You are on page 1of 35

a

r
X
i
v
:
0
8
0
1
.
1
5
1
2
v
1


[
m
a
t
h
.
C
V
]


9

J
a
n

2
0
0
8
Bergman Spaces Seminar
Lecture Notes
Department of Mathematics
University of Crete
Heraklion 2005
Contents
1 Introduction to Bergman Spaces 5
1.1 Basic Denitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2 Growth of A
p
Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 The Bergman Kernel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.4 Some Density Matters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2 The Bergman Projection 21
2.1 The Bergman Projection on A
p
() . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2 A Bounded Projection of L
1
() onto A
1
() . . . . . . . . . . . . . . . . . . . . . 25
2.3 A characterization of A
p
in terms of derivatives . . . . . . . . . . . . . . . . . . . 32
3
Chapter 1
Introduction to Bergman Spaces
1.1 Basic Denitions
We begin our discussion with the Bergman Spaces on the unit disc of the complex plane
= z C : [z[ < 1. (1.1)
These are dened as
A
p
() = f H() : |f|
p
A
p
()
=
_

[f(z)[
p
dm(z) < (1.2)
where dm(z) =
1

dA(z) is the two dimensional Lebesgue measure normalized in and H()


is the space of analytic functions on the unit disc.
However, the story begins with Stephan Bergman and around 1970 (The Kernel function
and Conformal Mapping) theres already been some progress in the study of the spaces
A
p
() = f H() : |f|
p
A
p
(())
=
_

[f(z)[
p
dA(z) < , 0 < p < (1.3)
where C is an open connected set, for the case p = 2. The interest was then focused on
the case = .
The study of the Bergman Spaces is inspired from that of the Hardy Spaces
H
p
() =
_
f H() : |f|
p
H
p
()
= sup
r[0,1)
1
2
_
2
0
[f(re
i
))[
p
d <
_
, 0 < p < (1.4)
and
H

() = f H() : |f|
H

()
= sup
z
[f(z)[ < (1.5)
since H
p
() A
p
(). Observe however that H

() = A

() so hencefotrth we will restrict


our interest to the case p < .
The study of H
p
spaces begun from Hardy between 1915 and 1930 and and then was
continued with great interest (around 1960, Lenarnt Carleson and then Shapiro and Shields
solved the so called universal interpolation problems).
5
6
As it turned out, the Bergman spaces A
p
() behave quite dierently from the Hardy
spaces, and the study of Bergman Spaces remained still until 1990. Then Hedenmalm (in A
2
)
and Duren, Khavinson, Shapiro, Sundberg, Sheip, Aleman (in A
p
) gave signicant results. As
a consequence, the study of the spaces A
p
() was rapidly evolved in the last 15 years. Lately
there is interest in the study of the spaces A
p
() and of the Bergman spaces on the unit ball.
1.2 Growth of A
p
Functions
We will see in this section a description of the growth of A
p
functions and some basic
consequences. We will state this result in the more general context of the spaces A
p
(), where
C is an open connected set.
Proposition 1.1. Let f A
p
(), 0 < p < . Then, for every z ,
[f(z)[
1

1
p
(z)

2
p
|f|
A
p
()
(1.6)
where (z) = dist(z, ).
Proof. Fix z and set = (z). We consider the disc (z, ) = : [ z[ < . Since
f is subharmonic in we have that
[f(z)[
p

1
2
_
2
0
[f(z +re
i
)[
p
d, 0 r <
and so
_

0
[f(z)[
p
rdr
_

0
1
2
_
2
0
[f(z +re
i
)[
p
d rdr.
Therefore,

2
[f(z)[
p

_
(z,)
[f()[
p
dA()
1

[f()[
p
dA()
which gives
[f(z)[

1
p

2
p
|f|
A
p
()
which is the desired result.
In the special case of the unit disc the statement of the theorem becomes
[f(z)[
1
(1 [z[)
2
p
|f|
A
p
()
. (1.7)
The analogous description of the growth of the derivatives of a function in A
p
() is con-
tained in the following Proposition.
Proposition 1.2. Let n be a positive integer greater than one and f A
p
(), 0 < p < .
Then
[f
(n)
(z)[
n!2
n
2
2
p
(1 [z[)
n+1+
2
p
|f|
A
p
()
(1.8)
7
Proof. Let r < 1 and set C = C : [[ =
1+r
2
. Using Cauchys integral formula for the
derivatives of the function f and the circle C we have that
[f
(n)
(z)[ =

n!
2i
_
C
f()
z
n+1
d

n!
2
_
C
[f()[
[ z[
n+1
[d[
n!
2
_
C
[f()[

[[ [z[

n+1
[d[

n!2
n+1
2(1 r)
n+1
_
C
[f(z)[[d[
n!2
n+1
2
2
p
2(1 r)
n+1+
2
p
|f|
A
p
()
2
_
1 +r
2
_

n!2
n
2
2
p
(1 r)
n+1+
2
p
|f|
A
p
()
.
Seting r = [z[ we get the desired estimate.
Let us take a look at the consequences implied by the growth of A
p
functions described in
Proposition 1.1.
First of all it is easy to see that convergence in A
p
() implies uniform convergence on the
compact sets of . While the proof is obvious, we will state this result as a lemma for future
reference.
Lemma 1.1. Let f
n

n
be a sequence of functions in A
p
() and f A
p
(). Suppose that f
n
converges to f in A
p
(). Then f
n
converges to f uniformly on the compact subsets of . As a
result, f
n
converges to f almost everywhere in .
We are now ready to show that the spaces A
p
(), 0 < p < , are complete. Of course,
| |
A
p
()
is a norm only when 1 p . When 0 < p < 1, A
p
() becomes a metric space by
dening d(f, g) =
_

[f(z) g(z)[
p
dm(z) = |f g|
p
A
p
()
as usual.
Theorem 1.1. The spaces A
p
(), 0 < p < , are complete.
Proof. It suces to show that A
p
() is a closed subspace of L
p
(). To this end, let f
n

be a sequence of functions in A
p
() and f L
p
() such that lim
n
|f
n
f|
p
L
p
()
= 0.
Then there exists a subsequence f
n
k
of f
n
which converges to f almost everywhere in .
Moreover, since f
n
converges in L
p
(), f
n
is Cauchy in L
p
(). Due to Lemma 1.1, f
n

is uniformly Cauchy on the compact subsets of . Therefore f


n
converges uniformly on the
compact subsets of to some function g. Since each f
n
is analytic in , g is also analytic in
the same set. It turns out that g must coincide with f almost everywhere in and hence that
f A
p
().
In what follows, we will use some classical notions from complex analysis as well as Montels
theorem. For the sake of completeness we shall digress a little and remind a few denitions as
well as the statement of Montels theorem.
Denition 1.1. Let G C be an open set and (, d) a complete metric space. We dene the
space of continuous functions dened o and taking their values in G as
C(G, ) = f : G [ f is continuous .
8
We usually consider the cases = C or = C

in order to avoid trivial cases for the


set C(G, ). For example, if G is an open connected set of C and = N = 1, 2, 3, . . ., then
C(G, ) contains just the constant functions on G.
The space C(G, C) given the metric of uniform convergence on the compact sets is a com-
plete metric space. The space
H(G) = f : G C [ f is analytic on G C(G, C)
is also a complete metric space, if given the same metric.
Denition 1.2. A set T C(G, ) is called normal if every sequence of elements in T has a
subsequence that converges in C(G, ) to some f C(G, ).
We now state without proof Montels theorem
Theorem 1.2. (Montel) Let T H(G) be a family of functions. Then T is normal if and
only if T is uniformly bounded on the compact subsets of G.
Using Montels theorem and the growth of A
p
functions well be able to show that every
sequence in A
p
, which is bounded with respect to the A
p
norm, has a subsequence that converges
pointwise to some function f A
p
.
Proposition 1.3. Let f
n
be a sequence of functions in A
p
() wich is uniformly bounded in
A
p
()
|f
n
|
A
p
()
M for every n N.
Then there exists a subsequence f
n
k
of f
n
and a function f A
p
() such that f
n
k

converges to f uniformly on the compact subsets of .


Proof. From Proposition 1.1 and the hypothesis we have that
[f
n
(z)[

1
p
dist(z, )

2
p
|f
n
|
A
p
()

1
p

2
p
M
for every z . It is easy to see now that the sequence f
n
is uniformly bounded on the
compact subsets of . According to Montels theorem, this is equivalent to f
n
being normal
on . By Denition 1.2, this means that there exists a subsequence f
n
k
of f
n
and a function
f C(, C) such that f
n
k
converges to f uniformly on the compact subsets of . It remains to
show that f is an element of A
p
(). But this is obvious since f must be analytic as a uniform
limit of analytic functions.
Remark. Baring in mind the growth condition (1.6)
[f(z)[
1

1
p
(z)

2
p
|f|
A
p
()
,
it is easy to see that A
p
(C) = 0. In order to avoid trivial cases like this one we must be a
little careful when chosing the set . In other words, has to be an open connected set such
that for every z one can nd an f A
p
() satisfying f(z) ,= 0. There are however several
open connected sets which give rise to non trivial spaces A
p
(). We give some examples
here.
9
(i) The unit disc and in general every bounded subset of C.
(ii) Let = C [0, ). Then the function : dened as
(z) =
_
i
1 +z
1 z
_
2
is a conformal mapping (one to one, onto) satisfying (0) = 1, (1) = 0 and (1) = .
Let
o
. We want to nd a function f A
p
() satisfying f(
o
) ,= 0. From (i), there exists
a function g A
p
() with g(
1
(
o
)) ,= 0. Dene f : C as
f(w) = g(
1
)(w) = (g
1
)(w).
Then
_

[f()[
p
dA() =
_

[g
1
()[
p
dA() =
_
()
[f(z)[
p
[

(z)[
2
dA(z)
=
_

[f(z)[
p
[

(z)[
2
dA(z) = 4
_

[f(z)[
p

i
1 +z
1 z
2
(1 z)
3

2
dA(z)
= 16
_

[f(z)[
p

1 +z
(1 z)
3

2
dA(z) < .
Since f(
o
) = g(
1
(
o
)) ,= 0, f is the function we were seeking for.
In general, if is an open, simply connected proper subset of C and
o
, there
always exists a unique one to one, analytic mapping from onto such that (
o
) = 0 and

(
o
) > 0. Then
_

(z)
2
p
[
p
dA(z) =
_

(z)[
2
dA(z) =
_

dA(z) < .
(iii) Let be an open connected. Suppose further that the boundary of , , has at least
one connected component S containing at least two points and that the the complement of S
in C

is not a singleton. Then denes a non trivial A


p
space.
To see this observe rst that since is closed and S is a component, S must also be
closed and of course connected. Dene

= C

S . Clearly

is open and connected


and therefore the discussion in (ii) yields that A
p
(

) is non trivial. Since

we are
nished.
1.3 The Bergman Kernel
For this section, well restrict our attention to the Hilbert space case p = 2. Let be an
open connected set such that the space A
p
() is non trivial. Fix a z and dene the point
evaluation functional
: A
2
() C, (f) = f(z).
The functional is bounded since for every f A
2
() we have that
[(f)[ = [f(z)[
1

1
p
(z)

2
p
|f|
A
2
()
10
from Poroposition 1.1 (remember that z is xed). From the Riesz representation theorem for
Hilbert spaces there exists a unique function K
z
A
2
() such that
(f) = f, K
z
)
A
2
()
=
_

f()K
z
()dA()
for every f A
2
().
Denition 1.3. The function K : C C C dened as
K(z, ) = K
z
()
is called the reproducing Kernel or the Bergman Kernel.
So the Bergman Kernel reproduces the values of every function in A
2
() by means of the
formula
f(z) =
_

f()K(z, )dA(). (1.9)


The following properties of the Bergman Kernel are simple consequences of the Denition
1.3.
(i) For every z, C we have that
K(z, ) = K
z
() = K

(z) = K(, z) (1.10)


Indeed,
K(, z) = K

(z) =
_

(w)K(z, w)dA(w) =
_

K(, w)K(z, w)dA(w)


=
_

K(, w)K(z, w)dA(w) =


_

K
z
(w)K(, w)dA(w) = K
z
()
= K(z, ).
(ii) The function K(z, ) is analytic with respect to z and counter analytic with respect to .
This is just a consequence of (1.10) since, by dention, the functions K
z
, K

are in A
2
() and
hence analytic.
(iii) For every z we have that
K(z, z) = |K(z, )|
2
A
2
()
. (1.11)
This is a just a simple calculation
K(z, z) = K(z, z) = K
z
(z) =
_

K
z
()K
z
()dA()
=
_

[K
z
()[
2
= |K(z, |
2
A
2
()
> 0.
(iv) Fix a z . Then, for every function f A
2
() satisfying f(z) = 1 we have the estimate
|f|
A
2
()
|K(z, )|
1
A
2
()
. (1.12)
11
To see this, use formula (1.9) to write
[f(z)[ =

f()K(z, )dA()

[f()[[K(z, )[dA()
|f|
2
A
2
()
_

[K(z, )[
2
dA() = |f|
2
A
2
()
|K(z, )|
2
A
2
()
.
It is easy to see that estimate (1.12) is sharp. Indeed, x a z and dene
f() =
K(, z)
K(z, z)
=
K(, z)
|K(z, )|
2
A
2
()
.
Then, f(z) = 1 and
|f|
2
A
2
()
=
1
|K(z, )|
2
A
2
()
_

[K(, z)[
2
dA()
=
|K(z, )|
2
A
2
()
|K(z, )|
4
A
2
()
= |K(z, )|
2
A
2
()
.
(v) The Bergman Kernel K(z, ) is the only function in A
2
() which reproduces the value of
every function f A
2
() in the sense of (1.9). This is just a consequence of the denition of
the Bergman Kernel by means of the Riesz representation theorem.
We next turn to the question of calculating the Bergman Kernel in A
2
(). We consider an
orthonormal base
n

n=0
of A
2
(). To simplify notation we write , ) for the inner product
in A
2
(). Then

n
,
m
) =
nm
=
_
1, if n = m
0, if n ,= m .
The usual Hilbert space identities hold:
(i) Fourier series: Every f A
2
() has a series representation
f =

n=0
c
n

n
, (1.13)
where c
n
= f,
n
) and the series in (1.13) converges with respect to the A
2
() norm. Now,
Lemma (1.1) implies that the partial sums

N
n=0
c
n

n
converge to f uniformly on the compact
subsets of .
(ii) Parsevals Identity: For every in f A
2
() we have that

n=0
[c
n
[
2
= |f|
2
A
2
()
. (1.14)
(iii) A formula for the Bergman Kernel: For every z, , the Bergman Kernel can be
calculated as
K(z, ) =

n=0

n
(z)
n
() (1.15)
where the series in (1.15) converges with respect to the A
2
() norm and hence uniformly on
the compact sets of .
12
To see this, write the series representation of the function K(, ) as in (1.13). The Fourier
coecients of the function K(, ) are
c
n
= K(, ),
n
) =
_

K(w, )
n
(w)dA(w) =
_

K(w, )
n
(w)dA(w)
=
_

K(, w)
n
(w)dA(w) =
n
().
Writing the series representation of the function K(, )
K(z, ) =

n=0
c
n

n
(z) =

n=0

n
(z)
n
()
we get (1.15).
The next step is to try to calculate the Bergman Kernel in the case of the unit disc = .
In order to do this we will employ formula (1.15). First of all we have to dene an orthonormal
base in A
2
().
Proposition 1.4. The set
n
(z) =

n + 1 z
n

n=0
is an orthonormal base of A
2
().
Proof. The proof will be done in two steps:
step 1 The set
n

n=0
is orthonormal.
Indeed, for every n, m N we have that

n
,
m
) =

n + 1

m + 1
_

z
n
z
m
dm(z)
=

n + 1

m + 1
_
1
0
_
2
0
(re
i
)
n
(re
i
)
m
1

drdr
=
1

n + 1

m+ 1
_
1
0
r
n+m+1
dr
_
2
0
e
i(nm)
d
=
_
1 , if n = m
0 , if n ,= m .
step 2 The set
n

n=0
is a base of A
2
().
It is equivalent to show that Parsevals formula
|f|
A
2
()
=

n=0
[f,
n
)[
2
(1.16)
holds true for every f A
2
(). If we write the expansion of f in its Taylor series
f(z) =

n=0
a
n
z
n
it is clear that we have to show that
|f|
A
2
()
=

n=0
[a
n
[
2
n + 1
. (1.17)
13
Consider the partial sums of the Taylor series of f, S
N
(f)(z) =

N
n=0
a
n
z
n
and the disc

of
radius , centered at the origin, where 0 < < 1,

= z C : [z[ < .
Then,
_

[S
N
(f)(z)[
2
dm(z) =
_

_
N

n=0
a
n
z
n
_
2
_
N

n=0
a
n
z
n
_
2
dm(z)
=
N

n=0
a
n
N

m=0
a
n
_

0
r
n+m+1
1

_
2
0
e
i(nm)
ddr
= 2
N

n=0
[a
n
[
2
_

0
r
2n+1
dr =
N

n=0
[a
n
[
2

2(n+1)
n + 1
.
Since S
N
(f) converges to f uniformly in

as n , it is clear that
lim
N
_

[S
N
(f)(z)[
2
dm(z) =
_

[f(z)[
2
dm(z).
However,
lim
N
N

n=0
[a
n
[
2

2(n+1)
n + 1
=

n=0
[a
n
[
2

2(n+1)
n + 1
and so, combining the last two formulas we get
_

[f(z)[
2
dm(z) =

n=0
[a
n
[
2

2(n+1)
n + 1
.
Letting 1

yields (1.17).
Having dened an orthonormal base of A
2
(),it is now easy to calculate an exact formula
for the Bergman Kernel. Indeed, employing (1.15), we have that
K(z, ) =

n=0

n
(z)
n
() =

n=0
(n + 1)(z)
n
=
1
(1 z)
2
.
We have actually established the following.
Theorem 1.3. (i) The Bergman Kernel in A
2
() is given by the formula
K(z, ) =
1
(1 z)
2
. (1.18)
(ii) For every function f A
2
() we have the representation
f(z) =
_

f()
(1 z)
2
dm() . (1.19)
We close this section with a theorem that relates the Bergman Kernels of two open connected
sets through a univalent mapping.
14
Theorem 1.4. Let , T be open and connected subsets of C and : T, (z) = w, be
a univalent mapping of onto T. Suppose further that J(w, ) is the Bergman Kernel in T.
Then, the Bergman Kernel in is given by
K(z, ) = J((z), ())

(z)

(). (1.20)
Proof. Dene the operator T : A
2
(T) A
2
() as
T(f)(z) = (f )(z)

(z).
It easy to check that T is an isometry. Indeed,
|f|
2
A
2
(D)
=
_
D
[f(w)[
2
dA(w) =
_
()
[f(w)[
2
dA(w) =
_

[f((z))[
2
[

(z)[
2
dA(w)
=
_

[T(f)(z)[
2
dA(z) = |Tf|
2
A
2
()
.
If g A
2
() then dene the function f on T by the formula
f(w) = g(
1
(w))(
1
)

(w).
Then f A
2
(T) by a simple change of variable and T(f) = g which shows that T is onto.
Now, consider any g A
2
() and f A
2
(T) such thath g = T(f). Then
f(w) =
_
D
J(w, )f()dA().
and replacing w by (z) in the above we get
f((z)) =
_
D
J((z), )f()dA()
for z . Making the change of variable = () results to
f((z)) =
_

J((z), ())f(())[

(z)[
2
dA()
and multiplying by

(z),
f((z))

(z) =
_

J((z), ())

(z)

() f(())

()dA().
Remembering that g(z) = T(f)(z) we get
g(z) =
_

J((z), ())

(z)

() g()dA()
which is the desired result.
An application of Theorem 1.4 is contained in the Corollary below.
15
Corollary 1.1. Let be an open, connected, proper subset of C and K(z, ) be the Bergman
Kernel for . From Riemanns mapping theorem we know that for every there exists a
conformal mapping of onto with () = 0 and

() > 0. Then, for every z ,

(z) =
_

K(, )
K(z, ). (1.21)
Proof. Using theorem 1.4, the Bergman Kernel of is written as
K(z, ) = J((z), ())

(z)

()
where
J(w, ) =
1

1
(1 w)
2
is the Bergman Kernel for the unit disc (the constant
1

is there because we have considered


the normalised Lebesgue measure on the unit disc). Combining tha last two relations we get
K(z, ) =
1

1
(1 (z)())
2

(z)

() =
1

(z)

()
since

() > 0 and () = 0. Therefore,


K(, ) =
1

())
2
and hence
K(z, ) =
1

(z)(K(, ))
1
2
which is just (1.21).
1.4 Some Density Matters
It is obvious from Proposition 1.4 that the polynomials are dense in A
2
(). In fact, some-
thing stronger is true.
Theorem 1.5. The set of polynomials is dense in A
2
(). Whatsmore, every function in A
2
()
can be approached in the A
2
() norm by the partial sums of its Taylor series, that is
lim
N
|S
N
(f) f|
A
2
()
= 0 (1.22)
where S
N
(f) =

N
n=0
a
n
z
n
is the partial sum of the Taylor series of f.
In order to prove this we shall need a simple lemma that relates the Taylor coecients of
f with its Fourier coecients with respect to the orthonormal base
n
.
Lemma 1.2. Suppose that f A
2
() has the Taylor expansion f(z) =

n=0
a
n
z
n
and that

n
=

n + 1 z
n
. Then, for every n N,
f,
n
) =
a
n

n + 1
(1.23)
16
Proof. Consider the disc

= z C : [z[ < and x some n N. Then, for N > n we have


that
_

S
N
(f)(z)
n
(z)dm(z) =
_

_
N

k=0
a
k
z
k
(z)
_

n
(z)dm(z)
=

n + 1
N

k=0
a
k
_

z
k
z
n
dm(z)
=

n + 1 2 a
n
_

0
r
2n+1
dr =

n + 1a
n

2n+1
n + 1
.
Since S
N
(f) converges to f, uniformly on

as N , we get
_

S
N
(f)(z)
n
(z)dm(z) =

n + 1a
n

2n+1
n + 1
.
Letting 1

we get (1.23).
Proof of Theorem 1.5. We write the Taylor series of f as
f(z) =

n=0
a
n
z
n
=

n=0
a
n

n + 1

n + 1z
n
=

n=0
a
n

n + 1

n
(z) .
Then, for N N,
_

[f(z) S
N
(f)(z)[
2
dm(z) = f S
N
(f), f S
N
(f))
= |f|
2
A
2
()
f, S
N
(f)) f, S
N
(f)) +|S
N
(f)|
2
A
2
()
= |f|
2
A
2
()

n=0
a
n
f, z
n
)
N

n=0
a
n
f, z
n
) +
N

n=0
[a
n
[
2
n + 1
= |f|
2
A
2
()

n=0
a
n

n + 1
f,
n
)
N

n=0
a
n

n + 1
f,
n
) +
N

n=0
[a
n
[
2
n + 1
.
Employing Lemma 1.2, the right hand side is equal to
|f|
2
A
2
()

n=0
[a
n
[
2
n + 1

N

n=0
[a
n
[
2
n + 1
+
N

n=0
[a
n
[
2
n + 1
= |f|
2
A
2
()

n=0
[a
n
[
2
n + 1
.
However,
|f|
2
A
2
()
=

n=0
[a
n
[
2
n + 1
from Parsevals identity and so
|f S
N
(f)|
A
2
()
0 as N
and this nishes the proof.
Remark. It is not true that the polynomials are dense in A
2
() for every simply connected
set C. Consider for example the set = [0, 1]. This is obviously simply connected
17
and gives rise to the space A
2
(). On one can consider the function f(z) = z
1
2
which is
clearly analytic square integrable. However, it is not possible to approach f by polynomials in
the A
2
() sense.
To see this, suppose that one could nd a sequence of polynomials P
n

nN
, such that
lim
n
|f P
n
|
A
2
()
= 0 .
Then the sequence P
n

nN
is Cauchy in A
2
() and hence in A
2
() since |P
n
P
m
|
A
2
()
=
|P
n
P
m
|
A
2
()
. This means that the sequence P
n

nN
is uniformly Cauchy on the compact
subsets of which in turn means that it converges uniformly on the compact subsets of
to some A
2
() function, say g. Since P
n

nN
also converges to f uniformly on the compact
subsets of , it turns out that f g in . But this means that f has an analytic extension to
the whole of , a contradiction.
We will be able to show next that the polynomials are dense in A
p
() for general p,
0 < p < . However, we wont be able to approach a function in A
p
() by the partial sums of
its Taylor series.
Theorem 1.6. The polynomials are dense in A
p
(), 0 < p < .
Proof. Let f A
p
(). We consider the function f

= f(z), 0 < < 1. The function f

is
analytic inside the disc

= z C : [z[ <
1

. We deduce that the partial sums of


the function f

, S
N
(f

), converge to f

uniformly on the compact subsets of

and hence
uniformly in . That is
S
N
(f

) f

uniformly in as N . (1.24)
It is immidiate from (1.24) that
lim
N
|S
N
(f

) f

|
p
A
p
()
= 0. (1.25)
It suces to show that
lim
1
|f f

|
p
A
p
()
= 0. (1.26)
Indeed, assuming for a moment (1.25), we have that
|f S
N
(f

)|
p
A
p
()
2
p
|f f

|
A
p
()
+ 2
p
|f

S
N
(f

))|
A
p
()
.
Let > 0. We chose close to 1 so that 2
p
|f f

|
A
p
()
<

2
(this is possible because of
(1.26)). Then, for N large enough, 2
p
|f

S
N
(f

))|
A
p
()
<

2
because of (1.25) and so
|f S
N
(f

)|
p
A
p
()
<
and so S
N
(f

) is the seeked for polynomial.


It remains to prove equation (1.26). We have that
|f f

|
p
A
p
()
=
_

[f(z) f

(z)[
p
dm(z)
=
_
1
0
1

_
2
0
[f(re
i
) f

(re
i
)[
p
drdr
18
For the inner integral we have the estimate
_
2
0
[f(re
i
) f

(re
i
)[
p
d
_
2
0
([f(re
i
)[ +[f

(re
i
)[)d
2
p
_
2
0
_
[f(re
i
)[
p
+[f

(re
i
)[
p
_
d
= 2
p
__
2
0
[f(re
i
)[
p
d +
_
2
0
[f(re
i
)[
p
d
_
2
p+1
_
2
0
[f(re
i
)[
p
d < .
The last inequality is due to the fact the function
F(r) =
_
2
0
[f(re
i
)[
p
d, 0 r < 1,
is an increasing function of r. Thus, since 0 < r < 1 (0 < < 1), we have that
F(r) < F(r).
This is a simple lemma which well prove after the proof of this theorem. Thus we get
_
2
0
[f(re
i
) f

(re
i
)[
p
d 2
p+1
_
2
0
[f(re
i
)[
p
d (1.27)
We show next that f

converges to f uniformly on the compact subsets of as 1. If


K is a compact subset of and z K then
[f(z) f(z)[ = [

n=0
a
n
(1
n
)z
n
[

n=0
[a
n
[[1
n
[[z[
n

n=0
[a
n
[[1
n
[M
n
for some M < 1. Now, Lebesgues dominated convergence theorem for series yields that f

converges to f uniformly on the compact subsets of as 1. Thus, for r < 1 xed, we get
lim
1
[f(re
i
) f

(re
i
)[ = 0
and so
lim
1
_
2
0
[f(re
i
) f

(re
i
)[d = 0. (1.28)
Combining (1.27) with (1.28) and Lebesgues dominated convergence theorem we get (1.26).
We now give the proof of the Lemma we already used in the proof of Theorem 1.6.
Lemma 1.3. Suppose g is a nonnegative subharmonic function in and 0 r < 1. Then the
function
F(r) =
1

_
2
0
g(re
i
)d.
is a decreasing function of r.
19
Proof. Since g is a subharmonic function, for every open connected set B with B , there
exists a function U, harmonic in B, such that g(z) = U(z) in B and g(z) U(z) in B.
Suppose that 0 r
1
< r
2
< 1. Set B = z C : [z[ < r
2
. Then
F(r
1
) =
1

_
2
0
g(r
1
e
i
)d
1

_
2
0
U(r
1
e
i
)d = 2U(0)
by the mean value theorem. Again by the mean value theorem we have that
2U(0) =
1

_
2
0
U(r
2
e
i
)d
and so
F(r
1
)
1

_
2
0
U(r
2
e
i
)d =
1

_
2
0
g(r
2
e
i
)d = F(r
2
)
which shows that F is increasing.
Remark. In theorem 1.6 we used the fact that the function
F(r) =
_
2
0
[f(re
i
)[
p
d, 0 r < 1,
is an increasing function of r. This is an immidiate consequence of lemma 1.3 since the function
[f(z)[
p
is a subharmonic function.
20
Chapter 2
The Bergman Projection
2.1 The Bergman Projection on A
p
()
Let us recall the formula of the main theorem of the previous chapter, that is Theorem 1.3.
The latter states that for every function f A
2
() we have the representation
f(z) =
_

f()
(1 z)
2
dm(). (2.1)
Although the discussion that led to Theorem 1.3 strongly depends on the Hilbert space structre
of A
2
(), one can try to see if the above formula has a meaning in a more general context.
This is the content of the following proposition.
Proposition 2.1. For f L
p
(), 1 p < , dene the function P(f) on as
P(f)(z) =
_

f()
(1 z)
2
dm(), z . (2.2)
Then,
(i) The function P(f) is a well dened analytic function on .
(ii) If in addition f A
p
(), 1 p < and z , we have that
P(f)(z) = f(z). (2.3)
Proof. Let us rst note that the integral in (2.2) is well dened for every f L
p
(), 1 p < .
Indeed, x a z and let 1 p < and q be the conjugate exponent of p, that is,
1
p
+
1
q
= 1.
Then,
_

[f()[
[1 z[
2
dm()
_
_

1
[1 z[
2q
dm()
_1
q
|f|
L
p
()
.
However, for z xed and , the function
1
(1z)
2
is a bounded function of and hence in
every L
q
(), 1 < q . Since the integral in (2.2) denes an analytic function on whenever
it exists, this proves (i).
For (ii) observe that the spaces A
p
() are nested so Theorem 1.3 implies that formula (2.3)
holds true for every function f A
p
(), 2 p < . Its easy to extend this formula to A
1
()
21
22
and hence to every A
p
(), 1 p < . Indeed, consider an f A
1
(). For (0, 1) dene
the function f

(z) = f(z). Now, f

() A
2
() and therefore we can write
f

(z) =
_

()
(1 z)
2
dm() =
1

f()
(1

z)
2

() d
where

is the characteristic function of the disc of radius centered at the origin. For every
and (0, 1) we have that
[f()[
[1

z[
2

()
1
[1 z[
2
[f()[ L
1
().
Empolying Lebesgues dominated convergence theorem yields formula (2.3) for f A
1
().
Denition 2.1. The linear operator P is called the Bergman Projection.
In the stronger L
2
() case, one can easily see that the Bergman Projection is the orthogonal
projection of L
2
() onto the closed subspace A
2
().
Proposition 2.2. The Bergman Projection is the orthogonal projection of L
2
() onto A
2
().
Proof. Since A
2
() is a closed subspace of L
2
(), there exists an orthogonal projection, say
P
o
, of L
2
() onto A
2
(). Then, for every f L
2
(), P
o
(f) A
2
() and so
P
o
(f)(z) = P
o
(f), K
z
) = f, P
o
(K
z
)) = f, K
z
) = P(f).
Consequently, P
o
coincides with P which shows the proposition.
When p ,= 2 there is no orthogonal projection. However, since P(L
p
()) A
p
(), it is
natural to ask whether the Bergman Projection is a bounded operator from L
p
() to A
p
()
which would also show that P(L
p
()) = A
p
().
Let us rst show that this is not the case when p = 1.
Proposition 2.3. The Bergman projection is not bounded from L
1
() to L
1
().
Proof. We will actually show that the adjoint operator of P, P

, is not a bounded operator on


L

(). We therefore need to nd a formula for the adjoint operator. To that end, consider
f L
1
() and h L

(). Writing down the denition of P

we have that
f, P

(h)) = P(f), h) =
_

P(f)(z)h(z)dm(z) =
_

_
_

f()
(1 z)
2
dm()
_
h(z) dm(z)
=
_

f()
_
_

h(z)
(1 z)
2
dm(z)
_
dm(),
where the last equality follows by applying Fubinis theorem. Thus, we have established the
formula
P

(h) =
_

h()
(1 z)
2
dm(), h L

().
23
Suppose now, for the sake of contradiction, that P is a bounded operator on L
1
(), that
is, that P B(L
1
())
1
. This is equivalent to saying that P

B(L

()). For z and


a (0, 1) dene the functions g
a
as
g
a
() =
(1 a)
2
[1 a[
2
.
Clearly g
a
L

() and |g
a
|
L

()
= 1. However,
P(g
a
)(a) =
_

g
a
()
(1 z)
2
dm() =
_

1
[1 a[
2
dm()
= 2
_
1
0
1
2
_
2
0
1
[1 are
i
[
2
d rdr = 2
_
1
0

n=0
a
2n
r
2n
rdr
= 2
_
1
0

n=0
a
2n
r
2n+1
dr =

n=0
1
n + 1
a
2n
= log
1
1 a
2
.
Now the hypothesis that P

B(L

) implies that
log
1
1 [a[
2
|P

g
a
|
L

()
|g
a
|
L

()
= 1.
Since this must hold for every a (0, 1) we ger a contradiction as a 1

.
Having got rided of the bad case p = 1 we will now show that the Bergman projection is
indeed a bounded operator from L
p
() onto A
p
() for all 1 < p < .
Theorem 2.1. The Bergman projection is a bounded linear operator from L
p
() onto A
p
()
for every 1 < p < .
Proof. We have already showed that P is onto since P(f) = f for every f A
p
() L
p
().
Now, x 1 < p < and an f L
p
(). It is clear that P(f) denes an anlytic function in the
unit disc so it remains to show that P(f) is in L
p
(). For z and q the conjugate exponent
of p we have the estimate
[P(f)(z)[
_

[f()[
[1 z[
2
dm() =
_

(1 [[
2
)

1
pq
[1 z[
2
q
[f()[(1 [[
2
)
1
pq
[1 z[
2
p
dm()

_
_

(1 [[
2
)

1
p
[1 z[
2
dm()
_1
q
_
_

(1 [[
2
)
1
q
[1 z[
2
[f()[
p
dm()
_1
p
= [I
1
(z)]
1
q
[I
2
(z)]
1
p
.
Taking p
th
powers and integrating we get
_

[P(f)(z)[
p
dm(z)
_

[I
1
(z)]
p
q
I
2
(z) dm(z). (2.4)
At this point we need to estimate integrals of the form
_

(1||
2
)

|1z|

dm() for suitable


choices of , R. Instead of doing so in the special case were interested in, we will state and
prove a general lemma which will come in handy in many cases through-out the text.
1
We denote by B(L) the set of bounded linear operators from L to L.
24
Lemma 2.1. Let s, t R with 1 < t < s. Then there exists a constant C = C(s, t), depending
only on s, t, such that
_

(1 [[
2
)
t2
[1 z[
s
dm() C(1 [z[
2
)
ts
, (2.5)
for every z .
Postponing the proof of this lemma for a while, lets see how we can apply it to complete
the proof of the theorem. For I
1
take t = 2
1
p
and s = 2 in the lemma above to get
I
1
(z) C(1 [z[
2
)

1
p
. Plugging this estimate into formula (2.4) we get
_

[P(f)(z)[
p
dm(z) C
_

(1 [z[
2
)

1
q
_
_

(1 [[
2
)
1
q
[1 z[
2
[f()[
p
dm()
_
dm(z)
= C
_

[f()[
p
(1 [[
2
)
1
q
_
_

(1 [z[
2
)

1
q
[1 z[
2
dm(z)
_
dm(),
where the last equation comes from an application of Fubinis theorem. Now, using Lemma 2.1
on more time with t = 2
1
q
and s = 2 we get
_

[P(f)(z)[
p
dm(z) C

[f()[
p
(1 [[
2
)
1
q
(1 [[
2
)

1
q
dm() = C

|f|
p
L
(
)
which nishes the proof.
We now prove Lemma 2.1.
Proof of Lemma 2.1. Let s, t R with 1 < t < s and set I =
_

(1||
2
)
t2
|1z|
s
dm(). Using the
fact that
1
(1z)
s
2
=

n=0
(n+
s
2
)
n!(
s
2
)
z
n

n
and writing I in polar coordinates we have
I =
_
1
0
1

_
2
0
1
[(1 zre
i
)
s
2
[
2
d(1 r
2
)
t2
r dr
= 2

n=0
_
(n +
s
2
)
n!(
s
2
)
_
2
_
1
0
r
2n+1
(1 r
2
)
t2
dr[z[
2n
=

n=0
_
(n +
s
2
)
n!(
s
2
)
_
2
_
1
0
r
n
(1 r)
t2
dr[z[
2n
=

n=0
_
(n +
s
2
)
n!(
s
2
)
_
2
(n + 1)(t 1)
(n +t)
[z[
2n
=
(t 1)
(
s
2
)
2

n=0
_
(n +
s
2
)
n!
_
2
n!
(n +t)
[z[
2n
.
Using standard estimates for the Gamma function we see that
_
(n+
s
2
)
n!
_
2
n!
(n+t)
is of the order
(n+st)
n!
. But this means that
I C(s, t)

n=0
(n +s t)
n!
[z[
2n
= C(1 [z[
2
)
ts
which is just the statement of the lemma.
25
2.2 A Bounded Projection of L
1
() onto A
1
()
The Bergman projection seems to be the natural operator that maps L
p
() onto A
p
(),
that is, it denes for every L
p
() function, its analytic counterpart on the unit disc and it
is the identity when restricted to A
p
(). The image of L
p
() under the Bergman projection
is exactly A
p
() which reects the fact that, well, its a projection! This nice theory however,
fails to provide with a bounded projection of L
1
() onto A
1
() since all these nice properties
hold for 1 < p < . We therefore set ourselves the task to nd out if such a projection exists on
L
1
() and, if it does, to describe it. As a sideresult, we will also dene the weighted Bergman
spaces which arise naturally in the seek of such a projection.
First, recall the representation formula
f(z) =
_

f()
(1 z)
2
dm(), z ,
for a nice function f, say f H

(). Although this identiy also holds for A


1
functions,
it does not dene a bounded operator. As a rst step, we will try to construct a family of
representation formulas, at least for nice functions. This family will be more general, in the
sense that it will include the above as a special case. Then we will see under what hypothesis
this new family may dene a bounded projection of L
1
() onto A
1
().
A rst remark is that it suces to represent a specic value of a function f, say f(0). It is
then easy to use a disc automorphism that carries 0 to any z and automatically obtain a
representation formula for the values of f at any z . In this spirit, we have the following
lemmas.
Lemma 2.2. Let f H

() and > 1. Then


f(0) = ( + 1)
_

f()(1 [[
2
)

dm(). (2.6)
Proof. First of all notice that the integral is well dened since f is bounded and > 1. Now,
if f(z) =

n=0
a
n
z
n
is the Taylor series of f, the right hand side of (2.6) equals
_

n=0
a
n

n
(1 [[
2
)

dm() =
_
1
0
_

n=0
1

_
2
0
e
in
d r
n
_
(1 r
2
)

rdr
= 2a
0
_
1
0
r(1 r
2
)

dr = f(0)
_
1
0
r(1 r)

dr
= f(0)
1
+ 1
.
Multiplying by + 1 we get the lemma.
We now use a disc automorphism to get a representation formula for f at any z .
Lemma 2.3. Let f H

() and > 1. Then


f(z) = ( + 1)
_

f()
(1 z)
+2
(1 [[
2
)

dm(). (2.7)
26
Proof. For z dene the disc automorphism
z
: as

z
(w) =
z w
1 zw
.
Clearly
z
(0) = z. Whatsmore, its easy to establish the properties

1
z
=
z
,

z
(w) =
1 [z[
2
(1 zw)
2
,
1 [
z
(w)[
2
= [

z
(w)[(1 [w[
2
).
Using these properties and Lemma 2.2, we can write, for any z ,
f(z) = (f
z
)(0) = ( + 1)
_

(f
z
)()(1 [[
2
)

dm()
= ( + 1)
_
z()
f()[

z
()[
2
(1 [
z
()[
2
)

dm()
= ( + 1)
_

f()
(1 [z[
2
)
2
[1 z[
4
(1 [z[
2
)

(1 [[
2
)

[1 z[
2
dm()
= ( + 1)(1 [z[
2
)
+2
_

f()
(1 [[
2
)

[1 z[
2+4
dm() .
Now, write the above identity for the function g, dened as g() = (1z)
+2
f(), in the place
of f to get
(1 [z[
2
)
+2
f(z) = ( + 1)(1 [z[
2
)
+2
_

f()(1 [[
2
)

(1 z)
+2
dm() .
This proves the lemma.
So formula (2.7) denes the family of representations we were seeking for, at least for nice
functions in H

(). The next step is to try to extend this formula to some bigger space
that will hopefully incude the spaces A
p
(), for 1 p < . This gives rise to the weighted
Bergman spaces.
Denition 2.2. For > 1 we dene the family of measures dm

(z) as
dm

(z) = ( + 1)(1 [z[


2
)

dm(z). (2.8)
Let 0 < p < . The weighted Bergman spaces A
p

() are dened as
A
p

() =
_
f H() : |f|
p
A
p

()
=
_

[f(z)[
p
dm

(z) <
_
. (2.9)
The spaces A
p

() consist of the functions that are in L


p

() = L
p
(, dm

) and are analytic


on the unit disc . Thus, the spaces A
p

() are closed subspaces of L


p

() and one can write


down a series of results analogous to the ones we have seen for the usual Bergman spaces. Let us
rst extend formula (2.7) to the case of A
p

() functions. This is the analogous of Proposition


2.1 for the weighted case.
27
Proposition 2.4. Let > 1 and f L
p

(), 1 p < . Dene the function P

(f) on
as
P

(f)(z) =
_

f()
(1 z)
+2
dm

(), z . (2.10)
Then,
(i) The function P

(f) is a well dened analytic function on .


(ii) If in addition f A
p

(), 1 p < and z , we have that


P

(f)(z) = f(z). (2.11)


The proof is just a repetition of the arguments in the proof of Proposition 2.1 and so its
omitted.
It should be clear by now that the operators P

play the role of the Bergman Projection in


the weighted case. One can easily see that P

is the orthogonal projection of L


2

() onto A
2

()
and that the operator P

: L
p

() A
p

() is bounded whenever 1 < p < . Remember


however that we seek to nd a bounded projection of L
1
() onto A
1
() in the non-weighted
case so this is not very helpful.
When 0, the weighted Bergman spaces contain the usual Bergman spaces, that is,
A
p
() A
p

(), and of course L


p
() L
p

() while the non-weighted case corresponds to


the value = 0. This means that formula (2.10) still holds if > 0 and f L
1
(). This is
more promising! What we really need is a thorough description of when the operator P

is a
bounded projection of L
p
() onto A
p
(). We already know this description when = 0 (this
is Theorem 2.1). The following Theorem gives an answer in the general case > 1.
Theorem 2.2. Let > 1 and 1 p < . For f L
p
() we dene the function P

(f) as
P

(f) = ( + 1)
_

f()
(1 z)
+2
(1 [[
2
)

dm().
Then P

is a bounded operator from L


p
() onto A
p
() if and only if p( + 1) > 1.
Remarks. (i) When = 0, this is just a repetition of the statement of Theorem 2.1, that is
that the Bergman Projection is bounded on L
p
() if and only if 1 < p < .
(ii) The theorem tells us that if > 0 then P

carries boundedly L
p
() onto A
p
() for any
1 p < . More specically this means that there exists a bounded projection of L
1
() onto
A
1
() which is what we were seeking for. Remember that this is not the case for the Hardy
space H
1
().
(iii) There exists an even more general version of this theorem that says that if 1 < , <
then the operator P

: L
p

() A
p

() is a bounded projection of L
p

() onto A
p

() if and
only if + 1 < ( + 1)p. Since we wont need the result in this generality, we will only give the
proof for the case = 0.
Proof. Case p = 1. Let us rst show that if P

is bounded on L
1
() if and only if > 0. To
that end we will use the fact that the adjoint operator of P

, P

, is bounded on L

() if and
only if P

is bounded on L
1
(). The operator P

is dened by means of the inner product


f, g) =
_

f(z)g(z)dm(z)
28
where f L
1
() and g L

(). That is, we dene the operator P

on L

() so that for
every pair of functions f L
1
() and g L

() we have that
P

(f), g) = f, P

(g)).
The left hand-side inner product can be calculated as follows.
P

(f), g) =
_

(f)()g()dm() =
_

f(z)
(1 z)
2+
dm

(z)g() dm()
=
_

f(z)
_

g()
(1 z)
2+
dm() dm

(z)
=
_

f(z)
_

( + 1)g()(1 [z[
2
)

(1 z)
2+
dm() dm

(z).
We have thus found an explicit formula for the operator P

,
P

(g)(z) =
_

( + 1)g()(1 [z[
2
)

(1 z)
2+
dm(). (2.12)
It is now easy to see that P

is bounded on L

() if and only if
sup
z
(1 [z[
2
)

dm()
[1 z[
2+
< (2.13)
(just test P

against the function g() =


(1z)
2+
|1z|
2+
).
Let us next show that equation (2.13) holds if and only if > 0. Indeed, suppose that
> 0. Then,
(1 [z[
2
)

dm()
[1 z[
2+
= (1 [z[
2
)

_
1
0
1

_
2
0
1
[1 zre
i
[
2+
d rdr
= (1 [z[
2
)

n=0
_
(n +

2
+ 1)
n!(

2
+ 1)
_
2
_
1
0
r
2n+1
dr[z[
2n
=
(1 [z[
2
)

n=0
_
(n +

2
+ 1)
n!(

2
+ 1)
_
2
1
n + 1
[z[
2n
C
(1 [z[
2
)

2
_
(

2
+ 1)
_
2

n=0
(n + 1)
1
[z[
2n
C

,
where C, C

are absolute constants. On the other hand, if = 0, we have that


_

1
[1 z[
2
dm() =

n=0
2
_
1
0
r
2n+1
dr[z[
2n
=

n=0
1
n + 1
[z[
2n
c log
1
1 [z[
,
for some absolute constant c. Finally, if 1 < < 0 then
(1 [z[
2
)

dm()
[1 z[
2+
(1 [z[
2
)

dm()
(1 +[z[)
2+

1
2
2+
(1 [z[
2
)

.
29
This shows that if P

is bounded on L
1
() then we must have that > 0.
Now let > 0. We will show that P

is bounded on L
1
(). We have that
_

[P

(f)(z)[dm(z) =
_

( + 1)
_

f()(1 [[
2
)

(1 z)
2+
dm()

dm(z)
( + 1)
_

[f()[
[1 z[
2+
(1 [[
2
)

dm()dm(z)
= ( + 1)
_

[f()[
_

1
[1 z[
2+
dm(z)(1 [[
2
)

dm().
However,
_

1
[1 z[
2+
dm(z) =
_

1
[(1 z)
2+
2
[
2
dm(z)
=
1
_
(

2
+ 1)
_
2

n=0
_
(n +

2
+ 1)
n!
_
2
1
n + 1
[[
2n
c
()
(

2
+ 1)
2

n=0
(n +)
n!()
[[
2n
=
()
(

2
+ 1)
1
(1 [[
2
)

.
Hence,
_

[P

(f)(z)[dm(z) C |f|
L
1
()
,
for some constant C depending only on . This nishes the case p = 1.
Case 1 < p < . Suppose now that P

B(L
p
(, dm)). This implies that P

B(L
q
(, dm))
where q =
p
p1
is the conjugate exponent of p. We will show that p( + 1) > 1. Suppose, for
the sake of contradiction, that p( + 1) 1 which is equivalent to
1
q
.
Remember that the operator P

is described by equation (2.12). Taking g() = 1


L
q
(, dm) we have
P

(g)(z) =
_

( + 1)(1 [z[
2
)

(1 z)
2+
dm()
= ( + 1)(1 [z[
2
)

_
1
0
1

_
2
0
1
(1 zre
i
)
2+
d rdr
= ( + 1)(1 [z[
2
)

_
1
0
1

_
2
0

n=0
(n + + 2)
n!( + 2)
r
n
z
n
e
in
drdr
= ( + 1)(1 [z[
2
)

_
1
0

n=0
1

_
2
0
e
in
dr
n
z
n
rdr
= ( + 1)(1 [z[
2
)

_
1
0
rdr = ( + 1)(1 [z[
2
)

.
However, the function (+1)(1[z[
2
)

is not in L
q
(, dm) and hence we have a contradiction.
Indeed, if <
1
q
then
( + 1)
q
_

(1 [z[
2
)
q
dm(z) = ( + 1)
q
_
1
0
x
q
dx
= ( + 1)
q
_
1
q + 1
_
1 lim
x0
1
x
1q
_
_
= .
30
On the other hand, if =
1
q
then
( + 1)
q
_

(1 [z[
2
)
1
dm(z) = ( + 1)
q
_
1
0
1
x
dx = .
In order to complete the proof we shall need a boundedness criterion known as Schurs Test.
Theorem 2.3. (Schurs Test) Let (X, ) be a measure space and K : XX R
+
a nonnegative
measurable function. For 1 < p < and f L
p
(X, d) we dene
T(f)(x) =
_
X
K(x, y)f(y)d(y), x X.
Suppose that there exists a positive constant C > 0 and a positive measurable function h on X
such that
(a)For almost every x X,
_
X
K(x, y)h(y)
q
d(y) Ch(x)
q
.
(b)For almost every y X,
_
X
K(x, y)h(x)
p
d(x) Ch(y)
p
.
Then, T B(L
p
(X, d)) and |T| C.
Remark. Suppose that (X, ) is a measure space and K : X X R
+
a nonnegative
measurable function. Suppose further that there exists some positive constant C and some
positive measurable function h, dened on X, such that
(i)
_
X
K(x, y)h(y)d(y) Ch(x) for -almost every x X.
(ii)
_
X
K(x, y)h(x)d(x) Ch(y) for -almost every y X.
Then, the operator T(f)(x) =
_
X
K(x, y)f(y)d(y) B(L
2
(X, d)) with |T| C.
We postpone the proof of Schurs test until after the end of the proof of Theorem 2.2
Lets x some p (1, ) and set q =
p
p1
. Suppose that p( + 1) > 1. We will now use
Schurs test in order to show that P

is a bounded operator from L


p
() to A
p
().
We set d() = (1 [[
2
)

dm(), h(z) = (1 [z[


2
)

1
pq
and K(z, ) =
1
|1z|
2+
. We have
that
_

K(z, )h()
q
d() =
_

1
[1 z[
2+
h()
q
(1 [[
2
)

dm()
=
_

(1 [[
2
)

1
p
[1 z[
2+
d().
31
However, since
1
(1z)
2+
2
=

n=0
(n+

2
+1)
n!(

2
+1)
z
n

n
, we get that,
_

(1 [[
2
)

1
p
[1 z[
+2
dm()

n=0
_
(n +

2
+ 1)
n!(

2
+ 1)
_
2
_
1
0
r
n
(1 r)

1
p
dr[z[
2n
=

n=0
_
(n +

2
+ 1)
n!(

2
+ 1)
_
2
(n + 1)( + 1
1
p
)
(n + + 2
1
p
)
[z[
2n
=
( +
1
q
)
(

2
+ 1)
2

n=0
_
(n +

2
+ 1)
n!
_
2
1
(n++1+
1
q
)
n!
[z[
2n
C
( +
1
q
)
(

2
+ 1)
2

n=0
(n + 1
1
q
)
n!
[z[
2n
= C
( +
1
q
)
(

2
+ 1)
2
1
(1 [z[)
1
p
= C

h(z)
q
,
for some constant C

, depending only on .
Using a similar calculation, we show that
_

K(z, )h(z)
p
d(z) C

h()
p
.
Now, Schurs test tells us that P

B(L
p
(), A
p
()) and that |P

| C

.
Proof of theorem 2.3. Let f L
p
(X, d). Then,
[T(f)(x)[
_
X
K(x, y)h(y)
1
h(y)
[f(y)[d(y).
H olders inequality now yields
[T(f)(x)[
__
X
K(x, y)h(y)
q
d(y)
_1
q
__
X
K(x, y)
[f(y)[
p
h(y)
p
d(y)
_1
p
C
1
q
h(x)
__
X
K(x, y)h(y)
p
[f(y)[
p
d(y)
_1
p
(2.14)
for -almost every x X. Integrating with respect to x and using Fubini we get
_
X
[T(f)(x)[
p
d(x) C
p
q
_
X
h(x)
p
_
X
K(x, y)h(y)
p
[f(y)[
p
d(y) d(x)
= C
p
q
_
X
_
X
K(x, y)h(x)
p
d(x)h(y)
p
[f(y)[
p
d(y)
C
p
q
+ 1
_
X
[f(y)[
p
d(y),
where the last inequality follows by (2.14). Thus we have that
|T(f)|
p
L
p
(Xd)
C
p
q
+1
|f|
p
L
p
(Xd)
.
Taking p-th roots copletes the proof.
32
2.3 A characterization of A
p
in terms of derivatives
Suppose that f is an analytic function and x some p [0, ). Let n be a positve integer.
We want to nd a condition for f
(n)
that assures that f A
p
(). We actually get something
better, that is, a characterization of the space A
p
() in terms of derivatives. Our main result
for this section is the following.
Theorem 2.4. Let n be a positive integer greater than 1 and 1 p < . Suppose that
f H(). Then f A
p
() if and only if (1 [z[
2
)f
(n)
(z) L
p
().
The proof will be done in several steps. Lets begin with the necessity of the condition of
Theorem 2.4.
Lemma 2.4. Let n be a positive integer greater than 1 and 1 p < . Suppose that f A
p
().
Then (1 [z[
2
)
n
f
(n)
L
p
().
Proof. Case p = 1. Suppose that f A
1
(). For = 1, equation (2.10) of Proposition 2.4
emplies
f(z) = P

(f)(z) =
_

f()
(1 z)
+2
dm

() = 2
_

(1 [[
2
)
(1 z)
3
f()dm().
Dierentiating n times we get
f
(n)
(z) = (n + 2)!
_

()
n
(1 [[
2
)
(1 z)
n+3
f()dm().
Thus,
_

(1 [z[
2
)
n
[f
(n)
(z)[dm(z) (n + 2)!
_

(1 [z[
2
)
n
_

(1 [[
2
)
[1 z[
n+3
[f()[dm()dm(z)
(n + 2)!
_

(1 [[
2
)[f()[
_

(1 [z[
2
)
n
[1 z[
n+3
dm(z)dm()
C
_

(1 [[
2
)[f()[
1
(1 [[)
dm()
= C|f|
A
1
()
.
Case 1 < p < . Suppose that f A
p
(). By Proposition 2.1 we have that
f(z) =
_

f()
(1 z)
2
dm().
Dierentiating n times we get
(1 [z[
2
)
n
f
(n)
(z) = (n + 1)!(1 [z[
2
)
n
_

()
n
f()
(1 z)
n+2
dm() = n!(P

n
S
n
)(f)(z),
where S
n
(f)(z) = z
n
f(z) and P

n
is described by equation (2.12). Consequently
_

[1 [z[
2
[
pn
[f
(n)
(z)[
p
dm(z) n!|P

n
| |S
n
| |f|
A
p
()
C
n
|f|
A
p
()
.
Remember that P

n
B(L
p
()) by Theorem 2.2.
33
To prove the suciency in Theorem 2.4 we need the following lemma.
Lemma 2.5. Let f H() and n be a positive integer greater than 1, such that
(i) (1 [z[
2
)
n
f
(n)
(z) L
1
(), and
(ii)f(0) = f

(0) = f

(0) = = f
(2n1)
(0) = 0.
Then, for every z ,
f(z) =
1
n!
_

(1 [[
2
)
n
f
(n)
()
()
n
(1 z)
2
dm().
Proof. First notice that condition (ii) implies
f
(n)
(z) =

m=2n
m(m1) . . . (mn + 1)a
m
z
mn
= z
n

m=2n
m(m1) . . . (mn + 1)a
m
z
m2n
.
As a result,
f
(n)
(z)
z
n
H().
For z , we write
_

(1 [[
2
)[f
(n)
()[
[1 z[
2
[
n
[
dm()
1
(1 [z[)
2
_

(1 [[
2
)
n

f
(n)
()

dm()

1
(1 [z[)
2
__
1
2
(1 [[
2
)
n

f
(n)
()

dm()
+
_
1
2
(1 [[
2
)
n

f
(n)
()

dm()
_
.
For the rst integral notice that sup
1
2

f
(n)
()

< . Hence
_
1
2
(1 [[
2
)
n
[f
(n)
()[dm() < .
On the other hand
_
1
2
(1 [[
2
)
n

f
(n)
()

dm() 2
n
_
1
2
(1 [[
2
)
n
[f
(n)
()[dm() <
due to hypothesis (i). This shows that the integral
_

(1||
2
)
n
f
(n)
()
()
n
(1z)
2
dm() exists. What we
actually showed is that the function F(z) =
(1|z|
2
)
n
f
(n)
(z)
z
n
is in L
1
(). But this means that
the integral
g(z) =
_

F()
(1 z)
2
dm() =
1
n!
_

(1 [[
2
)
n
f
(n)
()
(1 z)
2

n
dm()
34
denes an analytic function on . Hence
g
(n)
(z) = (n + 1)
_

(1 [[
2
)
n
f
(n)
()
(1 z)
n+2

n

n
dm()
= (n + 1)
_

(1 [[
2
)
n
f
(n)
()
(1 z)
n+2
dm() = P
n
(f
(n)
)(z).
Since f
(n)
A
1
n
(), part (ii) of Proposition 2.4 gives that P
n
(f
(n)
)(z) = f
(n)
(z). This means
that f
(n)
(z) = g
(n)
(z). However, for 0 k n 1, we have that
g
(k)
(z) =
(k + 1)!
n!
_

f
(n)()
(1 [[
2
)
k
(1 z)
n+2

n
dm().
But this implies
g
(k)
(0) =
(k + 1)!
n!
_

f
(n)()
(1 [[
2
)

nk
dm()
=
(k + 1)!
n!
_
1
0

m=2n
m(m1) . . . (mn + 1)r
m+k2n
1
2
_
2
0
e
i(mk)
d rdr
= 0 = f
(k)
(0).
Hence f = g and the proof is complete.
We are now ready to prove Theorem 2.4
Proof of Theorem 2.4. Let 1 p < and n be a postive integer, greater than 1. Suppose that
f H() is such that (1 [z[
2
)
n
f
(n)
(z) L
p
(). Suppose further that
f(0) = f

(0) = f

(0) = = f
(2n)
(0) = 0.
As weve seen in Lemma 2.5, if we set F(z) =
(1|z|
2
)
n
f
(n)
(z)
z
n
is in L
1
(), then
f(z) =
1
n!
_

(1 [[
2
)
n
f
(n)
()
(1 z)
2

n
dm() = P(f)(z).
If 1 < p < , then P B(L
p
(, dm)) and hence f A
p
().
Let p = 1. Then, for > 0, the operator
P

(f)(z) = ( + 1)
_

f()
(1 z)
+2
(1 [[
2
)dm()
is bounded on L
1
().
Bibliography
35

You might also like