CANONICAL BUNDLE FORMULA AND VANISHING THEOREM

Transcription

1 CANONICAL BUNDLE FORMULA AND VANISHING THEOREM OSAMU FUJINO Abstract. In this paper, we treat two different topics. We give sample computations of our canonical bundle formula. They help us understand our canonical bundle formula, Fujita Kawamata s semi-positivity theorem, and Viehweg s weak positivity theorem. We also treat Viehweg s vanishing theorem. We do not claim any originality of the results in this paper. Contents. Introduction 2. Sample computations of canonical bundle formula 2 3. Viehweg s vanishing theorem 8 References. Introduction In this paper, we treat two different topics. In Section 2, we give sample computations of our canonical bundle formula (cf. [FM]). The examples constructed in Section 2 help us understand our canonical bundle formula, Fujita Kawamata s semi-positivity theorem (cf. [K, Theorem 5]), and Viehweg s weak positivity theorem (cf. [V2, Theorem III and Theorem 4.]). There are no new results in Section 2. In Section 3, we treat Viehweg s vanishing theorem (cf. [V, Theorem IV]) which was obtained by Viehweg as a byproduct of his original proof of the Kawamata Viehweg vanishing theorem. It is a consequence of Bogomolov s vanishing theorem. There exists a generalization of Viehweg s vanishing theorem. See, for example, [EV, (2.3) Theorem] and [EV2, Corollary 5.2 d)]. Here, we quickly give a proof of the generalized Viehweg vanishing theorem as an application of the usual Kawamata Viehweg vanishing theorem. By our proof, we see that the Date: 2009/7/2, Version Mathematics Subject Classification. Primary 4F7; Secondary 4N30.

2 2 OSAMU FUJINO generalized Viehweg vanishing theorem is essentially the same as the Kawamata Viehweg vanishing theorem. There are no new results in Section 3. We note that Sections 2 and 3 are mutually independent. Although this paper contains no new results, we hope that the examples and the arguments will be useful. Notation. For a Q-divisor D = r j= d jd j such that D j is a prime divisor for every j and D i D j for i j, we define the round-down D = r j= d j D j, where for every rational number x, x is the integer defined by x < x x. The fractional part {D} of D denotes D D. In Section 3, κ (resp. ν) denotes the Kodaira dimension (resp. numerical Kodaira dimension). Acknowledgments. The author was partially supported by The Sumitomo Foundation, The Inamori Foundation, and by the Grant-in-Aid for Young Scientists (A) from JSPS. He thanks Takeshi Abe for answering my questions. We will work over C, the complex number field, throughout this paper. 2. Sample computations of canonical bundle formula We give sample computations of our canonical bundle formula obtained in [FM]. We will freely use the notation in [FM]. For the details of our canonical bundle formula, see [FM], [F, 3], and [F2, 3, 4, 5, and 6]. 2. (Kummer manifolds). Let E be an elliptic curve and E n the n- times direct product of E. Let G be the cyclic group of order two of analytic automorphisms of E n generated by an automorphism g : E n E n : (z,, z n ) ( z,, z n ). The automorphism g has 2 2n fixed points. Each singular point is terminal for n 3 and is canonical for n (Kummer surfaces). First, we consider q : E 2 /G E/G P, which is induced by the first projection, and g = q µ : Y P, where µ : Y E 2 /G is the minimal resolution of sixteen A -singularities. It is easy to see that Y is a K3 surface. In this case, it is obvious that g O Y (mk Y/P ) O P (2m) for every m. Thus, we can put L Y/P = D for any degree two Weil divisor D on P. We obtain K Y = g (K P + L Y/P ). Let Q i be the

3 CANONICAL BUNDLE FORMULA AND VANISHING THEOREM 3 branch point of E E/G P for i 4. Then we have 4 L ss Y/P = D ( 4 2 )Q i = D 2 Q i by the definition of the semi-stable part L ss Y/P. Therefore, we obtain Thus, but 4 K Y = g (K P + L ss Y/P + 2 Q i). L ss Y/P = D 2L ss Y/P = 2D 4 2 Q i 0 4 Q i 0. Note that L ss Y/P is not a Weil divisor but a Q-Weil divisor on P. 2.3 (Elliptic fibrations). Next, we consider E 3 /G and E 2 /G. We consider the morphism p : E 3 /G E 2 /G induced by the projection E 3 E 2 : (z, z 2, z 3 ) (z, z 2 ). Let ν : X E 3 /G be the weighted blow-up of E 3 /G at sixty-four (,, )-singularities. Thus 2 K X = ν K E 3 /G + 64 j= 2 E j, where E j P 2 is the exceptional divisor for every j. Let P i be an A -singularity of E 2 /G for i 6. Let ψ : X X be the blow-up of X along the strict transform of p (P i ), which is isomorphic to P, for every i. Then we obtain the following commutative diagram. Note that K X = φ K E 3 /G + E 3 /G φ:=ν ψ X p f E 2 /G µ Y 64 j= 2 E j + where E j is the strict transform of E j on X and F k is the ψ-exceptional prime divisor for every k. We can check that X is a smooth projective threefold. We put C i = µ (P i ) for every i. It can be checked that 6 k= F k,

4 4 OSAMU FUJINO C i is a ( 2)-curve for every i. It is easily checked that f is smooth outside 6 C i and that the degeneration of f is of type I0 along C i for every i. We renumber {E j } 64 j= as {Ej i }, where f(ej i ) = C i for every i 6 and j 4. We note that f is flat since f is equi-dimensional. Let us recall the following theorem (cf. [K2, Theorem 20] and [N, Corollary 3.2. and Theorem 3.2.3]). Theorem 2.4 (..., Kawamata, Nakayama,...). We have the following isomorphism. (f ω ) 2 O Y ( 6 6C i ), where ω O X (K ) = O X (K X f K Y ). The proof of Theorem 2.4 depends on the investigation of the upper canonical extension of the Hodge filtration and the period map. It is obvious that 6 2K X = f (2K Y + C i ) and 2mK X = f (2mK Y + m for all m since f C i = 2F i + 4 j= Ej i. Therefore, we have 2L 6 C i. On the other hand, f ω O Y ( L ). Note that Y is a smooth surface and f is flat. Since we have 6 O Y (2 L ) (f ω ) 2 O Y ( C i ) 6 6 2L 6 C i 2 L. 6C i ), Thus, L is a Weil divisor on Y. It is because the fractional part {L } is effective and linearly equivalent to zero. So, L is numerically equivalent to 6 2 C i. We have g Q i = 2G i + 4 j= Cj i. Here, we renumbered {C j } 6 j= as {Cj i }4 i,j= such that g(cj i ) = Q i for every i and j. More precisely, we put 2G i = g Q i 4 j= Cj i for every i. We note that we used the notations in 2.2. We consider A := g D 4 G i. Then A is a Weil divisor and 2A 6 C i. Thus, A is numerically

5 CANONICAL BUNDLE FORMULA AND VANISHING THEOREM 5 equivalent to 2 So, we have 6 C i. Since H (Y, O Y ) = 0, we can put L = A. L ss = g D 4 G i 6 j= 2 C j. We obtain the following canonical bundle formula. Theorem 2.5. The next formula holds. 6 K X = f (K Y + L ss + 2 C j), where L ss = g D 4 G i 6 j= C 2 j. We note that 2L ss 0 but Lss j= 0. The semi-stable part Lss is not a Weil divisor but a Q-divisor on Y. The next lemma is obvious since the index of K E 3 /G is two. We give a direct proof here. Lemma 2.6. H 0 (Y, L ) = 0. Proof. If there exists an effective divisor B on Y such that L B. Since B C i =, we have B C 2 i for all i. Thus B 6 C 2 i. This implies that B 6 C 2 i is an effective Q-divisor and is numerically equivalent to zero. Thus B = 6 C 2 i. It is a contradiction. We can easily check the following corollary. Corollary 2.7. We have f ω m In particular, f ω m { O Y ( 6 nc i) if m = 2n, O Y (L + 6 nc i) if m = 2n +. is not nef for any m. We can also check that { C if m is even, H 0 (Y, f ω m ) 0 if m is odd. 2.8 (Weak positivity). Let us recall the definition of Viehweg s weak positivity (cf. [V2, Definition.2] and [V4, Definition 2.]). Definition 2.9 (Weak positivity). Let W be a smooth quasi-projective variety and F a locally free sheaf on W. Let U be an open subvariety of W. Then, F is weakly positive over U if for every ample invertible sheaf H and every positive integer α there exists some positive integer

6 6 OSAMU FUJINO β such that S α β (F) H β is generated by global sections over U. This means that the natural map is surjective over U. H 0 (W, S α β (F) H β ) O W S α β (F) H β Remark 2.0 (cf. [V2, (.3) Remark. iii)]). In Definition 2.9, it is enough to check the condition for one invertible sheaf H, not necessarily ample, and all α > 0. For the details, see [V4, Lemma 2.4 a)]. Remark 2.. In [V3, Definition 3.], S α β (F) H β is only required to be generically generated. See also [M, (5.) Definition]. We explicitly check the weak positivity for the elliptic fibration constructed in 2.3 (cf. [V2, Theorem 4. and Theorem III] and [V4, Theorem 2.4 and Corollary 2.45]). Proposition 2.2. Let f : X Y be the elliptic fibration constructed in 2.3. Then f ω m is weakly positive over Y 0 = Y \ 6 C i. Let U be a Zariski open set such that U Y 0. Then f ω m is not weakly positive over U. Proof. Let H be a very ample Cartier divisor on E 2 /G and H a very ample Cartier divisor on Y such that L +H is very ample. We put H = O Y (µ H + H ). Let α be an arbitrary positive integer. Then S α (f ω m ) H O 6 Y (α nc i + µ H + H ) if m = 2n. When m = 2n +, we have S α (f ω m { ) H O Y (α 6 nc i + µ H + H + L + α 6 2 C i) if α is odd, O Y (α 6 nc i + µ H + H + α 6 2 C i) if α is even. Thus, S α (f ω m ) H is generated by global sections over Y 0 for every α > 0. Therefore, f ω m is weakly positive over Y 0. Let H be an ample invertible sheaf on Y. We put k = max(c j H). j Let α be a positive integer with α > k/2. We note that S 2α β (f ω m ) 6 H β (O Y (α mc i ) H) β.

7 CANONICAL BUNDLE FORMULA AND VANISHING THEOREM 7 If H 0 (Y, S 2α β (f ω m ) H β ) 0, then we can take 6 G (O Y (α mc i ) H) β. In this case, G C i < 0 for every i because α > k/2. Therefore, we obtain G 6 C i. Thus, S 2α β (f ω m ) H β is not generated by global sections over U for any β. This means that f ω m is not weakly positive over U. Proposition 2.2 implies that [V4, Corollary 2.45] is the best result. 2.3 (Semi-positivity). We give a supplementary example for Fujita Kawamata s semi-positivity theorem (cf. [K, Theorem 5]). For details of Fujita Kawamata s semi-positivity theorem, see, for example, [M, 5] and [F3, Section 5]. Definition 2.4. A locally free sheaf E on a projective variety V is semi-positive (or nef) if the tautological line bundle O PV (E)() is nef on P V (E). For details of semi-positive locally free sheaves, see [V4, Proposition 2.9]. Example 2.5. Let f : X Y be the elliptic fibration constructed in 2.3. Let Z := X C, where C is a smooth projective curve with the genus g(c) = r 2. Let π : Z C (resp. π 2 : Z X) be the first (resp. second) projection. We put h := f π 2 : Z Y. In this case, K Z = π K C π 2K X. Therefore, we obtain h ω m Z/Y = f π 2 (πω m C π 2ω m X ) ω m Y = (f ω m ) l, where l = dim H 0 (C, O C (mk C )). Thus, l = (2m )r 2m + if m 2 and l = r if m =. So, h ω Z/Y is a rank r 2 vector bundle on Y such that h ω Z/Y is not semi-positive. We note that h is smooth over Y 0 = Y \ 6 C i. We also note that h ω m Z/Y is weakly positive over Y 0 for every m by [V4, Theorem 2.4 and Corollary 2.45]. Example 2.5 shows that the assumption on the local monodromies around 6 C i is indispensable for Fujita Kawamata s semi-positivity theorem (cf. [K, Theorem 5 (iii)]). We close this section with a correction on [FM]. 2.6 (Correction). We give a remark on [FM, Section 4]. In [FM, 4.4], g : Y X is a log resolution of (X, ). However, it is better to assume that g is a log resolution of (X, (/b)b ) for the proof of [FM, Theorem 4.8].

8 8 OSAMU FUJINO 3. Viehweg s vanishing theorem In this section, we quickly give a proof of the generalized Viehweg vanishing theorem (cf. [EV, (2.3) Theorem]) as an application of the usual Kawamata Viehweg vanishing theorem. See also [EV2, Corollary 5.2 d)]. Our proof is different from the proofs given in [EV] and [EV2]. Theorem 3. (cf. [EV, (2.3) Theorem]). Let π : X S be a proper surjective morphism from a smooth variety X, L an invertible sheaf on X, and D an effective Cartier divisor on X such that SuppD is normal crossing. Assume that L N ( D) is π-nef for some positive integer N and that κ(x η, (L () ) η ) = m, where X η is the generic fiber of π, (L () ) η = L () Xη, and L () = L( D N ). Then we have R i π (L () ω X ) = 0 for i > dim X dim S m. We note that SuppD is not necessarily simple normal crossing. We only assume that SuppD is normal crossing. Remark 3.2. In Theorem 3., we assume that S is a point for simplicity. We note that κ(x, L () ) = m does not necessarily imply κ(x, L (i) ) = m for 2 i N, where L (i) = L i ( id N ). Therefore, the original arguments in [V] depending on Bogomolov s vanishing theorem do not seem to work in our setting. Let us recall the Kawamata Viehweg vanishing theorem. Although there are many formulations of the Kawamata Viehweg vanishing theorem, the following one is the most convenient one for various applications of the log minimal model program. Theorem 3.3 (Kawamata Viehweg s vanishing theorem). Let f : Y X be a projective morphism from a smooth variety Y and M a Cartier divisor on Y. Let be an effective Q-divisor on Y such that Supp is simple normal crossing and = 0. Assume that M (K X + ) is f-nef and f-big. Then for all i > 0. R i f O Y (M) = 0

9 CANONICAL BUNDLE FORMULA AND VANISHING THEOREM 9 We note that we can prove Theorem 3.3 without using Viehweg s vanishing theorem. See, for example, [KMM, Theorem -2-3]. The reader can find a log canonical generalization of the Kawamata Viehweg vanishing theorem in [F4, Theorem 2.48]. Remark 3.4. In the proof of [KMM, Theorem -2-3], when we construct completions π : X S with π X = π and a π-ample Q-divisor D with D X = D, it seems to be better to use Szabó s resolution lemma. It is because we have to make the support of the fractional part of D have only simple normal crossings. Remark 3.5. It is obvious that Theorem 3.3 is a special case of Theorem 3.. In this section, we give a quick proof of Theorem 3. only by using Theorem 3.3 and Hironaka s resolution. Therefore, Theorem 3. is essentially the same as Theorem 3.3. Let us start the proof of Theorem 3.. Proof. Without loss of generality, we can assume that S is affine. Let f : Y X be a proper birational morphism from a smooth quasiprojective variety Y such that Suppf D Exc(f) is simple normal crossing. We write K Y = f (K X + ( ε){ D N }) + E ε. Then F = E ε is an effective exceptional Cartier divisor on Y and independent of ε for 0 < ε. Therefore, the coefficients of F E ε are continuous for 0 < ε. Let L be a Cartier divisor on X such that L O X (L). We can assume that κ(x η, (L D ) N η) = m 0. Let Φ : X Z be the relative Iitaka fibration over S with respect to l(l D ), where l is a sufficiently large and divisible integer. We can N further assume that f (L D N ) Q ϕ A + E, where E is an effective Q-divisor such that SuppE Suppf D Exc(f) is simple normal crossing, ϕ = Φ f : Y Z is a morphism, and A is a ψ-ample Q-divisor on Z with ψ : Z S. Let i E i = SuppE Suppf D Exc(f) be the irreducible decomposition. We can write E ε = i aε ie i and E = i b ie i and note that a ε i is continuous for 0 < ε. We put ε = F E ε + εe. By the definition, we can see that every coefficient of ε is in [0, 2) for 0 < ε. Thus, ε is reduced. If a ε i < 0, then a ε i + for 0 < ε. Therefore, if N a ε i a ε i + εb i for 0 < ε, then a ε i > 0. Thus, F = F ε

10 0 OSAMU FUJINO is effective and f-exceptional for 0 < ε. On the other hand, (Y, { ε }) is obviously klt. We note that f (K X + L D N ) + F (K Y + { ε }) = f (K X + L D N ) + F f (K X + ( ε){ D N }) E ε (F E ε + εe) Q ( ε)f (L D N ) + εϕ A for a rational number ε with 0 < ε. We put M = f (K X + L D N ) + F. By combining the long exact sequence R i p O Y (M) R i p O Y (M + H) R i p O H (M + H) obtained from 0 O Y (M) O Y (M + H) O H (M + H) 0 for a p-ample general smooth Cartier divisor H on Y, where p = ψ ϕ = π f : Y S, and the induction on the dimension, we obtain R i p O Y (M) = R i p O Y (f (K X + L D N ) + F ) = 0 for every i > dim Y dim S m = dim X dim S m by Theorem 3.3 (cf. [V, Remark 0.2]). We note that M (K Y + { ε }) Q ( ε)f (L D N ) + εϕ A, (M + H) (K Y + { ε }) Q ( ε)f (L D N ) + εϕ A + H, and (M + H) H (K H + { ε } H ) Q ( ε)f (L D N ) H + εϕ A H. We also note that (H, { ε } H ) is klt and κ(h η, (ϕ A) Hη ) min{m, dim H η }. On the other hand, R i f O Y (M) = R i f O Y (f (K X + L D N ) + F ) = 0

11 CANONICAL BUNDLE FORMULA AND VANISHING THEOREM for every i > 0 by Theorem 3.3. We note that f O Y (f (K X + L D N ) + F ) O X (K X + L D N ) by the projection formula because F is effective and f-exceptional. Therefore, we obtain R i π O X (K X + L D N ) = Ri p O Y (M) = 0 for every i > dim X dim S m. We close this section with an obvious corollary. Corollary 3.6. Let X be an n-dimensional smooth complete variety, L an invertible sheaf on X. Assume that D L N for some positive integer N and that SuppD is simple normal crossing. Then we have for i > n κ(x, { D N }). H i (X, L () ω X ) = 0 References [EV] H. Esnault, E. Viehweg, Logarithmic de Rham complexes and vanishing theorems, Invent. Math. 86 (986), no., [EV2] H. Esnault, E. Viehweg, Lectures on vanishing theorems, DMV Seminar, 20. Birkhäuser Verlag, Basel, 992. [F] O. Fujino, Algebraic fiber spaces whose general fibers are of maximal Albanese dimension, Nagoya Math. J. 72 (2003), 27. [F2] O. Fujino, A canonical bundle formula for certain algebraic fiber spaces and its applications, Nagoya Math. J. 72 (2003), [F3] O. Fujino, Remarks on algebraic fiber spaces, J. Math. Kyoto Univ. 45 (2005), no. 4, [F4] O. Fujino, Introduction to the log minimal model program for log canonical pairs, preprint (2009). [FM] O. Fujino, S. Mori, A canonical bundle formula, J. Differential Geom. 56 (2000), no., [K] Y. Kawamata, Characterization of abelian varieties, Compositio Math. 43 [K2] (98), no. 2, Y. Kawamata, Kodaira dimension of certain algebraic fiber spaces, J. Fac. Sci. Univ. Tokyo Sect. IA Math. 30 (983), no., 24. [KMM] Y. Kawamata, K. Matsuda, and K. Matsuki, Introduction to the minimal model problem, Algebraic geometry, Sendai, 985, , Adv. Stud. Pure Math., 0, North-Holland, Amsterdam, 987. [M] [N] S. Mori, Classification of higher-dimensional varieties, Algebraic geometry, Bowdoin, 985 (Brunswick, Maine, 985), , Proc. Sympos. Pure Math., 46, Part, Amer. Math. Soc., Providence, RI, 987. N. Nakayama, Local structure of an elliptic fibration, Higher dimensional birational geometry (Kyoto, 997), , Adv. Stud. Pure Math., 35, Math. Soc. Japan, Tokyo, 2002.

12 2 OSAMU FUJINO [V] E. Viehweg, Vanishing theorems, J. Reine Angew. Math. 335 (982), 8. [V2] E. Viehweg, Weak positivity and the additivity of the Kodaira dimension for certain fibre spaces, Algebraic varieties and analytic varieties (Tokyo, 98), , Adv. Stud. Pure Math.,, North-Holland, Amsterdam, 983. [V3] E. Viehweg, Weak positivity and the additivity of the Kodaira dimension. II. The local Torelli map, Classification of algebraic and analytic manifolds (Katata, 982), , Progr. Math., 39, Birkhäuser Boston, Boston, MA, 983. [V4] E. Viehweg, Quasi-projective moduli for polarized manifolds, Ergebnisse der Mathematik und ihrer Grenzgebiete (3) [Results in Mathematics and Related Areas (3)], 30. Springer-Verlag, Berlin, 995. Department of Mathematics, Faculty of Science, Kyoto University, Kyoto , Japan address: