The numerical equivalence relation for height functions and ampleness and nefness criteria for divisors (Q2918787)
From MaRDI portal
| This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use this page instead for the normal view: The numerical equivalence relation for height functions and ampleness and nefness criteria for divisors |
scientific article; zbMATH DE number 6092626
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | The numerical equivalence relation for height functions and ampleness and nefness criteria for divisors |
scientific article; zbMATH DE number 6092626 |
Statements
The numerical equivalence relation for height functions and ampleness and nefness criteria for divisors (English)
0 references
10 October 2012
0 references
Heights, Weil height machine
0 references
ample divisors
0 references
effective divisors
0 references
intersection theory
0 references
The paper is about using the Weil height machine to obtain geometric information about line bundles on algebraic varieties, from arithmetic properties. For dimension one, the following result is already known:NEWLINENEWLINELet \(C\) be an algebraic curve and \(E_C, D_C\) divisors on \(C\) with \(\deg(D_C) \geq 1\), then NEWLINE\[NEWLINE\lim_{h_D(P) \rightarrow \infty} h_E(P)/ h_D(P)=\deg(E_C)/\deg(D_C).NEWLINE\]NEWLINE As a consequence \(E_C\) is ample (numerically effective) if the above limit is positive (non-negative). for higher dimension the author propose to use the fractional limit: NEWLINE\[NEWLINE\text{Flim}_D(E,U)=\liminf_{P \in U, h_D(P) \rightarrow \infty} h_E(P)/ h_D(P), NEWLINE\]NEWLINE where \(V\) is a projective algebraic variety, \(U \subset V\) is a subset of \(V\), and \(E,D\) are divisors on \(V\) with \(D\) ample. Using the geometric fact that for any ample divisors \(E,D\) we can always find a number \(m>0\) such that \(mD-E\) is again ample and a characterization of the kernel of the Weil height machine, the author obtains:NEWLINENEWLINENEWLINE(1) \(E\) is ample if and only if \(\text{Flim}_D(E,V)>0\),NEWLINENEWLINENEWLINE(2) \(E\) is numerically effective if and only if \(\text{Flim}_D(E,V) \geq 0\),NEWLINENEWLINENEWLINE(3) \(E\) is effective only if \(\text{Flim}_D(E,U) \geq 0\) for some open dense \(U \subset V\),NEWLINENEWLINENEWLINE(4) \(E\) is pseudo-effective only if \(\text{Flim}_D(E,U) \geq 0\) for some \(U\), which is an infinite intersection of open subsets of \(V\),NEWLINENEWLINENEWLINE(5) \(E\) is pseudo-effective if \(\text{Flim}_D(E,U) \geq 0\) for some open dense set \(U \subset V\),NEWLINENEWLINENEWLINE(6) suppose that any pseudo-effective divisor on \(V\) is linearly equivalent to a sum of an effective divisor and a numerically effective divisor; then \(E\) is pseudo-effective if and only if \(\text{Flim}_D(E,U) \geq 0\) for some open dense set \(U \subset V\).NEWLINENEWLINEIn the result above, pseudo-effective divisor refers to limits of a sequences of effective divisors. The question still stands whether or not we can drop the condition in (6) and find an open dense set for any pseudo-effective divisor \(E\).
0 references