ScalingStacks

Theorem 6.1 . [01D9]

Original official author HTML, exact retained edition. Historical TeX conversion verdicts remain unchanged. Cited-edition alignment and mathematical self-containment are not assessed.

Complete original source context · Original author HTML

Theorem 6.1.

Let KK be either 𝐂{\mathbf{C}} or a discretely valued field of residue characteristic zero, and let (X,L)(X,L) be a smooth projective polarized variety over KK. Then there exists a unique class PSH⁡(Lan)\operatorname{PSH}({L^{\mathrm{an}}}), the set of singular semipositive metrics, with the following properties:

  • •

    PSH⁡(Lan)\operatorname{PSH}({L^{\mathrm{an}}}) is a convex set which is closed under maxima and addition of constants;

  • •

    PSH⁡(Lan)∩C0​(Lan)=PSH0⁡(Lan)\operatorname{PSH}({L^{\mathrm{an}}})\cap C^{0}({L^{\mathrm{an}}})=\operatorname{PSH}^{0}({L^{\mathrm{an}}});

  • •

    if sis_{i}, 1≤i≤p1\leq i\leq p, are nonzero global sections of m​LmL for some m≥1m\geq 1, then ϕ:=1m​maxi​log⁡|si|∈PSH⁡(Lan)\phi:=\frac{1}{m}\max_{i}\log|s_{i}|\in\operatorname{PSH}({L^{\mathrm{an}}}); further, ϕ\phi is continuous iff the sections sis_{i} have no common zero.

  • •

    if (ϕj)(\phi_{j}) is an arbitrary family in PSH⁡(Lan)\operatorname{PSH}({L^{\mathrm{an}}}) that is uniformly bounded from above, then the usc regularization of supjϕj\sup_{j}\phi_{j} belongs to PSH⁡(Lan)\operatorname{PSH}({L^{\mathrm{an}}});

  • •

    if (ϕj)(\phi_{j}) is a decreasing net in PSH⁡(Lan)\operatorname{PSH}({L^{\mathrm{an}}}), then either ϕj→−∞\phi_{j}\to-\infty uniformly on Xan{X^{\mathrm{an}}}, or ϕj→ϕ\phi_{j}\to\phi pointwise on Xan{X^{\mathrm{an}}} for some ϕ∈PSH⁡(Lan)\phi\in\operatorname{PSH}({L^{\mathrm{an}}});

  • •

    Regularization: for every ϕ∈PSH⁡(Lan)\phi\in\operatorname{PSH}({L^{\mathrm{an}}}) there exists a decreasing sequence (ϕm)m=1∞(\phi_{m})_{m=1}^{\infty} of smooth/model metrics such that ϕm\phi_{m} converges pointwise to ϕ\phi on Xan{X^{\mathrm{an}}} as m→∞m\to\infty;

  • •

    Compactness: the space PSH⁡(Lan)/𝐑\operatorname{PSH}({L^{\mathrm{an}}})/{\mathbf{R}} is compact.

Original mathematics by the credited authors. Source-backed reader collection; mathematical self-containment is not assessed.