Positivity of line bundles and Newton-Okounkov bodies (Q683760): Difference between revisions

From MaRDI portal
Added link to MaRDI item.
Import240304020342 (talk | contribs)
Set profile property.
Property / MaRDI profile type
 
Property / MaRDI profile type: MaRDI publication profile / rank
 
Normal rank

Revision as of 00:58, 5 March 2024

scientific article
Language Label Description Also known as
English
Positivity of line bundles and Newton-Okounkov bodies
scientific article

    Statements

    Positivity of line bundles and Newton-Okounkov bodies (English)
    0 references
    0 references
    0 references
    9 February 2018
    0 references
    In the present paper, the authors provide an interesting link between the local and global positivity of line bundles and the geometry of Newton-Okounkov bodies. Let us recall that a Newton-Okounkov body \(\triangle_{Y_{\bullet}}(D)\) of a divisor \(D\) is a refined book-keeping device encoding the orders of vanishing along subvarieties of the ambient projective space \(X\). These subvarieties should build up the so-called admissible flag, i.e., \[ Y_{\bullet} \, : \, X = X_{0} \supseteq X_{1} \supseteq \dots \supseteq Y_{n} = \{\mathrm{pt}\}, \] where each element \(Y_{i}\) of \(Y_{\bullet}\) is irreducible of codimension \(i\) and smooth at \(Y_{n}\). Let us recall that for a big \(\mathbb{R}\)-divisor \(D\) we define the restricted base locus as \[ B_{-}(D) = \bigcup_{A} \, SB(D+A), \] where the sum is over all ample \(\mathbb{Q}\)-divisors \(A\), and the augmented base locus of an \(\mathbb{R}\)-divisor \(D\) is defined as \[ B_{+}(D) = \bigcap_{A}\, SB(B-A), \] where the intersection is taken over all ample \(\mathbb{Q}\)-divisors \(A\) in \(X\). Now we are ready to formulate the main results of the paper. Theorem A. Let \(D\) be a big \(\mathbb{R}\)-divisor in a smooth projective variety \(X\) of dimension \(n\), and let \(x \in X\). Then the following conditions are equivalent: i) \(x \notin B_{-}(D)\); ii) there exists an admissible flag \(Y_{\bullet}\) in \(X\) centered at \(x\) such that the origin \(0 \in \triangle_{Y_{\bullet}}(D)\); iii) the origin \(0\in \triangle_{Y_{\bullet}}(D)\) for every admissible flag \(Y_{\bullet}\) in \(X\) centered at \(x \in X\). Let us denote by \(\triangle_{\lambda} \subset \mathbb{R}^{n}\) the standard simplex of length \(\lambda\). Theorem B. In the setting of the above theorem, the following conditions are equivalent: i) \( x \notin B_{+}(D)\); ii) there exists an admissible flag \(Y_{\bullet}\) in \(X\) centered at \(x\) with \(Y_{1}\) ample such that \(\triangle_{\lambda} \subseteq \triangle_{Y_{\bullet}}(D)\) for some positive real number \(\lambda\); iii) for every admissible flag \(Y_{\bullet}\) in \(X\) there exists a real number \(\lambda > 0\) for which \(\triangle_{\lambda} \subseteq \triangle_{Y_{\bullet}}(D)\). In the last part of the paper the authors investigate relations between Nakayama's divisorial decomposition (or \(\sigma\)-decomposition) and Newton-Okounkov bodies in dimension greater than \(2\). Let us recall that the \(\sigma\)-decomposition of a pseudoeffective \(\mathbb{R}\)-divisor \(D\) allows us to write \(D = P_{\sigma}(D) + N_{\sigma}(D)\), where \(N_{\sigma}(D)\) is effective and \(\text{Supp}(N_{\sigma}(D))\) coincides with the divisorial part of \(D_{-}(D)\), and for all \(m\geq 0\) one has \(H^{0}(X, \mathcal{O}_{X}([mP_{\sigma}(D)]) \simeq H^{0}(X, \mathcal{O}_{X}([mD])\). However, on the contrary to the Zariski decomposition for surfaces, the positive part \(P_{\sigma}(D)\) is only movable. Before we formulate the result, let us define \[ \sigma_{\Gamma}(D) := \text{lim}_{\varepsilon \rightarrow 0^{+}} \text{inf}\{ \text{mult}_{\Gamma}(D') : D' \sim_{\mathbb{R}} D + \varepsilon A \text{ and } D' \geq 0 \}, \] where \(A\) is an ample divisor. Theorem C. Let \(X\) be a smooth projective variety, \(D\) a big \(\mathbb{R}\)-divisor, \(\Gamma\) a prime divisor, and \(Y_{\bullet} : Y_{0} = X \supset Y_{1} = \Gamma \supseteq Y_{2} \supseteq \dots \supseteq Y_{n} = \{\mathrm{pt}\}\) in \(X\). Then i) \(\triangle_{Y_{\bullet}}(D) \subseteq (\sigma_{\Gamma}(D), 0 , \dots , 0) + \mathbb{R}^{n}_{+}\); ii) \((\sigma_{\Gamma }(D), 0 , \dots , 0) \in \triangle_{Y_{\bullet}}(D)\), whenever \(x \in \Gamma\) is a very general point; iii) \(\triangle_{Y_{\bullet}}(D) = \nu_{Y_{\bullet}}(N_{\sigma}(D)) + \triangle_{Y_{\bullet}}(P_{\sigma}(D))\). Moreover, \(\triangle_{Y_{\bullet}}(D) = \triangle_{Y_{\bullet}}(P_{\sigma}(D))\) if only \(x \notin \text{Supp}(N_{\sigma}(D))\).
    0 references
    0 references
    ampleness
    0 references
    Newton-Okounkov bodies
    0 references
    positivity of line bundles
    0 references
    augmented and restricted base loci
    0 references

    Identifiers