Embedding non-projective Mori dream space (Q776217)

The author introduces the notion of weak Mori dream spaces, as natural generalisation of Mori dream spaces. Namely, the condition of projectivity is relaxed, while keeping the reasonable assumptions of \(\mathbb{Q}\)-factoriality, finite generation of the class group, and finite generation of the Cox ring. Two expected consequences are verified: MMP can be carried out for any divisor on weak Mori dream spaces, and that the closure of the movable cone of a weak Mori dream space admits a finite union of Nef cones that determine rational contractions from the variety, just as in the case of Mori dream spaces. However, the core of the paper is devoted to the analysis of ``neat embeddings'' into complete toric varieties. This is done by first taking a neat embedding into a \(\mathbb{Q}\)-factorial toric variety and then studying possible completions. It turns out that in taking such completion one may, naturally, increase the Picard number of the toric variety against the weak Mori dream space. However, it is verified that such discrepancy does not occur in certain cases, for example in the case of complete weak Mori dream spaces with Picard number at most two.