• 検索結果がありません。

Then, we have Z U log p q dVωU ≤ 1 e +CUVol(X, ω)(logCU + logC)

N/A
N/A
Protected

Academic year: 2021

シェア "Then, we have Z U log p q dVωU ≤ 1 e +CUVol(X, ω)(logCU + logC)"

Copied!
4
0
0

読み込み中.... (全文を見る)

全文

(1)

A MEMORANDUM ON THE INVARIANCE OF PLURIGENERA

(PRIVATE NOTE)

OSAMU FUJINO

Proposition 1. LetX be a complex manifold. Letp, qbe smooth non- negative functions on X such that q 6= 0 almost everywhere. Assume that

Vol(X, ω) = Z

X

1dVω <∞

and

Z

X

p

qdVω ≤C

holds for some constant C ≥1, wheredVω is a volume form on X. Let U ⊂X be any open set of X and dVωU a volume form on U such that dVωU ≤CUdVω for some constant CU ≥1. Then, we have

Z

U

log p

q

dVωU ≤ 1

e +CUVol(X, ω)(logCU + logC).

Moreover, letai be a smooth nonnegative function onX for1≤i≤k such that ai 6= 0 almost everywhere for 1≤ i≤ k−1. We put a0 ≡1 and a=ak. Assume that

Z

X

ai+1

ai

dVω ≤C

holds for 0≤i≤k−1. Then 1

k Z

U

loga dVωU ≤ 1

e +CUVol(X, ω)(logCU+ logC).

Date: 2006/3/28, Version 1.3.

The arguments in this note are indispensable when we obtain local uniform supre- mum norm estimates from the globalL2-estimates. I like intrinsic formulations.

1

(2)

2 OSAMU FUJINO

Proof. By Jensen’s inequality, Z

U

log p

q

dVωU

Vol(U, ωU) ≤log Z

U

p q

dVωU

Vol(U, ωU)

= log

1 Vol(U, ωU)

+ log

Z

U

p qdVωU

≤log

1 Vol(U, ωU)

+ log

Z

X

p

qCUdVω

≤log

1 Vol(U, ωU)

+ logCU + logC, where Vol(U, ωU) =R

U1dVωU <∞. Therefore, we have Z

U

log p

q

dVωU ≤Vol(U, ωU) log

1 Vol(U, ωU)

+ Vol(U, ωU)(logCU + logC).

Lemma 2. We consider g(x) = xlog 1x

for x > 0. Then g(x) ≤ g(1e) = 1e for any x > 0. Moreover, g(x) >0 for x < 1, g(1) = 0, and g(x)<0 for x >1. We note that g(x)→0 as x→0.

Proof. We have g(x) = −xlogx and g0(x) = −logx− 1. Thus, we

obtain the desired properties.

Therefore, we obtain Z

U

log p

q

dVωU ≤ 1

e +CUVol(X, ω)(logCU + logC).

The latter statement is obvious.

Theorem 3. InProposition 1, we further assume thatlogais a quasi- psh function. Let Y be a relatively compact open set of X. Then there exists a positive constant C0 such that

sup

x∈Y

1

kloga(x)≤C0 <∞.

Proof. Let Wα b Uα b X be relatively compact open sets of X for 1 ≤ α ≤ N. Assume that Y ⊂ SN

α=1Wα and Uα can be seen as a domain in Cn for any α, where n = dimX, and L can be trivialized on each Uα. Let dVωα be the Euclidean volume form on Uα ⊂Cn. We note that we can assume that dVωα ≤CαdVω onUα for some constant Cα ≥ 1 (if we need, we can shrink Uα). Take a point x ∈ Y. Then there is an α such that x ∈ Wα. On each Uα, we can further assume that loga = uα +vα, where uα is a psh function and vα is a smooth

(3)

A MEMO ON THE INVARIANCE OF PLURIGENERA 3

function. For every x ∈ Wα, there exists an open polydisk Ux whose center is x such thatUx bUα and

Vol(Ux, ωα) = Z

Ux

1dVωα =C0.

Note that the positive constant C0 is independent of x ∈Wα. By the sub-mean-value property of uα, we have

uα(x)≤ 1 Vol(Ux, ωα)

Z

Ux

uαdVωα = 1 C0

Z

Ux

uαdVωα. Note that

1

kvα(x)≤C1

for any x∈Wα and

1 kC0

Z

Ux

|vα|dVωα ≤C2

for any Ux, where C1 and C2 are positive constants independent of x and Ux. Thus, we obtain

1

k loga(x) = 1

kuα(x) + 1 kvα(x)

≤ 1 kC0

Z

Ux

uαdVωα +C1

≤ 1 kC0

Z

Ux

loga dVωα +C1+C2

≤ 1 C0

1

e +CαVol(X, ω)(logCα+ logC)

+C1+C2

=: Cα. Thus we have

sup

x∈Wα

1

kloga(x)≤Cα.

Then we obtain the upper bound C0 = maxαCα <∞.

Remark 4. LetL be a holomorphic line bundle onX andha smooth hermitian metric on L. Let sj be a holomorphic section of L onX for 1 ≤ j ≤ l. Then logPl

j=1|sj|2h is a quasi-psh function on X, where

|sj|h is the pointwise norm of sj with respect to h. For geometric applications, we use Theorem 3 for a=Pl

j=1|sj|2h.

Acknowledgments. I was partially supported by The Sumitomo Foun- dation and by the Grant-in-Aid for Young Scientists (A) ]17684001 from JSPS.

(4)

4 OSAMU FUJINO

Graduate School of Mathematics, Nagoya University, Chikusa-ku Nagoya 464-8602 Japan

E-mail address: [email protected]

参照

関連したドキュメント

Considering singular terms at 0 and permitting p 6= 2, Loc and Schmitt [17] used the lower and upper solution method to show existence of solution for (1.1) with the nonlinearity of

Lair and Shaker [10] proved the existence of large solutions in bounded domains and entire large solutions in R N for g(x,u) = p(x)f (u), allowing p to be zero on large parts of Ω..

We prove a continuous embedding that allows us to obtain a boundary trace imbedding result for anisotropic Musielak-Orlicz spaces, which we then apply to obtain an existence result

1 Miko laj Marciniak was supported by Narodowe Centrum Nauki, grant number 2017/26/A/ST1/00189 and.. Narodowe Centrum Bada´ n i Rozwoju, grant

Rhoudaf; Existence results for Strongly nonlinear degenerated parabolic equations via strong convergence of truncations with L 1 data..

In this paper, we study the existence and nonexistence of positive solutions of an elliptic system involving critical Sobolev exponent perturbed by a weakly coupled term..

Indeed, under the hypotheses from Example 8.3, we obtain (via the mountain pass theorem) the existence of a nontrivial solution for the problem (1.2), (1.3), while Example 8.4

It is not a bad idea but it means that since a differential field automorphism of L|[x 0 ] is given by a birational transformation c 7→ ϕ(c) of the space of initial conditions, we