THE MONGE-AMPÈRE OPERATOR AND GEODESICS IN THE SPACE OF KÄHLER POTENTIALS
D.H. Phong
*
and Jacob Sturm
†
November 27, 2006
1 Introduction
Let
be an
-dimensional compact complex manifold,
a positive holomorphic line bundle, and
the space of positively curved hermitian metrics on
. The purpose of this article is to prove that geodesics in the infinite-dimensional symmetric space
can be uniformly approximated by geodesics in the finite-dimensional symmetric spaces
, where
. Thus the
are becoming flat as
.
The motivation for this work comes from Donaldson's far reaching program [9, 10] relating the geometry of
to the existence and uniqueness of constant scalar curvature Kähler metrics. As advocated by S.T. Yau over the years,
should be approximated by
, and the properties of this approximation should be closely reflected in many basic questions of Kähler geometry. In particular, the condition of “stability” is one which concerns the growth of energy functionals along the geodesics of
. On the other hand, the existence and uniqueness of metrics of constant scalar curvature concerns the growth of energy functionals along the geodesics of
(see [11, 12, 24, 25, 26] ). Thus a good understanding of the relationship between these different types of geodesics is desirable.
In order to state the precise theorem, we need some notation: Let
be a smooth hermitian metric. If
we write
and for
, we denote by
the induced metric on
. The curvature of
is the
form on
defined locally by the formula
, where
is a local, nowhere vanishing holomorphic section. In particular,
. Let
If we fix
and let
we have a natural isomorphism
where
is identified with
so that
. Then
is an infinite dimensional manifold whose tangent space
at
is naturally identified with
.
Now let
and
and define a metric on
by
|
(1.1)
|
Donaldson [10] , Mabuchi [18] and Semmes [23] have shown that ( 1.1 ) defines a Riemannian metric which makes
into an infinite dimensional negatively curved symmetric space.
Furthermore, the geodesics of
in this metric are the paths
which satisfy the partial differential equation
|
(1.2)
|
The space
contains a canonical family of finite-dimensional negatively curved symmetric spaces
which are defined as follows: For
and for
an ordered basis of
, let
be the Kodaira embedding given by
. Then we have a canonical isomorphism
. Let
is the Fubini-Study metric on
and let
|
(1.3)
|
Note that the right side of ( 1.3 ) is independent of the choice of
. In particular
|
(1.4)
|
Let
Then
is a finite-dimensional negatively curved symmetric space sitting inside of
. It is well known that the
are topologically dense in
:
If
then there exists
such that
in the
topology. This follows from the Tian-Yau-Zelditch theorem on the density of states (see also Catlin [4] for corresponding results for the Bergman kernel). In fact, if
, then there is a canonical choice of the approximating sequence
: Let
be a basis of
which is orthonormal with respect to the metrics
. In other words,
|
(1.5)
|
Note that
is not unique: if
is orthonormal, so is
, for any
. Define
. Then Theorem 1 of [27] says that for
fixed, we have a
asymptotic expansion as
:
|
(1.6)
|
where the
are smooth functions on
defined locally by
. Let
and
. In particular, ( 1.3 ) and ( 1.6 ) imply that for each
|
(1.7)
|
where the norms are all taken with respect to
. Here, as before,
,
,
and
.
Now let
. It is known by the work of Chen [5] (see also more recent progress in Donaldson [13] and Chen-Tian [6] ) that there is a unique
geodesic
joining
to
which, according to a conjecture of Donaldson [10] , should be smooth.
Our main theorem says that
is a uniform limit of the smooth geodesics in
which join
to
.
To be precise, let
be the change of basis matrix defined by
.
Without loss of generality, we may assume that
is diagonal with entries
for some
. Let
where
is diagonal with entries
. Define
|
(1.8)
|
Then
is the smooth geodesic in
joining
to
.
Explicitly, using ( 1.3 ), we can also write
|
(1.9)
|
Theorem 1
Let
and
the unique
geodesic joining
to
.
Then
|
(1.10)
|
where, for any bounded function
, we define the upper envelope
of
by
.
Remark 1. The proof will show that
|
(1.11)
|
Remark 2. We note that the upper envelope
is independent of the choice of coordinate systems defining the balls
. It is the smallest upper semicontinuous function which is greater than or equal to
. If
is a sequence of plurisubharmonic functions which are locally uniformly bounded, then
is plurisubharmonic and equal to
almost everywhere. Similarly, we also define for later use the lower envelope
of a bounded function
by
. The function
is the largest lower semi-continuous function which is less than or equal to
.
To prove the theorem we first apply an observation of Donaldson [10] which shows that solving the geodesic equation on
is equivalent to solving the degenerate Monge-Ampère equation
on the manifold
, where
is the annulus
, the values of
are prescribed on the boundary of
by smooth rotationally symmetric data, and
is positive for every
. Here we are writing
where
and
is a smooth on
. We then use the Tian-Yau-Zelditch theorem to prove that the
geodesics are, in a certain sense, approximate solutions to ( 1.12 ). A key step is then proving that the limit of the sequence of approximate solutions is in fact a weak solution of the Monge-Ampère equation and that the weak solution thus obtained is unique. This is accomplished using the methods of pluripotential theory which were introduced and systematically developed by Bedford-Taylor in their fundamental work [1] and [2] . In recent years, this subject has been extended by Demailly [8] , Klimek [15] , Blocki [3] , and others. The uniqueness implies that the limit coincides with the
solution
. Now the uniqueness theorem for bounded open sets in
, which is due to Bedford-Taylor, requires smoothing techniques for rough solutions which do not generalize in a simple way to manifolds. It is quite possible that the regularization methods of Demailly, which have recently been successfully applied by Guedj and Zeriahi [14] in the setting of compact manifolds, will also apply in our setting. However, we use a different approach which exploits the particular structure of the weak solutions in our case. A particularly important ingredient for us is the existence of a vector field
, transversal to the boundary of
, with
|
(1.13)
|
where
is a sequence of approximate solutions to the Dirichlet problem for the Monge-Ampère equation (c.f. Theorem 3 and the proof of Theorem 6 below). For the application to Theorem 1, this hypothesis is a consequence of the a priori estimate
|
(1.14)
|
which is the one global estimate for the derivatives of
that we can actually obtain.
One approach to Donaldson's conjecture on the existence of smooth geodesics is the establishment of a priori estimates on the smooth approximate solutions
. As we just noted, the uniform
estimates for
and for
do hold. But extending these to higher derivative estimates seems rather hard, and at the same time, quite intriguing:
For example, the estimate which is needed for
can be carried out in certain special cases, and appears to be related to the central limit theorem in probability theory. In the last section, we shall make some remarks along these lines.
2 The volume estimate
In this section, we establish the basic properties of the functions
introduced in ( 1.9 ).
The properties of the functions
themselves are summarized in Lemma 1 below, while the properties of their corresponding Monge-Ampere measures are given in Theorem 2.
A first key ingredient is the following very general volume formula. Let
be a compact Kähler manifold of dimension
,
,
,
be the projection on the first factor, and
. Then
is a closed positive
-form on
such that
.
Let
be a smooth path joining
to
and define
by
|
(2.1)
|
and the corresponding
-form
by
|
(2.2)
|
In local coordinates
for
and
for
, if we identify a Kähler form
with the hermitian matrix
, then we can write
where
is the column vector
and
is the conjugate transpose of
. It follows that
|
(2.3)
|
and the condition
is equivalent to the geodesic equation
. Now
|
(2.4)
|
and we obtain the desired volume formula:
|
(2.5)
|
Note that ( 2.5 ) is only a true volume if
, but that ( 2.5 ) is valid even in the absence of this hypothesis.
It is perhaps noteworthy that
, where
, and
is precisely the functional whose critical points (with the constraint
, and
is a fixed function satisfying
), give Kähler-Einstein metrics.
Thus the preceding relation can be rewritten as
|
(2.6)
|
The asymptotic behavior of
along geodesics in
is known to be closely related to Chow-Mumford stability (see [28] , and also [19] , [20] ,[22] ). The simple formula ( 2.6 ) raises the intriguing possibility that the behavior of the Monge-Ampère operator
and geodesics in
can be linked even more directly to stability and constant scalar curvature Kähler metrics.
Fix now two potentials
and
in
, and let
,
be bases for
which are orthonormal with respect to the metrics induced by
and
. As in the Introduction, let
be the corresponding change of bases,
. We may assume that
is diagonal with eigenvalues
. We consider the functions
defined as in ( 1.9 ).
Lemma 1
Let
where
.
(a) For each
, let
be the extension of
to
as in ( 2.1 ), and let
is the corresponding
-form as in ( 2.2 ). Then
is a smooth positive
-form,
In particular, the
-form
form is a positive smooth measure on
; (b) There is a constant
which does not depend on
such that
|
(2.8)
|
(c) With the same constant
, we also have
|
(2.9)
|
Proof. Locally on
, we have
, where
is a holomorphic function. Since the logarithms of sums of squares of absolute values of holomorphic functions are plurisubharmonic in the usual sense on
, and since we can write
| |
|
(2.10)
|
where
is just the absolute value of
in a local trivialization, the positivity of
follows. Here we have simplified the notation by denoting
just by
.
We turn next to the proof of (b). Let
, and order the eigenvalues
so that
. Clearly,
|
(2.11)
|
By the Tian-Yau-Zelditch theorem, the expression in the middle tends to
as
.
Thus, if we set
and
, we have
and
, for
large enough.
To get inequalities in the opposite direction, we note that for each
,
|
(2.12)
|
This implies that for some
, we have
. Choosing
, we obtain for
large,
|
(2.13)
|
This shows that
. But
. Thus we conclude
|
(2.14)
|
To get the bound on
, interchange the roles of
and
, which changes
to
and
to
. Thus we have
|
(2.15)
|
and (b) is proved.
The first inequality in (c), namely that
is uniformly bounded, follows immediately from the bounds in (b) for
. To get the second inequality, we write
|
(2.16)
|
and hence
|
(2.17)
|
Applying (b) again, we find that
is uniformly bounded. Q.E.D.
We can now state and prove the following theorem, which is a key step in the construction of a generalized solution of the Monge-Ampère equation
:
Theorem 2
The volume of
approaches zero as
tends to infinity. More precisely,
|
(2.18)
|
where
is a constant which is independent of
. In particular, the positive measures
tend weakly to 0.
Proof. Recall that
has been evaluated and is given by ( 2.16 ). In particular, it is independent of the choice of metric on
. We now apply ( 2.5 ) to the path of Fubini-Study metrics
defined by ( 1.9 ) and obtain
|
(2.19)
|
Here
is defined by ( 2.1 ) with
replaced by
. Since
the formula ( 1.4 ) implies
and, observing that
and
, we can rewrite ( 2.19 ) as
|
(2.20)
|
Now observe that
On the other hand, ( 1.7 ) implies
and
. Moreover,
since the
are orthonormal with respect to
. Thus
and hence
|
(2.21)
|
Now the Riemann-Roch theorem implies
and thus we obtain from Lemma 1 that the right side of ( 2.21 ) is of the size
. Theorem 2 follows now from ( 2.20 ) and ( 2.21 ). Q.E.D.
3 Generalized solutions of the Monge-Ampère equation
In the previous section, we have seen that the functions
form a uniformly bounded sequence of functions whose Monge-Ampère operators
are positive measures on
which tend to 0. The Chern-Levine-Nirenberg inequality [7] implies that if any subsequence of the
's converges uniformly, then its limit
would satisfy the Monge-Ampère equation
in the generalized sense. A major problem is the fact that the bounds available to us at the present time (c.f. Lemma 1) are not strong enough to guarantee the existence of a uniformly convergent subsequence of the
's. Of course, weakly convergent subsequences can always be found. However, it is well-known that the Monge-Ampère operator
is not lower semi-continuous under weak limits [16] .
To circumvent these difficulties, we shall formulate and establish extensions of the classical convergence and uniqueness theorems of Bedford-Taylor for the Monge-Ampère operator for domains in
to the case of Kähler manifolds with boundary.
3.1 Convergence of approximate solutions
The first of these extensions is the following convergence theorem, where the key hypothesis is the existence of uniform
bounds for some transversal derivative of the
's at the boundary of the manifold
.Let
be a compact complex manifold with smooth boundary and
be the interior of
. Let
be a covering of
by a finite number of coordinate charts
.
Fix a smooth closed
-form
on
, and let
be smooth potentials for
on
, that is,
on
. We define the class of
-plurisubharmonic functions on
to be the class of functions
on
by
|
(3.1)
|
Note that this condition means that
is upper semi-continuous and satisfies the sub-mean value property. However, it is a stronger condition than the condition
by itself, since it is a pointwise condition, while the condition
depends only on the values of
almost everywhere.
Next, let
be a continuous function on
, and
a real, smooth closed
form on
. Set
,
, so that
.
Definition 1
Let
be an upper-semicontinuous function. We say
is a solution of the Dirichlet problem with boundary values
if
-
1.
-
2.
is continuous at
for all
and
.
-
3.
on
where
.
We wish to obtain a solution of the Dirichlet problem on
from a sequence of approximate solutions. Let
be a smooth real nowhere vanishing vector field on a neighborhood
of
which is transversal to
, in the sense that for
, the vector
is not tangent to
.
Theorem 3
Assume
. Let
have the properties:
-
1.
is uniformly bounded on
.
-
2.
is uniformly bounded on
;
-
3.
uniformly;
-
4.
There is a sequence
with the property
(i.e.
is decreasing and approaching zero) and
.
-
5.
Let
Then
is a solution of the Dirichlet problem with boundary values
.
Remark: Assumption 4. implies Assumption 3. On the other hand, after passing to a subsequence, Assumption 3. implies Assumption 4.Proof. We want to choose a sequence
with the property in such a way that
is monotonically decreasing on
. To do this, we just define
. Then
. Moreover, if
, then
|
(3.2)
|
Thus, replacing
by
, we may assume that
.
Let
Then
by Theorem 5.7 of [8] . Moreover, ( 3.2 ) implies that
in an open neighborhood of
. Thus
|
(3.3)
|
This follows from the following simple lemma:
Lemma 2
Let
with
and assume that
on some neighborhood of
. Then
Proof of Lemma 2 . To see this, let
be any compact subset such that
on
, and let
be a function with
on a neighborhood of
. Then, expanding
, we obtain
Here
. Since
on
, the lemma is proved.
Now assumption 5. of the theorem together with ( 3.3 ) implies that
weakly. On the other hand, since
monotonically, the Bedford-Taylor monotonicity theorem (Theorem 2.1 of [2] ; this is stated for domains in
, but generalizes in a straightforward fashion to manifolds) we have
weakly. Thus we have
. To finish the proof of the theorem, we must show that
is continuous at the boundary and has the right boundary values.
Let
. Choose
such that
. Extend
to a continuous function on a neighborhood
of
in such a way that
is constant on the flow lines of
. Then assumption 2. implies that if
is sufficiently small, then
|
(3.4)
|
and thus
. In particular,
and
is continuous at all points
. This proves the theorem. Q.E.D.
3.2 The domination principle for the Monge-Ampère operator
The proof of the extension of the Bedford-Taylor uniqueness theorem in this section follows closely the original arguments of [1, 2] , and especially the exposition of Blocki [3] .
When we consider generalized solutions of a partial differential equation, a particularly desirable property is their uniqueness. For bounded domains in
, the uniqueness of the generalized solution of the Dirichlet problem for the Monge-Ampère equation in the class
has been established by Bedford and Taylor [1] . It seems that this uniqueness theorem should extend as well to bounded domains in Kähler manifolds, at least if good smooth approximations of
-plurisubharmonic functions exist. Although there are now many powerful approximation theorems (see [8, 14] and references therein), we found it more convenient to extend the Bedford-Taylor uniqueness theorem to a situation adapted to the problem at hand, in the spirit of the earlier extension. The key hypothesis which we will exploit is a capacity zero condition.Recall the following notion of capacity of a set introduced by Bedford and Taylor [2] : If
is a Borel subset of a bounded domain
then
|
(3.5)
|
For our purposes, we can adapt this notion to Kähler manifolds as follows: Let
be a finite cover of
by coordinate neighborhoods. Then we say
if we can write
with
a Borel subset and
|
(3.6)
|
We say that
if
for every
.
Lemma 3
There is a constant
with the following property: If
is a Borel subset,
and
, then
|
(3.7)
|
if
. In particular, if
, then for all functions
, we have
|
(3.8)
|
Proof. Fix a smooth potential
on
such that
. Then
| |
|
(3.9)
|
Since the
are fixed, we have
and the lemma follows.
The following lemma follows immediately from the quasi-continuity theorem of Bedford-Taylor [2] :
Lemma 4
Let
. Then for every
, there is an open set
such that
and
is continuous on
.
Proof. Since
is a plurisubharmonic function on
, the quasi-continuity theorem of Bedford-Taylor ([2] , Theorem 3.5) implies that there is an open set
such that
is continuous on
and
. Let
. Then
is continuous on
and, by definition,
.
We also require a notion of “nearly continuous” functions:
Definition 2
We say that a bounded function
is “nearly continuous” if
-
1.
There exists a lower semi-continuous function
on
such that
;
-
2.
has capacity zero, that is,
;
-
3.
on
.
With this notion, we shall prove the following:
Theorem 4
Assume that
is a complex manifold of dimension
with smooth boundary, let
and let
be a real closed smooth
form on
satisfying
.
Let
be such that
on
. Assume as well:
-
1.
is continuous;
-
2.
There is a decreasing sequence
of nearly continuous functions in
such that
;
-
3.
For every
there is a compact set
such that
for
sufficiently large.
Then
|
(3.10)
|
Proof. We divide the proof into several steps.
Step 1. If we replace
by
, then we have
as
. Since for any positive measure
we have
whenever
is an increasing family of measurable sets, we may replace
by
. Thus we may assume that
|
(3.11)
|
Step 2. We prove the theorem under the assumption that
is continuous (in which case the hypotheses 2. and 3. are automatic, since we can take
for all
).
For
let
. Then
on a neighborhood of
. Since
, we can invoke Lemma 2 and conclude that
|
(3.12)
|
On the other hand,
on
and, by the Bedford-Taylor monotonicity theorem,
Since
is open, we obtain
This completes step 2.
Step 3. Now we treat the case where
itself is nearly continuous (in which case the hypotheses 2. and 3. are again automatic, since we can take
for all
). This step will be parallel to the argument in Step 2:
For
let
. Then
on the open set
(this set is open since
is lower semi-continuous). In particular, the Bedford-Taylor monotonicity theorem implies
weakly on
(as measures or as currents the two notions of weak convergence are equivalent). Thus we have
The first equality follows from the assumption that
has capacity zero and the inequality from the fact that
weakly. Here we are making strong use of the fact that
is open.
Next we claim that for all
:
|
(3.13)
|
To see this, let
be an open set containing
, which is relatively compact in
.
Then,
implies that
, and we have
, so
.
Since the set
is open (
is continuous and
is upper semi-continuous), we see that
in a neighborhood of
. Thus Lemma 2 implies
Since
and
are positive Borel measures which are finite on compact subsets of
, they are both regular. In particular
| |
|
(3.14)
|
where the inf is taken over all open sets
which contain
. This proves ( 3.13 ).
Step 4. We treat the general case. From step 1. we can assume that
.
Thus
on a neighborhood of
for
sufficiently large. By Lemma 4 , there is an open set
such that
(the capacity of
in
) and such that
is continuous on
. Choose
which is continuous
such that
on
. Then
implies
| |
|
(3.15)
|
where the last inequality follows from Lemma 3 , the fact that
and that the function
is bounded. Here the constant
depends only on the sup norm of
. Šince
is open and
, the Bedford-Taylor monotonicity theorem implies
Now
Thus
since our assumptions imply that the functions
are uniformly bounded. Next, the assumption 3. implies that the sets
are all contained in a relatively compact subset of
. Using Step 3,
and since
,
, we can conclude that
Since
is arbitrary:
Applying this last inequality to
and
, for some
:
Finally, taking the limit as
and noting that
, we obtain ( 3.16 ).
Next we prove another version of the domination theorem, but this time with the roles of
reversed:
Theorem 5
Assume that
is a complex manifold of dimension
with smooth boundary, let
and let
be a real closed smooth
form on
with the property:
. Let
satisfy
on
. Assume as well:
-
1.
is continuous;
-
2.
There is a decreasing sequence
of nearly continuous PSH functions on
such that
.
Then
|
(3.16)
|
Proof. The proof is parallel to that of Theorem 4 , although there are some important differences. We again divide it into several steps. Štep 1. As before, we may assume that
Moreover, we have
To see this, observe first that
in an open neighborhood of
. Also,
is open since
is lower semi-continuous and
is upper semi-continuous.
Since
on
, the set
is an open neighborhood of
. Thus
in an open neighborhood of
.
Step 2. Now we treat the case where
itself is nearly continuous (in which case hypotheses 2. and 3. are automatic, since we can take
for all
).
For
let
. Then
on the open set
(this set is open since
is upper semi-continuous). As before, the Bedford-Taylor monotonicity theorem implies
weakly and, making use of the fact that
is open, we get
| |
|
(3.17)
|
where
is the function which appears in the definition of the near continuity of
. Next we claim that for all
:
|
(3.18)
|
The proof is similar to that of ( 3.13 ), using
an open set containing
,
for
so that
. We use now the continuity of
and the lower semi-continuity of
to deduce that the set
is open, and
in a neighborhood of
. As before, Lemma 2 implies
|
(3.19)
|
and, using the fact that
and
are positive Borel measures which are finite on compact subsets of
,
| |
|
(3.20)
|
where the inf is taken over all open sets
which contain
. This proves ( 3.18 ).
Since
and
differ only on a set of capacity 0, we obtain the desired inequality.
Step 4. We treat the general case: From step 1. we can assume that
. Thus
on a neighborhood of
for
sufficiently large. Choose an open set
such that
(the capacity of
in
) and such that
is continuous on
.
Choose
which is continuous
such that
on
. Now
| |
|
(3.21)
|
where we have made use of step 3. and the fact that
on
to prove the first inequality. Now we have
where
is a compact set such that
so
| |
|
(3.22)
|
where the last inequality follows from the fact that
and that the function
is bounded. Here the constant
depends only on the sup norm of
.
Since
is compact, the Bedford-Taylor monotonicity theorem implies
Now
Thus
Since
is arbitrary:
Applying this last inequality to
and
, for some
:
Finally, taking the limit as
, we obtain ( 3.16 ).
We can now state and prove the uniqueness theorem which we need for the proof of Theorem 1:
Theorem 6
Let
be as in Theorem 4 , that is,
-
1.
is continuous;
-
2.
There is a decreasing sequence
of nearly continuous PSH functions on
such that
;
-
3.
For every
there is a compact set
such that
for
sufficiently large.
Assume that
on
and that
.
Then
.
Proof. We may assume, after replacing
and
by
and
for some constant
, that
and
are positive. We wish to show
. Assume not: Let
be such that
. Replacing
by
, we may assume that
on
.
Case 1.
. This implies
for some
. Let
and let
be a disk in some coordinate neighborhood of
. Then
has non-zero Lebesgue measure (in general, if
are psh functions such that
almost everywhere in a disk
, then
everywhere in
; this follows from local regularization). Now we have, using Theorem 4
| |
|
(3.23)
|
Case 2.
. This is treated exactly in the same was as in case 1 except that we use Theorem 5 instead of Theorem 4 . Q.E.D.
4 Proof of Theorem 1
We can give now the proof of Theorem 1.First, we apply Theorem 3 to construct a generalized solution of the Dirichlet problem for the Monge-Ampère equation
|
(4.1)
|
where
is defined by
and
. Define
, where
and
is defined as in ( 1.9 ). Let
. Then
and
by Lemma 1. We also have
by ( 1.7 ), so that
uniformly. Furthermore, Theorem 2 implies that
as
.
Thus, since
, we can apply Theorem 3, and conclude that
|
(4.2)
|
is a generalized solution of the desired Dirichlet problem ( 4.1 ).
Consider next the
geodesic
joining
to
in the space
of Kähler potentials.
Let
with
as before. We shall show that
. To do this, we would like to apply Theorem 6 with
|
(4.3)
|
First, we show that
is nearly continuous. Since
is smooth,
is lower semi-continuous.
Moreover,
has capacity 0, by Proposition 5.1 of [2] . It remains to show that
on
. The equation ( 1.7 ) implies
, for some constant
independent of
, if
and
is any derivative tangent to
. Thus, if
, there exists
such that if
then
|
(4.4)
|
for all
if
. Also,
implies that
|
(4.5)
|
We have then
, which implies that
|
(4.6)
|
and hence
|
(4.7)
|
if
,
. In particular,
for all
so
on
so
is indeed nearly continuous.
Thus the first two assumptions of Theorem 6 are satisfied.
In the next step, we will need the bound
|
(4.8)
|
which follows by taking the limit of ( 4.7 ) as
.
Next, we verify assumption 3: Since
on
we see that
is both upper and lower semi-continuous on
and thus
is continuous on
. Moreover, by Theorem 3,
on
so
is also continuous on
. Since
we see, by Dini's theorem, that
uniformly on
. Thus, for every
( 4.7 ) and ( 4.8 ) imply that for
for all
and
.
All the conditions of Theorem 6 are satisfied. We can thus conclude that
.
Finally, the uniform convergence of the functions
follows from their upper semi-continuity and the compactness of
. This is essentially Dini's theorem, and can be proven as follows. Assume that
is a sequence of upper semi-continuous functions, decreasing to a continuous limit
. For each
, the sets
form an open covering of
. Since
is compact, it admits a finite subcover, and since the sets are increasing as
increases, we must have
for some
. Q.E.D.
5 Remarks
We conclude with a few remarks.
Donaldson's conjecture says any two Kähler metrics can be connected by a smooth geodesic. One way to prove this conjecture is to establish a priori bounds on the derivatives of
. This was done in Lemma 1 for
and
. Let us now consider
:
|
(5.1)
|
Note that Lemma 1 implies
, but this is not strong enough, since we need a bound which is independent of
.
Define a random variable
whose probability distribution is given by
This is indeed a distribution since the total probability is
. Moreover,
is just the variance of
. Thus we need to prove that the variance of
is bounded by
. In the simplest case where
and the line bundle
and the metric
is the Fubini-Study metric, then an easy computation shows that
is just the binomial distribution with
trials, where the probability
of flipping heads is a function of
. As is well known, the variance of the binomial distribution is
, which is the bound we need.
In the case where
is the Fubini-Study metric on
, the eigenvalues of the change of basis matrix are just
. More generally, we can show that if
is a radially symmetric Kähler potential on
, and if
are the eigenvalues of the change of basis matrix, then
for some constant
, independent of
. From this one can show without difficulty that
is uniformly bounded.
The function
is equal almost everywhere to the
.
The convergence can also be guaranteed to take place in a Sobolev norm of positive order. Indeed, quite generally, the
-norm of
can be bounded by
if
is
-plurisubharmonic. Indeed,
|
(5.2)
|
and the right hand side can be bounded in turn by
|
(5.3)
|
(In fact, the same argument gives the following useful inequality
|
(5.4)
|
where
,
, is the familiar Aubin-Yau functional.) Returning to the problem at hand, we deduce that the
Sobolev norms of the functions
are uniformly bounded. The same is true for the
Sobolev norms of
for each
.
The functions
also satisfy an interesting Harnack inequality of Li-Yau type. Let
and let
. Then we claim
|
(5.5)
|
where
|
(5.6)
|
where
is the velocity in space at time
and the infimum is taken over all paths from
to
parametrized by
. To see this, let
. Then
, where
is the normal component of the holomorphic vector field on
generated by
[21] , and so
|
(5.7)
|
As in [17] , we can now choose a path
joining
to
where
and
. Then
| |
|
(5.8)
|
Completing the square, we obtain
|
(5.9)
|
References
-
Bedford, E. and B.A. Taylor, “The Dirichlet problem for a complex Monge-Ampre equation”, Invent. Math. 37 (1976), 1-44.
-
Bedford, E. and B.A. Taylor, “A new capacity for plurisubharmonic functions”, Acta Math. 149 (1982), 1-40.
-
Blocki, Z., “The complex Monge-Ampère operator and pluripotential theory”, lecture notes available from the author's website.
-
Catlin, D., “The Bergman kernel and a theorem of Tian”, Analysis and geometry in several complex variables (Katata, 1997), 1-23, Trends Math., Birkhäuser Boston, Boston, MA, 1999.
-
Chen, X.X., “The space of Kähler metrics”, J. Differential Geom. 56 (2000), 189-234.
-
Chen, X.X. and G. Tian, “Geometry of Kähler metrics and foliations by discs”, arXiv: math.DG / 0409433.
-
Chern, S.S., H. Levine, and L. Nirenberg, “Intrinsic norms on a complex manifold”, Global Analysis, Papers in honor of K. Kodaira, University of Tokyo Press (1969) 119-139.
-
Demailly, J.P., “Complex analytic and differential geometry”, book available from the author's website.
-
Donaldson, S.K., “Remarks on gauge theory, complex geometry and
-manifold topology”, Fields Medallists' lectures, 384-403, World Sci. Ser. 20th Century Math., 5, World Sci. Publishing, River Edge, NJ, 1997.
-
Donaldson, S.K., “Symmetric spaces, Kähler geometry, and Hamiltonian dynamics”, Amer. Math. Soc. Transl. 196 (1999) 13-33.
-
Donaldson, S.K., “Scalar curvature and projective imbeddings I”, J. Differential Geom. 59 (2001) 479-522.
-
Donaldson, S.K., “Scalar curvature and stability of toric varieties”, J. Differential Geom. 62 (2002), 289-349.
-
Donaldson, S.K., “Scalar curvature and projective imbeddings II”, arXiv: math.DG / 0407534.
-
Guedj, V. and A. Zeriahi, “Intrinsic capacities on compact Kähler manifolds”, arXiv: math.CV / 0401302.
-
Klimek, M., “Pluripotential theory”, London Mathematical Society monographs, New Series 6 (1991) Oxford University Press, New York.
-
Lelong, P., “Fonctions plurisousharmoniques et formes différentielles positives”, Gordon & Breach, Paris-London-New York (1968).
-
Li, P. and S.T. Yau, “On the parabolic kernel of the Schrödinger operator”, Acta Math. 156 (1986), 153-201.
-
Mabuchi, T., “Some symplectic geometry on compact Kähler manifolds”, Osaka J. Math. 24 (1987) 227-252.
-
Paul, S., “Geometric analysis of Chow Mumford stability”, Adv. Math. 182 (2004), 333-356.
-
Phong, D.H. and J. Sturm, “Stability, energy functionals, and Kähler-Einstein metrics”, Comm. Anal. Geometry 11 (2003) 563-597, arXiv: math.DG / 0203254.
-
Phong, D.H. and J. Sturm, “Scalar curvature, moment maps, and the Deligne pairing”, Amer. J. Math. 126 (2004) 693-712, arXiv: math.DG / 0209098.
-
Phong, D.H. and J. Sturm, “The Futaki invariant and the Mabuchi energy of a complete intersection”, Comm. Anal. Geometry 12 (2004) 321-343, arXiv: math.DG / 0312529.
-
Semmes, S., “Complex Monge-Ampère and symplectic manifolds”, Amer. J. Math. 114 (1992) 495-550.
-
Tian, G., “Kähler-Einstein metrics with positive scalar curvature”, Invent. Math. 130 (1997) 1-37.
-
Yau, S.T., “On the Ricci curvature of a compact Kähler manifold and the complex Monge-Ampère equation I”, Comm. Pure Appl. Math. 31 (1978) 339-411.
-
Yau, S.T., “Open problems in geometry”, Proc. Symp. Pure Math. 54, AMS Providence, RI (1993) 1-28.
-
Zelditch, S., “The Szegö kernel and a theorem of Tian”, Int. Math. Res. Notices 6 (1998) 317-331.
-
Zhang, S., “Heights and reductions of semi-stable varieties”, Compositio Math. 104 (1996) 77-105.