A Reider-type Result for Smooth Projective Toric Surfaces
Bach Le Tran

TL;DR
This paper establishes numerical criteria for when the adjoint line bundle on a smooth projective toric surface is nef or ample, based on the associated lattice polytope.
Contribution
It provides new necessary and sufficient conditions for the positivity of the adjoint series on toric surfaces using polytope analysis.
Findings
Criteria for nefness of |K_X+L|
Criteria for ampleness of |K_X+L|
Polytope-based numerical conditions
Abstract
Let be an ample line bundle over a smooth projective toric surface . Then corresponds to a very ample lattice polytope that encodes many geometric properties of . In this article, by studying , we will give some necessary and sufficient numerical criteria for the adjoint series to be either nef or (very) ample.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAlgebraic Geometry and Number Theory · Commutative Algebra and Its Applications · Historical Studies and Socio-cultural Analysis
A Reider-type Result for Smooth Projective Toric Surfaces
Bach Le Tran
School of Mathematics
University of Edinburgh
James Clerk Maxwell Building
Peter Guthrie Tait Road
Edinburgh EH9 3FD
Abstract.
Let be an ample line bundle over a smooth projective toric surface . Then corresponds to a very ample lattice polytope that encodes many geometric properties of . In this article, by studying , we will give some necessary and sufficient numerical criteria for the adjoint series to be either nef or (very) ample.
1. Introduction
The problem of determining whether a line bundle is nef or (very) ample is an important question in algebraic geometry. The Nakai-Moishezon criterion [12, 10] states that a Cartier divisor on a proper scheme over an algebraically closed field is ample if and only if for every closed integral subscheme of . For toric varieties, a special form of the criterion holds: if for every torus-invariant curve then is ample. Furthermore, if for every torus-invariant curve then is globally generated [7, 9, 11]. However, the question is more complicated when we consider the adjoint bundle . Namely, are there numerical conditions for so that is globally generated or ample? Fujita conjectured the following:
Conjecture 1.1** ([3]).**
Let be an -dimensional projective algebraic variety, smooth or with mild singularities, and an ample divisor on . Then
- (1)
For , is basepoint free. 2. (2)
For , is very ample.
The conjecture is true for toric varieties [4, 13]. For smooth surfaces, Fujita’s conjecture follows from Reider’s theorem [15].
In this article, we will present a combinatorial proof for a Reider-type result for smooth projective toric surfaces.
Proposition 1.2**.**
Let be a smooth projective toric surface not isomorphic to , and let be an ample line bundle on .
- (1)
The adjoint series is not base point free if and only if there exists an effective torus-invariant divisor such that
[TABLE] 2. (2)
The adjoint series is not ample if and only if there exists an effective torus-invariant divisor such that either
[TABLE]
Furthermore, if , then is not ample if and only if there exists an effective torus-invariant divisor such that either
[TABLE]
As a convention, in this article, we will follow the notations in [2]. In particular, we will always use to denote the ambient lattice if there is no confusions.
Acknowledgments
We would like to thank Milena Hering for suggesting the problem and for her invaluable guidance. We also want to thank Ivan Cheltsov for some of the comments.
2. Toric Surfaces Reviewed
Let be an ample line bundle over a projective toric variety corresponding to a polytope . Then we have a combinatorial interpretation of the intersection number where is any torus-invariant curve as follows.
Lemma 2.1** ([7, (1.4) and Page 457]).**
Let be an ample line bundle on a projective toric variety corresponding to a polytope . For a torus invariant curve , let be the corresponding edge on . Then is equal to the lattice length of , i.e.,
[TABLE]
For our purpose, we will need to use the classification of smooth projective toric surfaces: every smooth complete toric surfaces is a finite blowup of either , , or the Hirzebruch surface , where ( [2, Theorem 10.4.3]). Another important fact that we will use is that every ample line bundle on a smooth projective toric surface is also very ample.
Lemma 2.2** ([2, Theorem 6.1.15]).**
A line bundle on a smooth complete toric variety is ample if and only if it is very ample.
Smooth toric surfaces are interesting objects to work with; partially because of their computability. For example, we have the following lemma.
Lemma 2.3** ([2, Proposition 10.4.11]).**
Let be ray generators of a smooth complete fan in . Let be the smooth projective toric surface from and for . Let be the canonical divisor . Then
[TABLE]
where the are integers such that for all , where and .
The following corollary follows directly from [2, Lemma 10.4.1] and Lemma 2.3.
Corollary 2.4**.**
Let be ray generators of a smooth complete fan in . Let be the smooth projective toric surface from and for . Let be the canonical divisor . Then for ,
[TABLE]
We also know that the blowup of a toric variety corresponds to a subdivision of fan. Thus the number of generating rays of the fan corresponding to a toric surface increases after a blowup ([2, Proposition 3.3.15]).
Example 2.5*.*
Consider the Hirzebruch surface , , whose fan given by the following figure
The ray generators of are , , , and . Let the associated divisors be , , , and , respectively. By [2, Proposition 4.1.2],
[TABLE]
Thus , , and
[TABLE]
The maximal cones of are , , and as in Figure 1. Let . We compute the to be
[TABLE]
Then by [2, Lemma 6.1.13], is very ample if and only if . The nef cone of is given by
By [2, Lemma 10.4.1], we have , , , , .
Finally, we will make use of the Hodge’s Index Theorem:
Lemma 2.6** ([5, Theorem V.1.9]).**
Let be an ample divisor on a smooth projective surface . If is a divisor such that , then . The equality occurs if and only if is numerically equivalent to [math].
Corollary 2.7** ([5, Exercise V.1.9]).**
Let be an ample divisor on a smooth projective surface and an arbitrary divisor. Then
[TABLE]
Proof.
Since is ample, . Let . We have
[TABLE]
Then by Lemma 2.6, we must have . In other words,
[TABLE]
Since , it follows that . ∎
3. Toric Surfaces and Lattice Polygons
In this section, we review and prove some lemmas on lattice polygons that we will use to the proof of Proposition 1.2.
Lemma 3.1** ([1, Lemma 1]).**
Every lattice polygon with at least 5 edges has at least an interior lattice point.
Lemma 3.2**.**
Let be lattice points such that no three points are collinear. Then there exists a lattice point in .
Proof.
Let the coordinates of be for . By the pigeonhole principle, there must be such that and . Then the midpoint of is a lattice point. Since no three points in are collinear, it follows that . ∎
As a consequence, we obtain:
Lemma 3.3**.**
Let be a lattice polygon that has at least vertices and assume that one of its edges has lattice length . Then .
Proof.
It suffices to prove the lemma when is a lattice pentagon. Let , where are ordered clockwise in . Without loss of generality suppose that the lattice length of the edge joining and is ; i.e., there are other lattice points , , in between and .
Consider the polytope . Then by Lemma 3.1, there must be a lattice point in the interior of . Then lies in at most one of the segments , , , , . If lies in or if does not lie in any mentioned segments, consider the set of points . By Lemma 3.2, there must be another lattice point in that is not the same as the points listed before. If , then and . By Pick’s theorem [14],
[TABLE]
If , then and . Again, by Pick’s theorem,
[TABLE]
If lies in or then we get such a point from . If lies in or then we get from . The same argument follows and we proved the lemma. ∎
We will also need the following lemmas for the proof of Proposition 1.2.
Lemma 3.4**.**
Let be an ample line bundle over a smooth projective toric surface . Let be the fan of . Suppose that has rays . Then for any integer ,
[TABLE]
Proof.
Let be the polytope associated to . By Pick’s theorem ([14]) and since is ample so that for all ,
[TABLE]
where and are the sets of all boundary points and interior points of , respectively. By Lemma 2.1,
[TABLE]
Hence, combining (1) and (2) gives
[TABLE]
Since , by Lemma 3.1, . Therefore,
[TABLE]
∎
4. A Reider-type Result for Toric Surfaces
We will devote this section to prove Proposition 1.2. First of all, it is true for .
Lemma 4.1**.**
Proposition 1.2 holds for .
Proof.
Let be the fan of as follows.
By [2, Lemma 10.4.1], for all . Thus, we need to show that there exists such that in the first part and in the second part.
For any ample bundle on , if is not basepoint free, then there exists such that . Then By lemma 2.3,
[TABLE]
This implies , so that .
Now suppose that is not ample and . Then By lemma 2.3,
[TABLE]
This implies . Hence, either and or and . The conclusion follows. ∎
Secondly, we show that Proposition 1.2 holds for Hirzebruch surfaces.
Lemma 4.2**.**
Proposition 1.2 holds for , .
Proof.
Consider the Hirzebruch surface , as in Example 2.5. We have
[TABLE]
The canonical divisor of is given by
[TABLE]
Recall that , , , , and (cf. [2, Lemma 10.4.1]).
Let be an ample line bundle over . Then . We have two cases as follows.
- •
If then . For to be ample while is not nef, has to be of the form , . In this case, take , then
[TABLE]
For to be ample while is not ample, has to be of the form , or , or , where .
- (1)
If , take , then
[TABLE] 2. (2)
If , take , then
[TABLE] 3. (3)
If , take , then
[TABLE]
- •
: For to be ample but is not nef, has the form
[TABLE]
Take , then .
For to be ample but is not, has the form or , where .
- (1)
If , take , then
[TABLE] 2. (2)
If , take , then
[TABLE]
∎
Finally, we will give the proof for the final case of Proposition 1.2, when is an arbitrary blowup of or the Hirzebruch surface.
Proof of Proposition 1.2.
The sufficiency trivially holds by Corollary 2.4. We now prove the necessity.
By the classification of smooth projective toric surfaces, the proofs for the cases of (Lemma 4.1) and (Lemma 4.2), it suffices to prove the proposition in the case that the fan of has at least rays.
We first prove part 1. Suppose that is not basepoint free. Then there exists such that . Take . By Lemma 2.3,
[TABLE]
This implies , so since is ample,
[TABLE]
- •
If , then , which is a contradiction to the hypothesis that is ample.
- •
If , either or . But since is ample. Thus . The proposition holds for this case.
It remains to show that cannot be positive. Since the fan of contains at least 5 rays, by Lemma 3.4,
[TABLE]
In addition, it follows from Corollary 2.7 that
[TABLE]
Combining (5) with (3) and (4) yields
[TABLE]
This implies . The only possibility is . Then by (5), , which is impossible since . Therefore, it cannot be the case that .
We now prove the second part of the proposition. Suppose that is not ample, so there exists such that . Let . By Corollary 2.4,
[TABLE]
This implies ; hence,
[TABLE]
- •
If , then , so .
- •
If , either or .
Now we consider the case that . Since the fan of contains at least 5 rays, by Lemma 3.4,
[TABLE]
By Corollary 2.7,
[TABLE]
Since , then by (7), . Thus by (8), , so . It follows that . Hence, . This inequality combining with (6) and (7) yields
[TABLE]
This implies . The only possibilities are or .
- •
If then by (6). Since can only be either or , in this case. Furthermore, suppose that . If and then , a contradiction to (8).
- •
Now assume that . If , then by (7), and , a contradiction to (8). Now assume that . Then the polygon associated to has at least vertices and one of its edges has lattice length by Lemma 2.1. Hence, by Lemma 3.3. It follows that , a contradiction to (8).
The proposition follows. ∎
5. Some Applications
The following corollary gives an affirmative answer for a stronger form of Fujita’s conjecture (Conjecture 1.1) in case of smooth complete toric surfaces. Note that for -dimensional toric varieties, the Fujita’s conjecture is in fact a corollary of [4, Corollary 0.2] and [13, Theorem 1].
Corollary 5.1** ([4, 13]).**
Let be a smooth complete surface not isomorphic to . Let be an ample line bundle on such that for all toric invariant curve . Then is globally generated. If and for all toric invariant curve , then is very ample.
Proof.
Suppose that is not globally generated. By Proposition 1.2, there exists a toric invariant curve such that or , a contradiction. ∎
As a corollary, we have a stronger form of [8, Corollary 2.7] for smooth toric surfaces as follows.
Corollary 5.2**.**
If is an ample line bundle on a smooth complete toric surface not isomorphic to , then is nef, and is very ample.
Proof.
Take , then for any toric invariant curve , . By Proposition 1.2, is nef. Similarly, take , then , and . By Proposition 1.2, is very ample. ∎
Remark 5.3*.*
It would be interesting to see if we can apply the classification in Proposition 1.2 to the study of Iskovskikh-Shokurov conjecture [6] for conic bundles over smooth toric surfaces.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1[1] J.R. Arkinstall, Minimal requirements for Minkowski’s theorem in the plane i , Bulletin of the Australian Mathematical Society 22 (1980), no. 2, 259–274.
- 2[2] David A. Cox, John B. Little, and Henry K. Schenck, Toric varieties (graduate studies in mathematics) , American Mathematical Society, 2011.
- 3[3] Takao Fujita, On polarized manifolds whose adjoint bundles are not semipositive , Algebraic geometry (Tadao Oda, ed.), Adv. Stud. Pure Math., vol. 10, Sendai, North-Holland, Amsterdam, 1985, pp. 167–178.
- 4[4] Osamu Fujino, Notes on toric varieties from mori theoretic viewpoint , Tohoku Math. J. (2) 55 (2003), no. 4, 551–564.
- 5[5] Robin Hartshorne, Algebraic geometry , Springer-Verlag, New York, 1977, Graduate Texts in Mathematics, No. 52. MR MR 0463157 (57 #3116)
- 6[6] V A Iskovskikh and V V Shokurov, Birational models and flips , Russian Mathematical Surveys 60 (2005), no. 1, 27–94.
- 7[7] Robert Laterveer, Linear systems on toric varieties , Tohoku Math. J. (2) 48 (1996), no. 3, 451–458.
- 8[8] R. Lazarsfeld, Lectures on linear series , eprint ar Xiv:alg-geom/9408011, August 1994.
