structure of \(\mathrm{Eff}(X)\).