On the multicanonical systems of quasi-elliptic surfaces in characteristic 3
Toshiyuki Katsura

TL;DR
This paper investigates the properties of multicanonical systems on quasi-elliptic surfaces in characteristic 3, establishing that for m ≥ 5, these systems define a quasi-elliptic fiber space, with 5 being the minimal such value.
Contribution
It proves that for quasi-elliptic surfaces in characteristic 3, the multicanonical system |mK_S| induces a fiber space structure for all m ≥ 5, and 5 is the optimal bound.
Findings
|mK_S| defines a fiber space for m ≥ 5
The bound m=5 is optimal for such surfaces
Characterization of multicanonical systems in characteristic 3
Abstract
We consider the multicanonical systems of quasi-elliptic surfaces with Kodaira dimension in characteristic 3. We show that for any gives the structure of quasi-elliptic fiber space, and the number is best possible to give the structure for any such surfaces.
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 · Geometry and complex manifolds · Advanced Algebra and Geometry
On the multicanonical systems of
quasi-elliptic surfaces in characteristic 3
Toshiyuki Katsura
Faculty of Science and Engineering, Hosei University, Koganei-shi, Tokyo 184-8584, Japan
Abstract.
We consider the multicanonical systems of quasi-elliptic surfaces with Kodaira dimension in characteristic 3. We show that for any gives the structure of quasi-elliptic fiber space, and the number is best possible to give the structure for any such surfaces.
Research of the author is partially supported by Grant-in-Aid for Scientific Research (B) No. 15H03614.
Dedicated to Professor Piotr Pragacz on the occasion of his sixtieth birthday
1. Introduction
Let k be an algebraically closed field in characteristic , and let be an elliptic surface with Kodaira dimension over . Let be a canonical divisor of . We consider the multicanonical system . In Katsura-Ueno [4] and Katsura [3], we considered the following question:
Question 1.1*.*
Is there a positive integer such that if , then the multicanonical system gives a structure of elliptic surface for any elliptic surface over with ?
For the complex analytic case, Iitaka showed that and 86 is best possible (cf. Iitaka [2]). Namely, if is smaller than 86, then there exists an elliptic surface with such that does not give the structure of elliptic surface. In the case of algebraic elliptic surfaces, if the characteristic or , we showed that and 14 is best possible (Katsura-Ueno [4] and Katsura [3]). If , then we showed that and 12 is best possible (cf. Katsura [3]).
In this paper, we treat quasi-elliptic surfaces and consider the following similar question.
Question 1.2*.*
Is there a positive integer such that if , then the multicanonical system gives the structure of quasi-elliptic surface for any quasi-elliptic surface over with ?
Note that quasi-elliptic surfaces exist only in characterisitics and (cf. Bombieri-Mumford [1]). We show that if , then we have and the number is best possible. In characteristic , we have still some difficulties to determine the best possible number.
In Section 2, we summarize basic facts on quasi-elliptic surfaces. In Section 3, we examine the multicanonical system of quasi-elliptic surfaces in characteristic 3 and show our main theorem.
2. Some lemmas for quasi-elliptic surfaces
Let k be an algebraically closed field in characteristic and let be a quasi-elliptic surface defined over . Throughout this paper, we assume that any exceptional curve of the first kind is not contained in fibers. Such a surface exists if and only if is equal to 2 or 3, and the multiplicities of multiple fibers are all equal to (cf. Bombieri-Mumford [1]). We denote by the genus of the curve . In this section, we recall some facts on quasi-elliptic surfaces. We denote by the multiple fibers. Let be the torsion part of . Then, there exists a Cartier divisor on such that . The canonical divisor formula of is given by
[TABLE]
where with the length of the torsion part of and . Here, means linear equivalence. If is a tame multiple fiber, then we have . For details, see Bombieri-Mumford [1].
Lemma 2.1**.**
The Albanese variety of is isomorphic to the Jacobian variety of .
Proof.
Let be the Albanese mapping. If is a point, then by the universality of Albanese variety we see that the Jacobian variety of is also a point. Now, assume is not a point. Since the general fiber of is a rational curve with one cusp, the fibers are contracted by . Therefore, is a curve. We have, by the universality of Albanese variety, a commutative diagram:
[TABLE]
By this diagram, we have a morphism . Therefore, by the Stein factorization theorem, we see that is isomorphic to . Therefore, by the universality of Jacobian variety, we conclude (see also Katsura-Ueno [4], Lemma 3.4).
We find the following lemma and corollary in Lang [5] and Raynaud [6]. We give here an easy proof for the lemma.
Lemma 2.2**.**
Let be a quasi-elliptic surface over a non-singular complete curve with genus . Then, we have the inequality .
Proof.
By Noether’s formula and the self-intersection number of the first Chern class of , we have
[TABLE]
By Lemma 2.1, we have . Denoting by the Picard number of , we have also . Hence, we have .
Corollary 2.3**.**
(i)* If , then .*
(ii)* If , then .*
3. Multicanonical systems
In this section, let be an algebraically closed field of characteristic 3. Let be a quasi-elliptic surface defined over .
Example 3.1**.**
In characteristic 3, we consider the quasi-elliptic surface which is given by a non-singular complete model of the surface defined by
[TABLE]
Here, is a parameter of the base curve . By Lang [5] p.485, this surface has two tame multiple fibers at , and we have . We denote the two tame multiple fibers by and . The canonical divisor is given by
[TABLE]
Here, is a Cartier divisor on with such that , and is a general fiber of . Since we have , we see . Therefore, we have , and does not give the structure of quasi-elliptic surface. If , then we have , and gives the structure of quasi-elliptic surface.
We have the following theorem.
Theorem 3.2**.**
Assume that the characteristic . Then, for any quasi-elliptic surface with over and for any , the multicanonical system gives the unique structure of quasi-elliptic surface, and the number 5 is best possible.
Proof.
The method of the proof is similar to the one in Iitaka [2], Katsura-Ueno [4] and Katsura [3]. Since the Kodaira dimension is equal to 1, the structure of quasi-elliptic surface is unique. The Kodaira dimension of is equal to 1 if and only if
[TABLE]
Therefore, we need to find the least integer such that
[TABLE]
holds under the condition . Here, means the integral part of a real number . We have the following 6 cases:
Case (I)
Case (II-1)
Case (II-2)
Case (III-1)
Case (III-2)
Case (III-3)
We check under the condition for each case. In Case (I), by Lemma 2.2, we have . Hence, if , holds. In Case (II-1), if , holds. In Case (II-2), all multiple fibers are tame in this case, and we have at least one multiple fiber by . Since , holds for . In Case (III-1), holds for . In Case (III-2), since by Corollary 2.3, we have . Therefore, the number of wild fibers is less than or equal to . If there exists at least one tame multiple fiber, then holds for . If there exist no tame fibers and only one wild fiber, then by we have . Therefore, holds for . In Case (III-3), all multiple fibers are tame, and we have by . Therefore, holds for . The result on the best possible number in characteristic 3 follows from Example 3.1.
In characteristic 2, we can also consider a similar question to the one in characteristic 3. We have still difficulties to decide the best possible number. For example we need to solve the following question.
Question 3.3*.*
Does there exist a quasi-elliptic surface over an elliptic curve with only one tame multiple fiber and with in characteristic 2?
If there don’t exist such quasi-elliptic surfaces, then we can show that in characteristic 2, holds for and that the best possible number is equal to . Namely, we have in characteristic 2.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1[1] E. Bombieri and D. Mumford, Enriques’ classification of surfaces in char. p, II, In Complex Analysis and Algebraic Geometry (eds. by W.L. Baily, Jr., and T. Shioda), Iwanami Shoten, Tokyo, and Princeton Univ. Press, Princeton, NJ, 1977, 22-42.
- 2[2] S. Iitaka, Deformations of compact complex surfaces, II, J. Math. Soc. Japan , 22 (1970), 247-261.
- 3[3] T. Katsura, Multicanonical systems of elliptic surfaces in small characteristics, Compositio Math. , 97 (1995), 119-134.
- 4[4] T. Katsura and K. Ueno, On elliptic surfaces in characteristic p, Math. Ann. , 272 (1985), 291-330.
- 5[5] W. Lang, Quasi-elliptic surfaces in characteristic three, Ann. Scient. Ec. Norm. Sup. , 12 (1979), 473-500.
- 6[6] M. Raynaud, Surfaces elliptiques et quasi-elliptiques, manuscript.
