NEAR OPTIMAL BOUNDS IN FREIMAN'S THEOREM
NEAR OPTIMAL BOUNDS IN FREIMAN'S THEOREM
NEAR OPTIMAL BOUNDS IN FREIMAN'S THEOREM
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
<strong>NEAR</strong> <strong>OPTIMAL</strong> <strong>BOUNDS</strong> <strong>IN</strong> FREIMAN’S<br />
<strong>THEOREM</strong><br />
TOMASZ SCHOEN<br />
Abstract<br />
We prove that if for a finite set A of integers we have |A + A| ≤K|A|, thenAis 1+C(log K)−1/2<br />
contained in a generalized arithmetic progression of dimension at most K<br />
1+C(log K)−1/2<br />
and of size at most exp(K )|A| for some absolute constant C. We also<br />
discuss a number of applications of this result.<br />
1. Introduction<br />
Freiman’s theorem [8] is one of the most fundamental theorems in additive number<br />
theory. It asserts that if for a finite A ⊆ Z we have |A + A| ≤K|A|, thenA is<br />
a subset of a d-dimensional generalized arithmetic progression P1 +···+Pd with<br />
|P1|···|Pd| =|A|f , where d and f depend only on K (here Pi stands for usual<br />
arithmetic progressions). This result provides a complete description of sets for which<br />
the sumset is not much larger than the set itself. Freiman’s theorem has a plethora<br />
of deep applications to many important problems, such as the one recently described<br />
by Gowers in [9] and[11], who used it to get an effective version of Szemerédi’s<br />
theorem (for further works that make important use of Freiman’s result, see, e.g.,<br />
[6], [5], [25], [26]–[28]). However, most applications require a quantitative version of<br />
Freiman’s theorem. It is easy to see that one cannot do better than d(K) ≤ K − 1<br />
and f (K) = e O(K) . The estimates which followed from Freiman’s original proof<br />
were quite poor. The first useful estimates for d and f were provided by Ruzsa<br />
[19], whose ingenious argument combines techniques from different fields. Ruzsa’s<br />
bounds were good enough for many applications and his approach paved the way<br />
for further improvements. The next substantial progress was made by Chang [3],<br />
who showed that one can take d ≤ K 2+o(1) and f (K) ≤ exp(K 2+o(1) ). Her proof<br />
coupled Ruzsa’s main ideas with Chang’s so-called spectral lemma and covering<br />
lemma. Further improvements were due to Sanders [21, Theorem 1.3], who used a<br />
much-expanded version of Bogolyubov-Ruzsa’s lemma [20, Lemma 2.1] to obtain the<br />
best currently known bounds d(K) ≤ K 7/4+o(1) and f (K) ≤ exp(K 7/4+o(1) ). The main<br />
DUKE MATHEMATICAL JOURNAL<br />
Vol. 158, No. 1, c○ 2011 DOI 10.1215/00127094-1276283<br />
Received 16 February 2010. Revision received 13 July 2010.<br />
2010 Mathematics Subject Classification. Primary 11P70; Secondary 11B25.<br />
Author’s work supported in part by Ministerstwo Nauki i Szkolnictwa Wy˙zszego grant N N201 543538.<br />
1
2 TOMASZ SCHOEN<br />
result of this article is the following theorem, which gives nearly optimal estimates<br />
for d and f .<br />
<strong>THEOREM</strong> 1<br />
There exists an absolute constant C such that each set A ⊆ Z satisfying |A +<br />
A| ≤K|A| is contained in a generalized arithmetic progression of dimension at<br />
1+C(log K)−1/2<br />
most d(K) and size at most f (K)|A|, with d(K) ≤ K and f (K) ≤<br />
1+C(log K)−1/2<br />
exp(K ).<br />
Our approach is yet another follow-up of Ruzsa’s original argument, and our improvement<br />
relies on further refinement of Bogolyubov-Ruzsa’s lemma. A new addition<br />
to this setup is an elementary result, Lemma 3, which roughly says that one can<br />
find two sets “closely related” to A which have much better additive properties. As<br />
a consequence of our method, we infer that there is a proper arithmetic progression<br />
P of dimension Oε(Kε ) and size bounded from above by KO(1) |A| such that<br />
|A ∩ P |≥exp(−Oε(Kε ))|A| (for a precise statement, see Theorem 8 in Section 4).<br />
This fact has been already anticipated by Chang [3], Gowers [10], and Green and<br />
Tao [14, Conjecture 1.6], and is often required in applications. Remarkably, Green<br />
and Tao [14] proved that this result is equivalent to an inverse theorem for Gowers’s<br />
U 3-norm. It can also be considered a first step toward resolving Freiman-Ruzsa’s<br />
polynomial conjecture. We also note that, using Freiman-Bilu’s [1] argument (see also<br />
[3]), one cannot deduce from Theorem 1 that A is contained in a progression with<br />
1+C(log K)−1/2<br />
linear dimension and size at most exp(K )|A|. The reason is that we cannot<br />
guarantee our progression to be proper. Our proof can be modified to achieve it as<br />
well, but in this case the size of the progression goes up to exp(K2+o(1) )|A|.<br />
Finally, our method can be directly applied to Green-Ruzsa’s [13, Theorem 1.1]<br />
proof of Freiman’s theorem in the “torsion” case, giving the following statement, the<br />
proof of which we omit here (further improvement of this theorem will appear in [7]).<br />
<strong>THEOREM</strong> 2<br />
Let G be an arbitrary Abelian group, and let A ⊆ G satisfy |A + A| ≤K|A|. Then A<br />
is contained in a coset progression P +H of dimension d(K) ≤ (K+2) 3+C(log(K+2))−1/2<br />
and size f (K)|A| ≤exp((K + 2) 3+C(log(K+2))−1/2<br />
)|A| for some absolute constant C.<br />
This paper is organized as follows. Section 2 contains some basic definitions and facts<br />
that we use here. The main part of the paper is Section 3, where we prove Lemma 3,<br />
which is key to our argument. In order to keep the paper self-contained, we briefly<br />
outline the rest of Ruzsa’s argument (with Chang’s refinements) in Section 4. We<br />
conclude with some applications of our results and a few comments on our approach<br />
and its possible further developments.
<strong>NEAR</strong> <strong>OPTIMAL</strong> <strong>BOUNDS</strong> <strong>IN</strong> FREIMAN’S <strong>THEOREM</strong> 3<br />
2. Preliminaries<br />
This part of the article contains basic definitions and notation we will use later on.<br />
As usual, we set<br />
A + B ={a + b : a ∈ A, b ∈ B},<br />
and the k-fold sumset of A is denoted by kA.Byageneralized arithmetic progression<br />
of dimension d, we mean every set of the form P = P1+···+Pd, where P1,...,Pd are<br />
usual arithmetic progressions. The size of P is defined as the product |P1|···|Pd|.The<br />
dimension and size of progression P are denoted by dim(P ) and size(P ), respectively.<br />
If each x ∈ P has unique representation x = p1 +···+pd, pi ∈ Pi, thenwesay<br />
that P is proper. Then the cardinality of P is equal to its size.<br />
Let G, H be Abelian groups, and let A ⊆ G, B ⊆ H . We say that A is Fkisomorphic<br />
(i.e., Freiman isomorphic of order k) toB if there exists a bijective map<br />
ϕ : A → B such that<br />
if and only if<br />
x1 +···+xk = y1 +···+yk<br />
ϕ(x1) +···+ϕ(xk) = ϕ(y1) +···+ϕ(yk)<br />
for every x1,...,xk,y1,...,yk ∈ A.<br />
We call (x,x ′ ,y,y ′ ) an additive quadruple if x + y = x ′ + y ′ . The number of<br />
additive quadruples in X2 × Y 2 is denoted be E(X, Y ).<br />
In this paper, by Zn we always mean Z/nZ. The Fourier coefficients of the<br />
indicator function of a set X ⊆ Zn are defined by<br />
ˆX(s) = <br />
e −2πixs/n ,<br />
x∈X<br />
where s ∈ Zn. Parseval’s formula states that n−1<br />
s=0 | ˆX(s)| 2 =|X|n. We observe also<br />
that, for X, Y ⊆ Zn, wehave<br />
E(X, Y ) = 1 n−1<br />
| ˆX(s)|<br />
n<br />
2 | ˆY (s)| 2 . (1)<br />
s=0<br />
Our argument makes use of Plünnecke-Ruzsa’s inequality, which says, in particular<br />
that if |A + A| ≤K|A|, then for all k, l ∈ N, wehave<br />
|kA − lA| ≤K k+l |A|.
4 TOMASZ SCHOEN<br />
By α, we denote the distance from α to the nearest integer. For Ɣ ⊆ Zn and<br />
γ>0, the Bohr set B(Ɣ, γ ) is defined as<br />
B(Ɣ, γ ) ={x ∈ Zn : gx/n ≤γ for every g ∈ Ɣ}.<br />
3. The main lemma<br />
The key idea of our approach is to identify a relatively large subset of A that has<br />
better additive properties than A. One possible way would be to find A ′ ⊆ A such that<br />
|kA ′ |≤K εk |A ′ | for not too-big k. A result of this kind was proved in [17, Lemma<br />
1], but under a stronger assumption on A. Unfortunately, this method does not seem<br />
to work in our general setting. Let us remark, however, that Bogolyubov-Ruzsa’s<br />
lemma uses the number of additive quadruples rather than the size of the subset.<br />
This suggests the following approach: using the assumption |A + A| ≤K|A|, one<br />
should find for some small k a relatively dense set X, Y ⊆ kA − kA with E(X, Y )<br />
much bigger than one can expect from the size of A + A. As far as we know, the<br />
first such result was proved by Katz and Koester [15]. They showed that there is<br />
B ⊆ A ± A, |B| ≫|A|/K C with E(B,B) ≫|B| 3 /K 36/37 . Lemma 3 provides a<br />
much better estimate for E(X, Y ) in a “highly nonsymmetric” case, when one set is<br />
large and the other is small (a similar result for a diagonal case would give a much<br />
better result—see our concluding remarks below). Finally, let us also mention that<br />
Sanders in [22], and [23] (seealso[24] for related theorems), used a similar result<br />
for a different purpose—that of improving the bounds in many summed versions of<br />
Roth-Meshulam’s theorem.<br />
LEMMA 3<br />
Suppose that A is a subset of an Abelian group and that |A+A| ≤K|A|. Then for every<br />
ε>0, there are sets X ⊆ A and Y ⊆ A + A such that |X| ≥(2K2 ) −21/ε|A|,<br />
|Y |≥<br />
|A|, and E(X, Y ) ≥ K−2ε |X| 2 |Y |.<br />
Proof<br />
For a set B ⊆ A, denote by D = D(B) the set of all t ∈ B − B which has at least<br />
|B| 2 |B|2 |B|2<br />
≥ ≥<br />
2|B − B| 2|A − A| 2K2 |A|<br />
representations as b − b ′ , b,b ′ ∈ B (the last inequality follows from Plünnecke-<br />
Ruzsa’s inequality). Note that |D| ≥|B|/2.<br />
We first show that there exists a set B ⊆ A such that |B| ≥(2K 2 ) −21/ε +1 |A| and<br />
|A + Bt| ≥K −ε |A + B| (2)<br />
for all t ∈ D = D(B), where Bt = B ∩ (B + t). We construct B in the following<br />
iterative procedure. We start with A 1 = A. Now suppose that A l with
<strong>NEAR</strong> <strong>OPTIMAL</strong> <strong>BOUNDS</strong> <strong>IN</strong> FREIMAN’S <strong>THEOREM</strong> 5<br />
|Al |≥(2K2 ) −2l−1 +1 |A| has already been defined. If there exists t0 ∈ D(Al ) with<br />
|A + Al t0 |
6 TOMASZ SCHOEN<br />
Since |D ′ |≥|B|/2 ≥ (2K2 ) −21/ε|A|,<br />
the assertion follows for X = D ′ and Y =<br />
A + B. ✷<br />
Proof of Theorem 1<br />
In order to show Theorem 1, one needs to mimic Ruzsa-Chang’s proof using Lemma 3.<br />
In order to make the article self-contained, we briefly sketch their argument.<br />
Let n be a prime satisfying<br />
48|12A − 12A| >n>24|12A − 12A|.<br />
By Plünnecke-Ruzsa’s inequality, it follows that n ≤ 48K 24 |A|. By Ruzsa’s lemma<br />
(see [18, Theorem 2]), there exists a set A ′ ⊆ A such that |A ′ |≥|A|/12, whichis<br />
F12-isomorphic to T ⊆ Zn. Clearly, |T + T |≤12K|T |. Let X ⊆ T and Y ⊆ T + T<br />
be sets such that the assertion of Lemma 3 holds with<br />
Thus, we have<br />
ε = (log K) −1/2 .<br />
|X| ≥ 2(12K) 2 −2 1/ε<br />
|T |≥2 −10 (288K 2 ) −21/ε<br />
K −24 n.<br />
The next ingredient of our approach is Chang’s spectral lemma (see [3]and[13]).<br />
For Ɣ ⊆ Zn, define<br />
<br />
<br />
Span(Ɣ) = εgg : εg ∈{−1, 0, 1} .<br />
g∈Ɣ<br />
Then the spectral lemma can be stated as follows.<br />
LEMMA 4([3, Lemma 3.1])<br />
Let X ⊆ Zn, and let ={r ∈ Zn : | ˆX(r)| ≥λ|X|}. Then there is a set Ɣ ⊆ Zn such<br />
that |Ɣ| ≤2λ −2 log(n/|X|) and ⊆ Span(Ɣ).<br />
Let us emphasize at this point that Chang’s spectral lemma is absolutely crucial for<br />
our argument. While in Chang’s paper one can omit it and still get reasonable bounds<br />
d(K) ≤ K O(1) and f (K) ≤ exp(K o(1) ), without Lemma 4 our approach completely<br />
fails. In fact, as it is easy to see, Chang’s spectral lemma is especially useful if λ is<br />
much bigger than the density of X, and we apply it precisely in such a case using its<br />
full potential.<br />
Our goal is to get a version of Bogolyubov-Ruzsa’s lemma for sets X and Y .
<strong>NEAR</strong> <strong>OPTIMAL</strong> <strong>BOUNDS</strong> <strong>IN</strong> FREIMAN’S <strong>THEOREM</strong> 7<br />
LEMMA 5<br />
The set X + Y − X − Y contains a Bohr set B(Ɣ, γ ) such that |Ɣ| ≪K 3ε log K and<br />
γ ≫ K −3ε log −1 K.<br />
Proof<br />
Define<br />
= r ∈ Zn : | ˆX(r)| ≥K −ε |X|/2 .<br />
We will show that B = B(, 1/6) ⊆ X + Y − X − Y .Letr(x) be the number of<br />
representations of x in X + Y − X − Y .Ifx ∈ B, then, by (1) and Parseval’s formula,<br />
we have<br />
r(x) = 1 n−1<br />
| ˆX(s)|<br />
n<br />
2 | ˆY (s)| 2 e 2πixs/n<br />
≥ 1<br />
n<br />
≥ 1<br />
2n<br />
s=0<br />
<br />
s∈<br />
<br />
s∈<br />
s=0<br />
| ˆX(s)| 2 | ˆY (s)| 2 cos(2πxs/n) − 1<br />
n<br />
| ˆX(s)| 2 | ˆY (s)| 2 − 1<br />
n<br />
<br />
| ˆX(s)| 2 | ˆY (s)| 2<br />
s∈<br />
<br />
| ˆX(s)| 2 | ˆY (s)| 2<br />
s∈<br />
≥ 1 n−1<br />
| ˆX(s)|<br />
2n<br />
2 | ˆY (s)| 2 − 3 <br />
| ˆX(s)|<br />
2n<br />
2 | ˆY (s)| 2<br />
s∈<br />
> E(X, Y )/2 − 3<br />
8n K−2ε |X| 2<br />
n−1<br />
| ˆY (s)| 2<br />
s=0<br />
≥ E(X, Y )/2 − (3/8)K −2ε |X| 2 |Y | > 0.<br />
By Chang’s spectral lemma (Lemma 4), there is a set Ɣ such that |Ɣ| ≪<br />
K 2ε log(n/|X|) ≪ K 3ε log K and ⊆ Span(Ɣ), so that<br />
B Ɣ, 1/(6|Ɣ|) ⊆ B(, 1/6) ⊆ X + Y − X − Y. ✷<br />
Remark. We proved Bogolyubov-Ruzsa’s lemma in a standard way, but one can easily<br />
strengthen the assertion by removing one summand. Indeed, by (3), we have<br />
1 n−1<br />
ˆX(s)| ˆY (s)|<br />
n<br />
s=0<br />
2 e −2πib0s/n −ε<br />
≥ K |X||Y |.<br />
Therefore, using a similar argument as in the proof of Lemma 5, we infer that a shift<br />
of B(Ɣ, γ ) is contained in X + Y − Y.
8 TOMASZ SCHOEN<br />
Let us first recall Ruzsa’s lemma on the geometry of numbers (see [20, Lemma<br />
3.2]).<br />
LEMMA 6<br />
Let B(Ɣ, γ ) be a Bohr set in Zn. Then there is a proper progression P ⊆ B(Ɣ, γ ) of<br />
dimension |Ɣ| and size at least (γ/|Ɣ|) |Ɣ| n.<br />
Thus by Lemmas 5 and 6, one can find a proper progression P ⊆ X + Y − X − Y ⊆<br />
3T − 3T with dim(P ) ≤|Ɣ| ≪K 3ε log K and<br />
size(P ) ≥ exp −K 3ε (log K) 3/2+o(1) n.<br />
Let P ′ betheimageofP under a considered F12-isomorphism between A ′ and T .It<br />
induces an F2-isomorphism between P and P ′ , which implies that P ′ is also proper.<br />
For the last step of the argument we need Chang’s covering lemma [3,Step4].<br />
LEMMA 7<br />
If |A + A| ≤K|A| and |B + A| ≤L|B|, then there are sets S1,...,Sk each of size<br />
≤ 2K such that A ⊆ B − B + (S1 − S1) +···+(Sk − Sk) and k ≤ 1 + log(KL).<br />
Now, we have<br />
so that<br />
|P ′ + A| ≤|5A − 4A| ≤K 9 |A| ≤K 9 exp K 3ε (log K) 3/2+o(1) |P ′ |<br />
A ⊆ P ′ − P ′ + (S1 − S1) +···+(Sk − Sk)<br />
with k ≤ K 3ε (log K) 5/2+o(1) . Finally, A is contained in a progression Q with<br />
and<br />
1+C(log K)−1/2<br />
dim(Q) ≤ dim(P ) + 2kK ≤ K<br />
size(Q) ≤ 2 dim(P ′ 1+C(log K)−1/2<br />
) ′ 2kK K<br />
size(P )3 ≤ e |A|<br />
for some constant C. This completes the proof of Theorem 1. <br />
4. Some applications<br />
We focus on the consequences of a new version of Bogolyubov-Ruzsa’s lemma in<br />
Freiman’s theorem as one of the most important applications. However, as an immediate<br />
consequence of the results obtained in Section 3, we obtain the following step in<br />
the direction of Freiman-Ruzsa’s polynomial conjecture, which may also have other<br />
applications.
<strong>NEAR</strong> <strong>OPTIMAL</strong> <strong>BOUNDS</strong> <strong>IN</strong> FREIMAN’S <strong>THEOREM</strong> 9<br />
<strong>THEOREM</strong> 8<br />
There exists an absolute constant C such that, if A ⊆ Z satisfies |A+A| ≤K|A|, then<br />
C(log K)−1/2<br />
there is a proper generalized arithmetic progression P with dim(P ) ≤ K<br />
and size(P ) ≤ 48K24 C(log K)−1/2<br />
|A| such that |A ∩ P |≥exp(K )|A|.<br />
Using Theorem 8 and Theorem 1, one can directly improve many results based on<br />
Freiman’s theorem; we mention only some of them here. The most exciting case is<br />
Gowers’s theorem; however, here the effect is not so impressive. It is proved in [11,<br />
Theorem 18.6] that, for every k ∈ N, the density of a subset of {1,...,n} which does<br />
not contain any k-term arithmetic progression does not exceed<br />
(log log n) −ck ,<br />
where ck = 2−2k+9. Applying our version of Bogolyubov-Ruzsa’s lemma, one can at<br />
most increase the value of ck. On the other hand, one can deduce from Lemma 5 (and<br />
in view of the Remark) that, for each set A ⊆{1,...,n} of density δ =|A|/n, one<br />
can find an arithmetic progression of length<br />
1/δ)1/2<br />
e−c(log<br />
cδn<br />
in 5A for some absolute constant c>0, beating the previous best bound cδn cδ1/3<br />
obtained by Sanders [20]. Another interesting consequence is a result related to a<br />
problem considered by Konyagin and Łaba [16, Corollary 3.7]. Inserting our estimates<br />
in Sanders’s proof (see [21, Theorem 1.2]), we infer that there exists c>0 such that,<br />
for every finite set A ⊆ R and every transcendental α ∈ R, wehave<br />
|A + α · A| ≥(log |A|) c log log |A| |A|.<br />
Let us also formulate our version of Bogolyubov-Ruzsa’s lemma for F n 2<br />
follows.<br />
<strong>THEOREM</strong> 9<br />
There are absolute constants C1,C2 such that, if A ⊆ Fn 2<br />
to read as<br />
has density δ, then6A<br />
√<br />
contains a subspace H of codimension C1 exp(C2 log(1/δ)).<br />
We refer the reader to [12] for other consequences of Lemma 3 in F n 2 . The last<br />
application provides progress toward Ruzsa’s conjecture, which states that, for every<br />
finite set of squares A, wehave|A + A| ≫|A| 2−ε . Currently, the best estimate<br />
|A + A| ≥(log |A|) 1/12 |A| is due to Chang [4].
10 TOMASZ SCHOEN<br />
<strong>THEOREM</strong> 10<br />
There is a positive constant c such that, for every finite set A ⊆{1 2 , 2 2 ,...}, we have<br />
|A + A| ≥(log |A|) c log log |A| |A|.<br />
Proof<br />
Set |A + A| =K|A|. By Theorem 8, there is a proper generalized arithmetic pro-<br />
C(log K)−1/2<br />
gression P = P1 +···+Pd with d ≤ K and size(P ) ≤ 48K24 |A| such<br />
that<br />
K)−1/2<br />
−KC(log<br />
|A ∩ P |≥e |A|.<br />
Suppose that |P1| =maxi |Pi| ≥|P | 1/d , and write P = <br />
x∈P2+···+Pd (P1 + x). Now,<br />
we make use of a theorem of Bombieri and Zannier [2, Theorem 1] stating that an<br />
arithmetic progression of length k contains O(k2/3+ε ) squares. Therefore, we have<br />
|A ∩ P |≤ <br />
|(P1 + x) ∩ A| ≪ <br />
|P1| 3/4 ≤ (48K 24 |A|) 1−1/(4d) ,<br />
hence<br />
x∈P2+···+Pd<br />
x∈P2+···+Pd<br />
K ≥ (log |A|) c log log |A| ,<br />
and the assertion follows. ✷<br />
It would be interesting to strengthen Lemma 3 since, clearly, each such improvement<br />
would result in better bounds in Freiman’s theorem. The argument presented here has<br />
one major weakness—it works only for small sets X (one would prefer to have an<br />
analog of this result which is also valid for sets X of roughly the same order as Y ).<br />
Indeed, if one could prove a version of Lemma 3 with |X| ∼|Y | and, say ε = 1/20,<br />
then Balog-Szemerédi-Gowers’s theorem would give |X ′ +X ′ |≤K 1−c |X ′ |, for some<br />
X ′ ⊆ X and c>0. Then, by iterating this process, one would most probably get a<br />
result very close to Freiman-Ruzsa’s polynomial conjecture.<br />
References<br />
[1] Y. BILU, “Structure of sets with small sumset” in Structure Theory of Set Addition,<br />
Astérisque 258, Soc. Math. France, Paris, 1999, 77 – 108. MR 1701189<br />
[2] E. BOMBIERI and U. ZANNIER, A note on squares in arithmetic progressions, II, Atti<br />
Accad. Naz. Lincei Cl. Sci. Fis. Mat. Natur. Rend. Lincei (9) Mat. Appl. 13<br />
(2002), 69 – 75. MR 1949479
<strong>NEAR</strong> <strong>OPTIMAL</strong> <strong>BOUNDS</strong> <strong>IN</strong> FREIMAN’S <strong>THEOREM</strong> 11<br />
[3] M.-C. CHANG, A polynomial bound in Freiman’s theorem, Duke Math. J. 113 (2002),<br />
399 – 419. MR 1909605<br />
[4] ———, On problems of Erdős and Rudin, J. Funct. Anal. 207 (2004), 444 – 460.<br />
MR 2032997<br />
[5] ———, On sum-product representations in Zq, J. Eur. Math. Soc. (JEM) 8 (2006),<br />
435 – 463. MR 2250167<br />
[6] ———, Some consequences of the polynomial Freiman-Ruzsa conjecture, C. R. Math.<br />
Acad. Sci. Paris 347 (2009), 583 – 588. MR 2532910 2101, 2, 6, 8911<br />
[7] K. CWAL<strong>IN</strong>A and T. SCHOEN, Linear bound on dimension in Green-Ruzsa’s theorem,<br />
in preparation.<br />
[8] G. A. FREĬMAN, Foundations of a Structural Theory of Set Addition, Trans. Math.<br />
Monog. 37, Amer. Math. Soc., Providence, 1973. MR 0360496<br />
[9] W. T. GOWERS, A new proof of Szemerédi’s theorem for arithmetic progressions of<br />
length four, Geom. Funct. Anal. 8 (1998), 529 – 551. MR 1631259<br />
[10] ———, “Rough structure and classification” in GAFA 2000 (Tel Aviv, 1999), Geom.<br />
Funct. Anal. 2000, 79 – 117. MR 1826250<br />
[11] ———, A new proof of Szemerédi’s theorem, Geom. Funct. Anal. 11 (2001),<br />
465 – 588. MR 1844079<br />
[12] B. GREEN, “Finite field models in additive combinatorics” in Surveys in Combinatorics<br />
2005, London Math. Soc. Lecture Note Ser. 327, Cambridge Univ. Press,<br />
Cambridge, 2005, 1 – 27. MR 2187732<br />
[13] B. GREEN and I. Z. RUZSA, Freiman’s theorem in an arbitrary abelian group, J. Lond.<br />
Math. Soc. (2) 75 (2007), 163 – 175. MR 2302736<br />
[14] B. GREEN and T. TAO, An equivalence between inverse sumset theorems and inverse<br />
conjectures for the U 3 norm, Math. Proc. Cambridge. Philos. Soc. 149 (2010),<br />
1 – 19. MR 2651575<br />
[15] N. H. KATZ and P. KOESTER, On additive doubling and energy, preprint,<br />
arXiv:0802.4371v1 [math.CO]<br />
[16] S. KONYAG<strong>IN</strong> and I. ŁABA, Distance sets of well-distributed planar sets for polygonal<br />
norms, Israel J. Math. 152 (2006), 157 – 175. MR 2214458<br />
[17] T. ŁUCZAK and T. SCHOEN, On a problem of Konyagin, Acta Arith. 134 (2008),<br />
101 – 109. MR 2429639<br />
[18] I. Z. RUZSA, Arithmetical progressions and the number of sums, Period. Math. Hungar.<br />
25 (1992), 105 – 111. MR 1200845<br />
[19] ———, Generalized arithmetical progressions and sumsets, Acta Math. Hungar. 65<br />
(1994), 379 – 388. MR 1281447<br />
[20] T. SANDERS, Additive structures in sumsets, Math. Proc. Cambridge. Philos. Soc. 144<br />
(2008), 289 – 316. MR 2405891<br />
[21] ———, Appendix to Roth’s theorem on progressions revisited by J. Bourgain, J. Anal.<br />
Math. 104 (2008), 193 – 206. MR 2403434<br />
[22] ———, On a nonabelian Balog-Szemerédi-type lemma, J. Aust. Math. Soc. 89 (2010),<br />
127 – 132. MR 2727067<br />
[23] ———, Structure in sets with logarithmic doubling, to appear in Canad. Math. Bull.,<br />
preprint, arXiv:1002.1552v1 [math.CA]
12 TOMASZ SCHOEN<br />
[24] I. SHKREDOV and S. YEKHAN<strong>IN</strong>, Sets with large additive energy and symmetric sets,<br />
J. Combin. Theory Ser. A 118 (2011), 1086 – 1093.<br />
[25] B. SUDAKOV, E. SZEMERÉDI,andV. H. VU, On a question of Erdős and Moser, Duke<br />
Math. J. 129 (2005), 129 – 155. MR 2155059<br />
[26] E. SZEMERÉDI and V. H. VU, Long arithmetic progressions in sum-sets and the number<br />
of x-sum-free sets, Proc. Lond. Math. Soc. (3) 90 (2005), 273 – 296. MR 2142128<br />
[27] ———, Finite and infinite arithmetic progressions in sumsets, Ann. of Math. (2) 163<br />
(2006), 1 – 35. MR 2195131<br />
[28] ———, Long arithmetic progressions in sumsets: Thresholds and bounds, J. Amer.<br />
Math. Soc. 19 (2006), 119 – 169. MR 2169044 21121, 992, 62494611,<br />
8, 91, 944411 1<br />
Faculty of Mathematics and Computer Science, Adam Mickiewicz University, 61-614 Poznań,<br />
Poland; schoen@amu.edu.pl
SUR LA NON-DENSITÉ DES PO<strong>IN</strong>TS ENTIERS<br />
PASCAL AUTISSIER<br />
Résumé<br />
On donne des résultats de non-densité pour les points entiers sur des variétés affines,<br />
dans l’esprit de la conjecture de Lang-Vojta. En particulier, soit X une variété projective<br />
de dimension d ≥ 2 sur un corps de nombres K (resp., sur C). Soit H la somme de<br />
2d diviseurs amples sur X qui se coupent proprement. On montre que tout ensemble<br />
de points quasi-entiers (resp., toute courbe entière) sur X − H est non Zariski-dense.<br />
Abstract<br />
We give nondensity results for integral points on affine varieties, in the spirit of the<br />
Lang-Vojta conjecture. In particular, let X be a projective variety of dimension d ≥ 2<br />
over a number field K (resp., over C). Let H be the sum of 2d properly intersecting<br />
ample divisors on X. We show that any set of quasi-integral points (resp., any integral<br />
curve) on X − H is not Zariski-dense.<br />
1. Introduction<br />
On s’intéresse ici aux solutions à coordonnées (quasi-)entières de systèmes d’équations<br />
polynomiales à coefficients dans un corps de nombres K: on prouve qu’une variété<br />
projective sur K privée de suffisamment d’hypersurfaces n’a pas beaucoup de points<br />
(quasi-)entiers. On a également les résultats analogues en géométrie hyperbolique:<br />
une telle variété (sur C) n’a pas de courbe entière non dégénérée.<br />
Plus précisément, soit S un ensemble fini de places de K. On note OKS l’anneau<br />
des S-entiers de K. Cet article s’inscrit dans le cadre d’un problème important de la<br />
géométrie arithmétique.<br />
CONJECTURE 1.1 ([6, Conjecture 4.2], [13, Conjecture 3.4.3])<br />
Soit X une variété projective lisse sur K de diviseur canonique KX.SoitD un diviseur<br />
effectif sur X, à croisements normaux. Posons Y = X − D. On suppose KX + D big<br />
(par exemple, ample) sur X. Alors tout ensemble E ⊂ Y (K) S-entier sur Y est non<br />
Zariski-dense dans Y .<br />
DUKE MATHEMATICAL JOURNAL<br />
Vol. 158, No. 1, c○ 2011 DOI 10.1215/00127094-1276292<br />
Received 14 December 2009. Revision received 16 July 2010.<br />
2010 Mathematics Subject Classification. Primary 14G25; Secondary 11J97, 11G35.<br />
13
14 PASCAL AUTISSIER<br />
L’hypothèse sur KX + D est une condition de positivité de ce diviseur. Notons que<br />
cette conjecture est encore largement ouverte: le cas où X = P2 K n’est par exemple pas<br />
connu. Levin en a proposé des cas particuliers intéressants, lorsque D a suffisamment<br />
de composantes irréductibles.<br />
CONJECTURE 1.2 ([8, Conjecture 5.3A])<br />
Soit X une variété projective sur K de dimension d ≥ 1. Soient D1,...,Dr des<br />
diviseurs effectifs big sur X qui se coupent proprement. Posons Y = X−D1∪···∪Dr.<br />
On suppose r ≥ d + 2. Alors aucun ensemble S-entier sur Y n’est Zariski-dense<br />
dans Y .<br />
Pour une explication du lien entre 1.1 et 1.2, on pourra consulter [8, paragraphe 14.2].<br />
Un théorème classique de Siegel montre cette conjecture lorsque X est une courbe.<br />
Le cas des surfaces lisses est obtenu par Levin [8, théorème 6.2A.c].<br />
Par ailleurs, c’est une conséquence directe du théorème du sous-espace (voir<br />
Schmidt [11], Schlickewei [10]) lorsque X = P d K et les Di sont des hyperplans<br />
en position générale. Plus généralement, cet énoncé est connu de Vojta [15, corollaire<br />
0.3] pour X lisse et r ≥ d + ρ + 1, oùρ désigne le nombre de Picard de<br />
X K.<br />
Dans un article récent, Corvaja, Levin, et Zannier [2] démontrent la conjecture<br />
1.2 avec d 2 − d + 1 diviseurs Di amples sur X lisse (d ≥ 3). On se propose ici<br />
d’améliorer ce résultat en considérant un nombre de diviseurs linéaire en d (au lieu<br />
de quadratique).<br />
THÉORÈME 1.3<br />
Soit X une variété projective sur K de dimension d ≥ 2. Soient D1,...,D2d des<br />
diviseurs effectifs amples sur X qui se coupent proprement. Posons Y = X − D1 ∪<br />
···∪D2d. Alors tout ensemble E ⊂ Y (K) S-entier sur Y est non Zariski-dense dans<br />
Y .<br />
On l’obtient comme corollaire d’un critère (théorème 2.11) qui donne des conditions<br />
géométriques de non-Zariski-densité des points S-entiers.<br />
La démonstration s’inspire de la méthode introduite par Corvaja et Zannier dans<br />
le cas des courbes (voir [3]) et des surfaces (voir [4]). Cette méthode, dont l’ingrédient<br />
arithmétique principal est le théorème du sous-espace, a ensuite été étendue en dimension<br />
supérieure par Levin [8], l’auteur [1], et par Corvaja, Levin, et Zannier [2].<br />
L’idée nouvelle ici est de considérer un faisceau cohérent codant les ordres<br />
d’annulation optimaux en les Di et de prouver un théorème de convexité (théorème<br />
3.6) sur les sections globales de ce faisceau.<br />
À titre d’application du critère 2.11, on montre en outre l’énoncé suivant.
SUR LA NON-DENSITÉ DES PO<strong>IN</strong>TS ENTIERS 15<br />
THÉORÈME 1.4<br />
Soit X une variété projective sur K de dimension d ≥ 2. Soient D1,...,Dd+2 des<br />
diviseurs effectifs non nuls sur X qui se coupent proprement. Posons L = d+2 i=1 Di<br />
et Y = X − D1 ∪···∪Dd+2. On suppose que L − (d + 1)Di est ample pour tout<br />
i ∈{1,...,d+ 2}. Alors aucun ensemble S-entier sur Y n’est Zariski-dense dans Y .<br />
L’hypothèse sur les L − (d + 1)Di est vérifiée lorsque les Di vivent dans un cône<br />
“suffisamment étroit” du groupe de Néron-Severi de X.<br />
Par ailleurs, Vojta [13] adéveloppé un “dictionnaire” entre la géométrie diophantienne<br />
et la théorie de Nevanlinna: l’étude des points S-entiers sur les variétés sur<br />
K est mise en analogie avec l’étude des courbes entières sur les variétés complexes.<br />
À titre d’exemple, le théorème de Siegel sur les courbes est l’analogue du théorème<br />
de Picard.<br />
Pour l’étayer, on obtient aussi le critère (théorème 2.12) qui, dans ce dictionnaire,<br />
correspond au théorème 2.11.<br />
La section 2 décrit les critères de quasi-hyperbolicité, qui sont démontrés à la<br />
section 4. La section 3 donne le théorème de convexité 3.6 (ainsi qu’un lemme<br />
d’algèbre locale prouvé à la section 6). On en déduit les théorèmes 1.3 et 1.4 à la<br />
section 5.<br />
2. Définitions et énoncés<br />
Soit K un corps. Commençons par quelques définitions de géométrie.<br />
Conventions<br />
On appelle variété sur K tout schéma quasi-projectif et géométriquement intègre sur<br />
K. Le mot “diviseur” sous-entend “diviseur de Cartier”.<br />
Soit X une variété projective sur K de dimension d ≥ 1. Lorsque L est un<br />
diviseur sur X tel que h 0 (X, L) ≥ 1, ondésigne par BL le lieu de base de Ɣ(X, L) et<br />
par L : X − BL → P(Ɣ(X, L)) le morphisme défini par Ɣ(X, L). Pour tout diviseur<br />
effectif D sur X, on note 1D la section globale de OX(D) qu’il définit.<br />
Définition 2.1<br />
Un diviseur L sur X est dit grand lorsque h 0 (X, L) ≥ 1 et L est génériquement fini.<br />
Définition 2.2<br />
Soient D1,...,Dr des diviseurs effectifs sur X. On dit que D1,...,Dr se coupent<br />
proprement lorsque pour toute partie I non vide de {1,...,r}, la section globale<br />
(1Di)i∈I de <br />
i∈I OX(Di) est régulière (autrement dit, pour tout x ∈ <br />
i∈I Di, en<br />
notant ϕi une équation locale de Di en x, les(ϕi)i∈I forment une suite régulière de<br />
l’anneau local OX,x).
16 PASCAL AUTISSIER<br />
Remarque 2.3<br />
Supposons X de Cohen-Macaulay (par exemple, lisse sur K); alors d’après [5, lemme<br />
A.7.1], les diviseurs D1,...,Dr se coupent proprement si et seulement si pour toute<br />
partie I non vide de {1,...,r}, le fermé <br />
i∈I Di est purement de codimension #I<br />
dans X (éventuellement vide).<br />
Définition 2.4<br />
Lorsque L est un diviseur sur X tel que q = h 0 (X, L) ≥ 1 et E un diviseur effectif<br />
non nul sur X, on pose<br />
α(L, E) = 1<br />
q<br />
<br />
h 0 (X, L − kE).<br />
Introduisons ensuite les notions d’hyperbolicité étudiées dans cet article.<br />
k≥1<br />
Convention 2.5<br />
Lorsque K est un corps de nombres et v une place de K,ondésigne par Kv le complété<br />
de K en v, et on normalise la valeur absolue ||v de sorte que |x|v =|x| [Kv:R] si v est<br />
archimédienne et |p|v = p −[Kv:Qp] si v est p-adique.<br />
Lorsque K est un corps de nombres et S un ensemble fini de places de K, on note<br />
OK,S l’anneau des S-entiers de K (i.e., l’ensemble des x ∈ K tels que |x|v ≤ 1 pour<br />
toute place finie v/∈ S).<br />
Soit K un corps de nombres.<br />
Définition 2.6<br />
Soient Y une variété sur K, K ′ une extension finie de K,etS un ensemble fini de places<br />
de K ′ . Un ensemble E ⊂ Y (K ′ ) est dit S-entier sur Y lorsqu’il existe un OK ′ ,S-schéma<br />
intègre et quasi-projectif Y de fibre générique YK ′ tel que E ⊂ Y(OK ′ ,S).<br />
Définition 2.7<br />
Soient Y une variété sur K,etK ′ une extension finie de K. Un ensemble E ⊂ Y (K ′ )<br />
est dit quasi-entier sur Y lorsqu’il existe un ensemble fini S de places de K ′ tel que E<br />
soit S-entier sur Y .<br />
Définition 2.8<br />
Soit Y une variété sur K. On dit que Y est arithmétiquement quasi-hyperbolique<br />
lorsqu’il existe un fermé Z = Y tel que pour toute extension finie K ′ de K et tout<br />
ensemble quasi-entier E ⊂ Y (K ′ ) sur Y , l’ensemble E − Z(K ′ ) soit fini.
SUR LA NON-DENSITÉ DES PO<strong>IN</strong>TS ENTIERS 17<br />
Définition 2.9<br />
Soit U une variété complexe. Une courbe entière sur U est une application holomorphe<br />
f : C → U non constante.<br />
Définition 2.10<br />
Soit U une variété complexe. On dit que U est Brody quasi-hyperbolique lorsqu’il<br />
existe un fermé Z = U tel que pour toute courbe entière f sur U, onaitf (C) ⊂ Z.<br />
Terminons cette section avec les énoncés principaux de ce travail. Ces critères de<br />
quasi-hyperbolicité généralisent (et simplifient) des résultats de Levin [8, théorèmes<br />
8.3A et 8.3B] et de l’auteur (voir [1,théorèmes 3.3 et 3.5]).<br />
THÉORÈME 2.11<br />
Soit X une variété projective sur K de dimension d ≥ 1. Soient D1,...,Dr des<br />
diviseurs effectifs non nuls sur X qui se coupent proprement. Posons Y = X − D1 ∪<br />
···∪Dr. Soit m ≥ 1 un entier. On suppose que le diviseur L = m r i=1 Di est grand<br />
sur X et que α(L, Di) >mpour tout i ∈{1,...,r}. AlorsYest arithmétiquement<br />
quasi-hyperbolique.<br />
THÉORÈME 2.12<br />
Soit V une variété complexe projective de dimension d ≥ 1. Soient D1,...,Dr des<br />
diviseurs effectifs non nuls sur V qui se coupent proprement. Posons U = V − D1 ∪<br />
···∪Dr. Soit m ≥ 1 un entier. On suppose que le diviseur L = m r i=1 Di est<br />
grand sur V et que α(L, Di) >mpour tout i ∈{1,...,r}. AlorsUest Brody<br />
quasi-hyperbolique.<br />
3. Préliminaires<br />
Soient K un corps et W un K-espace vectoriel de dimension finie.<br />
Définitions<br />
Une filtration de W est une famille décroissante F = (Fx)x∈R+ de sous-espaces<br />
vectoriels de W telle que Fx ={0} pour tout x assez grand. Une base B de W est<br />
dite adaptée à la filtration F lorsque B ∩ Fx est une base de Fx pour tout réel x ≥ 0.<br />
LEMME 3.1<br />
Soient F et G deux filtrations de W . Il existe alors une base de W adaptée à F et à<br />
G .<br />
Démonstration<br />
C’est le lemme 3.2 de [4]. <br />
Soit r un entier ≥ 1. On note ici l’ensemble des t = (t1,...,tr) ∈ Rr + tels que<br />
t1 +···+tr = 1.
18 PASCAL AUTISSIER<br />
Définition 3.2<br />
Une partie N de N r est dite saturée lorsque a + b ∈ N pour tout a ∈ N r et tout<br />
b ∈ N.<br />
Soit A un anneau local noethérien. Soit (ϕ1,...,ϕr) une suite régulière de A. Pour<br />
toute partie saturée N de Nr ,ondésigne par I(N) l’idéal de A engendré par l’ensemble<br />
{ϕ b1<br />
1 ,...,ϕbr r , b ∈ N}. On aura besoin d’un résultat d’algèbre locale (qui ne sera<br />
démontré qu’àlasection6).<br />
LEMME 3.3<br />
Soient N et M deux parties saturées de N r . On a alors l’égalité d’idéaux<br />
I(N) ∩ I(M) = I(N ∩ M).<br />
Remarque 3.4<br />
On utilisera dans la suite le cas particulier suivant du lemme 3.3: pour t ∈ et<br />
x ∈ R+, notons N(t,x) l’ensemble des b ∈ Nr tels que t1b1 +···+trbr ≥ x. Ona<br />
alors l’inclusion<br />
I N(t,x) ∩ I N(u,y) ⊂ I N(λt + (1 − λ)u,λx+ (1 − λ)y) <br />
pour tous t, x, u, y, et tout λ ∈ [0, 1], puisque N(t,x) ∩ N(u,y) ⊂ N(λt + (1 −<br />
λ)u,λx+ (1 − λ)y).<br />
Maintenant, soit X une variété projective sur K de dimension d ≥ 1. SoitLun diviseur sur X tel que q = h0 (X, L) ≥ 1. SoientD1,...,Dr des diviseurs effectifs<br />
sur X qui se coupent proprement, tels que r i=1 Di soit non vide.<br />
Introduisons quelques notations. Soit t ∈ . Pour x ∈ R+, ondéfinit l’idéal<br />
J(t,x) de OX par<br />
J(t,x) = <br />
b∈N(t,x)<br />
OX<br />
<br />
−<br />
r<br />
biDi<br />
i=1<br />
et on pose F (t)x = Ɣ(X, J(t,x) ⊗ L). Observons que (F (t)x)x∈R+ est une filtration<br />
de Ɣ(X, L). Pour tout s ∈ Ɣ(X, L) −{0}, on note µt(s) = sup{y ∈ R+ | s ∈ F (t)y}.<br />
On pose finalement<br />
F ( t ) = 1<br />
+∞<br />
h<br />
q<br />
0 X, J( t,x) ⊗ L dx.<br />
Remarque 3.5<br />
Soit B = (s1,...,sq) une base de Ɣ(X, L). On a alors l’inégalité<br />
F ( t ) ≥ 1<br />
q<br />
+∞<br />
0<br />
0<br />
# F ( t )x ∩ B dx = 1<br />
q<br />
<br />
,<br />
q<br />
µt(sk),<br />
k=1
SUR LA NON-DENSITÉ DES PO<strong>IN</strong>TS ENTIERS 19<br />
qui est même une égalité siB est adaptée à la filtration F ( t ).<br />
Voici le résultat principal de cette section.<br />
THÉORÈME 3.6<br />
L’application F : → R+ ainsi définie est concave. On a en particulier F ( t ) ≥<br />
mini α(L, Di) pour tout t ∈ .<br />
Démonstration<br />
Soient t ∈ , u ∈ et λ ∈ [0, 1]. Il s’agit de montrer que<br />
F λt + (1 − λ)u ≥ λF ( t ) + (1 − λ)F ( u ).<br />
Le lemme 3.1 construit une base B = (s1,...,sq) de W = Ɣ(X, L) adaptée aux<br />
filtrations F ( t ) et F ( u ).<br />
Soit (x,y) ∈ R2 + . Le lemme 3.3 (voir aussi remarque 3.4) fournit l’inclusion<br />
d’idéaux J(t,x) ∩ J(u,y) ⊂ J(λt + (1 − λ)u,λx+ (1 − λ)y), puisque les diviseurs<br />
D1,...,Dr se coupent proprement. On a en particulier l’inclusion d’espaces vectoriels<br />
F ( t )x ∩ F ( u )y ⊂ F λt + (1 − λ) u <br />
λx+(1−λ)y .<br />
Soit s ∈ W −{0}. Ce qui précède implique que s appartient à F (λt + (1 −<br />
λ)u)λx+(1−λ)y pour tout x
20 PASCAL AUTISSIER<br />
Choisissons une section rationnelle s de L définie et non nulle en P . On pose<br />
1<br />
h ˆL(P ) =−<br />
[K ′ <br />
ln s(P )v,<br />
: Q]<br />
où v parcourt l’ensemble des places de K ′ .Ceréel ne dépend pas du choix de s.<br />
On va utiliser la version suivante du théorème du sous-espace.<br />
PROPOSITION 4.2<br />
On suppose L grand sur X. Notons q = h0 (X, L). Soient s1,...,sl des sections non<br />
nulles engendrant Ɣ(X, L). Soit ε>0. Il existe alors un fermé Z = X tel que pour<br />
toute extension finie K ′ de K et tout ensemble fini S de places de K ′ , l’ensemble des<br />
points P ∈ (X − Z)(K ′ ) vérifiant<br />
<br />
v∈S<br />
max<br />
J ∈L<br />
<br />
j∈J<br />
v<br />
ln sj(P ) −1<br />
v ≥ (q + qε)[K′ : Q]h ˆL(P ) (1)<br />
est fini, où L désigne l’ensemble des parties J de {1,...,l} telles que (sj)j∈J soit une<br />
base de Ɣ(X, L).<br />
Démonstration<br />
Posons P = P(Ɣ(X, L)).Désignons par π : X ′ → X l’éclatement de X relativement<br />
à BL et par E = π −1 (BL) le diviseur exceptionnel. En notant L ′ = OX ′(−E) ⊗ π ∗L, on a un morphisme L ′ : X′ → P génériquement fini qui prolonge L. Il existe donc<br />
un fermé Z1 = X ′ tel que L ′ |X ′ −Z1 soit à fibres finies.<br />
On munit L0 = OP(1) d’une métrique adélique ( ′ v )v telle que la section globale<br />
1E de ∗ L ′L∨0 ⊗ π ∗L vérifie<br />
sup<br />
X ′ 1E<br />
v<br />
′<br />
v ≤ 1<br />
pour toute place v. On applique alors la version de Vojta [14, théorème 0.3, reformulation<br />
3.4] du théorème du sous-espace.<br />
Il existe une réunion finie H de K-hyperplans de P P q−1<br />
K telle que pour toute<br />
extension finie K ′ de K et tout ensemble fini S de places de K ′ , l’ensemble des points<br />
P ∈ (P − H )(K ′ ) vérifiant<br />
<br />
ln sj(P ) ′−1<br />
v<br />
v∈S<br />
max<br />
J ∈L<br />
j∈J<br />
≥ (q + qε)[K′ : Q]hL0 ˆ (P )<br />
est fini.<br />
Prenons Z = BL ∪ π(Z1 ∪ −1<br />
L ′ (H )). Alors pour toute extension finie K′ de K<br />
et tout ensemble fini S de places de K ′ , l’ensemble des points P ∈ (X − Z)(K ′ )<br />
vérifiant (1) est fini.
SUR LA NON-DENSITÉ DES PO<strong>IN</strong>TS ENTIERS 21<br />
Montrons donc les critères de quasi-hyperbolicité delasection2.<br />
Démonstration du théorème 2.11<br />
On reprend en l’améliorant la démonstration du [1,théorème 3.3]. On procède en deux<br />
étapes: dans la première, on construit un fermé Z = X candidat à contenir “presque<br />
tous les points entiers”; dans la seconde, on prouve que Y est arithmétiquement quasihyperbolique.<br />
Étape 1<br />
Posons ε = (1/4m)(min α(L, Di) − m) et q = h<br />
i 0 (X, L), et fixons un entier b ≥<br />
(1/ε + 3)r. On choisit aussi une base B0 de Ɣ(X, L).<br />
Désignons par P l’ensemble des parties I non vides de {1,...,r} telles que<br />
<br />
i∈I Di soit non vide. Soit I ∈ P . On note I l’ensemble des a = (ai)i ∈ NI tels<br />
que <br />
i∈I ai = b. Soita∈I . Pour x ∈ R+, ondéfinit l’idéal J(x) de OX par<br />
J(x) = <br />
− <br />
b<br />
OX<br />
où la somme porte sur les b ∈ NI tels que<br />
<br />
aibi ≥ bx,<br />
i∈I<br />
i∈I<br />
biDi<br />
et on pose F (I,a)x = Ɣ(X, J(x) ⊗ L). Le lemme 3.1 fournit une base BI,a de<br />
Ɣ(X, L) adaptée à la filtration F (I,a).<br />
On munit chaque faisceau OX(Di) d’une métrique adélique. Appliquons le<br />
théorème du sous-espace (proposition 4.2) avec {s1,...,sl} =B0 ∪ <br />
I,a BI,a (remarquons<br />
que cette réunion est finie puisque P et les I le sont).<br />
Il existe un fermé Z = X tel que pour toute extension finie K ′ de K et tout<br />
ensemble fini S de places de K ′ , l’ensemble des points P ∈ (X − Z)(K ′ ) vérifiant<br />
l’inégalité (1) est fini.<br />
Étape 2<br />
Soient K ′ une extension finie de K et S un ensemble fini de places de K ′ contenant<br />
les places archimédiennes. Soit E ⊂ Y (K ′ ) un ensemble S-entier sur Y . Raisonnons<br />
par l’absurde en supposant E − Z(K ′ ) infini. Choisissons une suite injective (Pn)n≥0<br />
d’éléments de E − Z(K ′ ).<br />
Quitte à extraire, on peut supposer (par compacité) que pour tout v ∈ S, la suite<br />
(Pn)n≥0 converge dans X(K ′ v ) vers un yv ∈ X(K ′ v ).<br />
Pour tout v ∈ S, on note Iv l’ensemble des i ∈{1,...,r} tels que yv ∈ Di.<br />
Quitte à extraire de nouveau, on peut supposer que pour tout v ∈ S tel que Iv soit non
22 PASCAL AUTISSIER<br />
vide et tout i ∈ Iv, la suite<br />
<br />
ln 1Di(Pn)v<br />
<br />
ln 1Dj (Pn)v<br />
j∈Iv<br />
converge vers un tvi ∈ [0, 1]. Observons que l’on a <br />
i∈Iv tvi = 1.<br />
Fait: Soitv∈S. Il existe une base (s1v,...,sqv) de Ɣ(X, L) contenue dans<br />
{s1,...,sl} telle que l’on ait la minoration suivante pour tout n ≥ 0:<br />
−<br />
q<br />
ln skv(Pn)v ≥−(q + 2qε)ln1L(Pn)v − O(1), (2)<br />
k=1<br />
où le O(1) est indépendant de n.<br />
Prouvons ce fait. Si Iv est vide, on prend {s1v,...,sqv} =B0 et on obtient la<br />
minoration (2) en remarquant que ln 1L(Pn)v = O(1) (puisque yv /∈ L).<br />
On suppose maintenant Iv non vide. On a donc Iv ∈ P . Choisissons un a v =<br />
(avi)i ∈ Iv tel que avi ≤ (b + r)tvi pour tout i ∈ Iv. On prend alors {s1v,...,sqv} =<br />
BIv,a v .Vérifions que ce choix convient.<br />
Soit s ∈ Ɣ(X, L) −{0}. Notons µ(s) le plus grand rationnel µ tel que s ∈<br />
F (Iv,av )µ. Enécrivant localement s = <br />
b fb<br />
<br />
i∈Iv 1bi<br />
Di au voisinage de yv, on<br />
trouve<br />
− ln s(Pn)v ≥−max<br />
b<br />
Par définition des tvi, ona<br />
−bi ln 1Di(Pn)v ≥−<br />
n≥0<br />
<br />
bi ln 1Di(Pn)v − O(1).<br />
i∈Iv<br />
avibi<br />
b + r<br />
− mε<br />
r<br />
<br />
ln 1Dj (Pn)v<br />
pour tout n assez grand et tout i ∈ Iv.<br />
On en déduit l’inégalité (pour tout n ≥ 0)<br />
<br />
µ(s)b<br />
<br />
− ln s(Pn)v ≥− − mε ln 1Dj (Pn)v − O(1).<br />
b + r<br />
On écrit cette inégalité pour s = skv, puis on somme sur k. En observant que l’on a<br />
q<br />
µ(skv) =<br />
k=1<br />
j∈Iv<br />
j∈Iv<br />
+∞<br />
dim F (Iv,av )x dx (remarque 3.5)<br />
0<br />
≥ min α(L, Di)q (théorème 3.6)<br />
i<br />
b + r<br />
= (1 + 4ε)qm ≥ (q + 3qε)m,<br />
b
SUR LA NON-DENSITÉ DES PO<strong>IN</strong>TS ENTIERS 23<br />
on obtient alors<br />
−<br />
q<br />
ln skv(Pn)v ≥−(q + 2qε)m <br />
ln 1Dj (Pn)v − O(1).<br />
k=1<br />
Le fait énoncé (2) s’en déduit en remarquant que ln 1Dj (Pn)v = O(1) pour tout<br />
j/∈ Iv.<br />
Maintenant, l’ensemble E est S-entier sur Y , donc pour tout n ≥ 0, ona<br />
[K ′ : Q]h ˆL(Pn) =− <br />
ln 1L(Pn)v + O(1).<br />
v∈S<br />
j∈Iv<br />
En utilisant la minoration (2), on trouve (pour tout n ≥ 0)<br />
− <br />
v∈S<br />
q<br />
ln skv(Pn)v ≥ (q + 2qε)[K ′ : Q]h ˆL(Pn) − O(1).<br />
k=1<br />
D’où une contradiction avec (1). <br />
Démonstration du théorème 2.12<br />
On reprend de la même manière la démonstration du [1, théorème 3.5], qui utilise<br />
la version de Vojta [16, théorème 2] du théorème de Cartan au lieu du théorème du<br />
sous-espace. Les détails sont laissés au lecteur. <br />
5. Démonstration des théorèmes 1.3 et 1.4<br />
Soient K un corps de nombres et X une variété projective sur K de dimension<br />
d ≥ 2. Pour tous diviseurs L1,...,Ld sur X,ondésigne par <br />
L1 ···Ld leur nombre<br />
d’intersection.<br />
Définition 5.1<br />
Un diviseur L sur X est dit big lorsque lim inf<br />
n→+∞ (1/nd )h 0 (X, nL) > 0.<br />
Remarque 5.2<br />
D’après le lemme de Kodaira [7, p. 141], le diviseur L est big si et seulement si nL<br />
est grand pour tout entier n assez grand.<br />
Définition 5.3<br />
Soit L un diviseur big sur X. On dit que L est presque ample lorsqu’il existe un entier<br />
n ≥ 1 tel que BnL soit vide.<br />
On montre ici de légères extensions des théorèmes 1.3 et 1.4 de l’introduction.
24 PASCAL AUTISSIER<br />
THÉORÈME 1.3BIS<br />
Soient D1,...,D2d des diviseurs effectifs presque amples sur X qui se coupent<br />
proprement. Alors X − D1 ∪···∪D2d est arithmétiquement quasi-hyperbolique.<br />
Démonstration<br />
D’après [1, théorème 4.4], il existe (m1,...,m2d) ∈ (N ∗ ) 2d tel qu’en posant L =<br />
2d<br />
i=1 miDi, onait<br />
lim inf<br />
n→+∞<br />
1<br />
n α(nL, miDi) > 1 pour tout i ∈{1,...,2d}.<br />
On en déduit le résultat en appliquant le théorème 2.11 (les diviseurs<br />
m1D1,...,m2dD2d se coupent encore proprement). <br />
THÉORÈME 1.4BIS<br />
Soient D1,...,Dr des diviseurs effectifs non nuls sur X qui se coupent proprement,<br />
avec r ≥ d + 2. Posons L = r i=1 Di et Y = X − D1 ∪···∪Dr. On suppose<br />
que L − (d + 1)Di est ample pour tout i ∈{1,...,r}.AlorsYest arithmétiquement<br />
quasi-hyperbolique.<br />
Démonstration<br />
D’après le lemme 5.4 ci-dessous, on a lim inf<br />
n→+∞ (1/n)α(nL, Di) > 1 pour tout i ∈<br />
{1,...,r}. On obtient le résultat en appliquant le théorème 2.11. <br />
LEMME 5.4<br />
Soient L un diviseur ample sur X et D un diviseur effectif non nul sur X. Soit t ≥ 1<br />
un entier tel que L − tD soit ample. On a alors les minorations<br />
lim inf<br />
n→+∞<br />
1<br />
α(nL, D) ≥<br />
n<br />
t<br />
(d + 1) L d<br />
d<br />
j=0<br />
L d−j (L − tD) j > t<br />
d + 1 .<br />
Démonstration<br />
La formule de Hirzebruch-Riemann-Roch donne que χ(X, nL−kD) est une fonction<br />
polynomiale en (n, k) dont on peut expliciter la composante homogène dominante.<br />
Pour tout (n, k) tel que 0 ≤ k ≤ tn,onaχ(X, nL − kD) = (1/d!)〈(nL −<br />
kD) d 〉+O(n d−1 ).<br />
Par ailleurs, pour tout n assez grand et tout k ∈{0,...,tn}, onah i (X, nL −<br />
kD) = 0 pour tout i ≥ 1, puisque L et L − tD sont amples. On a en particulier<br />
h 0 (X, nL − kD) = (1/d!)〈(nL − kD) d 〉+O(n d−1 ).
SUR LA NON-DENSITÉ DES PO<strong>IN</strong>TS ENTIERS 25<br />
On trouve ainsi les estimations suivantes:<br />
tn<br />
k=1<br />
h 0 (X, nL − kD) = 1<br />
d!<br />
= 1<br />
d!<br />
= 1<br />
d!<br />
tn<br />
k=1<br />
d<br />
j=0 k=1<br />
d<br />
j=0<br />
〈(nL − kD) d 〉+O(n d−1 ) <br />
tn<br />
Or un calcul montre (par multilinéarité) la formule<br />
d<br />
j=0<br />
C j (−1) j<br />
d−j j<br />
d L D<br />
j + 1 t j+1 = t<br />
d + 1<br />
C j<br />
d 〈Ld−j D j 〉n d−j (−k) j + O(n d )<br />
C j<br />
d 〈Ld−j D j 〉 (−1)j<br />
j + 1 t j+1 n d+1 + O(n d ).<br />
d<br />
〈L d−j (L − tD) j 〉.<br />
D’où la première inégalité del’énoncé.<br />
La deuxième inégalité s’en déduit facilement: les diviseurs L et L − tD sont<br />
amples, donc on a 〈L d−j (L − tD) j 〉 > 0 pour tout j ∈{1,...,d}. <br />
6. Démonstration du lemme 3.3<br />
Soit A un anneau local noethérien. On aura besoin d’un résultat classique dû à Krull.<br />
LEMME 6.1<br />
Soient I et J deux idéaux de A avec J = A. On a alors l’égalité<br />
<br />
(I + J k ) = I.<br />
k≥1<br />
Démonstration<br />
Il suffit d’appliquer [9, corollaire 11.D.2, p. 69], dans l’anneau A/I. <br />
Soit (ϕ1,...,ϕr) une suite régulière de A. Rappelons que pour toute partie saturée N<br />
de Nr ,ondésigne par I(N) l’idéal 〈ϕ b1<br />
1 ···ϕbr r ,b∈ N〉 de A. Le lemme 3.3 est une<br />
conséquence directe de l’énoncé suivant.<br />
LEMME 6.2<br />
Soit N une partie saturée de N r . On a alors l’égalité d’idéaux<br />
Démonstration<br />
Notons ici I1(N) = <br />
I(N) = <br />
c∈N r −N<br />
〈ϕ c1+1<br />
1<br />
c∈Nr −N 〈ϕc1+1 1 ,...,ϕcr +1<br />
r<br />
j=0<br />
,...,ϕ cr +1<br />
r<br />
〉.<br />
〉. Vérifions d’abord l’inclusion
26 PASCAL AUTISSIER<br />
I(N) ⊂ I1(N):sib ∈ N et c ∈ Nr − N, alors il existe un indice i tel que bi ≥ ci + 1,<br />
donc ϕ b1<br />
1 ,...,ϕbr r ∈ ϕci+1 i A ⊂〈ϕc1+1 1 ,...,ϕcr +1<br />
r 〉.<br />
On prouve l’inclusion I1(N) ⊂ I(N) par récurrence sur r. Le cas r = 1 est<br />
facile (avec le lemme 6.1 si N est vide). Supposons r ≥ 2 et le résultat au cran r − 1.<br />
Montrons par une (deuxième) récurrence sur k que I1(N) ⊂ I(N) + ϕk 1A pour tout<br />
k ∈ N. Sik = 0, cette inclusion est évidente. Supposons k ≥ 1 et le résultat au cran<br />
k − 1. Soitsun élément de I1(N). Ils’écrit s = x + ϕ k−1<br />
1 a avec un x ∈ I(N) et un<br />
a ∈ A.<br />
Posons Nk = {b = (b2,...,br) ∈ Nr−1 | (k − 1,b2,...,br) ∈ N}. Soit<br />
b ∈ N r−1 − Nk. Alorsϕ k−1<br />
1 a = s − x appartient à I1(N) donc en particulier à<br />
〈ϕk 1 ,ϕb2+1 2 ,...,ϕbr +1<br />
r 〉. Autrement dit, il existe a ′ ∈ A tel que ϕ k−1<br />
1 a − ϕk 1a′ ∈<br />
〈ϕ b2+1<br />
2 ,...,ϕbr +1<br />
r 〉. Puisque (ϕ b2+1<br />
2 ,...,ϕbr +1<br />
r ,ϕ1) est une suite régulière de A, on<br />
obtient a − ϕ1a ′ ∈〈ϕ b2+1<br />
2 ,...,ϕbr +1<br />
r 〉 en simplifiant par ϕ k−1<br />
1 .Réduisons modulo ϕ1<br />
on trouve dans A ′ = A/ϕ1A que ā ∈〈¯ ϕ2 b2+1 ,..., ϕr ¯ br +1 〉.<br />
On a ainsi que ā ∈ <br />
b∈Nr−1 〈 ¯ −Nk ϕ2 b2+1 ,..., ϕr ¯ br +1 〉.Or(¯ ϕ2,..., ϕr) ¯ est une suite<br />
régulière de A ′ , donc ā ∈〈¯ ϕ2 b2 ··· ϕr ¯ br , b ∈ Nk〉 par la première hypothèse de<br />
récurrence. On écrit a = y + ϕ1a1 avec y ∈〈ϕ b2<br />
2 ···ϕbr r ,b∈ Nk〉 et a1 ∈ A. Ilen<br />
découle l’égalité s = x + ϕ k−1<br />
1 y + ϕk 1a1 avec ϕ k−1<br />
1 y ∈〈ϕk−1 1 ϕb2 2 ···ϕbr r ,b∈ Nk〉 ⊂<br />
I(N), qui implique bien que s est dans I(N) + ϕk 1A. D’où l’inclusion I1(N) ⊂ <br />
k∈N (I(N) + ϕk 1A). On conclut par le lemme 6.1. <br />
Remerciements. Je remercie Yuri Bilu pour m’avoir fourni la référence [2]. Je remercie<br />
également les rapporteurs pour leurs suggestions pertinentes.<br />
References<br />
[1] P. AUTISSIER, Géométrie, points entiers et courbes entières, Ann. Sci. Éc. Norm.<br />
Supér. (4) 42 (2009), 221 – 239. MR 2518077 14, 17, 21, 23, 24<br />
[2] P. CORVAJA, A. LEV<strong>IN</strong> et U. ZANNIER, Integral points on threefolds and other<br />
varieties, Tohoku Math. J. 61 (2009), 589 – 601. MR 2598251 14, 26<br />
[3] P. CORVAJA et U. ZANNIER, A subspace theorem approach to integral points on curves,<br />
C. R. Math. Acad. Sci. Paris 334 (2002), 267 – 271. MR 1891001 14<br />
[4] ———, On integral points on surfaces, Ann. of. Math. (2) 160 (2004), 705 – 726.<br />
MR 2123936 14, 17<br />
[5] W. FULTON, Intersection Theory, 2nd ed., Ergeb. Math. Grenzgeb. 3, Springer, Berlin,<br />
1998. MR 1644323 16<br />
[6] S. LANG, Number Theory, III: Diophantine Geometry, Encyclopaedia Math. Sci. 60,<br />
Springer, Berlin, 1991. MR 1112552 13
SUR LA NON-DENSITÉ DES PO<strong>IN</strong>TS ENTIERS 27<br />
[7] R. LAZARSFELD, Positivity in Algebraic Geometry, I: Classical Setting—Line Bundles<br />
and Linear Series, Ergeb. Math. Grenzgeb. 48, Springer, Berlin, 2004.<br />
MR 2095471 23<br />
[8] A. LEV<strong>IN</strong>, Generalizations of Siegel’s and Picard’s theorems, Ann. of Math. (2) 170<br />
(2009), 609 – 655. MR 2552103 14, 17<br />
[9] H. MATSUMURA, Commutative Algebra, 2nd ed., Math. Lecture Note Ser. 56,<br />
Benjamin/Cummings, Reading, Mass., 1980. MR 0575344 25<br />
[10] H. P. SCHLICKEWEI, The p-adic Thue-Siegel-Roth-Schmidt theorem, Arch. Math.<br />
(Basel) 29 (1977), 267 – 270. MR 0491529 14<br />
[11] W. M. SCHMIDT, Diophantine approximation, Lecture Notes in Math. 785, Springer<br />
Berlin, 1980. MR 0568710 14<br />
[12] J.-P. SERRE, Lectures on the Mordell-Weil Theorem, 3rd ed., Aspects Math., Friedr.<br />
Vieweg and Sons, Braunschweig, 1997. MR 1757192<br />
[13] P. VOJTA, Diophantine Approximations and Value Distribution Theory, Lecture Notes<br />
in Math. 1239, Springer, Berlin, 1987. MR 0883451 13, 15<br />
[14] ———, A refinement of Schmidt’s subspace theorem, Amer. J. Math. 111 (1989),<br />
489 – 518. MR 1002010 20<br />
[15] ———, Integral points on subvarieties of semiabelian varieties, I, Invent. Math. 126<br />
(1996), 133 – 181. MR 1408559 14<br />
[16] ———, On Cartan’s theorem and Cartan’s conjecture, Amer. J. Math. 119 (1997),<br />
1 – 17. MR 1428056 23<br />
[17] S. ZHANG, Small points and adelic metrics,J.AlgebraicGeom.4 (1995), 281 – 300.<br />
MR 1311351 19<br />
Institut de Mathématiques de Bordeaux, Université Bordeaux I, 33405 Talence CEDEX,<br />
France; pascal.autissier@math.u-bordeaux1.fr
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS<br />
CASIM ABBAS<br />
Abstract<br />
Giroux showed that every contact structure on a closed 3-dimensional manifold is<br />
supported by an open book decomposition. We extend this result by showing that the<br />
open book decomposition can be chosen in such a way that the pages are solutions to<br />
a homological perturbed holomorphic curve equation.<br />
Contents<br />
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29<br />
2. Existence and local foliations . . . . . . . . . . . . . . . . . . . . . . . 36<br />
3. From local foliations to global ones . . . . . . . . . . . . . . . . . . . 52<br />
4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77<br />
Appendix. Some local computations near the punctures . . . . . . . . . . . 78<br />
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80<br />
1. Introduction<br />
This paper is the starting point of a larger program by the author, Hofer, and Lisi<br />
investigating a perturbed holomorphic curve equation in the symplectization of a 3dimensional<br />
contact manifold (see [6], [7]). One aim of this program is to provide<br />
an alternative proof of the Weinstein conjecture in dimension 3 as outlined in [4]<br />
complementing Taubes’s gauge theoretical proof (see [33], [34]). A special case of<br />
this paper’s main result has been used in the proof of the Weinstein conjecture for<br />
planar contact structures in [4]. Another reason for studying this equation is to construct<br />
foliations by surfaces of section with nontrivial genus. This is usually impossible to<br />
do with the unperturbed holomorphic curve equation since solutions generically do<br />
not exist.<br />
Consider a closed 3-dimensional manifold M equipped with a contact form λ.<br />
This is a 1-form which satisfies λ ∧ dλ = 0 at every point of M. We denote the<br />
associated contact structure by ξ = ker λ, and we denote the Reeb vector field by Xλ.<br />
DUKE MATHEMATICAL JOURNAL<br />
Vol. 158, No. 1, c○ 2011 DOI 10.1215/00127094-1276301<br />
Received 20 July 2009. Revision received 8 July 2010.<br />
2010 Mathematics Subject Classification. Primary 53D35; Secondary 35J62.<br />
29
30 CASIM ABBAS<br />
Recall that the Reeb vector field is defined by the two equations<br />
iXλdλ = 0 and iXλλ = 1.<br />
Definition 1.1 (Open book decomposition)<br />
Assume that K ⊂ M is a link in M and that τ : M\K → S 1 is a fibration so<br />
that the fibers Fϑ = τ −1 (ϑ) are interiors of compact embedded surfaces ¯Fϑ with<br />
∂ ¯Fϑ = K, where ϑ is the coordinate along K. We also assume that K has a tubular<br />
neighborhood K × D, D ⊂ R 2 being the open unit disk, such that τ restricted to<br />
K × (D\{0}) is given by τ(ϑ, r, φ) = φ, where (r, φ) are polar coordinates on D.<br />
Then we call τ an open book decomposition of M, the link K is called the binding of<br />
the open book decomposition, and the surfaces Fϑ are called the pages of the open<br />
book decomposition.<br />
It is a well-known result in 3-dimensional topology that every closed 3-dimensional<br />
orientable manifold admits an open book decomposition. Indeed, Alexander proved<br />
the following theorem in 1923 (see [10], [30]):<br />
<strong>THEOREM</strong> 1.2<br />
Every closed, orientable manifold M of dimension 3 is diffeomorphic to<br />
W (h) ∪Id (∂W × D 2 ),<br />
where D 2 is the closed unit disk in R 2 ,whereW is an orientable surface with boundary,<br />
and where h : W → W is an orientation-preserving diffeomorphism which restricts<br />
to the identity near ∂W.HereW (h) denotes the manifold obtained from W × [0, 2π]<br />
by identifying (x,0) with (h(x), 2π).<br />
The above decomposition is an open book decomposition, and the pages are given by<br />
Fϑ := (W ×{ϑ}) ∪Id (∂W × Iϑ), 0 ≤ ϑ
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 31<br />
dλ ′ induces an area form on each fiber Fϑ with K consisting of closed orbits of the<br />
Reeb vector field Xλ ′,andλ′ orients K as the boundary of (Fϑ,dλ ′ ).<br />
We refer to a contact form λ ′ above as a Giroux contact form. Note that λ ′ is not unique<br />
and that it is in general different from the original contact form λ. The following<br />
theorem by Giroux guarantees existence of such open book decompositions, and it<br />
contains a uniqueness statement as well (see also [16, Proposition 2]).<br />
<strong>THEOREM</strong> 1.4 ([16, Theorem 3])<br />
Every co-oriented contact structure ξ = ker λ on a closed 3-dimensional manifold is<br />
supported by some open book. Conversely, if two contact structures are supported by<br />
the same open book, then they are diffeomorphic.<br />
In the topological category, it is possible to modify an open book decomposition<br />
such that the pages of the new decomposition have lower genus at the expense of<br />
increasing the number of connected components of K. It was not known for some<br />
time whether a similar statement could also be made in the context of supporting open<br />
book decompositions. In particular, it was unclear whether every contact structure<br />
was supported by an open book decomposition whose pages were punctured spheres<br />
(planar pages). The author and his collaborators could resolve the Weinstein conjecture<br />
for contact forms inducing a planar contact structure in 2005 (see [4]). So the question<br />
of whether all contact structures are planar became a priority, which prompted Etnyre<br />
to address it in [14]. He showed that overtwisted contact structures always admit<br />
supporting open book decompositions with planar pages, but many contact structures<br />
do not. Since then, planar open book decompositions have become an important tool<br />
in contact geometry.<br />
In this paper, we will prove that every contact structure has a supporting open book<br />
decomposition such that the pages solve a homological perturbed Cauchy-Riemann<br />
type equation which we now describe after introducing some notation. We write<br />
πλ = π : TM → ξ for the projection along the Reeb vector field Xλ. Fix a complex<br />
multiplication J : ξ → ξ so that the map ξ ⊕ ξ → R, definedby<br />
(h, k) → dλ(h, J k),<br />
defines a positive definite metric on the fibers. We call such complex multiplications<br />
compatible (with dλ). The equation of interest here is the following nonlinear firstorder<br />
elliptic system. The solutions consist of 5-tuplets (S,j,Ɣ,ũ, γ ) where (S,j)<br />
is a closed Riemann surface with complex structure j, Ɣ ⊂ S is a finite subset,<br />
ũ = (a,u) : ˙S → R × M is a proper map with ˙S = S \ Ɣ,andγ is a 1-form on S so
32 CASIM ABBAS<br />
that<br />
Here the energy E(ũ) is defined by<br />
⎧<br />
⎪⎨<br />
π ◦ Tu◦ j = J ◦ π ◦ Tu on ˙S<br />
(u<br />
⎪⎩<br />
∗λ) ◦ j = da + γ on ˙S<br />
dγ = d(γ ◦ j) = 0 on S<br />
E(ũ) < ∞.<br />
E(ũ) = sup<br />
ϕ∈<br />
<br />
˙S<br />
ũ ∗ d(ϕλ),<br />
(1.1)<br />
where consists of all smooth maps ϕ : R → [0, 1] with ϕ ′ (s) ≥ 0 for all s ∈ R.<br />
Note that equation (1.1) reduces to the usual pseudoholomorphic curve equation<br />
in the symplectization R × M if we set γ = 0. The following proposition, which is a<br />
modification of a result by Hofer [17], shows that solutions to problem (1.1) approach<br />
cylinders over periodic orbits of the Reeb vector field.<br />
PROPOSITION 1.5<br />
Let (M,λ) be a closed 3-dimensional manifold equipped with a contact form λ.Then<br />
the associated Reeb vector field has periodic orbits if and only if the associated PDE<br />
problem (1.1) has a nonconstant solution.<br />
Proof<br />
Let (S,j,Ɣ,ũ, γ ) be a nonconstant solution of (1.1). If Ɣ = ∅, then the results in [17]<br />
imply that, near a puncture, the solution is asymptotic to a periodic orbit (see also [3]<br />
for a complete proof). Here we use the fact that γ is exact near the punctures. The<br />
aim now is to show that, in the absence of punctures, the map a is constant while the<br />
image of u lies on a periodic Reeb orbit. Assume that Ɣ =∅.Since<br />
we find, after applying d, that<br />
u ∗ λ =−da ◦ j − γ ◦ j,<br />
ja =−d(da ◦ j) = u ∗ dλ.<br />
In view of the equation π ◦ Tu◦ j = J ◦ π ◦ Tu, we see that u ∗ dλ is a nonnegative<br />
integrand. Applying Stokes’s theorem, we obtain <br />
S u∗ dλ = 0, implying that<br />
π ◦ Tu≡ 0.
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 33<br />
Hence a is a harmonic function on S and therefore constant. So far, we also know that<br />
the image of u lies on a Reeb trajectory, and it remains to show that this trajectory is<br />
actually periodic.<br />
Let τ : ˜S → S be the universal covering map. The complex structure j lifts to a<br />
complex structure ˜j on ˜S. Now pick smooth functions f, g on ˜S such that<br />
dg = τ ∗ γ =: ˜γ, −df = τ ∗ (γ ◦ j) = ˜γ ◦ ˜j.<br />
Then the map u ◦ τ : ˜S → M satisfies<br />
(u ◦ τ) ∗ λ = df.<br />
Theimageofu◦ τ lies on a trajectory x of the Reeb vector field in view of<br />
<br />
D(u ◦ τ)(z)ζ = Df (z)ζ · Xλ (u ◦ τ)(z) ,<br />
hence (u ◦ τ)(z) = x(h(z)) for some smooth function h on ˜S, and it follows that, after<br />
maybe adding a constant to f ,wehave<br />
(u ◦ τ)(z) = x f (z) .<br />
The function f does not descend to S. If it did, it would have to be constant since it<br />
is harmonic. On the other hand, this would imply that u is constant in contradiction to<br />
our assumption that it is not. Therefore, there is a point q ∈ S and two lifts z0,z1 ∈ ˜S<br />
such that f (z0) >f(z1).Letℓ : S 1 → S be a loop which lifts to a path α :[0, 1] → ˜S<br />
with α(0) = z0 and α(1) = z1. Considering the map<br />
v := u ◦ ℓ : S 1 −→ M,<br />
we see that v(t) = (u ◦ τ ◦ α)(t) = x f (α(t)) and x(f (z0)) = x(f (z1)), that is, the<br />
trajectory x is a periodic orbit. Hence the image of u is a periodic orbit for the Reeb<br />
vector field. <br />
The following is the main result of this paper.<br />
<strong>THEOREM</strong> 1.6<br />
Let M be a closed 3-dimensional manifold, and let λ ′ be a contact form on M. Then<br />
the following holds for a suitable contact form λ = fλ ′ ,wheref is a positive function<br />
on M. There exists a smooth family (S,jτ ,Ɣτ , ũτ = (aτ ,uτ ),γτ )τ∈S 1 of solutions to<br />
(1.1) for a suitable compatible complex structure J :kerλ → ker λ such that<br />
• all maps uτ have the same asymptotic limit K at the punctures, where K is a<br />
finite union of periodic trajectories of the Reeb vector field Xλ;
34 CASIM ABBAS<br />
• uτ ( ˙S) ∩ uτ ′( ˙S) =∅ if τ = τ ′ ;<br />
• M\K = <br />
τ∈S 1 uτ ( ˙S);<br />
• the projection P onto S 1 defined by p ∈ uτ ( ˙S) ↦→ τ is a fibration;<br />
• the open book decomposition given by (P,K) supports the contact structure<br />
ker λ, and λ is a Giroux form.<br />
Here is a very brief outline of the argument. The reader is invited to skip forward to<br />
Section 4 to see in more detail how all the partial results of this paper are tied together<br />
to prove the main result. In Section 2, we find a Giroux contact form which has a<br />
certain normal form near the binding. Following an argument by Wendl ([39], [38]),<br />
we will then almost be able to turn the Giroux leaves into solutions of (1.1) without<br />
harmonic form except for the fact that we have to accept a confoliation form instead<br />
of a contact form. Pick one of these Giroux leaves as a starting point. The next step is<br />
to prove a result which permits us to perturb the Giroux leaf into a genuine solution<br />
of (1.1) while simultaneously perturbing the confoliation form slightly into a contact<br />
form. This is where the harmonic form in (1.1) is required. We actually obtain a local<br />
family of nearby solutions, not just one. In Section 3, we prove a compactness result<br />
which extends the local family of solutions into a global one. The remarkable fact is<br />
that there is a compactness result in the context of this paper, although there is none<br />
in general for the perturbed holomorphic curve equation. The special circumstances<br />
in this paper imply a crucial a priori bound which implies that a sequence of solutions<br />
has a pointwise convergent subsequence with a measurable limit. The objective is then<br />
to show that the regularity of this limit is much better, that it is actually smooth.<br />
We consider two solutions (S,j,Ɣ,ũ, γ ) and (S ′ ,j ′ ,Ɣ ′ , ũ ′ ,γ ′ ) equivalent if there<br />
exists a biholomorphic map φ :(S,j) → (S ′ ,j ′ ) mapping Ɣ to Ɣ ′ (preserving the<br />
enumeration) so that ũ ′ ◦ φ = ũ. We will often identify a solution (S,j,Ɣ,ũ, γ ) of<br />
(1.1) with its equivalence class [S,j,Ɣ,ũ, γ ].WenotethatwehaveanaturalR-action<br />
on the solution set by associating to c ∈ R and [S,j,Ɣ,ũ, γ ] the new solution<br />
c + [S,j,Ɣ,ũ, γ ] = [S,j,Ɣ,(a + c, u),γ], ũ = (a,u).<br />
A crucial concept for our discussion is the notion of a finite energy foliation F .<br />
Definition 1.7 (Finite energy foliation)<br />
A foliation F of R × M is called a finite energy foliation if every leaf F is the image<br />
of an embedded solution [S,j,Ɣ,ũ, γ ] of the equations (1.1), that is,<br />
F = ũ( ˙S),
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 35<br />
so that u( ˙S) ⊂ M is transverse to the Reeb vector field, and for every leaf F ∈ F we<br />
also have c + F ∈ F for every c ∈ R; that is, the foliation is R-invariant.<br />
We recall the concept of a global surface of section. Let M be a closed 3-manifold,<br />
and let X be a nowhere-vanishing smooth vector field.<br />
Definition 1.8 (Surface of section)<br />
(a) A local surface of section for (M,X) consists of an embedded compact surface<br />
⊂ M with boundary, so that ∂ consists of a finite union of periodic orbits<br />
(called the binding orbits). In addition, the interior ˙ = \ ∂ is transverse<br />
to the flow.<br />
(b) A local surface of section is called a global surface of section if, in addition,<br />
every orbit other than a binding orbit hits ˙ in forward and backward time.<br />
Furthermore, the globally defined return map : ˙ → ˙ has a bounded<br />
return time; that is, there exists a constant c>0 so that every x ∈ ˙ hits ˙<br />
again in forward time not exceeding c.<br />
Using Proposition 2.5 below, the existence part of Giroux’s theorem can be rephrased<br />
as follows.<br />
<strong>THEOREM</strong> 1.9<br />
Let M be a closed orientable 3-manifold, and let ˜λ be a contact form on M. Then<br />
there exists a smooth function f : M → (0, ∞) so that the contact form λ = f ˜λ has<br />
a Reeb vector field admitting a global surface of section.<br />
Existence results for finite energy foliations with a given contact form λ are hard to<br />
come by since they usually have striking consequences. In [20] for example, Hofer,<br />
Wysocki, and Zehnder show that every compact strictly convex energy hypersurface<br />
S in R 4 carries either two or infinitely many closed characteristics. The proof relies on<br />
constructing a special finite energy foliation. In special cases they were established by<br />
Hofer, Wysocki, and Zehnder [22] and by Wendl ([38], [40]). Proofs usually require<br />
a starting point, that is, a finite energy foliation for a slightly different situation as the<br />
given one. Then some kind of continuation argument is employed where all kinds of<br />
things can and do happen to the original foliation. In [22], the authors start with an<br />
explicit finite energy foliation for the round 3-dimensional sphere S 3 ⊂ R 4 which is<br />
then deformed. Wendl’s papers also use a rather special manifold as a starting point.<br />
The main result of this paper, Theorem 1.6, provides a starting finite energy foliation<br />
for any closed 3-dimensional contact manifold (M,ker λ) since it is obtained from<br />
deforming the leaves of Giroux’s open book decomposition. The pages are usually<br />
not punctured spheres, and generically there are no pseudoholomorphic curves on
36 CASIM ABBAS<br />
punctured surfaces with genus which are transverse to the Reeb vector field. This<br />
makes the introduction of the harmonic form in (1.1) a necessity. The price to be paid<br />
is that compactness issues are more complicated.<br />
Wendl [39] published a proof of Theorem 1.6 for the special case where ker λ is<br />
a planar contact structure, that is, where the surfaces ˙S are punctured spheres. This<br />
result was outlined in [4]. Regardless of whether the contact structure is planar or not,<br />
there are two main steps in the proof: existence of a solution and compactness of a<br />
family of solutions. While the author established the compactness part for Theorem<br />
1.6 long before [4] appeared, we will use the same argument described by Wendl in<br />
[39] for the existence part since it simplifies the proof considerably.<br />
The main theorem of this article was the first step in the proof of the Weinstein<br />
conjecture for the planar case in [4]. Recall that the Weinstein conjecture [38, p. 358]<br />
states the following: Every Reeb vector field X on a closed contact manifold M admits<br />
a periodic orbit.<br />
In fact, Weinstein added the additional hypothesis that the first cohomology group<br />
H 1 (M,R) with real coefficients vanishes, but there seems to be no indication that this<br />
additional hypothesis is needed.<br />
Moreover, Theorem 1.6 is also the starting point for the construction of global<br />
surfaces of section in the forthcoming paper [7]. Another application will be an<br />
alternative proof of the Weinstein conjecture in dimension 3 (see [7]) as outlined in<br />
[4]. This complements Taubes’s recent proof of the Weinstein conjecture in dimension<br />
3 using a perturbed version of the Seiberg-Witten equations (see [33], [34]). The<br />
main issue with the homological perturbed holomorphic curve equation (1.1) isthat<br />
there is no natural compactification of the space of solutions unless the harmonic<br />
forms are uniformly bounded. In the forthcoming papers [6] and[7] the lack of compactness<br />
is investigated, and bounds for the harmonic forms are derived in particular<br />
cases.<br />
2. Existence and local foliations<br />
2.1. Local model near the binding orbits<br />
We use the same approach as in [39] and[38] to prove existence of a solution to<br />
(1.1). Given a closed contact 3-manifold (M,ξ), Giroux’s theorem implies that there<br />
is an open book decomposition as in Theorem 1.2 supporting ξ. On the other hand,<br />
any other contact structure ξ ′ supported by the same open book is diffeomorphic to<br />
ξ. Starting with an open book decomposition for M, we construct a contact structure<br />
supported by Giroux contact form λ which has a certain normal form near the<br />
binding.
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 37<br />
Definition 2.1<br />
Let θ ∈ S 1 = R/2πZ denote polar coordinates on the unit disk D ⊂ R 2 by (r, φ),<br />
and let γ1,γ2 :[0, +∞) → R be smooth functions. A 1-form<br />
λ = γ1(r) dθ + γ2(r) dφ<br />
is called a local model near the binding if the following conditions are satisfied:<br />
(1) the functions γ1,γ2,andγ2(r)/r 2 are smooth if considered as functions on the<br />
disk D (in particular, γ ′ 1 (0) = γ ′ 2 (0) = γ2(0) = 0);<br />
(2) µ(r) := γ1(r)γ ′ 2 (r) − γ ′ 1 (r)γ2(r) > 0 if r>0;<br />
(3) γ1(0) > 0 and γ ′ 1 (r) < 0 if r>0;<br />
(4) limr→0(µ(r)/r) = γ1(0)γ ′′<br />
2 (0) > 0;<br />
(5) κ := (γ ′′ ′′<br />
1 (0)/γ 2 (0)) /∈ Z and κ ≤−1/2;<br />
(6) A(r) = (1/µ 2 (r))(γ ′′<br />
2 (r)γ ′ ′′<br />
1 (r) − γ 1 (r)γ ′ 2<br />
We explain some of the conditions above. First, since<br />
(r)) is of order r for small r>0.<br />
λ ∧ dλ = µ(r)dθ ∧ dr ∧ dφ = µ(r)<br />
dθ ∧ dx ∧ dy,<br />
r<br />
the form λ is a contact form on S 1 × D. The Reeb vector field is given by<br />
X(θ,r,φ) = γ ′ 2 (r)<br />
µ(r)<br />
∂<br />
∂θ − γ ′ 1 (r)<br />
µ(r)<br />
∂<br />
∂φ<br />
The trajectories of X all lie on tori Tr = S 1 × ∂Dr:<br />
We compute<br />
and<br />
=: α(r) ∂<br />
∂θ<br />
+ β(r) ∂<br />
∂φ .<br />
θ(t) = θ0 + α(r) t, φ(t) = φ0 + β(r) t. (2.1)<br />
γ<br />
lim α(r) = lim<br />
r→0 r→0<br />
′′<br />
2 (r)<br />
µ ′ (r)<br />
γ<br />
lim β(r) =−lim<br />
r→0 r→0<br />
′′<br />
1 (r)<br />
µ ′ (r)<br />
= γ ′′<br />
2 (0)<br />
Recalling that ∂ ∂ ∂<br />
= x − y , we obtain for r = 0<br />
∂φ ∂y ∂x<br />
X = 1 ∂<br />
γ1(0) ∂θ ,<br />
γ1(0)γ ′′<br />
1<br />
=<br />
(0) γ1(0)<br />
2<br />
=− γ ′′<br />
1 (0)<br />
γ1(0)γ ′′<br />
2 (0).
38 CASIM ABBAS<br />
that is, the central orbit has minimal period 2πγ1(0). If the ratio α(r)/β(r) is irrational<br />
then the torus Tr carries no periodic trajectories. Otherwise, Tr is foliated with periodic<br />
trajectories of minimal period<br />
τ = 2πm<br />
α<br />
= 2πn<br />
β ,<br />
where α/β = m/n or β/α = n/m for suitable integers m, n (choose whatever makes<br />
sense if either α or β is zero). We calculate<br />
dα<br />
lim<br />
r→0 dr<br />
γ<br />
= lim<br />
r→0<br />
′′<br />
2 (r)µ(r) − γ ′ 2 (r)µ′ (r)<br />
µ 2 (r)<br />
γ<br />
= lim<br />
r→0<br />
′′′<br />
2 (r)<br />
2µ ′ − lim<br />
(r) r→0<br />
= γ ′′′<br />
= 0<br />
since µ ′′ (0) = γ1(0)γ ′′′<br />
coordinates on the disk, we get<br />
2 (0)<br />
2µ ′ (0)<br />
µ ′′ (r)<br />
2µ ′ γ<br />
(r)<br />
′ 2 (r)<br />
r<br />
′′′ ′′<br />
γ1(0)γ 2 (0)γ 2<br />
− (0)<br />
2(µ ′ (0)) 2<br />
2 (0) and µ′ (0) = γ1(0)γ ′′<br />
2<br />
X(θ,x,y) = α(x,y) ∂<br />
∂θ<br />
− β(x,y) y ∂<br />
∂x<br />
r<br />
µ(r)<br />
(0) > 0. Converting to Cartesian<br />
+ β(x,y) x ∂<br />
∂y ,<br />
and linearizing the Reeb vector field along the center orbit yields<br />
⎛<br />
0 0 0<br />
⎞<br />
DX(θ,0, 0) = ⎝ 0 0 −β(0) ⎠ .<br />
0 β(0) 0<br />
The linearization of the Reeb flow is given by<br />
⎛<br />
1 0 0<br />
⎞<br />
Dφt(θ,0, 0) = ⎝ 0 cos β(0)t − sin β(0)t ⎠ (2.2)<br />
0 sin β(0)t cos β(0)t<br />
with<br />
The spectrum of (t) is given by<br />
(t) = e β(0)tJ , J =<br />
σ (t) ={e ±iβ(0)t }.<br />
<br />
0 −1<br />
.<br />
1 0
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 39<br />
The binding orbit has period 2πγ1(0), since<br />
and it is nondegenerate and elliptic.<br />
′′ γ 1<br />
γ1(0)β(0) =− (0)<br />
γ ′′<br />
2<br />
(0) /∈ Z,<br />
Example 2.2<br />
For the contact form Tdθ+(1/k)(xdy−ydx) = Tdθ+(r 2 /k)dφ the central orbit<br />
S 1 ×{0} is degenerate, but<br />
is a local model near the binding if<br />
In this case,<br />
and we note that<br />
If<br />
λ = (1 − r 2 )(T dθ+ r2<br />
k dφ)<br />
k, T > 0, kT /∈ Z, and kT ≥ 1<br />
2 .<br />
µ(r) = 2rT<br />
k (1 − r2 ) 2 > 0 and<br />
A(r) = 1<br />
µ 2 (r)<br />
γ ′′<br />
2<br />
(r)γ ′<br />
1<br />
α(r)<br />
β(r) =−γ ′ 2 (r)<br />
γ ′ 1<br />
′′ ′<br />
(r) − γ 1 (r)γ 2 (r) =<br />
1 − 2r2<br />
=<br />
(r) kT<br />
γ ′′<br />
1 (0)<br />
γ ′′<br />
2<br />
m<br />
=<br />
n<br />
(0) =−kT,<br />
4kr<br />
T (1 − r2 .<br />
) 4<br />
for integers n, m, then the invariant torus Tr is foliated with periodic orbits. The<br />
case m = 0 is only possible if r = 1/ √ 2.Ifr is sufficiently small, then |m| ≥2.<br />
Indeed, we would otherwise be able to find sequences rl ↘ 0 and {nl} ⊂Z such that<br />
kT/(1 − 2r 2 l ) = nl, which is impossible. The binding orbit has period 2πT while<br />
the periodic orbits close to the binding orbit have much larger periods equal to<br />
τ = 2πTm (1 − r2 ) 2<br />
.<br />
1 − 2r2 Example 2.3<br />
Consider the contact form λ = T (1 − r 2 )dθ + (r 2 /k)dφ on S 1 × D. Itisalsoa<br />
local model near the binding if k, T > 0, kT ≥ 1/2, andkT is not an integer. We
40 CASIM ABBAS<br />
even have A(r) ≡ 0. In contrast to Example 2.2, if kT /∈ Q, then the invariant tori Tr<br />
carry no periodic orbits. If kT = n/m ∈ Q but not in Z, all invariant tori are foliated<br />
with periodic orbits of period 2πmT with |m| ≥2 while the binding orbit has period<br />
2πT. The function A(r) is identically zero. This is the contact form on the irrational<br />
ellipsoid in R 4 .<br />
The following proposition is essentially due to Wendl. The construction in the proof<br />
was used by Thurston and Winkelnkemper [35] to show existence of contact forms on<br />
closed 3-manifolds.<br />
PROPOSITION 2.4 ([39, Proposition 1])<br />
Let M be a 3-dimensional manifold given by an open book decomposition<br />
M = W (h) ∪Id (∂W × D 2 )<br />
as described in Theorem 1.2. We denote the pages by<br />
Fα := (W ×{α}) ∪Id (∂W × Iα), 0 ≤ α
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 41<br />
([0,ε] × ( ˙<br />
nS1 ), {0} ×( ˙<br />
nS1 )), where we take an n-fold disjoint union of circles<br />
S1 ≈ R/2πZ according to the number n of components of ∂W.<br />
We<br />
<br />
claim that there is an area form on W that satisfies<br />
• = 2πn,<br />
W<br />
• |C = dt ∧ dθ.<br />
Indeed, start with any area form ′ so that <br />
W ′ = 2πn.Thenwehave ′ |C =<br />
f ′ (t,θ)dt∧dθ with a positive smooth function f ′ (after switching signs if necessary).<br />
Now pick a new smooth positive function f which is equal to some constant c if<br />
t ≤ (1/3)ε and which agrees with f ′ if t ≥ (2/3)ε so that the resulting area form<br />
still satisfies <br />
= 2πn. Do one component of ∂W at a time. Rescaling the<br />
W<br />
t-coordinate, we may assume that c = 1.<br />
Let α1 be any 1-form on W which equals (1 + t) dθ near ∂W. Then by Stokes’s<br />
theorem we obtain<br />
<br />
<br />
<br />
( − dα1) = 2πn− α1 = 2πn+ dθ = 0.<br />
W<br />
∂W<br />
∂W<br />
The 2-form −dα1 on W is closed and vanishes near ∂W. Then there exists a 1-form<br />
β on W with<br />
dβ = − dα1<br />
and β ≡ 0 near ∂W.Nowdefineα2 := α1 + β.Thenα2 satisfies the following:<br />
• dα2 is an area form on W inducing the same orientation as , (2.3)<br />
• α2 = (1 + t) dθ near ∂W. (2.4)<br />
The set of 1-forms on W satisfying (2.3) and (2.4) is therefore nonempty and also<br />
convex. We define the following 1-form on W × [0, 2π], where α is any 1-form on<br />
W satisfying (2.3) and (2.4):<br />
˜α(x,τ):= τα(x) + (2π − τ)(h ∗ α)(x).<br />
This 1-form descends to the quotient W (h), and the restriction to each fiber of the<br />
fiber bundle W (h) π → S 1 satisfies condition (2.3). Moreover, since h ≡ Id near ∂W,<br />
we have ˜α(x,τ) = 2π(1 + t) dθ for all (x,τ) = ((t,θ),τ) near ∂W(h) = ∂W × S 1 .<br />
Let dτ be a volume form on S 1 . We claim that<br />
λ1 := −δ ˜α + π ∗ dτ
42 CASIM ABBAS<br />
are contact forms on W (h) whenever δ>0 is sufficiently small. Pick (x,τ) ∈ W(h),<br />
and let {u, v, w} be a basis of T(x,τ)W (h) with π∗u = π∗v = 0. Then<br />
(λ1 ∧ dλ1)(x,τ)(u, v, w)<br />
= δ 2 (˜α ∧ d ˜α)(x,τ)(u, v, w) − δ [dτ(π∗w) d ˜α(x,τ)(u, v)]<br />
= 0<br />
for sufficiently small δ>0,anddλ1 is a volume form on W . Now we have to continue<br />
the contact forms λ1 beyond ∂W(h) ≈ ∂W × S1 onto ∂W × D2 . At this point it is<br />
convenient to change coordinates. We identify C × S1 with ∂W × (D2 1+ε \D2 1 ), where<br />
D2 ρ is the 2-disk of radius ρ. Using polar coordinates (r, φ) on D2 1+ε with 0 ≤ φ ≤ 2π<br />
and 0 0 and γ ′<br />
1 (r) < 0 if r>0,<br />
hence the curves γ = γδ have to turn counterclockwise in the first quadrant starting<br />
at the point (γ1(0), 0) and later connecting with (−δ(1 − ε0), 1). In the case where<br />
δ>0, the Reeb vector fields are given by<br />
Xδ(θ,r,φ) = γ ′ 2 (r)<br />
µ(r)<br />
∂<br />
∂θ − γ ′ 1 (r)<br />
µ(r)<br />
∂<br />
∂φ ,
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 43<br />
and, in particular,<br />
Xδ(θ,r,φ) = ∂<br />
∂φ for r ≥ 1 − ε0, (2.5)<br />
which implies that the Reeb vector fields Xδ converge as δ ↘ 0. In addition to λδ being<br />
contact forms for δ>0, we also want the given open book decomposition to support<br />
ker λδ, hence Xδ needs to be transverse to the pages of the open book decomposition<br />
which is equivalent to γ ′ 1 (r) = 0. A curve γ (r) fulfilling these conditions can clearly<br />
be constructed. <br />
The following result shows that we can always assume that a Giroux contact form is<br />
equal to any of the forms provided by Proposition 2.4.<br />
PROPOSITION 2.5<br />
Let M be a closed 3-dimensional manifold with contact structure ξ. Then, for every<br />
δ>0, there is a diffeomorphism ϕδ : M → M such that ker λδ = ϕ∗ξ where λδ is<br />
given by Proposition 2.4.<br />
Proof<br />
Existence of an open book decomposition supporting ξ follows from the existence<br />
part of Giroux’s theorem. On the other hand, Proposition 2.4 yields contact forms λδ<br />
such that ker λδ is also supported by the same open book decomposition as ξ for any<br />
δ>0. By the uniqueness part of Giroux’s theorem, ξ and ker λδ are diffeomorphic. <br />
It follows from our previous construction of the forms λδ that λ0 satisfies λ0 ∧dλ0 > 0<br />
on ∂W ×D1−ε0 and that λ0 = dφ otherwise. For δ → 0, the Reeb vector fields Xδ will<br />
converge to some vector field X0, which is the Reeb vector field of λ0 if r
44 CASIM ABBAS<br />
the standard complex structure on the cylinder [0, +∞) × S 1 after introducing polar<br />
coordinates near the punctures. Moreover, all the punctures are positive ∗ the family<br />
(ũα)0≤α≤2π is a finite energy foliation, and the curves ũα are ˜Jδ-holomorphic near the<br />
punctures.<br />
Proof<br />
We parameterize the leaves of the open book decomposition uα : ˙S → M, 0 ≤ α<<br />
2π, and we assume that they look as follows near the binding:<br />
uα :[0, +∞) × S 1 −→ S 1 × D1,<br />
uα(s, t) = t,r(s)e iα ,<br />
(2.6)<br />
where r are smooth functions with lims→∞ r(s) = 0 to be determined shortly. We<br />
use the notation (r, φ) for polar coordinates on the disk D = D1. We identify some<br />
neighborhood U of the punctures of ˙S with a finite disjoint union of half-cylinders<br />
[0, +∞) × S1 . Recall that the binding orbit is given by<br />
<br />
t<br />
<br />
x(t) = , 0, 0 , 0 ≤ t ≤ 2πγ1(0)<br />
γ1(0)<br />
and that it has minimal period T = 2πγ1(0). We define smooth functions aα : ˙S → R<br />
by<br />
s<br />
aα(z) := 0 γ1<br />
<br />
′ ′ r(s ) ds if z = (s, t) ∈ [0, +∞) × S1 ⊂ U,<br />
0 if z/∈ U<br />
so that<br />
u ∗<br />
α λ0 ◦ j = daα,<br />
where j is a complex structure on ˙S which equals the standard structure i on [0, +∞)×<br />
S 1 (i.e., near the punctures). We want to turn the maps ũα = (aα,uα) : ˙S → R × M<br />
into ˜J0-holomorphic curves for a suitable almost-complex structure ˜J0 on R × M.<br />
Recall that the contact structure is given by<br />
<br />
∂ ∂<br />
ker λδ = Span{η1,η2} = Span , −γ2(r)<br />
∂r ∂θ<br />
We define complex structures Jδ :kerλδ → ker λδ by<br />
<br />
Jδ(θ,r,φ) − γ2(r) ∂<br />
∂θ<br />
∂<br />
<br />
+ γ1(r) := −<br />
∂φ<br />
1<br />
h(r)<br />
∗ A puncture pj is called positive for the curve (aα,uα) if limz→pj aα(z) =+∞.<br />
∂<br />
<br />
+ γ1(r) .<br />
∂φ<br />
∂<br />
∂r<br />
(2.7)
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 45<br />
and<br />
Jδ(θ,r,φ) ∂<br />
∂r<br />
<br />
:= h(r) − γ2(r) ∂<br />
∂θ<br />
∂<br />
<br />
+ γ1(r) ,<br />
∂φ<br />
where h :(0, 1] → R\{0} are suitable smooth functions. Also recall that γ1,γ2<br />
depend on δ away from the binding orbit. We want Jδ to be compatible with dλδ,that<br />
is, we want<br />
dλδ(η1,Jη1) = h(r) µ(r) > 0 and dλδ(η2,Jη2) = µ(r)<br />
> 0<br />
h(r)<br />
so that h(r) > 0. We also demand that Jδ extends smoothly over the binding {r = 0}.<br />
Expressing the vectors η1 and η2 in Cartesian coordinates, we have<br />
and<br />
η1 = 1<br />
r<br />
<br />
x ∂ ∂<br />
+ y<br />
∂x ∂y<br />
η2 =−γ2(r) ∂<br />
∂θ + γ1(r) x ∂<br />
∂y − γ1(r) y ∂<br />
∂x .<br />
We introduce the following generators of the contact structure:<br />
and<br />
We compute from this<br />
Now<br />
ε1 := γ1(r) ∂<br />
∂y<br />
= yγ1(r)<br />
r<br />
ε2 := γ1(r) ∂<br />
∂x<br />
= xγ1(r)<br />
r<br />
− xγ2(r)<br />
r 2<br />
η1 + x<br />
η2<br />
r2 <br />
∂<br />
∂θ<br />
yγ2(r)<br />
+<br />
r2 ∂<br />
∂θ<br />
η1 − y<br />
η2.<br />
r2 η1 = 1<br />
rγ1(r) (yε1 + xε2), η2 = xε1 − yε2.<br />
Jδε1 = yγ1(r)h(r)<br />
η2 −<br />
r<br />
x<br />
r2h(r) η1<br />
<br />
1<br />
=<br />
r xyγ1(r)h(r)<br />
xy<br />
−<br />
r3 <br />
ε1<br />
h(r)γ1(r)<br />
<br />
1<br />
−<br />
r y2 x<br />
γ1(r)h(r) +<br />
2<br />
r3 <br />
γ1(r)h(r)<br />
ε2
46 CASIM ABBAS<br />
and<br />
Jδε2 = xγ1(r)h(r)<br />
η2 +<br />
r<br />
y<br />
r2h(r) η1<br />
<br />
= − 1<br />
r xyγ1(r)h(r)<br />
xy<br />
+<br />
r3 <br />
ε2<br />
h(r)γ1(r)<br />
<br />
1<br />
+<br />
r x2 y<br />
γ1(r)h(r) +<br />
2<br />
r3 <br />
ε1.<br />
γ1(r)h(r)<br />
Inserting x = r cos φ, y = r sin φ, and demanding that the limit is φ-independent as<br />
r → 0, we arrive at the condition that rh(r) γ1(r) ≡±1 for small r. Recalling that<br />
we need h>0, we obtain<br />
h(r) = 1<br />
rγ1(r)<br />
for small r.<br />
As usual, we continue Jδ to an almost-complex structure ˜Jδ on R × M by setting<br />
˜Jδ(θ,r,φ) ∂<br />
∂τ<br />
:= Xδ(θ,r,φ),<br />
where τ denotes the coordinate in the R-direction. We emphasize that ˜Jδ also makes<br />
sense for δ = 0. We now arrange r(s) in (2.6) such that the Giroux leaves ũα = (aα,uα)<br />
become ˜J0-holomorphic curves. ∗ We compute for r ≤ 1 − ε0<br />
∂sũα + ˜J0(uα)∂tũα = γ1(r) ∂ ∂<br />
+ r′<br />
∂τ ∂r + <br />
˜J0(uα)<br />
∂<br />
<br />
∂θ<br />
= γ1(r) ∂ ∂<br />
+ r′<br />
∂τ ∂r + ˜J0(uα) γ1(r) Xδ(uα) <br />
<br />
+ ˜J0(uα)<br />
∂<br />
<br />
− γ1(r)Xδ(uα)<br />
∂θ<br />
′ ∂<br />
= r<br />
∂r + ′<br />
˜J0(uα)<br />
γ 1 (r)<br />
<br />
γ1(r)<br />
µ(r)<br />
∂ ∂<br />
<br />
− γ2(r)<br />
∂φ ∂θ<br />
<br />
= r ′ − γ ′ 1 (r)<br />
<br />
∂<br />
µ(r)h(r) ∂r ,<br />
hence the Giroux leaves satisfy the equation if we choose r to be a solution of the<br />
ordinary differential equation<br />
r ′ (s) =<br />
γ ′ 1 (r(s))<br />
µ(r(s)) h(r(s)) .<br />
∗ The calculation shows that we can make them Jδ-holomorphic for all δ ≥ 0 near the binding.
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 47<br />
Note that r ′ (s) < 0. We also choose h(r) ≡ 1 for r ≥ 1 − ε0. We continue the<br />
almost-complex structures Jδ :kerλδ → ker λδ (which were only defined near the<br />
binding) smoothly to all of M. Away from the binding we have Xδ = ∂/∂φ, andwe<br />
extend Jδ as before to T (R × M). Away from the binding, if δ = 0, wehavethat<br />
ker λ0 coincides with the tangent spaces of the pages of the open book decomposition.<br />
Because aα is constant away from the binding, the solutions ũα which we constructed<br />
near the binding fit together smoothly with the pages of the open book decomposition<br />
and solve the holomorphic curve equation for the almost-complex structure J0. <br />
Remark 2.7<br />
Near the binding orbit the function r(s) satisfies a differential equation of the form<br />
and<br />
r ′ (s) = r(s) r(s) := γ ′ 1 (r(s))γ1(r(s))<br />
µ(r(s))<br />
lim<br />
r→0<br />
(r) = γ ′′<br />
1 (0)<br />
γ ′′<br />
2<br />
(0) =: κ.<br />
r(s),<br />
Writing r(s) = c(s)e κs , the function c(s) satisfies c ′ (s) = ((r(s)) − κ)c(s), and<br />
hence it is a decreasing function which converges to a constant as s →+∞.<br />
We return to Examples 2.2 and 2.3, and we compute r(s) for large s. The differential<br />
equation in the case of Example 2.3 for large s is<br />
so that<br />
r ′ (s) = γ ′ 1 (r(s))γ1(r(s))<br />
r(s) =−kT<br />
µ(r(s))<br />
1 − r 2 (s) r(s),<br />
r(s) =<br />
1<br />
√ 1 + ce 2kT s ,<br />
where c is a constant. In Example 2.2, the differential equation reads<br />
and solutions satisfy<br />
r ′ (s) = γ ′ 1 (r(s))γ1(r(s))<br />
r(s) =−<br />
µ(r(s))<br />
kT<br />
1 − r2 (s) r(s),<br />
r(s) = ce −kT s e (1/2)r2 (s) .<br />
2.2. Functional analytic setup and the implicit function theorem<br />
In the following theorem we prove the existence of a smooth family of solutions near a<br />
given solution. In Proposition 2.6, we constructed a finite energy foliation for the data
48 CASIM ABBAS<br />
(λ0,J0) with vanishing harmonic form. The form λ0, however, is only a confoliation<br />
form. We produce solutions for the perturbed data (λδ,Jδ), and harmonic forms appear<br />
if the surface S is not a sphere. The key result is an application of the implicit function<br />
theorem in a suitable setting.<br />
<strong>THEOREM</strong> 2.8<br />
Assume one of the following.<br />
(1) Let (a0,u0) : ˙S → R × M be one of the ˜J0-holomorphic curves described in<br />
Proposition 2.6 with complex structure j0 on S (we refer to such u0 as a Giroux<br />
leaf) and confoliation form λ0.<br />
(2) Let ( ˙S,j0,a0,u0,γ0) be a solution of the differential equation (1.1) for some<br />
dλ0-compatible complex structure J0 :kerλ0 → ker λ0 which, near the binding<br />
orbit, agrees with (2.7), and where λ0 is a contact form which is a local<br />
model near the binding. Assume that u0 is an embedding and that it is of the<br />
form u0 = φg(v0), whereg : S → R is a smooth function, φ is the flow of the<br />
Reeb vector field, and where v0 : ˙S → M is a Giroux leaf as in Proposition<br />
2.6.<br />
(3) Let Jδ be a smooth family of dλδ-compatible complex structures also agreeing<br />
with (2.7) near the binding orbit, where (λδ)−ε
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 49<br />
Note that this is well defined because u is transverse to the Reeb vector field so that<br />
πTu(z) :TzS → ker λ(u(z)) is an isomorphism. By the second equation of (1.1), we<br />
then have to solve the equation df ◦ jf + u ∗ 0 λ ◦ jf = da + γ for a,f,γ on ˙S which<br />
is equivalent to the equation<br />
¯∂jf (a + if ) = u ∗<br />
0 λ ◦ jf − i(u ∗<br />
0 λ) − γ − i(γ ◦ jf ). (2.9)<br />
Recall that we are looking for a of the form a = a0 + b, where b is a suitable realvalued<br />
function defined on the whole surface S. We obtain the differential equation<br />
¯∂jf (b + if ) = u ∗<br />
0 λ ◦ jf − i(u ∗<br />
0 λ) − ¯∂jf a0 − γ − i(γ ◦ jf ), (2.10)<br />
and it follows from a straightforward calculation (see the appendix) that all expressions<br />
on the right-hand side of (2.10) are bounded near the punctures; in particular, they<br />
are contained in the spaces L p (T ∗ S ⊗ C) for any p. This is what the assumption<br />
κ ≤−(1/2) from Definition 2.1 is needed for. We work in the function space b +if ∈<br />
W 1,p (S,C), where p>2. For any complex structure j on S, the space L p (T ∗ S ⊗ C)<br />
of complex-valued 1-forms of class L p decomposes into complex linear and complex<br />
antilinear forms (with respect to j). We use the notation<br />
L p (T ∗ S ⊗ C) = L p (T ∗ S ⊗ C) 1,0<br />
j ⊕ Lp (T ∗ S ⊗ C) 0,1<br />
j .<br />
The operator b + if ↦→ ¯∂jf (b + if ) is then a section in the vector bundle<br />
L p (T ∗ S ⊗ C) 0,1 :=<br />
<br />
b+if ∈W 1,p (S,C)<br />
{b + if }×L p (T ∗ S ⊗ C) 0,1<br />
jf → W 1,p (S,C).<br />
This vector bundle is of course trivial, but here are some explicit local trivializations<br />
for f, g ∈ W 1,p (S,R) sufficiently close to each other:<br />
If we write<br />
we see that fg is invertible with<br />
fg : Lp (T ∗S ⊗ C) 0,1<br />
jf −→Lp (T ∗S ⊗ C) 0,1<br />
jg<br />
τ ↦−→ τ + i(τ ◦ jg).<br />
τ + i(τ ◦ jg) = τ ◦ (IdTS − jf ◦ jg),<br />
−1<br />
fg τ = τ ◦ (IdTS − jf ◦ jg) −1 .<br />
(2.11)<br />
It follows from the Hodge decomposition theorem that every cohomology class [σ ] ∈<br />
H 1 (S,R) has a unique harmonic representative ψj(σ ) ∈ H 1 j (S), where H 1 j (S) is
50 CASIM ABBAS<br />
defined as<br />
H 1<br />
j (S) := γ ∈ E 1 (S) dγ = 0, d(γ ◦ j) = 0 <br />
(2.12)<br />
and where E 1 (S) denotes the space of all (smooth) real-valued 1-forms on S, and<br />
we write E 0,1 (S) = E 0,1<br />
j (S) for the space of complex antilinear 1-forms on S with<br />
respect to j, that is, complex-valued 1-forms σ such that iσ + σj = 0. Note that<br />
our definition coincides with the set of closed and co-closed 1-forms on S. Moreover,<br />
by elliptic regularity, we may also consider Sobolev forms. We identify H 1 (S,R)<br />
with R 2g , and we consider the following parameter-dependent section in the bundle<br />
L p (T ∗ S ⊗ C) 0,1 → W 1,p (S,C)<br />
F : W 1,p (S,C) × R 2g −→ L p (T ∗ S ⊗ C) 0,1<br />
F (b + if, σ ):= ¯∂jf (b + if ) − u ∗<br />
0λ ◦ jf + i(u ∗<br />
0λ) + ¯∂jf a0 + ψjf (σ ) + i <br />
ψjf (σ ) ◦ jf<br />
(2.13)<br />
with jf as in (2.8). Recalling that z ↦→ jf (z) may not be differentiable, we interpret<br />
the equation d(γ ◦ jf ) = 0 in the sense of weak derivatives. The solution set of (2.10)<br />
is then the zero set of F . We consider the real parameter δ which we dropped from<br />
the notation, fixed at the moment. For g ≡ 0 and b + if small in the W 1,p -norm,<br />
we consider the composition ˆF (b + if, σ ) = fg(F (b + if, σ )). Its linearization<br />
in the point (b + if, σ ) = (0,σ0), where σ0 is defined by ψj0(σ0) = γ0 and where<br />
F (0,σ0) = 0,is<br />
where<br />
D ˆF (0,σ0) :W 1,p (S,C) × R 2g −→ L p (T ∗ S ⊗ C) 0,1<br />
j0<br />
D ˆF (0,σ0)(ζ,σ) = ¯∂j0ζ + ψj0(σ ) + i(ψj0(σ ) ◦ j0) + Lζ,<br />
L : W 1,p (S,C) → W 1,p (T ∗ S ⊗ C) 0,1<br />
j0 ↩→ L p (T ∗ S ⊗ C) 0,1<br />
j0<br />
is the compact linear map<br />
where<br />
Lζ =− 1<br />
2 u∗<br />
0 λ ◦ (Aζ + j0Aζj0) + i<br />
2 u∗<br />
0 λ ◦ (j0Aζ − Aζj0) + Bζ + iBζj0,<br />
Bζ = d<br />
<br />
<br />
<br />
dτ<br />
τ=0<br />
ψjτ k(σ0), ζ = h + ik,
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 51<br />
and where<br />
Aζ = d<br />
<br />
<br />
jτ<br />
dτ k = h (πλTu0)<br />
τ=0<br />
−1 [J (u0)DXλ(u0)<br />
− DXλ(u0)J (u0) + DJ(u0)Xλ(u0)](πλTu0).<br />
The linear term L therefore does not contribute to the Fredholm index of D ˆF (0,σ0).<br />
We note that the linear map ζ ↦→ Lζ only depends on the imaginary part of ζ .We<br />
claim that the operator<br />
W 1,p (S,C) × R 2g −→ L p (T ∗ S ⊗ C) 0,1<br />
j0<br />
(ζ,σ) ↦−→ ¯∂j0ζ + ψj0(σ ) + i(ψj0(σ ) ◦ j0)<br />
is a surjective Fredholm operator of index 2. Then we would have ind(D ˆF (0,σ0)) = 2<br />
as well. Here is the argument: The Riemann-Roch theorem asserts that the kernel and<br />
the cokernel of the Cauchy-Riemann operator ¯∂j (acting on smooth complex-valued<br />
functions on S) are both finite-dimensional and that<br />
<br />
dimR ker ¯∂j<br />
0,1<br />
− dimR E (S)/Im ¯∂j = 2 − 2g,<br />
where g is the genus of the surface S. The only holomorphic functions on S are the<br />
constant functions, hence E0,1 (S)/Im ¯∂j has dimension 2g.<br />
On the other hand, the vector space H 1 j (S) of all (real-valued) harmonic 1-forms<br />
on S also has dimension 2g (see [15]). We now consider the linear map<br />
: H 1<br />
j (S) −→ E0,1 (S)/Im ¯∂j<br />
(γ ):= [γ + i(γ ◦ j)],<br />
where [ . ] denotes the equivalence classes of (0, 1)-forms. Assume that (γ ) = [0],<br />
that is, that there is a complex-valued smooth function f = u+iv on S such that ¯∂jf =<br />
γ +i(γ ◦j).Sinceγ is a harmonic 1-form, we conclude that d(dv◦j) = d(du◦j) = 0<br />
(i.e., both u and v are harmonic). Since there are only constant harmonic functions<br />
on S we obtain γ = 0 (i.e., is injective and also bijective). Hence, (0, 1)-forms<br />
γ + i(γ ◦ j) with γ ∈ H 1 j (S) make up the cokernel of ¯∂j : C ∞ (S,C) → E 0,1 (S).<br />
This proves the claim that the operator D ˆF (0,σ0) is Fredholm of index 2. We<br />
now show that the operator D ˆF (0,σ0) is surjective. Using the decomposition<br />
L p (T ∗ S ⊗ C) 0,1<br />
j0 = R(¯∂j0) ⊕ H 1<br />
j0 (S)
52 CASIM ABBAS<br />
and denoting the corresponding projections by π1,π2, we see that it suffices to prove<br />
the surjectivity of the operator<br />
T : W 1,p (S,C) → R(¯∂j0)<br />
Tζ := ¯∂j0ζ + π1(Lζ ),<br />
which is a Fredholm operator of index 2. Assume that ζ ∈ ker T .Unlessζ ≡ 0, the<br />
set {z ∈ S | ζ (z) = 0} consists of finitely many points by the similarity principle (see<br />
[23]), and the local degree of each zero is positive. On the other hand, the sum of all<br />
the local degrees has to be zero, hence elements in the kernel of T are nowhere zero.<br />
Actually, if h + ik ∈ ker T ,thenevenk is nowhere zero because h + c + ik ∈ ker T<br />
for any real constant c since the zero-order term L only depends on the imaginary part<br />
of ζ . Therefore, ∗ dim ker T ≤ 2, and since the Fredholm index of T equals 2, we<br />
actually have dim ker T = 2. This proves the surjectivity of T and also of D ˆF (0,σ0)<br />
so that the set M of all pairs (b + if, γ ) solving the differential equation (2.10) isa<br />
2-dimensional manifold with T(0,γ0)M = ker D ˆF (0,γ0). If we add a real constant to<br />
b + if , then we obtain again a solution of (2.10). If we divide M by this R-action,<br />
then we obtain a 1-dimensional family of solutions (ũτ )−ε
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 53<br />
where the map r satisfies for every (i, j) ∈ N 2 a decay estimate of the form<br />
with Mij and d positive constants.<br />
|∇ i<br />
s ∇j<br />
t r(s, t)| ≤Mij e −ds<br />
Our situation is less general than in [31], so we will explain the notation in the context<br />
of this paper. The setup is a manifold M with contact form λ and contact structure<br />
ξ = ker λ. Consider a periodic orbit ¯P of the Reeb vector field Xλ with period T ,and<br />
we may assume here that T is its minimal period. We introduce P (t) := ¯P (Tt/2π)<br />
such that P (0) = P (2π). IfJ : ξ → ξ is a dλ-compatible complex structure, then<br />
the set of all ˜J -holomorphic half-cylinders<br />
ũ = (a,u) :[R,∞) × S 1 → R × M, S 1 = R/2πZ<br />
for which |a(s, t) − Ts/2π| and |u(s, t) − P (t)| decay at some exponential rate (in<br />
local coordinates near the orbit P (S 1 )) is denoted by M(P,J). Note that it is assumed<br />
here that the domain [R,∞) × S 1 is endowed with the standard complex structure. A<br />
smooth map U :[R,∞)×S 1 → P ∗ ξ for which U(s, t) ∈ ξP (t) is called an asymptotic<br />
representative of ũ if there is a proper embedding ψ :[R,∞) × S 1 → R × S 1<br />
asymptotic to the identity so that<br />
ũ ψ(s, t) = Ts/2π, exp P (t) U(s, t) , ∀ (s, t) ∈ [R,∞) × S 1<br />
(exp is the exponential map corresponding to some metric on M, e.g., the one induced<br />
by λ and J ). Every ũ ∈ M(P,J) has an asymptotic representative (see [31]). The<br />
asymptotic operator AP,J is defined as follows:<br />
(AP,Jh)(t) :=− T<br />
2π J P (t) d<br />
<br />
<br />
ds<br />
s=0<br />
<br />
Dφ−s(φs(P (t)))h(φs(P (t))) ,<br />
where φs is the flow of the Reeb vector field and where h is a section in P ∗ ξ →<br />
S 1 . Because the Reeb flow preserves the splitting TM = R Xλ ⊕ ξ we have also<br />
(AP,Jh)(t) ∈ ξP (t).<br />
We compute the asymptotic operator AP,J for the binding orbit<br />
¯P (t) =<br />
<br />
t<br />
<br />
, 0, 0 ∈ S<br />
γ1(0) 1 × R2 .<br />
Recall that the above periodic orbit has minimal period T = 2πγ1(0). Using<br />
<br />
φs P (t) = φs(t,0, 0) = t + s<br />
<br />
, 0, 0 ,<br />
γ1(0)
54 CASIM ABBAS<br />
formula (2.2) for the linearization of the Reeb flow with h(t) = (0,ζ(t),η(t)), and<br />
the fact that J (t,0, 0) ∂ ∂<br />
∂<br />
= and J (t,0, 0) ∂x ∂y ∂y =−∂ , we compute<br />
∂x<br />
′ h (t)<br />
(AP,Jh)(t) =−γ1(0)J (t,0, 0)<br />
γ1(0) +<br />
<br />
0 β(0) ζ (t)<br />
−β(0) 0 η(t)<br />
where J0 =<br />
if<br />
0 −1<br />
1 0<br />
=−J0 h ′ (t) − γ1(0)β(0) h(t)<br />
=−J0h ′ (t) + κh(t),<br />
<br />
and κ = γ ′′ ′′<br />
1 (0)/γ 2 (0) ∈ (−1, 0). Hence λ ∈ σ (AP,J) precisely<br />
h ′ (t) = (λ − κ) J0 h(t) and h(2π) = h(0)<br />
(i.e., σ (AP,J) ={κ + l | l ∈ Z}), and the largest negative eigenvalue is given by κ.<br />
The corresponding eigenspace consists of all constant vectors h(t) ≡ const ∈ R 2 .<br />
The eigenspace for the eigenvalues κ + l consists of all<br />
h(t) = e J0lt h0, with h0 ∈ R 2 .<br />
<strong>THEOREM</strong> 3.2 (Compactness)<br />
Let λ be a contact form on M which is a local model near the binding (of the Giroux<br />
leaf v0), and let J :kerλ→ker λ be a dλ-compatible complex structure. Consider a<br />
smooth family of solutions (S,jτ ,aτ ,uτ ,γτ ,J)0≤τ
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 55<br />
Remark 3.3<br />
We may assume without loss of generality that v0 ≡ u0 and f0 ≡ 0.Ifz ∈ ˙S,thenwe<br />
denote by T (z) > 0 the positive return time of the point u0(z); that is, we have<br />
T (z) := inf T>0 φT (u0(z)) ∈ u0( ˙S) < +∞.<br />
We claim that the return time z ↦→ T (z) extends continuously over the punctures of<br />
the surface, and that therefore there is an upper bound<br />
T := sup T (z) < ∞.<br />
z∈ ˙S<br />
Using (2.1)and(2.6), we note that, asymptotically near the punctures, φT (u0(s, t)) ∈<br />
S 1 × R 2 has the following structure:<br />
φT (u0(s, t)) = t + α(r(s))T, r(s) exp[i(α0 + β(r(s))T )] ,<br />
where r(s) is a strictly decreasing function, α0 is some constant, and α(r),β(r) are<br />
suitable functions for which the limits limr→0 β(r) and limr→0 α(r) exist and are not<br />
zero. Hence, if T = T (u0(s, t)) is the positive return time at the point u0(s, t), then<br />
T u0(s, t) =<br />
and therefore the limit for s →+∞exists.<br />
2π<br />
|β(r(s))| ,<br />
The remainder of this section is devoted to the proof of Theorem 3.2. We recall<br />
that the functions aτ and fτ satisfy the Cauchy-Riemann type equation (2.9)whichis<br />
¯∂jτ (aτ + ifτ ) = u ∗<br />
0 λ ◦ jτ − i(u ∗<br />
0 λ) − γτ − i(γτ ◦ jτ ),<br />
where the complex structure jτ is given by (2.8)or<br />
jτ (z) = πλTu0(z) −1 Tφfτ (z)(u0(z)) −1 · J φfτ (z)(u0(z)) <br />
Tφfτ (z) u0(z) πλTu0(z),<br />
and that γτ is a closed 1-form on S with d(γτ ◦ jτ ) = 0.<br />
The following L ∞ -bound is the crucial ingredient for the compactness result. We<br />
claim that<br />
sup<br />
0≤τ
56 CASIM ABBAS<br />
Restricting any of the solutions to a simply connected subset U ⊂ ˙S, we can<br />
write γτ = dhτ for a suitable function hτ : U → R, and the maps<br />
ũτ : U → R × M, ũτ = (aτ + hτ ,uτ )<br />
are ˜J -holomorphic curves. If two such curves ũτ and ũτ ′ have an isolated intersection,<br />
then the corresponding intersection number is positive (see [28], [5], or [27] for<br />
positivity of (self-)intersections for holomorphic curves). We claim that<br />
u0( ˙S) ∩ uτ ( ˙S) =∅, ∀ 0 0 is different from u0( ˙S), we conclude that if uτ and u0 intersect,<br />
then the intersection point of the corresponding holomorphic curves ũτ and ũ0 must<br />
be isolated. But on the other hand, this implies that uτ ′ and u0 would also intersect for<br />
all τ ′ sufficiently close to τ by positivity of the intersection number showing that the<br />
set O is open.<br />
We conclude from the above that we have a sequence τk ↘ ˜τ and points pk,qk ∈ ˙S<br />
such that uτk(pk) = u0(qk). Passing to a suitable subsequence, we may assume<br />
convergence of the sequences (pk)k∈N and (qk)k∈N to points p, q ∈ S. Because of<br />
u˜τ ( ˙S) ∩ u0( ˙S) =∅the points p, q must be punctures, and they have to be equal<br />
z0 = p = q ∈ S\ ˙S. The reason for this is the following. The maps uτk,u0 are<br />
asymptotic near the punctures to a disjoint union of finitely many periodic Reeb<br />
orbits which are not iterates of other periodic orbits. Also, different punctures always<br />
correspond to different periodic orbits. This follows from Giroux’s result and our<br />
constructions in Section 2 of this paper.
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 57<br />
We now derive a contradiction using Siefring’s result. The harmonic forms γτk in<br />
equation (1.1) are defined on all of S. Hence they are exact on some open neighborhood<br />
U of the puncture z0,andγτk = dhτk for suitable functions hτk on U and similarly γ˜τ =<br />
dh˜τ . We may also assume that j˜τ |U = jτk|U = j0 after changing local coordinates<br />
near z0. Then on the set U, the maps ũτk = (aτk + hτk,uτk) and ũ0 = (a0 + h0,u0) are<br />
holomorphic curves with ũτk(pk) = ũ0(qk) while the images of ũ˜τ and ũ0 have empty<br />
intersection. Now let<br />
U˜τ ,Uτk, U0 :[R,∞) × S 1 → R 2<br />
be asymptotic representatives of the holomorphic curves ũ˜τ , ũτk, ũ0, respectively.<br />
Invoking Theorem 3.1 and our subsequent computation of the asymptotic operator<br />
and its spectrum, we obtain the following asymptotic formulas<br />
Uτ (s, t) − U0(s, t) = e λτ s eτ (t) + rτ (s, t) , τ = ˜τ,τk, s ≥ Rτ , (3.2)<br />
where Rτ > 0 is some constant and where λτ < 0 is some negative eigenvalue of<br />
the asymptotic operator AP,J. It is of the form λτ = κ + lτ , where lτ is an integer,<br />
κ = γ ′′<br />
1<br />
(0)/γ ′′<br />
2 (0) is not an integer, and where eτ (t) = e J0lτ t hτ , hτ ∈ R 2 \{0} is an<br />
eigenvector corresponding to the eigenvalue λτ = κ + lτ . Note that the above formula<br />
applies since Uτ − U0 cannot vanish identically. We will actually show that lτ ≡ 0.<br />
The asymptotic representative U0 is given by<br />
u0(s, t) = t,r(s)e iα0<br />
<br />
= t,U0(s, t) ,<br />
using equation (2.6), and we recall that r(s) = c(s)eκs , where c(s) → c∞ > 0 as<br />
s →+∞. An asymptotic representative of ũτ , however, is given by an expression<br />
such as<br />
<br />
uτ ψ(s, t) = t,Uτ (s, t) ,<br />
where ψ :[R,∞) × S 1 → R × S 1 is a proper embedding converging to the identity<br />
map as s →+∞. Writing (s ′ ,t ′ ) = ψ(s, t), we get using equations (2.1) forthe<br />
Reeb flow<br />
Uτ (s, t) = c(s ′ )e κs′<br />
e i(α0+β(r(s ′ ))fτ (s ′ ,t ′ ))<br />
= e κs eτ + rτ (s, t) .<br />
The asymptotic formula for Uτ a priori allows for other decay rates, but κ is the only<br />
possible one. Dividing by e κs and passing to the limit s →+∞, we obtain<br />
eτ = c∞e iα0 e iβ(0)fτ (∞) ,
58 CASIM ABBAS<br />
where fτ (∞) = lims→+∞ fτ (s, t) which is independent of t since fτ extends continuously<br />
over the punctures. Hence the difference Uτ − U0 has decay rate λτ ≡ κ as<br />
claimed unless the two eigenvectors eτ and e0 agree, which is equivalent to<br />
fτ (∞) ∈ 2π<br />
β(0) Z<br />
or τ = ˜τ in our case. The maps Uτ − U0 satisfy a Cauchy-Riemann type equation<br />
to which the similarity principle applies so that, for every zero (s, t) of Uτ − U0, the<br />
map σ ↦→ (Uτ − U0)(s + ɛ cos σ, t + ɛ sin σ ) has positive degree for small ɛ>0.<br />
The Cauchy-Riemann type equation mentioned above is derived in [31, Section 5.3]<br />
as well as in [1, Section 3] in a slightly different context, and also in [21]. If R is<br />
sufficiently large, then the map<br />
S 1 → S 1 , t ↦→ Wτ (R,t) := Uτ − U0<br />
|Uτ − U0| (R,t)<br />
is well defined, and it has degree lτ because the remainder term rτ (s, t) decays<br />
exponentially in s. Zeros of Uτ − U0 contribute in the following way: if R ′ 0 so large that<br />
degW˜τ (R ′ , ·) = l˜τ < 0.Forτ sufficiently close to ˜τ,wealsohavedegWτ (R ′ , ·) = l˜τ .<br />
On the other hand, we have degWτ (R,·) = 0 for R>R ′ sufficiently large. Equation<br />
(3.3) implies that the map Uτ − U0 must have zeros in [R ′ ,R] × S 1 to account<br />
for the difference in degrees, but we know that there are none for τ < ˜τ. This<br />
contradiction shows that l˜τ = 0 is impossible. Choose again R ′ > 0 so large that<br />
degW˜τ (R ′ , ·) = 0. The degree does not change if we slightly alter τ. In particular,<br />
we have degWτ (R ′ , ·) = 0 for τ > ˜τ close to ˜τ as well. For R>>R ′ ,wehave<br />
degW˜τ (R,·) = 0, and we recall that<br />
(Uτk − U0)(sk,tk) = 0, τk ↘ ˜τ<br />
for a suitable sequence (sk,tk) with sk →+∞and that the set of zeros of Uτk − U0 is<br />
discrete. This, however, contradicts equation (3.3) since the zeros have positive orders.<br />
Summarizing, we have shown that the assumption O = ∅leads to a contradiction<br />
which implies the a priori bound (3.1).<br />
The monotonicity of the functions fτ in τ and the bound (3.1) imply that the<br />
functions fτ converge pointwise to a measurable function fτ0 as τ ↗ τ0. Wealso
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 59<br />
know that fτ0 L ∞ ( ˙S) ≤ T . We then obtain a complex structure jτ0 on ˙S by<br />
jτ0(z) = πλTu0(z) −1 Tφfτ (z)(u0(z)) 0 −1 × J φfτ (z)(u0(z)) 0 <br />
Tφfτ (z) u0(z) 0 πλTu0(z).<br />
By definition, the complex structure jτ0 is also of class L∞ and jτ (z) → j1(z)<br />
pointwise. Our task is to improve the regularity of the limit fτ0 and the character of<br />
the convergence fτ → fτ0 . We also have to establish convergence of the functions aτ<br />
for τ ↗ τ0. The complex structures jτ are of course all smooth, but the limit jτ0 might<br />
only be measurable.<br />
3.1. The Beltrami equation<br />
For the reader’s convenience, we briefly summarize a few classical facts from the<br />
theory of quasiconformal mappings (see [8], [9]). The punctured surface ˙S carries<br />
metrics gτ , also of class L ∞ for τ = τ0 and smooth otherwise, so that<br />
In fact, gτ is given by<br />
gτ (z) jτ (z)v, jτ (z)w = gτ (z)(v, w), for all v, w ∈ Tz ˙S.<br />
gτ (z)(v, w) = dλ uτ (z) πλTuτ (z)v, J(uτ (z))πλTuτ (z)w .<br />
In the case τ = τ0, we replace πλTuτ (z) by Tφfτ 0 (z)(u0(z))πλTu0(z). Wehave<br />
sup τ gτ L ∞ ( ˙S) < ∞ and gτ → gτ0 pointwise as τ ↗ τ0. Our considerations about<br />
the regularity of the limit are of local nature, so we may replace ˙S with a ball B ⊂ C<br />
centered at the origin. Denoting the metric tensor of gτ by (g τ kl )1≤k,l≤2, wedefinethe<br />
following complex-valued smooth functions:<br />
and we note that<br />
µτ (z) :=<br />
1<br />
2 (gτ 11 (z) − gτ 22 (z)) + igτ 12 (z)<br />
1<br />
2 (gτ 11 (z) + gτ 22 (z)) + gτ 11 (z)gτ 22 (z) − (gτ ,<br />
12 (z))2<br />
sup µτ L∞ ( ˙S) < 1<br />
τ<br />
and that µτ → µτ0 pointwise. We view the functions µτ as functions on the whole<br />
complex plane by trivially extending them beyond B. Then they are also τ-uniformly<br />
bounded in L p (C) for all 1 ≤ p ≤∞and µτ → µτ0 in Lp (C) for 1 ≤ p
60 CASIM ABBAS<br />
for τ
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 61<br />
We used here the notation ∂ = (1/2)(∂s + i∂t) and ∂ = (1/2)(∂s − i∂t). Properties<br />
(1) and (2) above should be understood in the sense of distributions. The proof of (4)<br />
involves the Calderón-Zygmund inequality and the Riesz-Thorin convexity theorem<br />
(see [25] and[32]). Following [9], we define Bp with p>2 to be the space of all<br />
locally integrable functions on the plane which have weak derivatives in Lp (C),vanish<br />
in the origin, and which satisfy a global Hölder condition with exponent 1 − 2<br />
p .For<br />
u ∈ Bp, we then define a norm by<br />
|u(z2) − u(z1)|<br />
uBp := sup<br />
z1=z2 |z2 − z1| 1−(2/p) +∂uLp (C) +∂uLp (C)<br />
so that Bp becomes a Banach space. We usually choose p > 2 such that<br />
cp sup τ µτ L ∞ (C) < 1, where cp is the constant from item (4) above.<br />
<strong>THEOREM</strong> 3.4 ([9, Theorem 1])<br />
Assume that p>2 such that cp sup τ µτ L ∞ (C) < 1.Ifσ ∈ L p (C), then the equation<br />
has a unique solution u = uµ,σ ∈ Bp.<br />
∂u = µ∂u+ σ<br />
For the existence part of the theorem, one first solves the following fixed-point problem<br />
in L p (C):<br />
This is possible because the map<br />
q = Ɣ(µq) + Ɣσ.<br />
L p (C) −→ L p (C)<br />
q ↦−→ Ɣ(µq + σ )<br />
is a contraction in view of cpµL ∞ (C) < 1. Then<br />
u := A(µq + σ )<br />
is the desired solution. The following estimate is also derived in [9]:<br />
qL p (C) ≤ c ′<br />
p σ L p (C)<br />
with c ′ p = cp/(1 − cpµL ∞ (C)), which follows from<br />
qL p (C) ≤Ɣ(µq)L p (C) +ƔσL p (C)<br />
≤ cpµL ∞ (C)qL p (C) + cpσ L p (C).<br />
(3.4)
62 CASIM ABBAS<br />
Recalling our original situation, we have the following result which shows that<br />
there is some sort of conformal mapping for j1 on the ball B.<br />
<strong>THEOREM</strong> 3.5 ([9, Theorem 4])<br />
Let µ : C → C be an essentially bounded measurable function with µ|C\B ≡ 0 and<br />
p>2 such that cpµL ∞ (C) < 1. Then there is a unique map α : C → C with<br />
α(0) = 0 such that<br />
∂α = µ∂α<br />
in the sense of distributions with ∂α − 1 ∈ L p (C).<br />
The desired map α is given by<br />
α(z) = z + u(z),<br />
where u ∈ Bp solves the equation ∂u = µ∂u + µ. In particular, α ∈ W 1,p (B).<br />
Lemma 8 in [9] states that α : C → C is a homeomorphism. We can apply the<br />
theorem to all the µτ , 0 2 such that cp sup n µnL ∞ (C) < 1 and any compact set<br />
B ⊂ C.<br />
Proof<br />
We first estimate with g ∈ Lp (C) and z = 0<br />
|Ag(z)| = 1<br />
<br />
z<br />
<br />
<br />
g(ξ) dξ d¯ξ<br />
2π C ξ(ξ − z)<br />
≤ |z|<br />
2π gLp <br />
1<br />
<br />
<br />
(C) <br />
ξ(ξ − z)<br />
≤ CpgLp 2<br />
1−<br />
(C) |z| p ,<br />
L p<br />
p−1 (C)<br />
(3.5)
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 63<br />
where the last estimate holds in view of<br />
<br />
|ξ(ξ − z)| −p/(p−1) dξ d¯ξ ζ =z−1 <br />
ξ<br />
=<br />
C<br />
If q solves q = Ɣ(µq + µ),then<br />
C<br />
=|z| 2−2p/(p−1)<br />
¯∂(αn − α) = µn ∂(αn − α) − µ∂α+ µn ∂α<br />
|z 2 ζ 2 − z 2 ζ | −p/(p−1) |z| 2 dζ d¯ζ<br />
<br />
|ζ (ζ − 1)| −p/(p−1) dζ d¯ζ .<br />
C <br />
2πCp<br />
<br />
= µn ∂(αn − α) + µn − µ + (µn − µ)Ɣ(µq + µ),<br />
that is, the difference αn − α again satisfies an inhomogeneous Beltrami equation. By<br />
Theorem 3.4, wehave<br />
αn − α = A(µnqn + λn),<br />
where λn = µn − µ + (µn − µ)Ɣ(µq + µ) and where qn ∈ L p (C) solves qn =<br />
Ɣ(µnqn + λn). Combining this with (3.5)and(3.4), we obtain<br />
|αn(z) − α(z)| ≤Cp µnqn + λnL p (C) |z| 1−2/p<br />
≤ (Cp sup µnL<br />
n<br />
∞ (C) · c ′<br />
pλnLp (C) + Cp λnLp (C)) |z| 1−2/p . (3.6)<br />
Since µn − µLp (C) → 0 and (µn − µ)Ɣ(µq + µ)Lp (C) → 0 by Lebesgue’s<br />
theorem we also have λnLp (C) → 0 and therefore αn → α uniformly on compact<br />
sets. Since ¯∂(αn − α) = µn ∂(αn − α) + λn and αn − α = A(µnqn + λn), we verify<br />
that<br />
and<br />
∂(αn − α) = Ɣ(µnqn + λn) = qn<br />
¯∂(αn − α) = µnqn + λn.<br />
Invoking (3.4) once again, we see that both ∂(αn − α)L p (C) and ¯∂(αn − α)L p (C)<br />
can be bounded from above by a constant times λnL p (C) which converges to<br />
zero. <br />
We also need some facts concerning the classical case where µ ∈ C k,α (BR(0)),<br />
BR(0) ={z ∈ C ||z|
64 CASIM ABBAS<br />
<strong>THEOREM</strong> 3.7<br />
Let µ, γ, δ ∈ C k,α (BR ′(0)) with 0
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 65<br />
for a suitable constant c>0 depending on α and R. This is only implicitly proved in<br />
[11], so we sketch the proof of this inequality. We have<br />
and<br />
(T2 − T1)h(z) = A (µ2 − µ1)∂h + (γ2 − γ1)h (z)<br />
− zƔ (µ2 − µ1)∂h + (γ2 − γ1)h (0),<br />
∂ (T2 − T1)h (z) = Ɣ (µ2 − µ1)∂h + (γ2 − γ1)h (z)<br />
− Ɣ (µ2 − µ1)∂h + (γ2 − γ1)h (0),<br />
¯∂ (T2 − T1)h (z) = (µ2 − µ1)(z)∂h(z) + (γ2 − γ1)(z)h(z).<br />
We need [13, (21)–(24)]. Adapted to our notation, they look as follows with z, z1,z2 ∈<br />
BR(0):<br />
Recalling that<br />
and<br />
|(Ah)(z2) − (Ah)(z1)|<br />
|z2 − z1| α<br />
|(Ah)(z)| ≤4RhC 0 (BR(0))<br />
|(Ɣh)(z)| ≤ 2α+1<br />
α Rα hC 0,α (BR(0))<br />
|(Ɣh)(z2) − (Ɣh)(z1)|<br />
|z2 − z1| α<br />
≤ Cα hC0,α (BR(0)).<br />
≤ 2hC 0 (BR(0)) + 2α+2<br />
α Rα hC 0,α (BR(0))<br />
hC 1,α (BR(0)) := hC 0 (BR(0)) +∂hC 0,α (BR(0)) +¯∂hC 0,α (BR(0))<br />
and that the Hölder norm satisfies<br />
kC 0,α (BR(0)) := kC0 |k(z2) − k(z1)|<br />
(BR(0)) + sup<br />
z1=z2 |z2 − z1| α<br />
hkC 0,α (BR(0)) ≤ C hC 0,α (BR(0)) kC 0,α (BR(0))<br />
for a suitable constant C depending only on α and R, the asserted inequality for the<br />
operator norm of T2 − T1 follows. In the same way, we obtain<br />
g2 − g1C 1,α (BR(0)) ≤ c δ2 − δ1C 0,α (BR(0)).
66 CASIM ABBAS<br />
Since<br />
w2 − w1C 1,α (BR(0)) ≤(T2 − T1)w2C 1,α (BR(0))<br />
+ θ w2 − w1C 1,α (BR(0)) +g2 − g1C 1,α (BR(0))<br />
and since θ
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 67<br />
The integral <br />
d(fγ) vanishes by Stokes’s theorem since fγ is a smooth 1-form on<br />
S<br />
the closed surface S. The form da∧γ ◦jf is not smooth on S, but the integral vanishes<br />
anyway for the following reason. As we have proved in the appendix, the form γ ◦ jf<br />
is bounded near the punctures, and hence in local coordinates near a puncture it is of<br />
the form<br />
σ = F (w1,w2)dw1 + F2(w1,w2)dw2, w1 + iw2 ∈ C,<br />
where F1,F2 are smooth (except possibly at the origin) but bounded. Passing to polar<br />
coordinates via<br />
φ :[0, ∞) × S 1 −→ C\{0}<br />
φ(s, t) = e −(s+it) = w1 + iw2,<br />
we see that φ ∗ σ has to decay at the rate e −s for large s. The form da has γ1(r(s)) ds<br />
as its leading term. Computing the integral <br />
Ɣ a(γ ◦ jf ) over small loops Ɣ around<br />
the punctures and using Stokes’s theorem, we conclude that the contribution from<br />
neighborhoods<br />
<br />
of the punctures can be made arbitrarily small. Therefore, the integral<br />
˙S da ∧ γ ◦ jf must vanish.<br />
If is a volume form on S, then we may write u∗ 0λ ∧ γ = g · for a suitable<br />
smooth function g. Defining<br />
<br />
|u<br />
˙S<br />
∗<br />
<br />
0λ ∧ γ | := |g| ,<br />
˙S<br />
we have<br />
γ 2<br />
L2 ,jf =<br />
<br />
<br />
u<br />
˙S<br />
∗<br />
<br />
<br />
0λ ∧ γ <br />
<br />
≤ |u ∗<br />
0λ ∧ γ |<br />
˙S<br />
≤u ∗<br />
0 λL 2 ,jf γ L 2 ,jf ,<br />
which implies the assertion. <br />
We resume the proof of the compactness result, Theorem 3.2. All the considerations<br />
which follow are local. The task is to improve the regularity of the limit fτ0 and the<br />
nature of the convergence fτ → fτ0 . Because the proof is somewhat lengthy, we<br />
organize it in several steps. For τ
68 CASIM ABBAS<br />
be the conformal transformations as in Section 2, that is,<br />
Tατ (z) ◦ i = jτ (ατ (z)) ◦ Tατ (z), z ∈ B.<br />
The L ∞ -bound (3.1) on the family of functions (fτ ) and the above L 2 -bound imply<br />
convergence of the harmonic forms α ∗ τ γτ after maybe passing to a subsequence.<br />
PROPOSITION 3.9<br />
Let τ ′ k be a sequence converging to τ0, and let B ′ = Bε ′(0) with B′ ⊂ B. Then there is<br />
a subsequence (τk) ⊂ (τ ′ k ) such that the harmonic 1-forms α∗ τk γτk converge in C∞ (B ′ ).<br />
Proof<br />
First, the harmonic 1-forms α ∗ τ γτ satisfy the same L 2 -bound as in Proposition 3.8:<br />
α ∗<br />
τ γτ 2<br />
L2 (B) =<br />
<br />
<br />
=<br />
<br />
=<br />
B<br />
B<br />
Uτ<br />
α ∗<br />
τ γτ ◦ i ∧ α ∗<br />
τ γτ<br />
α ∗<br />
τ (γτ ◦ jτ ) ∧ α ∗<br />
τ γτ<br />
γτ ◦ jτ ∧ γτ<br />
≤u ∗<br />
0 λL 2 ,jτ<br />
≤ C,<br />
where C is a constant depending only on λ and u0 since<br />
We write<br />
sup jτ L∞ ( ˙S) < ∞.<br />
τ<br />
α ∗<br />
τ γτ = h 1<br />
τ ds + h2<br />
τ dt,<br />
where hk τ , k = 1, 2 are harmonic and bounded in L2 (B) independent of τ. Ify ∈ B<br />
and BR(y) ⊂ BR(y) ⊂ B, then the classical mean-value theorem<br />
h k 1<br />
τ (y) =<br />
πR2 <br />
h k<br />
τ (x)dx<br />
implies that, for any ball Bδ = Bδ(y) with Bδ ⊂ Bδ ⊂ B, we have the rather generous<br />
estimate<br />
h k<br />
τ C0 (Bδ(y)) ≤ 1<br />
√ h<br />
πδ k<br />
τ L2 √<br />
C<br />
(B) ≤ √ .<br />
πδ<br />
BR(y)
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 69<br />
With y ∈ B and with ν = (ν1,ν2) being the unit outer normal to ∂Bδ(y), weget<br />
∂sh k 1<br />
τ (y) =<br />
πδ2 <br />
∂sh<br />
Bδ(y)<br />
k<br />
τ (x)dx<br />
= 1<br />
πδ2 <br />
div(h<br />
Bδ(y)<br />
k<br />
τ , 0) dx<br />
= 1<br />
πδ2 <br />
h k<br />
τ ν1 ds<br />
and<br />
|∇h k 1<br />
τ (y)| =<br />
πδ 2<br />
∂Bδ(y)<br />
<br />
<br />
<br />
h<br />
∂Bδ(y)<br />
k<br />
τ νds<br />
≤ 2<br />
δ hk<br />
τ C 0 (Bδ(y))<br />
so that, for B ′ = Bε ′, r being the radius of B, andδ = r − ε′ ,wehave<br />
∇h k<br />
τ C 0 (B ′ ) ≤ 2√ C<br />
√ πδ 2 .<br />
By iterating this procedure on nested balls we obtain τ-uniform C 0 (B ′ )-bounds on all<br />
derivatives. Convergence as stated in the proposition then follows from the Ascoli-<br />
Arzela theorem. <br />
3.3. A uniform L p -bound for the gradient<br />
The first step of the regularity story is showing that the gradients of aτ + ifτ are<br />
uniformly bounded in L p (B ′ ) for some p>2 and for any ball B ′ with B ′ ⊂ B. It will<br />
soon become apparent why this gradient bound is necessary. Since we do not have a<br />
lot to start with, the proof will be indirect. Recall the differential equation (2.9)<br />
where<br />
We set<br />
¯∂jτ (aτ + ifτ ) = u ∗<br />
0 λ ◦ jτ − i(u ∗<br />
0 λ) − γτ − i(γτ ◦ jτ ),<br />
jτ (z) = πλTu0(z) −1 Tφfτ (z)(u0(z)) −1 ×J φfτ (z)(u0(z)) <br />
Tφfτ (z) u0(z) πλTu0(z).<br />
<br />
<br />
<br />
φτ (z) := aτ (z) + ifτ (z), z ∈ Uτ
70 CASIM ABBAS<br />
so that, for z ∈ B, wehave<br />
<br />
∂(φτ ◦ ατ )(z) = ∂jτ φτ ατ (z) ◦ ∂sατ (z)<br />
= u ∗<br />
0λ ◦ jτ − i(u ∗<br />
0λ)ατ (z) ◦ ∂sατ (z)<br />
<br />
− (α ∗<br />
τ γτ )(z) · ∂<br />
+ i(α∗ τ<br />
∂s γτ )(z) · ∂<br />
<br />
∂t<br />
and<br />
=: ˆFτ (z) + ˆGτ (z)<br />
=: ˆHτ (z),<br />
(3.8)<br />
sup ˆFτ L<br />
τ<br />
p (B) < ∞ for some p>2 (3.9)<br />
since ατ → ατ0 in W 1,p (B) and supτ jτ L∞ < ∞. Wealsohave<br />
sup ˆGτ C<br />
τ<br />
k (B ′ ) < ∞ (3.10)<br />
for any ball B ′ ⊂ B ′ ⊂ B and any integer k ≥ 0 in view of Proposition 3.9 (the<br />
proposition asserts uniform convergence after passing to a suitable subsequence, but<br />
uniform bounds on all derivatives are established in the proof). We claim now that,<br />
for every ball B ′ ⊂ B ′ ⊂ B, there is a constant CB ′ > 0 such that<br />
∇(φτ ◦ ατ )L p (B ′ ) ≤ CB ′, ∀ τ ∈ [0,τ0). (3.11)<br />
Arguing indirectly, we may assume that there is a sequence τk ↗ τ0 such that<br />
Now define<br />
∇(φτk ◦ ατk)L p (B ′ ) →∞ for some ball B ′ ⊂ B ′ ⊂ B. (3.12)<br />
εk := inf ε>0 ∃ x ∈ B ′ : ∇(φτk ◦ ατk)L p (Bε(x)) ≥ ε 2/p−1 ,<br />
which are positive numbers since ε (2/p)−1 →+∞. Because we assumed (3.12) we<br />
must have infk εk = 0, hence we will assume that εk → 0. Otherwise, if ε0 =<br />
(1/2) infk εk > 0, then we cover B ′ with finitely many balls of radius ε0, andwe<br />
would get a k-uniform L p -bound on each of them, contradicting (3.12). We claim that<br />
∇(φτk ◦ ατk)L p (Bε k (x)) ≤ ε (2/p)−1<br />
k , ∀ x ∈ B ′ . (3.13)<br />
Otherwise, we could find y ∈ B ′ so that<br />
∇(φτk ◦ ατk)Lp (Bε (y)) >ε k (2/p)−1<br />
k ,
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 71<br />
and we would still have the same inequality for a slightly smaller ε ′ k 0, and we define for z ∈ BR(0) the<br />
functions<br />
ξk(z) := (φτk ◦ ατk) xk + εk(z − xk) ,<br />
which makes sense if k is sufficiently large. The transformation<br />
: x ↦−→ xk + εk(x − xk)<br />
satisfies (B1(xk)) = Bεk(xk) and (B1(y)) ⊂ Bεk(xk + εk(y − xk)) so that<br />
<br />
|∇(φτk ◦ ατk)(x)|<br />
Bε (xk) k p dx = ε 2<br />
<br />
|∇(φτk k<br />
◦ ατk)<br />
B1(xk)<br />
xk + εk(z − xk) | p dz<br />
= ε 2<br />
<br />
k ε −p<br />
k |∇ξk(z)| p dz<br />
and<br />
B1(xk)<br />
∇ξkL p (B1(xk)) = ε 1−(2/p)<br />
k ∇(φτk ◦ ατk)L p (Bε k (xk)) (3.15)<br />
= 1<br />
by (3.14) and, for any y for which ξk|B1(y) is defined and for large enough k, wehave<br />
∇ξkL p (B1(y)) ≤ ε 1−(2/p)<br />
k ∇(φτk ◦ ατk)L p (Bε k (xk+εk(y−xk))) ≤ 1 (3.16)<br />
by (3.13). The functions ξk satisfy the equation<br />
¯∂ξk(z) = εk ˆFτk<br />
<br />
xk + εk(z − xk) + εk ˆGτk<br />
<br />
xk + εk(z − xk) =: Hτk(z) (3.17)
72 CASIM ABBAS<br />
and, for every R>0, wehave<br />
sup ∇ξkL<br />
k<br />
p (BR(0)) < ∞, ∇ξkL p (B2(0)) ≥ 1 (3.18)<br />
because of (3.16) and(3.15) sinceB1(xk) ⊂ B2(0) for large k. The upper bound<br />
on ∇ξkL p (BR(0)) depends on how many balls B1(y) are needed to cover BR(0). We<br />
compute for ρ>0<br />
HτkLp (Bρ(xk)) = ε 1−(2/p)<br />
k ˆHτkL p (Bρε (xk)),<br />
k<br />
with ˆHτk as in (3.8). We conclude from p>2, (3.9), and (3.10) that<br />
for any R>0 as k →∞. Defining<br />
HτkL p (BR(0)) −→ 0<br />
X l,p :={ψ ∈ W l,p (B,C 2 ) | ψ(0) = 0, ψ(∂B) ⊂ R 2 },l≥ 1, B⊂ C aball,<br />
the Cauchy-Riemann operator<br />
¯∂ : X l,p −→ W l−1,p (B,C 2 )<br />
is a bounded bijective linear map. By the open mapping principle, we have the following<br />
estimate:<br />
ψl,p,B ≤ C ∂ψl−1,p,B, ∀ ψ ∈ X l,p . (3.19)<br />
Let R ′ ∈ (0,R). Now pick a smooth function β : R 2 → [0, 1] with β|B R ′ (0) ≡ 1 and<br />
supp(β) ⊂ BR(0). Define<br />
We note that<br />
ζk(z) := Re ξk(z) − ξk(0) + iβ(z) Im ξk(z) − ξk(0) .<br />
sup Im(ξk)L<br />
k<br />
p (BR(0)) ≤ CR<br />
with a suitable constant CR > 0 because of the uniform bound<br />
sup Im(ξk)L<br />
k,R<br />
∞ (BR(0)) < ∞.
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 73<br />
Using (3.19), we then obtain<br />
because of<br />
ξk − ξk(0)1,p,B R ′ (0) ≤ζk1,p,BR(0)<br />
≤ C ∂ζkL p (BR(0))<br />
≤ CR ′(HτkL p (BR(0)) +∇ξkL p (BR(0))<br />
+Im(ξk) − Im(ξk)(0)L p (BR(0)))<br />
∂ζk = Hτk + i(β − 1)∂ Im(ξk) + i ∂β Im(ξk) − Im(ξk)(0) .<br />
(3.20)<br />
Hence the sequence (ξk − ξk(0)) is uniformly bounded in W 1,p (BR ′(0)), and in particular,<br />
it has a subsequence which converges in Cα 2<br />
(BR ′(0)) for 0
74 CASIM ABBAS<br />
W 1,p<br />
loc (C) since the sequence (ξk − ξk) depends on the choice of the ball BR(0). Our<br />
sequence ξk − ξk(0) has a convergent subsequence on any ball.<br />
3.4. Convergence in W 1,p (B ′ )<br />
Pick a sequence τk ↗ τ0. We claim that the sequence ( ˆFτk) converges in L p (B) maybe<br />
after passing to a suitable subsequence (recall that, so far, we only have the uniform<br />
bound (3.9)). The functions ˆFτk converge pointwise almost everywhere after passing<br />
to some subsequence. Indeed, the sequence {(u∗ 0λ ◦ jτk − i(u∗ 0λ))ατ (z)} converges<br />
k<br />
already pointwise since jτk and ατk do (recall that the sequence (ατk) converges in<br />
W 1,p (B) and therefore uniformly). The sequence (∂sατk) converges in Lp (B) and<br />
therefore pointwise almost everywhere after passing to a suitable subsequence. Then<br />
by Egorov’s theorem, for any δ>0, there is a subset Eδ ⊂ B with |B\Eδ| ≤δ so that<br />
the sequence ˆFτk converges uniformly on Eδ. Letαbe the Lp-limit of the sequence<br />
(∂sατk), andletε>0. We introduce<br />
<br />
C := 2 sup (u<br />
0≤τ0 sufficiently small such that<br />
αL p (B\Eδ) ≤ ε<br />
3 C .<br />
Now choose k0 ≥ 0 so large that, for all k ≥ k0, wehave<br />
∂sατk − αL p (B) ≤ ε<br />
3 C<br />
Then, if k, l ≥ k0, wehave<br />
and ˆFτk − ˆFτlL ∞ (Eδ) ≤ ε<br />
3 |B| .<br />
ˆFτk − ˆFτlL p (B) ≤ˆFτk − ˆFτlL p (Eδ) +ˆFτk − ˆFτlL p (B\Eδ)<br />
proving the claim.<br />
≤|Eδ|ˆFτk − ˆFτlL ∞ (Eδ) + 2 sup ˆFτkL<br />
k≥k0<br />
p (B\Eδ)<br />
≤|B|ˆFτk − ˆFτl p<br />
L ∞ (Eδ)<br />
≤ ε<br />
+ C · sup ∂sατkL<br />
k≥k0<br />
p (B\Eδ)<br />
Recalling that φτ = aτ + ifτ and that the family fτ satisfies a uniform L ∞ -bound, we<br />
have<br />
sup Im(φτ ◦ ατ )L<br />
τ<br />
∞ (B) < ∞.
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 75<br />
Now pick three balls B ′′′ ⊂ B ′′ ⊂ B ′ ⊂ B such that the closure of one is contained<br />
in the next. Our aim is to establish W 1,p (B ′′′ )-convergence of a subsequence of the<br />
sequence (φτk ◦ ατk). By Proposition 3.10, we have a uniform Lp (B ′ )-bound on the<br />
gradient. If β : R2 → [0, 1] is a smooth function with supp (β) ⊂ B ′ and β|B ′′ ≡ 1<br />
and if<br />
ζτ = Re φτ ◦ ατ − φτ (0) + iβ Im φτ ◦ ατ − φτ (0) ,<br />
then we proceed in the same way as in (3.20), and we obtain<br />
ϕk1,p,B ′′ ≤ C ( ˆHτkL p (B ′ ) +∇(φτk ◦ ατk)L p (B ′ ) +Im(ϕk)L p (B ′ )),<br />
where we wrote<br />
ϕk := φτk ◦ ατk − (φτk ◦ ατk)(0),<br />
and where C>0 is a constant depending only on p, B ′ ,andB ′′ . The sequence (ϕk) is<br />
then uniformly bounded in W 1,p (B ′′ ) by Proposition 3.10, and it converges in L p (B ′′ )<br />
after passing to a suitable subsequence. We now use the regularity estimate<br />
ϕl − ϕk1,p,B ′′′ ≤ C ( ˆFτl − ˆFτkL p (B ′′ ) (3.22)<br />
+ˆGτl − ˆGτkL p (B ′′ ) +ϕl − ϕkL p (B ′′ )).<br />
Since the right-hand side converges to zero as k, l →∞, we obtain the following.<br />
PROPOSITION 3.12<br />
For every ball B ′ ⊂ B ′ ⊂ B, the sequence (φτk ◦ ατk − φτk(0)) has a subsequence<br />
which converges in W 1,p (B ′ ).<br />
3.5. Improving regularity using both the Beltrami and Cauchy-Riemann equations<br />
In order to improve the convergence of the conformal transformations ατk , we need<br />
to improve the convergence of the maps µτk → µτ0 and the regularity of its limit.<br />
It is known that the inverses α−1 of the conformal transformations ατk also satisfy a<br />
τk<br />
Beltrami equation (see [9])<br />
where<br />
∂α −1<br />
τ = ντ ∂α −1<br />
τ ,<br />
ντ (z) =− ∂ατ (α −1<br />
τ (z))<br />
∂ατ (α −1<br />
τ (z))µτ<br />
α −1<br />
τ (z)
76 CASIM ABBAS<br />
(follows from 0 = ∂(α−1 τ ◦ ατ ) = ∂α−1 τ (ατ )∂ατ + ∂α−1 τ (ατ )∂ατ ). After passing to a<br />
suitable subsequence, we may assume that ∂ατk and ∂ατk converge pointwise almost<br />
everywhere since they converge in Lp (B). Hence we may assume that the sequence<br />
(ντk) also converges pointwise almost everywhere. We also have<br />
|ντ (z)| ≤ <br />
µτ<br />
−1<br />
α (z) ,<br />
and hence ντ satisfies the same L ∞ -bound as µτ . By Lemma 3.6, we conclude that<br />
α −1<br />
τk<br />
−→ α−1<br />
1 in W 1,p (B)<br />
with the same p>2 as in Lemma 3.6 applied to the functions µτ . After passing to<br />
some subsequence, the sequence<br />
(ϕk) := φτk − aτk(0) ◦ ατk<br />
converges in W 1,p (B ′ ) for any ball B ′ ⊂ B by Proposition 3.12. Indeed, the expression<br />
φτk ◦ ατk − φτk(0) and ϕk differ by a constant term ifτk(0), but the sequence (ifτk(0))<br />
has a convergent subsequence.<br />
We would like to derive a decent notion of convergence for the sequence (ϕτk◦α −1<br />
τk ),<br />
but the space W 1,p is not well behaved under compositions. The composition of two<br />
functions of class W 1,p is only in W 1,p/2 . Since we cannot choose p>2 freely, we<br />
rather carry out the argument in Hölder spaces. By the Sobolev embedding theorem<br />
and Rellich compactness, we may assume that the sequences (ϕk) and (α−1) converge<br />
τk<br />
in C0,α (B ′ ) for any ball B ′ ⊂ B ′ ⊂ B and 0
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 77<br />
spaces, that is,<br />
ϕl − ϕkC k+1,γ (B ′′′ ) ≤ C ( ˆFτl − ˆFτkC k,γ (B ′′ )<br />
+ˆGτl − ˆGτkC k,γ (B ′′ ) +ϕl − ϕkC k,γ (B ′′ )).<br />
The sequence ( ˆFτk) now converges in the C0,α2-norm, and the sequence ( ˆGτk) converges<br />
in any Hölder norm. We obtain with the above regularity estimate C1,α2-convergence of<br />
the sequence (ϕk), and composing with α−1 yields C1,α4-convergence<br />
of (fτk) and (µτl).<br />
τk<br />
Invoking Theorem 3.7 again then improves the convergence of the transformations<br />
ατk, α−1 to C τk 2,α4.<br />
We now iterate the procedure using the regularity estimate for the<br />
Cauchy-Riemann operator in Hölder space and the estimate for the Beltrami equation<br />
in Theorem 3.7.<br />
Theorem 3.2 follows if we apply the implicit function theorem to the limit solution<br />
(S,jτ0, ũτ0 = (aτ0,uτ0),γτ0), hence we obtain the same limit for every sequence {τk},<br />
and we obtain convergence in C∞ .<br />
4. Conclusion<br />
The following remarks tie together the loose ends and prove the main result, Theorem<br />
1.6. We start with a closed 3-dimensional manifold with contact form λ ′ . Giroux’s<br />
theorem, Theorem 1.4, then permits us to change the contact form λ ′ to another<br />
contact form λ such that ker λ = ker λ ′ and such that there is a supporting open book<br />
decomposition with binding K consisting of periodic orbits of the Reeb vector field<br />
of λ. Invoking Proposition 2.4, we construct a family of 1-forms (λδ)0≤δ
78 CASIM ABBAS<br />
is finite, the images of uτ and u0 must agree for some sufficiently large τ, concluding<br />
the proof. <br />
Appendix. Some local computations near the punctures<br />
In this appendix, we present some local computations needed for the proof of Theorem<br />
2.8. The issue is to show that the 1-forms<br />
u ∗<br />
0 λ ◦ jf − da0 and u ∗<br />
0 λ + da0 ◦ jf<br />
are bounded on ˙S. We obtain in the second case of the theorem<br />
u ∗<br />
0 λ ◦ jf − da0 = u ∗<br />
0 λ ◦ (jf − jg) + γ0<br />
= dg ◦ (jf − jg) + γ0 + v ∗<br />
0 λ ◦ (jf − jg).<br />
The first case can be treated as a special case: here the objective is to show that the<br />
1-form v ∗ 0 λ ◦ (jf − i) = v ∗ 0 λ ◦ (jf − j0) is bounded near the punctures. We again drop<br />
the subscript δ in the notation since we are only concerned with a local analysis near<br />
the binding, and all the forms λδ are identical there. We use coordinates (θ,r,φ) near<br />
the binding. The contact structure is then generated by<br />
η1 = ∂<br />
∂r = (0, 1, 0), η2<br />
∂<br />
=−γ2<br />
∂θ<br />
+ γ1<br />
∂<br />
∂φ = (−γ2, 0,γ1).<br />
The projection onto the contact planes along the Reeb vector field is then given by<br />
πλ(v1,v2,v3) = 1<br />
µ (v1γ ′ ′<br />
1 + v3γ 2 ) η2 + v2 η1, with µ = γ1γ ′ ′<br />
2 − γ 1γ2, and the flow of the Reeb vector field is given by<br />
where<br />
φt(θ,r,φ) = θ + α(r)t,r,φ + β(r)t ,<br />
α(r) = γ ′ 2 (r)<br />
µ(r)<br />
and β(r) =− γ ′ 1 (r)<br />
µ(r) .<br />
The linearization of the flow Tφτ (θ,r,φ) preserves the contact structure. In the basis<br />
{η1,η2} it is given by<br />
<br />
1 0<br />
Tφτ (θ,r,φ) =<br />
with A(r) =<br />
τA(r) 1<br />
1<br />
µ 2 ′′ ′ ′′ ′<br />
γ 2 (r)γ 1 (r) − γ 1 (r)γ 2<br />
(r)<br />
(r) .
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 79<br />
The complex structure(s) we chose earlier in (2.7) had the following form near the<br />
binding with respect to the basis {η1,η2}:<br />
<br />
<br />
J (θ,r,φ) =<br />
0 −rγ1(r)<br />
0<br />
.<br />
1<br />
rγ1(r)<br />
The induced complex structure jτ on the surface is then given by<br />
jτ (z) = [πλTv0(z)] −1 <br />
[Tφτ v0(z) ] −1 J φτ (v0(z)) <br />
Tφτ v0(z) πλTv0(z).<br />
With v0(s, t) = (t,r(s),α),wefindthat<br />
so that<br />
and<br />
jτ =<br />
<br />
πλTv0(s, t) =<br />
r ′ (s) 0<br />
0<br />
γ ′ 1 (s)<br />
µ(r(s))<br />
−τA(r)rγ1(r) − rγ1(r)γ ′ 1 (r)<br />
r ′ µ(r)<br />
r ′ µ(r)<br />
rγ1(r)γ ′ 1 (r)(1 + τ 2 A 2 (r)r 2 γ 2 1<br />
(r)) τA(r)rγ1(r)<br />
<br />
−τA(r)rγ1(r) −1<br />
=<br />
1 + τ 2A2 (r)r2γ 2 <br />
1 (r) τA(r)rγ1(r)<br />
<br />
<br />
−1 0<br />
= j0 + τA(r) γ1(r)<br />
τA(r)γ1(r) 1<br />
<br />
jτ − jσ = A(r)rγ1(r)(τ − σ )<br />
using the fact that r(s) satisfies the differential equation<br />
With v ∗ 0 λ = γ1(r) dt, we obtain<br />
r ′ (s) = γ ′ 1 (r(s))γ1(r(s))r(s)<br />
.<br />
µ(r(s))<br />
<br />
<br />
−1 0<br />
(τ + σ )A(r)rγ1(r) 1<br />
v ∗<br />
0λ ◦ (jτ − jσ )|(s,t) = (τ − σ )A r(s) r(s)γ 2<br />
<br />
1 r(s)<br />
· (τ + σ )A r(s) <br />
r(s)γ1 r(s) ds + dt .<br />
<br />
,
80 CASIM ABBAS<br />
Converting from coordinates (s, t) on the half-cylinder to Cartesian coordinates x +<br />
iy = e −(s+it) in the complex plane, we get, with ρ = x 2 + y 2 ,<br />
ds =− 1<br />
(xdx+ ydy) and dt =−1 (xdy− ydx).<br />
ρ2 ρ2 Recall that r(s) = c(s)e κs , where c(s) is a smooth function which converges to a<br />
constant as s →+∞, and we assumed that κ ≤−1/2 and that κ /∈ Z. Another<br />
assumption was that A(r) = O(r). Hence r(s) is close to ρ −κ if s is large (and ρ is<br />
small) and A(r(s)) = O(ρ −κ ). Also recall that γ1(r(s)) = O(1). Summarizing, we<br />
need the expression<br />
A r(s) r(s)ρ −2 ρ = O(ρ −2κ−1 )<br />
to be bounded, which amounts to κ ≤−1/2. The same argument applied to the form<br />
dg ◦ (jτ − jσ ) leads to the same conclusion.<br />
Acknowledgments. I am very grateful to Richard Siefring and Chris Wendl for explaining<br />
some of their work to me. Their results are indispensable for the arguments<br />
in this article. I would also like to thank Samuel Lisi for the numerous discussions we<br />
had about the subject of this article.<br />
References<br />
[1] C. ABBAS, Pseudoholomorphic strips in symplectizations, II: Fredholm theory and<br />
transversality, Comm. Pure Appl. Math. 57 (2004), 1 – 58. MR 2007355 58<br />
[2] ———, Pseudoholomorphic strips in symplectizations, III: Embedding properties and<br />
compactness, J. Symplectic Geom. 2 (2004), 219 – 260. MR 2108375<br />
[3] ———, Introduction to compactness results in symplectic field theory, in preparation.<br />
32<br />
[4] C. ABBAS, K. CIELIEBAK,andH. HOFER, The Weinstein conjecture for planar contact<br />
structures in dimension three, Comment. Math. Helv. 80 (2005), 771 – 793.<br />
MR 2182700 29, 31, 36<br />
[5] C. ABBAS and H. HOFER, Holomorphic curves and global questions in contact<br />
geometry, in preparation. 56<br />
[6] C. ABBAS, H. HOFER,andS. LISI, Renormalization and energy quantization in Reeb<br />
dynamics, in preparation. 29, 36<br />
[7] ———, Some applications of a homological perturbed Cauchy-Riemann equation,<br />
in preparation. 29, 36<br />
[8] L. V. AHLFORS, Lectures on Quasiconformal Mappings, Van Nostrand Math. Stud. 10,<br />
D. Van Nostrand, Toronto, 1966. 59, 60<br />
[9] L. V. AHLFORS and L. BERS, Riemann’s mapping theorem for variable metrics, Ann. of<br />
Math. (2) 72 (1960), 385 – 404. MR 0115006 59, 60, 61, 62, 75
HOLOMORPHIC OPEN BOOK DECOMPOSITIONS 81<br />
[10] J. W. ALEXANDER, A lemma on systems of knotted curves, Proc. Natl. Acad. Sci. USA<br />
9 (1923), 93 – 95. 30<br />
[11] L. BERS, Riemann Surfaces, lectures, New York University, 1957 – 1958.<br />
[12] L. BERS and L. NIRENBERG, “On a representation theorem for linear elliptic systems<br />
with discontinuous coefficients and its applications” in Convegno internazionale<br />
sulle equazioni lineari alle derivate parziali, Trieste (1954), 111 – 140,<br />
Cremonese, Rome, 1955. MR 0076981 60, 63, 64, 65, 66 60<br />
[13] S.-S. CHERN, An elementary proof of the existence of isothermal parameters on a<br />
surface, Proc. Amer. Math. Soc. 6 (1955), 771 – 782. MR 0074856 60, 63, 64,<br />
65<br />
[14] J. B. ETNYRE, Planar open book decompositions and contact structures,Int.Math.<br />
Res. Not. IMRN 2004, no. 79, 4255 – 4267. MR 2126827 31<br />
[15] O. FORSTER, Lectures on Riemann Surfaces, Grad. Texts in Math. 81, Springer, New<br />
York, 1981. MR 0648106 51<br />
[16] E. GIROUX,“Géométrie de contact: De la dimension trois vers les dimensions<br />
supérieures” in Proceedings of the International Congress of Mathematicians,<br />
Vol. II (Beijing, 2002), Higher Ed. Press, Beijing, 2002, 405 – 414. MR 1957051<br />
30, 31<br />
[17] H. HOFER, Pseudoholomorphic curves in symplectizations with applications to the<br />
Weinstein conjecture in dimension three, Invent. Math. 114 (1993), 515 – 563.<br />
MR 1244912 32<br />
[18] ———, “Holomorphic curves and real three-dimensional dynamics” in GAFA 2000<br />
(Tel Aviv, 1999), Geom. Funct. Anal. 2000, Special Volume, Part II, 674 – 704.<br />
MR 1826268<br />
[19] H. HOFER, K. WYSOCKI,andE. ZEHNDER, Properties of pseudoholomorphic curves in<br />
symplectizations, II: Embedding controls and algebraic invariants, Geom. Funct.<br />
Anal. 5 (1995), 270 – 328. MR 1334869<br />
[20] ———, The dynamics on three-dimensional strictly convex energy surfaces,<br />
Ann. of Math. (2) 148 (1998), 197 – 289. MR 1652928 35<br />
[21] ———, Properties of pseudoholomorphic curves in symplectizations,. III: Fredholm<br />
theory, Progr. Nonlinear Differential Equations Appl. 35,Birkhäuser, Basel,<br />
1999, 381 – 475. MR 1725579 58<br />
[22] ———, Finite energy foliations of tight three-spheres and Hamiltonian dynamics,<br />
Ann. of Math. (2) 157 (2003), 125 – 255. MR 1954266 35<br />
[23] H. HOFER and E. ZEHNDER, Symplectic Invariants and Hamiltonian Dynamics,<br />
Birkhäuser Advanced Texts: Basler Lehrbücher, Birkhäuser, Basel, 1994.<br />
MR 1306732 52<br />
[24] C. HUMMEL, Gromov’s Compactness Theorem for Pseudoholomorphic Curves, Prog.<br />
Math. 151,Birkhäuser, Basel, 1997. MR 1451624<br />
[25] P. D. LAX, Functional Analysis, John Wiley, New York, 2001. MR 1892228 61<br />
[26] L. LICHTENSTE<strong>IN</strong>, Zur Theorie der konformen Abbildung nichtanalytischer,<br />
singularitätenfreier Flächenstücke auf ebene Gebiete, J. Krak. Auz. (1916),<br />
192 – 217. 60
82 CASIM ABBAS<br />
[27] D. MCDUFF and D. SALAMON, J-holomorphic Curves and Symplectic Topology,Amer.<br />
Math. Soc. Colloq. Publ. 52, Amer. Math. Soc., Providence, 2004. MR 2045629<br />
56<br />
[28] M. J. MICALLEF and B. WHITE, The structure of branch points in minimal surfaces and<br />
in pseudoholomorphic curves, Ann. of Math. (2) 141 (1995), 35 – 85.<br />
MR 1314031 56<br />
[29] C. B. MORREY, On the solutions of quasi-linear elliptic partial differential equations,<br />
Trans. Amer. Math. Soc. 43, no. 1 (1938), 126 – 166. MR 1501936 60<br />
[30] D. ROLFSEN, Knots and Links, Mathematics Lecture Series 7, Publish or Perish,<br />
Berkeley, Cali., 1976. MR 0515288 30<br />
[31] R. SIEFR<strong>IN</strong>G, Relative asymptotic behavior of pseudoholomorphic half-cylinders,<br />
Comm. Pure Appl. Math. 61 (2008), 1631 – 1684. MR 2456182 52, 53, 58<br />
[32] E. M. STE<strong>IN</strong>, Singular Integrals and Differentiability Properties of Functions, Princeton<br />
Math. Ser. 30, Princeton Univ. Press, Princeton, 1970. MR 0290095 61<br />
[33] C. H. TAUBES, The Seiberg-Witten equations and the Weinstein conjecture, Geom.<br />
Topol. 11 (2007), 2117 – 2202. MR 2350473 29, 36<br />
[34] ———, The Seiberg-Witten equations and the Weinstein conjecture, II: More closed<br />
integral curves of the Reeb vector field, Geom. Topol. 13 (2009), 1337 – 1417.<br />
MR 2496048 29, 36<br />
[35] W. P. THURSTON and H. E. W<strong>IN</strong>KELNKEMPER, On the existence of contact forms,Proc.<br />
Amer. Math. Soc. 52 (1975), 345 – 347. MR 0375366 40<br />
[36] A. J. TROMBA, Teichmüller Theory in Riemannian Geometry, Lectures Math. ETH<br />
Zürich, Birkhäuser, Basel, 1992. MR 1164870<br />
[37] A. WE<strong>IN</strong>STE<strong>IN</strong>, On the hypotheses of Rabinowitz’ periodic orbit theorems,J.<br />
Differential Equations 33 (1979), 353 – 358. MR 0543704<br />
[38] C. WENDL, Finite energy foliations on overtwisted contact manifolds, Geom. Topol. 12<br />
(2008) 531 – 616. MR 2390353 34, 35, 36<br />
[39] ———, Open book decompositions and stable Hamiltonian structures, Expo. Math.<br />
28 (2010), 187 – 197. MR 2671115 34, 36, 40<br />
[40] ———, Strongly fillable contact manifolds and J-holomorphic foliations, Duke Math.<br />
J. 151 (2010), 337 – 387. MR 2605865 35<br />
Department of Mathematics, Michigan State University, East Lansing, Michigan 48824, USA;<br />
abbas@math.msu.edu
CALDERÓN <strong>IN</strong>VERSE PROBLEM WITH PARTIAL<br />
DATA ON RIEMANN SURFACES<br />
COL<strong>IN</strong> GUILLARMOU and LEO TZOU<br />
Abstract<br />
On a fixed smooth compact Riemann surface with boundary (M0,g), we show that,<br />
for the Schrödinger operator g + V with potential V ∈ C 1,α (M0) for some α>0,<br />
the Dirichlet-to-Neumann map N |Ɣ measured on an open set Ɣ ⊂ ∂M0 determines<br />
uniquely the potential V . We also discuss briefly the corresponding consequences for<br />
potential scattering at zero frequency on Riemann surfaces with either asymptotically<br />
Euclidean or asymptotically hyperbolic ends.<br />
Contents<br />
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83<br />
2. Harmonic and holomorphic Morse functions on a Riemann surface . . . 86<br />
3. Carleman estimate for harmonic weights with critical points . . . . . . . 95<br />
4. Complex geometric optics on a Riemann surface . . . . . . . . . . . . . 99<br />
5. Identifying the potential . . . . . . . . . . . . . . . . . . . . . . . . . . 106<br />
6. Inverse scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111<br />
Appendix. Boundary determination . . . . . . . . . . . . . . . . . . . . . . 114<br />
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118<br />
1. Introduction<br />
The problem of determining the potential in the Schrödinger operator by boundary<br />
measurement goes back to Calderón [8]andGelfand[12]. Mathematically, it amounts<br />
to asking if one can detect some data from boundary measurement in a domain (or<br />
manifold) with boundary. The typical model to have in mind is the Schrödinger<br />
operator P := g + V , where g is a metric and V is a potential; then we define the<br />
Cauchy data space by<br />
C := (u|∂,∂νu|∂); u ∈ H 1 (), u∈ ker P ,<br />
DUKE MATHEMATICAL JOURNAL<br />
Vol. 158, No. 1, c○ 2011 DOI 10.1215/00127094-1276310<br />
Received 17 September 2009. Revision received 22 July 2010.<br />
2010 Mathematics Subject Classification. Primary 35R30; Secondary 58J32.<br />
Guillarmou’s work partially supported by National Science Foundation grant DMS-0635607.<br />
Tzou’s work partially supported by National Science Foundation grant DMS-0807502.<br />
83
84 GUILLARMOU and TZOU<br />
where ∂ν is the interior pointing normal vector field to ∂ and where H 1 () =<br />
W 1,2 () is the space of L2 functions with one derivative in L2 .<br />
The first natural question is the following full data inverse problem: does the<br />
Cauchy data space determine uniquely the metric g and/or the potential V ?Ina<br />
sense, the most satisfying known results are those obtained when the domain ⊂ Rn is already known and g is the Euclidean metric, for which the identification of V<br />
has been proved in dimension n>2 by Sylvester and Uhlmann [32] and Novikov<br />
[30] (where reconstruction is also obtained), and very recently in dimension 2 by<br />
Bukgheim [6] when the domain is simply connected. In a recent work [16], we<br />
proved the identification of V on a Riemann surface with boundary from full data<br />
measurement. A related question is the conductivity problem which consists in taking<br />
V = 0 and replacing g by −divσ ∇, where σ is a positive definite symmetric tensor.<br />
An elementary observation shows that the problem of recovering a sufficiently smooth<br />
isotropic conductivity (i.e., σ = σ0Id for a function σ0) is contained in the abovementioned<br />
problem of recovering a potential V . For domain of R2 , Nachman [29]<br />
used the ¯∂ techniques to show that the Cauchy data space determines the conductivity.<br />
Recently a new approach developed by Astala and Päivärinta in [2] improved this<br />
result to assuming that the conductivity is only an L∞ scalar function on a simply<br />
connected domain. This was later generalized to L∞ anisotropic conductivities by<br />
Astala, Lassas, and Päivärinta in [3]. We notice that there still are rather few results<br />
in the direction of recovering the Riemannian manifold (,g) when V = 0, for<br />
instance the surface case by Lassas and Uhlmann [24] (seealso[4], [17]), the real<br />
analytic manifold case by Lassas, Taylor, and Uhlmann [23] (seealso[15] forthe<br />
Einstein case), the case of manifolds admitting limiting Carleman weights and in a<br />
same conformal class by Dos Santos Ferreira, Kenig, Salo, and Uhlmann [9]. The<br />
second natural, but harder, problem is the partial data inverse problem: if Ɣ1 and Ɣ2<br />
are open subsets of ∂, does the partial Cauchy data space for P<br />
CƔ1,Ɣ2 := (u|Ɣ1,∂νu|Ɣ2); u ∈ H 1 <br />
(∂M0), Pu= 0, u= 0in∂ \ Ɣ1<br />
determine the domain , the metric, the potential? For a fixed domain of R n ,the<br />
recovery of the potential if n>2 with partial data measurements was initiated by<br />
Bukhgeim and Uhlmann [7] and later improved by Kenig, Sjöstrand, and Uhlmann<br />
[21] to the case where Ɣ1 and Ɣ2 are respectively open subsets of the “front” and<br />
“back” ends of the domain. We refer the reader to the references for a more precise<br />
formulation of the problem. In dimension 2, the recent works of Imanuvilov, Uhlmann,<br />
and Yamamoto [19] solves the problem for fixed domains of R 2 in the case when<br />
Ɣ1 = Ɣ2 and when the potential are in C 2,α () for some α>0. In this article, we<br />
address the same question when the background domain is a fixed Riemann surface<br />
with boundary. We prove the following recovery result.
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 85<br />
<strong>THEOREM</strong> 1.1<br />
Let (M0,g) be a smooth compact Riemann surface with boundary, let Ɣ ⊂ ∂M0 be<br />
an open subset of the boundary, and let g be the nonnegative Laplacian on M0.For<br />
α ∈ (0, 1),letV1,V2 ∈ C1,α (M0) be two potentials and, for i = 1, 2,let<br />
C Ɣ<br />
i =: (u|Ɣ,∂νu|Ɣ); u ∈ H 1 (M0), (g + Vi)u = 0, u = 0on∂M0 \ Ɣ <br />
(1)<br />
be the respective Cauchy partial data spaces. If C Ɣ 1 = C Ɣ 2 ,thenV1 = V2.<br />
Here the space C1,α (M0) is the usual Hölder space for α ∈ (0, 1). Notice that when<br />
g + Vi do not have L2 eigenvalues for the Dirichlet condition, the statement above<br />
can be given in terms of Dirichlet-to-Neumann operators. Since ˆg = e−2ϕg when<br />
ˆg = e2ϕg for some function ϕ, it is clear that in the statement in Theorem 1.1, we<br />
need only to fix the conformal class of g instead of the metric g (or equivalently,<br />
we need only to fix the complex structure on M0). In particular, the smoothness<br />
assumption of the Riemann surface with boundary is not really essential since we<br />
can change it conformally to make it smooth; for the Cauchy data space, this just<br />
has the effect of changing the potential conformally (all we need is that this new<br />
potential be C1,α ). Observe also that Theorem 1.1 implies that, for a fixed Riemann<br />
surface with boundary (M0,g), the Dirichlet-to-Neumann map on Ɣ for the operator<br />
u →−divg(γ ∇gu) determines the isotropic conductivity γ if γ ∈ C3,α (M0) in the<br />
sense that two conductivities giving rise to the same Dirichlet-to-Neumann map are<br />
equal. This is a standard observation by transforming the conductivity problem to a<br />
potential problem with potential V := (gγ 1/2 )/γ 1/2 . So our result also extends that<br />
of Henkin and Michel [17] in the case of isotropic conductivities. Notice also that<br />
reconstruction methods for isotropic conductivities are obtained in a recent work of<br />
Henkin and Novikov [18] for full boundary data measurements on Riemann surfaces.<br />
The method to identify the potential follows [6] and[19] and is based on the<br />
construction of a large set of special complex geometric optic solutions of (g +<br />
V )u = 0, also called Faddeev-type solutions and introduced first by Faddeev in [10].<br />
More precisely, if Ɣ0 = ∂M0 \ Ɣ is the set where we do not know the Dirichletto-Neumann<br />
operator, then we construct solutions of the form u = Re e/h (a +<br />
r(h)) + eRe()/hs(h) with u|Ɣ0 = 0, where h>0 is a small parameter, and a<br />
are holomorphic functions on (M0,g) independent of h,and||r(h)||L2 = O(h) while<br />
||s(h)||L2 = O(h3/2 | log h|) as h → 0. The idea of [6] to reconstruct V (p) for p ∈ M0<br />
is to take with a nondegenerate critical point at p and then use stationary phase as<br />
h → 0. In our setting, the function needs to be purely real on Ɣ0 and Morse with a<br />
prescribed critical point at p. One of our main contributions is a geometric construction<br />
of the holomorphic Carleman weights satisfying such conditions. We should point<br />
out that we use a method quite different from that in [19] to construct this weight,<br />
and we believe that our method simplifies their construction even in their case. A
86 GUILLARMOU and TZOU<br />
Carleman estimate on the surface for this degenerate weight needs to be proved, and<br />
we follow the ideas of [19]. We manage to improve the regularity of the potential to<br />
C1,α instead of C2,α in [19]. Since it did not seem to be available in the literature in<br />
our geometric setting, we also provide a short proof in the appendix of the fact that the<br />
partial Cauchy data space C Ɣ determines a potential V ∈ C0,α (M0) on Ɣ (for α>0).<br />
An interesting fact about this proof is that it uses our Carleman estimate, and it allows<br />
rather weak assumptions on the regularity of V .<br />
In Section 6, we obtain two inverse scattering results as a corollary of Theorem<br />
1.1, first for partial data scattering at zero frequency for + V on asymptotically<br />
hyperbolic surfaces with potential decaying at the conformal infinity, and second for<br />
full data scattering at zero frequency for + V with V compactly supported on an<br />
asymptotically Euclidean surface.<br />
Another straightforward corollary in the asymptotically Euclidean case full data<br />
setting is the recovery of a compactly supported potential from the scattering operator<br />
at a positive frequency. The proof is essentially the same as for the operator Rn + V<br />
once we know Theorem 1.1, so we omit it.<br />
2. Harmonic and holomorphic Morse functions on a Riemann surface<br />
2.1. Riemann surfaces<br />
We start by recalling a few elementary definitions and results about Riemann surfaces<br />
(see, e.g., [11] for more details). Let (M0,g0) be a compact, oriented, connected,<br />
smooth Riemannian surface with boundary ∂M0. The surface M0 can be considered<br />
as a subset of a larger oriented Riemannian surface with boundary (M,g), extending<br />
smoothly the metric g0 to M. The conformal class of g and the orientation on<br />
the surface M induce a structure of Riemann surface with boundary (i.e., a surface<br />
equipped with a complex structure via holomorphic charts zα : Uα → C). The Hodge<br />
star operator ⋆ (well defined since M is oriented) acts on the cotangent bundle T ∗M, its eigenvalues are ±i, and the respective eigenspaces T ∗<br />
1,0<br />
M := ker(⋆ + iId) and<br />
T ∗<br />
0,1 M := ker(⋆ − iId) are subbundles of the complexified cotangent bundle CT ∗ M,<br />
and the splitting CT ∗M = T ∗<br />
∗<br />
1,0M ⊕ T0,1M holds as complex vector spaces. Since ⋆<br />
is conformally invariant on 1-forms on M, the complex structure depends only on the<br />
conformal class of g. In holomorphic coordinates z = x + iy in a chart Uα, one has<br />
⋆(udx + vdy) =−vdx + udy and<br />
T ∗<br />
1,0 M|Uα C dz, T ∗<br />
0,1 M|Uα C d ¯z,<br />
where dz = dx + idy and d ¯z = dx − idy. We define the natural projections induced<br />
by the splitting of CT ∗ M as<br />
π1,0 : CT ∗ M → T ∗<br />
1,0 M, π0,1 : CT ∗ M → T ∗<br />
0,1 M.
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 87<br />
The exterior derivative d defines the de Rham complex 0 → 0 → 1 → 2 → 0,<br />
where k := kT ∗M denotes the real bundle of k-forms on M. Let us denote Ck the complexification of k . Then the ∂ and ¯∂ operators can be defined as differential<br />
operators ∂ : C 0 → T ∗<br />
1,0 M and ¯∂ : C0 → T ∗<br />
0,1<br />
M by<br />
∂f := π1,0 df, ¯∂ := π0,1 df, (2)<br />
where they satisfy d = ∂ + ¯∂ and where they are expressed in holomorphic coordinates<br />
by<br />
∂f = ∂zf dz,<br />
¯∂f = ∂¯zf d¯z,<br />
with ∂z := (∂x − i∂y)/2 and ∂¯z := (∂x + i∂y)/2. Similarly, one can define the ∂ and<br />
¯∂ operators from C 1 to C 2 by setting<br />
∂(ω1,0 + ω0,1) := dω0,1,<br />
¯∂(ω1,0 + ω0,1) := dω1,0<br />
if ω0,1 ∈ T ∗<br />
0,1M and ω1,0 ∈ T ∗<br />
1,0M. In coordinates, this is simply<br />
∂(udz+ vd¯z) = ∂v ∧ d ¯z,<br />
¯∂(udz+ vd¯z) = ¯∂u ∧ dz.<br />
There is a natural operator, the Laplacian acting on functions and defined by<br />
f := −2i ⋆¯∂∂f = d ∗ d,<br />
where d ∗ is the adjoint of d through the metric g and where ⋆ is the Hodge star operator<br />
mapping 2 to 0 and induced by g as well.<br />
2.2. Maslov index and boundary-value problem for the ∂ operator<br />
In this section, we consider the setting where M0 is an oriented Riemann surface with<br />
boundary ∂M0 and Ɣ ⊂ ∂M0 is an open subset, and we let Ɣ0 = ∂M0 \ Ɣ be its<br />
complement in ∂M0. We can identify the boundary with a disjoint union of circles<br />
∂M0 = m i=1 ∂iM0, where each ∂iM0 S1 .SinceƔwill be the piece of the boundary<br />
where we know the Cauchy data space, it is sufficient to assume that Ɣ is a connected<br />
nonempty open segment of ∂1M0 = S1 , which we now do. Following [26], we adopt<br />
the following notation: let E → M0 be a complex line bundle with complex structure<br />
J : E → E, andletD : C∞ (M0,E) → C∞ (M0,T∗ 0,1 ⊗ E) be a Cauchy-Riemann<br />
operator with smooth coefficients on M0 acting on sections of the bundle E. Observe<br />
that in the case when E = M0 × C is the trivial line bundle with the natural complex<br />
structure on M0,thenDcan be taken to be the operator ∂ introduced in (2). For q>1,<br />
we define<br />
DF : W ℓ,q<br />
F (M0,E) → W ℓ−1,q (M0,T ∗<br />
0,1 M0 ⊗ E),
88 GUILLARMOU and TZOU<br />
where F ⊂ E |∂M0 is a totally real subbundle (i.e., a subbundle such that JF ∩ F is<br />
the zero section) and where DF is the restriction of D to the Lq-based Sobolev space<br />
with ℓ derivatives and boundary condition F<br />
W ℓ,q<br />
F (M0,E):= ξ ∈ W ℓ,q (M0,E) ξ(∂M0) ⊂ F .<br />
When q = 2, we will use the standard notation H ℓ (M0) := W ℓ,2 (M0) for L 2 -based<br />
Sobolev spaces. The boundary Maslov index for a totally real subbundle F ⊂ E∂M0<br />
of a complex vector bundle is defined in generality in [26, Appendix C.3]; we only<br />
recall the definition for our setting.<br />
Definition 2.2.1<br />
Let E = M0 × C and ∂M0 = m j=1 ∂iM0 be a disjoint union of m circles. The<br />
boundary Maslov index µ(E,F) is the degree of the map ρ ◦ : ∂M0 → ∂M0,<br />
where<br />
|∂iM0 : S 1 ∂iM0 → GL(1, C)/GL(1, R)<br />
is the natural map assigning to z ∈ S 1 the totally real subspace Fz ⊂ C, where<br />
GL(1, C)/GL(1, R) is the space of totally real subbundles of C, and where ρ :<br />
GL(1, C)/GL(1, R) → S 1 is defined by ρ(A.GL(1, R)) := A 2 /|A| 2 .<br />
In this setting, we have the following boundary-value Riemann-Roch theorem stated<br />
in [26, Theorem C.1.10].<br />
<strong>THEOREM</strong> 2.2.2<br />
Let E → M0 be a complex line bundle over an oriented compact Riemann surface<br />
with boundary, and let F ⊂ E |∂M0 be a totally real subbundle. Let D be a smooth<br />
Cauchy-Riemann operator on E acting on W ℓ,q (M0,E) for some q ∈ (1, ∞) and<br />
ℓ ∈ N. Then we have the following.<br />
(1) The following operators are Fredholm:<br />
DF : W ℓ,q<br />
F (M0,E) → W ℓ−1,q (M0,T ∗<br />
0,1 M0 ⊗ E),<br />
D ∗<br />
F<br />
ℓ,q ∗<br />
: WF (M0,T0,1M0 ⊗ E) → W ℓ−1,q (M0,E).<br />
(2) The real Fredholm index of DF is given by<br />
Ind(DF ) = χ(M0) + µ(E,F),<br />
where χ(M0) is the Euler characteristic of M0 and where µ(E,F) is the<br />
boundary Maslov index of the subbundle F .
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 89<br />
(3) If µ(E,F) < 0, thenDF is injective, while if µ(E,F) + 2χ(M0) > 0, the<br />
operator DF is surjective.<br />
As an application, we obtain the following (here and in what follows, H m (M0) :=<br />
W m,2 (M0)).<br />
COROLLARY 2.2.3<br />
(i) For q > 1 and k ∈ N0, letω∈ W k,q (M0,T∗ 0,1M0). Then there exists u ∈<br />
W k+1,q (M0) real-valued on Ɣ0 such that ¯∂u = ω.<br />
(ii) For m>1/2, letf∈ H m (∂M0) be a real-valued function. Then there exists<br />
a holomorphic function v ∈ H m+(1/2) (M0) such that Re(v)|Ɣ0 = f .<br />
(iii) For k ∈ N and q>1, the space of W k,q (M0) holomorphic functions on M0<br />
which are real-valued on Ɣ0 is infinite-dimensional.<br />
Proof<br />
(i) Let L ∈ N be arbitrarily large, and let Ɣ be a connected nonempty open segment<br />
of one connected component ∂1M0 = S 1 of ∂M0. ThenƔ can be defined in a<br />
coordinate θ (respecting the orientation of the boundary) by Ɣ ={θ ∈ S 1 |<br />
0 0.SinceL can be taken as<br />
large as we want, this achieves the proof of (i).<br />
(ii) Let w ∈ H m+(1/2) (M0) be a real function with boundary value f on ∂M0.Then<br />
by (i), there exists R ∈ H m+(1/2) (M0) such that i ¯∂R =−¯∂w and R purely real<br />
on Ɣ0; thus v := iR + w is holomorphic such that Re(v) = f on Ɣ0.<br />
(iii) Taking the subbundle F as in the proof of (i), we have that dim ker DF =<br />
χ(M0) + 2L if L satisfies 2χ(M0) + 2L >0, and since L can be taken as large<br />
as we like, this concludes the proof.
90 GUILLARMOU and TZOU<br />
LEMMA 2.2.4<br />
Let {p0,p1,...,pn} ⊂M0 be a set of n + 1 disjoint points. Let c1,...,cK ∈ C, let<br />
N ∈ N, and let z be a complex coordinate near p0 such that p0 ={z = 0}. Thenif<br />
p0 ∈ int(M0), there exists a holomorphic function f on M0 with zeros of order at least<br />
N at each pj such that f is real on Ɣ0 and f (z) = c0 +c1z+···+cKz K +O(|z| K+1 )<br />
in the coordinate z.Ifp0 ∈ ∂M0, then the same is true except that f is not necessarily<br />
real on Ɣ0.<br />
Proof<br />
First, using linear combinations and induction on K, it suffices to prove the lemma for<br />
any K and c0 =···=cK−1 = 0, which we now show. Consider the subbundle F as<br />
in the proof of (i) in Corollary 2.2.3. The Maslov index µ(E,F) is given by 2L, and so<br />
for each N ∈ N, one can take L large enough to have µ(F,E)+2χ(M0) ≥ 2N(1+n).<br />
Therefore, by Theorem 2.2.2, the dimension of the kernel of ∂F will be greater than<br />
2(n + 1)N. Now, since for each pj and complex coordinate zj near pj the map<br />
u → (u(pj),∂zju(pj),...,∂N−1 zj u(pj)) ∈ CN is linear, this implies that there exists<br />
a nonzero element u ∈ ker DF which has zeros of order at least N at all pj.<br />
First, assume that p0 ∈ int(M0) and that we want the desired Taylor expansion<br />
at p0 in the coordinate z. In the coordinate z, one has u(z) = αzM + O(|z| M+1 ) for<br />
some α = 0 and M ≥ N. Define the function rK(z) = χ(z) cK<br />
α z−M+K , where χ(z) is<br />
a smooth cutoff function supported near p0 and which is 1 near p0 ={z = 0}. Since<br />
M ≥ N>1, this function has a pole at p0 and trivially extends smoothly to M0\{p0},<br />
which we still call rK. Observe that the function is holomorphic in a neighborhood<br />
of p0 but not at p0 where it is only meromorphic, so that in M0 \{p0}, ∂rK is a<br />
smooth and compactly supported section of T ∗<br />
0,1 M0, and therefore trivially extends<br />
smoothly to M0 (by setting its value to be zero p0) to a 1-form denoted ωK. Bythe<br />
surjectivity assertion in Corollary 2.2.3, there exists a smooth function RK satisfying<br />
∂RK =−ωK,andRK|Ɣ0 ∈ R.WenowhavethatRK + rK is a holomorphic function<br />
on M\{p0} meromorphic with a pole of order M − K at p0, and in coordinate z one<br />
has z M−K (RK(z) + rK(z)) = cK + O(|z|). Setting f = u(RK + rK), wehavethe<br />
desired holomorphic function. Note that f also vanishes to order N at all p1,...,pn<br />
since u vanishes. This achieves the proof. <br />
Now, if p0 ∈ ∂M0, we can consider a slightly larger smooth domain of M containing<br />
M0, and we apply the result above.<br />
2.3. Morse holomorphic functions with prescribed critical points<br />
The main result of this section is the following.<br />
PROPOSITION 2.3.1<br />
Let p be an interior point of M0, and let ɛ > 0 be small. Then there exists a<br />
holomorphic function on M0 which is Morse on M0 (up to the boundary) and
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 91<br />
real-valued on Ɣ0, which has a critical point p ′ at distance less than ɛ from p and<br />
such that Im((p ′ )) = 0.<br />
Let O be a connected open set of M such that Ō is a smooth surface with boundary,<br />
with M0 ⊂ Ō ⊂ M and Ɣ0 ⊂ ∂Ō.Fixk>2 a large integer; we denote by Ck ( Ō) the<br />
Banach space of Ck real-valued functions on Ō. Then the set of harmonic functions<br />
on Ō which are in the Banach space Ck ( Ō) (and smooth in O by elliptic regularity)<br />
is the kernel of the continuous map : Ck ( Ō) → Ck−2 ( Ō), and so it is a Banach<br />
subspace of Ck ( Ō). The set H ⊂ Ck ( Ō) of harmonic functions u in Ck ( Ō) such that<br />
there exists v ∈ Ck ( Ō) harmonic with u + iv holomorphic on O is a Banach subspace<br />
of Ck ( Ō) of finite codimension. Indeed, let {γ1,...,γN} be a homology basis for O;<br />
then we have<br />
defined by<br />
H = ker L with L :ker∩C k ( Ō) → CN<br />
L(u) :=<br />
<br />
1<br />
<br />
∂u .<br />
πi γj<br />
j=1,...,N<br />
For all Ɣ ′ 0 ⊂ ∂M0 such that the complement of Ɣ ′ 0<br />
We now show the following.<br />
LEMMA 2.3.2<br />
HƔ ′ 0 :={u ∈ H; u|Ɣ ′ 0<br />
contains an open subset, we define<br />
= 0}.<br />
The set of functions u ∈ HƔ ′ 0<br />
intersection of open dense sets) in HƔ ′ 0 with respect to the Ck (<br />
which are Morse in O is residual (i.e., a countable<br />
Ō) topology.<br />
Proof<br />
We use an argument very similar to that used by Uhlenbeck [34]. We start by defining<br />
m : O × HƔ ′ 0 → T ∗O by (p, u) ↦→ (p, du(p)) ∈ T ∗ p O. This is clearly a smooth map,<br />
linear in the second variable, and moreover mu := m(., u) = (·,du(·)) is Fredholm<br />
since O is finite-dimensional. The map u is a Morse function if and only if mu is<br />
transverse to the zero section, denoted T ∗<br />
0 O,ofT∗O, that is, if<br />
Image(Dpmu) + Tmu(p)(T ∗<br />
0 O) = Tmu(p)(T ∗ O) ∀p ∈ O such that mu(p) = (p, 0),<br />
which is equivalent to the fact that the Hessian of u at critical points is nondegenerate<br />
(see, e.g., [34, Lemma 2.8]). We recall the following transversality theorem.
92 GUILLARMOU and TZOU<br />
<strong>THEOREM</strong> 2.3.3 ([34, Theorem 2])<br />
Let m : X × HƔ ′ 0 → W be a Ck map, where X, HƔ ′ , and W are separable Banach<br />
0<br />
manifolds with W and X of finite dimension. Let W ′ ⊂ W be a submanifold such<br />
that k>max(1, dim X − dim W + dim W ′ ).Ifmis transverse to W ′ , then the set<br />
{u ∈ HƔ ′ 0 ; mu is transverse to W ′ } is dense in HƔ ′ , and more precisely, it is a residual<br />
0<br />
set.<br />
We want to apply it with X := O, W := T ∗O,andW ′ := T ∗<br />
0 O, and the map m is<br />
defined above. We have thus proved Lemma 2.3.2 if one can show that m is transverse<br />
to W ′ .Let(p, u) be such that m(p, u) = (p, 0) ∈ W ′ . Then, identifying T(p,0)(T ∗O) with TpO ⊕ T ∗ p O, one has<br />
D(p,u)m(z, v) = z, dv(p) + Hessp(u)z ,<br />
where Hesspu is the Hessian of u at the point p, viewed as a linear map from<br />
TpO to T ∗ p O. To prove that m is transverse to W ′ , we need to show that (z, v) →<br />
(z, dv(p) + Hessp(u)z) is surjective from TpO ⊕ HƔ ′ 0 to TpO ⊕ T ∗ p O,whichis<br />
realized, for instance, if the map v → dv(p) from HƔ ′ 0 to T ∗ p O is surjective. But from<br />
Lemma 2.2.4, we know that there exist holomorphic functions v and ˜v on O such that<br />
v and ˜v are purely real on Ɣ ′ 0 . Clearly, the imaginary parts of v and ˜v belong to HƔ ′ 0 .<br />
Furthermore, for a given complex coordinate z near p ={z = 0}, we can arrange<br />
them to have series expansion v(z) = z + O(|z| 2 ) and ˜v(z) = iz + O(|z| 2 ) around<br />
the point p. We see, by coordinate computation of the exterior derivative of Im(v) and<br />
Im(v ), thatdIm(v)(p) and d Im(˜v)(p) are linearly independent at the point p. This<br />
shows our claim and ends the proof of Lemma 2.3.2 by using Theorem 2.3.3. <br />
We now proceed to show that the set of all functions u ∈ HƔ ′ 0<br />
degenerate critical points on Ɣ ′ 0 is also residual.<br />
LEMMA 2.3.4<br />
For all p ∈ Ɣ ′ 0 and k ∈ N, there exists a holomorphic function u ∈ Ck (<br />
Im(u)|Ɣ ′ 0<br />
= 0 and ∂u(p) = 0.<br />
such that u has no<br />
Ō) such that<br />
Proof<br />
The proof is quite similar to that of Lemma 2.2.4. By Lemma 2.2.4, we can choose a<br />
holomorphic function v ∈ Ck ( Ō) such that v(p) = 0 and Im(v)|Ɣ ′ = 0; then either<br />
0<br />
∂v(p) = 0 and we are done, or ∂v(p) = 0. Assume now the second case, and let<br />
M ∈ N be the order of p as a zero of v. By the Riemann mapping theorem, there is<br />
a conformal mapping from a neighborhood Up of p in Ō to a neighborhood {|z| <<br />
ɛ, Im(z) ≥ 0} of the real line Im(z) = 0 in C, and one can assume that p ={z = 0}<br />
in these complex coordinates. Take r(z) = χ(z)z−M+1 , where χ ∈ C∞ 0 (|z| ≤ɛ) is a
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 93<br />
real-valued function with χ(z) = 1 in {|z|
94 GUILLARMOU and TZOU<br />
At a point (p, u) such that b(p, u) = 0, a simple computation yields that the differential<br />
D(p,u)b : TpƔ ′ 0 × HƔ ′ 0 → T(p,∂νu(p))(N ∗ Ɣ ′ 0 ) is given by (w, u′ ) ↦→ (w, ∂τ ∂νu(p)w +<br />
∂νu ′ (p)). This observation combined with Lemma 2.3.4 shows that, for all (p, u) ∈<br />
Ɣ ′ 0 × HƔ ′ 0 such that b(p, u) = (p, 0), b is transverse to N ∗ 0 Ɣ′ 0 at (p, 0). Now we can<br />
apply Theorem 2.3.3; with X = Ɣ ′ 0 , W = N ∗Ɣ ′ 0 ,andW ′ = N ∗ 0 Ɣ′ 0 , we see that the set<br />
{u ∈ HƔ ′ 0 ; bu is transverse to N ∗ 0 Ɣ′ 0 } is residual in HƔ ′ . In view of Lemma 2.3.5, we<br />
0<br />
deduce the following.<br />
LEMMA 2.3.7<br />
The set of functions u ∈ HƔ ′ 0 such that u has no degenerate critical point on Ɣ′ 0 is<br />
residual in HƔ ′ 0 .<br />
Observing the general fact that finite intersection of residual sets remains residual, the<br />
combination of Lemmas 2.3.7 and 2.3.2 yields this corollary.<br />
COROLLARY 2.3.8<br />
The set of functions u ∈ HƔ ′ 0<br />
points on Ɣ ′ 0 is residual in HƔ ′ 0 with respect to the Ck (<br />
dense.<br />
which are Morse in O and have no degenerate critical<br />
Ō) topology. In particular, it is<br />
We are now in a position to give a proof of the main proposition of this section.<br />
Proof of Proposition 2.1<br />
As explained above, choose O in such a way that Ō is a smooth surface with boundary,<br />
containing M0, sothatƔ0⊂∂O andsothatOcontains ∂M0\Ɣ0. LetƔ ′ 0 be an open<br />
and such<br />
subset of the boundary of Ō such that the closure of Ɣ0 is contained in Ɣ ′ 0<br />
that ∂Ō\Ɣ′ 0 =∅.Letpbe an interior point of M0. By Lemma 2.2.4, there exists a<br />
holomorphic function f = u + iv on Ō such that f is purely real on Ɣ′ 0 , v(p) = 1,<br />
and df (p) = 0 (thus v ∈ HƔ ′ 0 ).<br />
By Corollary 2.3.8, there exists a sequence (vj)j of Morse functions vj ∈ HƔ ′ 0<br />
such that vj → v in C k (M0) for any fixed k large. By the Cauchy integral formula,<br />
there exist harmonic conjugates uj of vj such that fj := uj + ivj → f in C k (M0).<br />
Let ɛ>0 be small, and let U ⊂ O be a neighborhood containing p andnoother<br />
critical points of f , and with boundary a smooth circle of radius ɛ. In complex local<br />
coordinates near p, we can identify ∂f and ∂fj to holomorphic functions on an open<br />
set of C. Then by Rouche’s theorem, it is clear that ∂fj has precisely one zero in U<br />
and that vj never vanishes in U if j is large enough. Fix to be one of the fj for j<br />
large enough. By construction, is Morse in O and has no degenerate critical points<br />
on Ɣ0 ⊂ Ɣ ′ 0 . We notice that, since the imaginary part of vanishes on all of Ɣ′ 0 ,itis<br />
clear from the reflection principle applied after using the Riemann mapping theorem
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 95<br />
(as in the proof of Lemma 2.3.5) that no point on Ɣ0 ⊂ Ɣ ′ 0 can be an accumulation<br />
point for critical points. Now ∂M0\Ɣ0 is contained in the interior of O, and therefore<br />
no points on ∂M0\Ɣ0 can be an accumulation point of critical points. Since is Morse<br />
in the interior of O, there are no degenerate critical points on ∂M0\Ɣ0. This ends the<br />
proof. <br />
3. Carleman estimate for harmonic weights with critical points<br />
In this section, we prove a Carleman estimate using harmonic weight with nondegenerate<br />
critical points in a way similar to [19]. Let = ϕ + iψ be a holomorphic Morse<br />
function with no critical points on Ɣ0, and which is purely real on Ɣ0. In particular,<br />
this implies that ∂νϕ|Ɣ0 = 0.<br />
PROPOSITION 3.1<br />
Let (M0,g) be a smooth Riemann surface with boundary, and let ϕ : M0 → R be a<br />
C k (M0) harmonic Morse function for k large. Then for all V ∈ L ∞ (M0) there exist<br />
C>0 and h0 > 0 such that, for all h ∈ (0,h0) and u ∈ C ∞ (M0) with u|∂M0 = 0,we<br />
have<br />
1<br />
h u2<br />
L 2 (M0)<br />
1<br />
+ u|dϕ|2<br />
h2 L2 (M0) +du2 L2 2<br />
(M0) +∂νuL2 (Ɣ0)<br />
<br />
≤ C e −ϕ/h (g + V )e ϕ/h u 2<br />
L2 1 2<br />
(M0) + ∂νuL h 2 <br />
(Ɣ) ,<br />
where ∂ν is the exterior unit normal vector field to ∂M0.<br />
We start by modifying the weight as follows. If ϕ0 := ϕ : M0 → R is a real-valued<br />
harmonic Morse function with critical points {p1,...,pN} in the interior of M0,then<br />
we let ϕj : M0 → R be harmonic functions such that pj is not a critical point of<br />
ϕj for j = 1,...,N, their existence is ensured by Lemma 2.2.4. Forallɛ>0, we<br />
define the convexified weight ϕɛ := ϕ − (h/2ɛ)( N j=0 |ϕj| 2 ). To prove the estimate,<br />
we will localize in charts j covering the surfaces. These charts will be taken so that<br />
if j ∩ ∂M0 = ∅,thenj∩∂M0 S1 is a connected component of ∂M0. Moreover,<br />
by Riemann’s mapping theorem (e.g., [25, Lemma 3.2]), this chart can be taken to be<br />
a neighborhood of |z| =1 in {z ∈ C; |z| ≤1} and such that the metric g is conformal<br />
to the Euclidean metric |dz| 2 .<br />
LEMMA 3.2<br />
Let be a chart of M0 as above, and let ϕɛ : → R be as above. Then there are<br />
constants C, C ′ > 0 such that, for all u ∈ C ∞ (M0) supported in and h>0 small<br />
(3)
96 GUILLARMOU and TZOU<br />
enough, we have the estimate<br />
C<br />
ɛ u2 L2 <br />
(M0) + C′ − Im(〈∂τ u, u〉L2 (∂M0)) + 1<br />
h<br />
<br />
∂M0<br />
|u| 2 ∂νϕɛdvg<br />
≤e −ϕɛ/h¯∂e ϕɛ/h 2<br />
uL2 (M0) , (4)<br />
where ∂ν and ∂τ denote, respectively, the exterior pointing normal vector fields and<br />
its rotation by an angle +π/2.<br />
Proof<br />
We use complex coordinates z = x +iy in the chart , where u is supported. Since the<br />
Lebesgue measure dxdy is bounded below and above by dvg, g is conformal to |dz| 2 ,<br />
and the boundary terms in (4) depend only on the conformal class, it then suffices to<br />
prove the estimates with respect to dxdy and the Euclidean metric. We thus integrate<br />
by parts with respect to dxdy and we have<br />
4e −ϕɛ/h¯∂e ϕɛ/h<br />
<br />
2 <br />
u = ∂x + i∂yϕɛ<br />
<br />
u + i∂y +<br />
h<br />
∂xϕɛ<br />
<br />
<br />
u<br />
h<br />
2<br />
<br />
<br />
= ∂x + i∂yϕɛ<br />
<br />
<br />
u<br />
h<br />
2 <br />
<br />
+ i∂y + ∂xϕɛ<br />
<br />
<br />
u<br />
h<br />
2<br />
+ 2<br />
<br />
ϕɛ|u|<br />
h <br />
2 − 1<br />
2 ∂xϕɛ.∂x|u| 2 − 1<br />
<br />
2<br />
∂yϕɛ.∂y|u|<br />
2<br />
+ 2<br />
<br />
∂νϕɛ|u|<br />
h ∂M0<br />
2 <br />
<br />
− 2 ∂xRe(u).∂yIm(u)<br />
M0<br />
− ∂xIm(u).∂yRe(u) <br />
(5)<br />
<br />
<br />
= ∂x + i∂yϕɛ<br />
<br />
<br />
u<br />
h<br />
2 <br />
<br />
+ i∂y + ∂xϕɛ<br />
<br />
<br />
u<br />
h<br />
2<br />
+ 1<br />
<br />
ϕɛ|u|<br />
h <br />
2<br />
+ 1<br />
<br />
∂νϕɛ|u|<br />
h ∂M0<br />
2 <br />
+ 2 ∂τ Re(u).Im(u),<br />
∂M0<br />
where := −(∂ 2 x + ∂2 y ), where ∂ν is the exterior pointing normal vector field to the<br />
boundary, and where ∂τ is the tangent vector field to the boundary (i.e., ∂ν rotated<br />
with an angle π/2) for the Euclidean metric |dz| 2 .Then〈uϕɛ,u〉=(h/ɛ)(|dϕ0| 2 +<br />
|dϕ1| 2 +···+|dϕN| 2 )|u| 2 ,sinceϕj are harmonic, so the proof follows from the fact<br />
that |dϕ0| 2 +|dϕ1| 2 +···+|dϕN| 2 is uniformly bounded away from zero. <br />
Themainsteptogofrom(4) to(3) is the following lemma, whose proof is a slight<br />
modification of the one in [19, Proposition 5.3].
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 97<br />
LEMMA 3.3<br />
With the same assumptions as in Proposition 3.1, and if is either an interior chart<br />
of (M0,g0) or a chart containing a whole boundary-connected component, then there<br />
are positive constants C, C ′ such that for all ɛ>0 small, for all 0 0,<br />
e −ϕɛ/h ϕɛ/h 2 C<br />
<br />
e u ≥<br />
ɛ<br />
du 2 + 1<br />
2<br />
u|dϕɛ|<br />
h2 + 2<br />
h 〈∂xu, u∂xϕɛ〉+ 2<br />
h 〈∂yu,<br />
<br />
u∂yϕɛ〉<br />
+ C ′<br />
<br />
(B∂τA − A∂τ B)|∂νu| 2 + 1<br />
h<br />
∂M0<br />
<br />
∂M0<br />
|∂νu| 2 ∂νϕɛ<br />
<br />
.<br />
(7)
98 GUILLARMOU and TZOU<br />
Using the fact that u is real-valued, that ϕ is harmonic, and that N<br />
j=0 |dϕj| 2 is<br />
uniformly bounded away from zero, we see that<br />
2<br />
h 〈∂xu, u∂xϕɛ〉+ 2<br />
h 〈∂yu, u∂yϕɛ〉 = 1<br />
h 〈u, uϕɛ〉 ≥ C<br />
ɛ u2<br />
for some C>0, and therefore, by possibly modifying C>0, wehave<br />
e −ϕɛ/h ϕɛ/h 2 C<br />
<br />
e u ≥ du<br />
ɛ<br />
2 + 1<br />
h2 u|dϕɛ| 2 + C<br />
ɛ u2<br />
<br />
+ boundary terms. (8)<br />
Now if the diameter of the support of u is chosen small (with size depending only on<br />
|Hessϕ0|(p)) with a unique critical point p of ϕ0 inside, one can use integration by<br />
parts and the fact that the critical point is nondegenerate to obtain<br />
¯∂u 2 + 1<br />
h2 u|∂ϕ0| 2 ≥ 1<br />
<br />
<br />
∂¯z(u<br />
h<br />
2 <br />
<br />
)∂zϕ0dxdy<br />
≥ 1<br />
<br />
<br />
u<br />
h<br />
2 ∂2 z ϕ0<br />
<br />
<br />
dxdy<br />
≥ C′′<br />
h u2<br />
(9)<br />
for some C ′′ > 0. Clearly, the same estimate holds trivially if does not contain a<br />
critical point of ϕ0. Using a partition of unity (θj)j in and absorbing terms of the<br />
form ||u¯∂θj|| 2 into the right-hand side, one obtains (9) for any function u supported<br />
in and vanishing at the boundary. Thus, combining with (8), there are positive<br />
constants C, C0,C1 such that, for h small enough, we have<br />
C<br />
<br />
du<br />
ɛ<br />
2 + 1<br />
≥ C1<br />
ɛ<br />
h 2 u|dϕɛ| 2 + C<br />
ɛ u2<br />
<br />
<br />
du 2 + 1<br />
h 2 u|dϕ0| 2 + 1<br />
h u2<br />
Combining now with (8)gives<br />
≥ C<br />
<br />
du<br />
ɛ<br />
2 + 1<br />
<br />
.<br />
h2 u|dϕ0| 2 − C0<br />
ɛ<br />
<br />
u2<br />
2<br />
e −ϕɛ/h<br />
<br />
ϕɛ/h 2<br />
C1<br />
e u ≥ du<br />
ɛ<br />
2 + 1<br />
h2 u|dϕ|2 + 1<br />
h u2<br />
<br />
+ boundary terms.<br />
Let us now discuss the boundary terms in (7). If ϕj aretakensothat∂νϕj = 0 on Ɣ0,<br />
then ∂νϕɛ = 0 on Ɣ0 and ∂νϕɛ = ∂νϕ + O(h/ɛ) on Ɣ, and thus<br />
<br />
1<br />
|∂νu|<br />
h<br />
2 |∂νϕɛ| ≤ C2<br />
<br />
|∂νu|<br />
h<br />
2<br />
∂M0<br />
for some constant C2 > 0. We finally claim that B∂τ A − A∂τ B = 1 on ∂M0 ∩ .<br />
Indeed, since the chart near a connected component can be taken to be an interior<br />
neighborhood of the circle |z| =1 in C, one has A + iB = e −it , where t ∈ S 1<br />
parameterize the boundary component, so that B∂τ A − A∂τ B = 1 since ∂τ = ∂t<br />
Ɣ
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 99<br />
for the Euclidean metric. Combining all these estimates and possibly modifying the<br />
constants achieves the proof. <br />
Proof of Proposition 3.1<br />
Using triangular inequality and absorbing the term ||Vu|| 2 into the left-hand side of<br />
(3), it suffices to prove (3) with g instead of g + V .Letv ∈ C ∞ 0 (M0); by Lemma<br />
3.3, we have that there exist constants C, C ′ ,C1,C2 > 0 such that<br />
C<br />
<br />
1<br />
ɛ h e−ϕɛ/h 2 1<br />
v +<br />
h2 e−ϕɛ/h 2 1<br />
v|dϕ| +<br />
h2 e−ϕɛ/hv|dϕɛ| 2 +e −ϕɛ/h<br />
<br />
2<br />
dv<br />
≤ <br />
j<br />
+e −ϕɛ/h ∂νv 2<br />
Ɣ0<br />
C ′<br />
ɛ<br />
≤ C1<br />
≤ C2<br />
1<br />
h e−ϕɛ/h χjv 2 + 1<br />
h 2 e−ϕɛ/h χjv|dϕ| 2 + 1<br />
h 2 e−ϕɛ/h χjv|dϕɛ| 2<br />
+e −ϕɛ/h<br />
d(χjv) 2<br />
<br />
+e −ϕɛ/h<br />
∂νvƔ0<br />
<br />
e −ϕɛ/h<br />
g(χjv) 2 +e −ϕɛ/h<br />
∂νv 2<br />
<br />
Ɣ<br />
j<br />
−ϕɛ/h<br />
e gv 2 +e −ϕɛ/h 2 −ϕɛ/h 2 −ϕɛ/h<br />
v +e dv +e ∂νv 2 <br />
Ɣ ,<br />
where (χj)j is a partition of unity associated to the complex charts j on M0. Since<br />
constants on both sides are independent of ɛ and h, we can take ɛ small enough so that<br />
C2e −ϕɛ/h v 2 + C2e −ϕɛ/h dv 2 can be absorbed to the left side. Now set v = e ϕɛ/h w<br />
with w|∂M0 = 0. Then we have (for some new constant C>0)<br />
1<br />
h w2 + 1<br />
h2 w|dϕ|2 + 1<br />
h2 w|dϕɛ| 2 +dw 2 +∂νω 2<br />
Ɣ0<br />
≤ C e −ϕɛ/h<br />
ge ϕɛ/h 2<br />
w +∂νω 2 <br />
Ɣ<br />
Finally, fix ɛ>0,setu := e 1 N ɛ j=0 |ϕj | 2<br />
w, and use the fact that e 1 N ɛ j=0 |ϕj | 2<br />
is independent<br />
of h and is bounded uniformly away from zero and above; we then obtain the desired<br />
estimate for 0 0, with phase a Carleman weight (here a Morse holomorphic function), and<br />
such that the phase has a nondegenerate critical point at p, in order to apply the stationary<br />
phase method. In general, complex geometric optic solutions are 1-parameter<br />
families of solutions of the equation (g + V )u = 0 which are perturbations of
100 GUILLARMOU and TZOU<br />
certain exponentially growing solutions (with respect to the parameter) of the free<br />
equation gu = 0. They were introduced by Faddeev in [10] and are also called<br />
Faddeev-type solutions. The oscillating part of the solutions is what will allow us to<br />
recover information on the perturbation V .<br />
In this section, the potential V has the regularity V ∈ C 1,α (M0) for some α>0.<br />
Choose p ∈ int(M0) such that there exists a holomorphic function = ϕ +iψ which<br />
is Morse on M0, C k in M0 for large k ∈ N, and such that ∂(p) = 0 and has<br />
only finitely many critical points in M0. Furthermore, we ask that is purely real<br />
on Ɣ0. By Proposition 2.3.1, such points p form a dense subset of M0. Given such a<br />
holomorphic function, the purpose of this section is to construct solutions u on M0 of<br />
( + V )u = 0 of the form<br />
u = e /h (a + ha0 + r1) + e /h (a + ha0 + r1) + e ϕ/h r2 with u|Ɣ0 = 0 (10)<br />
for h>0 small, where a is holomorphic and u ∈ C k (M0) for large k ∈ N, a0 ∈<br />
H 2 (M0) is holomorphic, and moreover where a(p) = 0 and a vanishes to high order<br />
at all other critical points p ′ ∈ M0 of . Furthermore, we ask that the holomorphic<br />
function a is purely imaginary on Ɣ0. The existence of such a holomorphic function<br />
is a consequence of Lemma 2.2.4. Given such a holomorphic function on M0, we<br />
consider a compactly supported extension to M, still denoted a.<br />
The remainder terms r1,r2 will be controlled as h → 0 and have particular<br />
properties near the critical points of . More precisely, r2 will be a OL 2(h3/2 | log h|)<br />
and r1 will be of the form hr12 + oL 2(h), where r12 is independent of h, which can<br />
be used to obtain sufficient information from the stationary phase method in the<br />
identification process.<br />
4.1. Construction of r1<br />
We will construct r1 to satisfy<br />
e −/h (g + V )e /h (a + r1) = OL2(h| log h|)<br />
and r1 = r11 + hr12. WeletGbe the Green operator of the Laplacian on the smooth<br />
surface with boundary M0 with Dirichlet condition, so that gG = Id on L2 (M0).In<br />
particular, this implies that ¯∂∂G = (i/2)⋆−1 , where ⋆−1 is the inverse of ⋆ mapping<br />
functions to 2-forms. We extend a to be a compactly supported Ck function on M0,<br />
and we will search for r1 ∈ H 2 (M0) satisfying ||r1||L2 = O(h) and<br />
e −2iψ/h ∂e 2iψ/h r1 =−∂G(aV ) + ω + OH 1(h| log h|), (11)<br />
where ω is a smooth holomorphic 1-form on M0. Indeed, using the fact that is<br />
holomorphic, we have<br />
e −/h ge /h =−2i ⋆¯∂e −/h ∂e /h =−2i ⋆¯∂e −(1/h)(− ¯) ∂e (1/h)(− ¯)<br />
=−2i ⋆¯∂e −2iψ/h ∂e 2iψ/h ;
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 101<br />
applying −2i ⋆¯∂ to (11), we obtain (note that ∂G(aV ) ∈ C 2,α (M0) by elliptic<br />
regularity)<br />
e −/h (g + V )e /h r1 =−aV + OL2(h| log h|).<br />
We choose ω to be a smooth holomorphic 1-form on M0 such that at all critical<br />
points p ′ of in M0, the form b := ∂G(aV ) − ω with value in T ∗<br />
1,0M0 vanishes<br />
to the highest possible order. Writing b = b(z)dz in local complex coordinates, b(z)<br />
is C2+α by elliptic regularity, and we have −2i∂¯zb(z) = aV ; therefore, ∂z∂¯zb(p ′ ) =<br />
∂2 ¯z b(p′ ) = 0 at each critical point p ′ = p by construction of the function a. Therefore,<br />
we deduce that at each critical point p ′ = p, ∂G(aV ) has Taylor series expansion<br />
2 j=0 cjzj + O(|z| 2+α ). That is, all the lower-order terms of the Taylor expansion of<br />
∂G(aV ) around p ′ are polynomials of z only.<br />
LEMMA 4.1.1<br />
Let {p0,...,pN} be finitely many points on M0, and let θ be a C2,α section of T ∗<br />
1,0M0. Then there exists a Ck holomorphic function f on M0 with k ∈ N large, such that f<br />
vanishes to high order at the points {p1,...,pN} and ω = ∂f satisfies the following:<br />
in complex local coordinates z near p0 , one has ∂ℓ z θ(p0) = ∂ℓ z ω(p0) for ℓ = 0, 1, 2,<br />
where θ = θ(z) dz and ω = ω(z) dz.<br />
Proof<br />
This is a direct consequence of Lemma 2.2.4. <br />
Applying this to the form ∂G(aV ) and using the observation made above, we can<br />
construct a C k holomorphic form ω such that in local coordinates z centered at a critical<br />
point p ′ of (i.e., p ′ ={z = 0} in this coordinate), for b =−∂G(aV )+ω = b(z) dz<br />
we have<br />
|∂ m<br />
¯z ∂ℓ<br />
z b(z)| =O(|z|2+α−ℓ−m ) for ℓ + m ≤ 2 if p ′ = p,<br />
|b(z)| =O(|z|) if p ′ = p.<br />
Now, we let χ1 ∈ C ∞ 0 (M0) be a cutoff function supported in a small neighborhood<br />
Up of the critical point p and identically 1 near p, while χ ∈ C ∞ 0 (M0) is defined<br />
similarly with χ = 1 on the support of χ1. We construct r1 = r11 + hr12 in two steps:<br />
first, we construct r11 to solve (11) locally near the critical point p of , and second,<br />
we construct the global correction term r12 away from p by using the extra vanishing<br />
of b in (12) at the other critical points.<br />
We define locally in complex coordinates centered at p and containing the support<br />
of χ<br />
r11 := χe −2iψ/h R(e 2iψ/h χ1b),<br />
(12)
102 GUILLARMOU and TZOU<br />
<br />
−1 where Rf (z) :=−(2πi) R2(1/¯z − ¯ξ)fd¯ξ ∧ dξ for f ∈ L∞ compactly supported<br />
is the classical Cauchy-Riemann operator inverting locally ∂z (r11 is extended by zero<br />
outside the neighborhood of p). The function r11 is in C3+α (M0), and we have<br />
e −2iψ/h ∂(e 2iψ/h r11) = χ1<br />
−∂G(aV ) + ω + η<br />
with η := e −2iψ/h R(e 2iψ/h χ1b)∂χ. (13)<br />
We then construct r12 by observing that b vanishes to order 2 + α at critical points of<br />
other than p (from (12)), and ∂χ = 0 in a neighborhood of any critical point of ψ,<br />
so we can find r12 satisfying<br />
2ir12∂ψ = (1 − χ1)b.<br />
This is possible since both ∂ψ and the right-hand side are valued in T ∗<br />
1,0 M0, and∂ψ<br />
has finitely many isolated zeroes on M0; indeed, r12 is then a function in C 2,α (M0 \ P )<br />
where P := {p1,...,pN} is the set of critical points other than p, and it extends to a<br />
C 1,α (M0) function which satisfies in local complex coordinates z near each pj<br />
|∂ β<br />
¯z ∂ γ<br />
z r12(z)| ≤C|z − pj| 1+α−β−γ , β + γ ≤ 2,<br />
where we used also the fact that ∂ψ can locally be considered as a holomorphic<br />
function with a zero of order 1 at each pj. This implies that r1 ∈ H 2 (M0), andwe<br />
have<br />
e −2iψ/h ∂(e 2iψ/h r1) = b + h∂r12 + η =−∂G(aV ) − ω + h∂r12 + η.<br />
Now the first error term ||∂r12||H 1 (M0) is bounded by<br />
||∂r12||H 1 (1<br />
− χ1)b(z)<br />
<br />
<br />
(M0) ≤ C <br />
∂zψ(z)<br />
H 2 (Up)<br />
<br />
≤ C<br />
for some constant C, where we used the fact that (1 − χ1)b(z)/∂zψ(z) is in H 2 (Up)<br />
and is independent of h. To deal with the η term, we need the following.<br />
LEMMA 4.1.2<br />
The estimates<br />
hold true wherer12 solves 2ir12∂ψ = b.<br />
||η||H 2 = O(| log h|),<br />
||η||H 1 ≤ O(h| log h|),<br />
||r1||L2 = O(h),<br />
||r1 − hr12||L2 = o(h)
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 103<br />
Proof<br />
We start by observing that<br />
||r1 − hr12||L2 =<br />
||r1||L2 ≤<br />
<br />
<br />
χe<br />
−2iψ/h R(e 2iψ/h χ1b) − h χ1b<br />
<br />
<br />
<br />
2i∂zψ<br />
<br />
<br />
χe<br />
−2iψ/h R(e 2iψ/h χ1b) − h χ1b<br />
<br />
<br />
<br />
2i∂zψ<br />
L 2 (Up)<br />
L 2 (Up)<br />
,<br />
+ h||r12||L 2 (M0).<br />
The first term is estimated in [19, Proposition 2.7] and it is a o(h), while ||r12||L 2<br />
is independent of h. Now we are going to estimate the H 2 norms of η. Locally in<br />
complex coordinates z centered at p (i.e., p ={z = 0}), we have<br />
η(z) =−∂zχ(z)e −2iψ(z)/h<br />
<br />
e 2iψ(ξ)/h 1 dξ1dξ2<br />
χ1(ξ)b(ξ) ,<br />
¯z − ¯ξ π<br />
ξ = ξ1 +iξ2. (15)<br />
C<br />
Since b is C2,α in U, we decompose b(ξ) =〈∇b(0),ξ〉+b(ξ) using Taylor’s formula,<br />
so we haveb(0) = ∂ξb(0) = 0 and we split the integral (15) with 〈∇b(0),ξ〉 andb(ξ).<br />
Since the integrand with the 〈∇b(0),ξ〉 term is smooth and compactly supported in ξ<br />
(recall that χ1 = 0 on the support of ∂zχ), we can apply stationary phase to get<br />
<br />
<br />
∂zχ(z)e −2iψ(z)/h<br />
<br />
e<br />
C<br />
2iψ(ξ)/h<br />
1<br />
¯z − ¯ξ<br />
χ1(ξ)〈∇b(0),ξ〉 dξ1dξ2<br />
π<br />
<br />
<br />
≤ Ch 2<br />
uniformly in z. Nowsetbz(ξ) = ∂zχ(z)χ1(ξ)b(ξ)/(z − ξ) which is C2,α in ξ and<br />
smooth in z. Letθ∈ C∞ 0 ([0, 1)) be a cutoff function which is equal to 1 near zero,<br />
and set θh(ξ) := θ(|ξ|/h). Then integrating by parts, we have<br />
<br />
C<br />
e 2iψ(ξ)/hbz(ξ)dξ1dξ2 = h 2<br />
<br />
<br />
− h<br />
e<br />
supp(χ1)<br />
2iψ(ξ)/h ∂¯ξ<br />
supp(χ1)<br />
e 2iψ(ξ)/h θh(ξ)∂ξ<br />
(14)<br />
<br />
1 − θh(ξ)<br />
2i∂¯ξψ ∂ξ<br />
bz(ξ)<br />
<br />
dξ1dξ2<br />
2i∂ξψ<br />
bz(ξ)<br />
<br />
dξ1dξ2. (16)<br />
2i∂ξψ<br />
Using polar coordinates with the fact that bz(0) = 0, it is easy to check that the<br />
second term in (16) is bounded uniformly in z by Ch 2 . To deal with the first term,<br />
we use bz(0) = ∂ξ bz(0) = ∂¯ξ bz(0) = 0 and a straightforward computation in polar<br />
coordinates shows that the first term of (16) is bounded uniformly in z by Ch 2 | log(h)|.<br />
We conclude that<br />
||η||L 2 ≤ C||η||L ∞ ≤ Ch2 | log h|.<br />
It is also direct to see that the same estimates hold with a loss of h −2 for any derivatives<br />
in z, ¯z of order less or equal to 2, since they only hit the χ(z) factor, the (¯z − ¯ξ) −1
104 GUILLARMOU and TZOU<br />
factor, or the oscillating term e −2iψ(z)/h . So we deduce that<br />
||η||H 2 = O(| log h|)<br />
and this ends the proof. <br />
We summarize the result of this section with the following.<br />
LEMMA 4.1.3<br />
Let k ∈ N be large, and let ∈ C k (M0) be a holomorphic function on M0 which is<br />
Morse in M0 with a critical point at p ∈ int(M0). Leta ∈ C k (M0) be a holomorphic<br />
function on M0 vanishing to high order at every critical point of other than p.Then<br />
there exists r1 ∈ H 2 (M0) such that ||r1||L2 = O(h) and<br />
e −/h ( + V )e /h (a + r1) = OL2(h| log h|).<br />
4.2. Construction of a0<br />
We have constructed the correction term r1 which solves the Schrödinger equation to<br />
order h as stated in Lemma 4.1.3. In this section, we construct a holomorphic function<br />
a0 which annihilates the boundary value of the solution on Ɣ0. In particular, we have<br />
the following.<br />
LEMMA 4.2.1<br />
There exists a holomorphic function a0 ∈ H 2 (M0) independent of h such that<br />
and<br />
e −/h ( + V )e /h (a + r1 + ha0) = OL2(h| log h|)<br />
[e /h (a + r1 + ha0) + e /h (a + r1 + ha0)]|Ɣ0 = 0.<br />
Proof<br />
First, notice that h −1 r1|∂M0 = r12|∂M0 ∈ H 3/2 (∂M0) is independent of h. Since is<br />
purely real on Ɣ0 and a is purely imaginary on Ɣ0, we see that this lemma amounts to<br />
constructing a holomorphic function a0 ∈ H 2 (M0) with the boundary condition<br />
Re(r12) + Re(a0) = 0 on Ɣ0.<br />
To construct a0, it suffices to use item (ii) in Corollary 2.2.3. <br />
4.3. Construction of r2<br />
The goal of this section is to complete the construction of the complex geometric optic<br />
solutions by the following proposition.
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 105<br />
PROPOSITION 4.3.1<br />
There exist solutions to ( + V )u = 0 with boundary condition u|Ɣ0 = 0 of the<br />
form (10) with r1, a0 constructed in the previous sections and r2 satisfying r2L2 =<br />
O(h3/2 | log h|).<br />
This is a consequence of the following lemma (which follows from the Carleman<br />
estimate obtained above).<br />
LEMMA 4.3.2<br />
If V ∈ L ∞ (M0) and f ∈ L 2 (M0), then for all h>0 small enough, there exists a<br />
solution v ∈ L 2 to the boundary-value problem<br />
satisfying the estimate<br />
e ϕ/h (g + V )e −ϕ/h v = f, v|Ɣ0 = 0,<br />
vL 2 ≤ Ch1/2 f L 2.<br />
Proof<br />
The proof is identical to the proof of [19, Proposition 2.2]; we repeat the argument<br />
here for the convenience of the reader. Define for all h>0 the vector space A :=<br />
{u ∈ H 1 0 (M0); (g + V )u ∈ L2 (M0), ∂νu |Ɣ= 0} equipped with the scalar product<br />
<br />
(u, w)A := e −2ϕ/h (gu + Vu)(gw + Vw)dvg.<br />
M0<br />
Since ψ is constant along Ɣ0 and since ϕ + iψ is holomorphic, then ∂νϕ = 0 on Ɣ0.<br />
Therefore, we may apply the Carleman estimate of Proposition 3.1 to the weight ϕ<br />
to assert that the space A is a Hilbert space equipped with the scalar product above.<br />
By using the same estimate, the linear functional L : w → <br />
M0 e−ϕ/hfwdvg on A<br />
is continuous and its norm is bounded by h1/2 ||f ||L2. By the Riesz theorem, there is<br />
an element u ∈ A such that (u, .)A = L and with norm bounded by the norm of L.<br />
It remains to take v := e−ϕ/h (gu + Vu), which solves (g + V )e−ϕ/hv = e−ϕ/hf and, in addition, satisfies the desired norm estimate. Furthermore, since<br />
<br />
e −ϕ/h <br />
v(g + V )w dvg = e −ϕ/h fwdvg<br />
M0<br />
for all w ∈ A, by Green’s theorem we have<br />
<br />
e −ϕ/h <br />
v∂νw dvg = 0 =<br />
∂M0<br />
Ɣ0<br />
M0<br />
e −ϕ/h v∂νw dvg<br />
for all w ∈ A. Thisimpliesthatv = 0 on Ɣ0.
106 GUILLARMOU and TZOU<br />
Proof of Proposition 4.1<br />
We note that<br />
( + V ) e /h (a + r1 + ha0) + e/h (a + r1 + ha0) + e ϕ/h <br />
r2 = 0<br />
if and only if<br />
e −ϕ/h ( + V )e ϕ/h r2 =−e −ϕ/h ( + V ) e /h (a + r1 + ha0) + e /h (a + r1 + ha0) .<br />
By Lemma 4.2.1, the right-hand side of the above equation is OL2(h| log h|). Therefore,<br />
using Lemma 4.3.2, one can find such r2 which satisfies<br />
r2L 2 ≤ Ch3/2 | log h|, r2 |Ɣ0= 0.<br />
Since the ansatz e /h (a + r1 + ha0) + e /h (a + r1 + ha0) is arranged to vanish on<br />
Ɣ0, the solution<br />
u = e /h (a + r1 + ha0) + e /h (a + r1 + ha0) + e ϕ/h r2<br />
vanishes on Ɣ0 as well. <br />
5. Identifying the potential<br />
We now assume that V1,V2 ∈ C 1,α (M0) are two potentials, with α>0, such that<br />
the respective Cauchy data spaces C Ɣ 1 , C Ɣ 2 for the operators g + V1 and g + V2<br />
on Ɣ ⊂ ∂M0 are equal. We may also assume that V1 = V2 on Ɣ by boundary<br />
identification, a fact which is proved below in the appendix. Let Ɣ0 = ∂M0 \ Ɣ be the<br />
complement of Ɣ in ∂M0, and possibly by taking Ɣ slightly smaller, we may assume<br />
that Ɣ0 contains an open set. Let p ∈ M0 be an interior point of M0 such that, using<br />
Proposition 2.3.1, we can choose a holomorphic Morse function = ϕ + iψ on M0<br />
with purely real on Ɣ0, C k in M0 for some large k ∈ N, with a critical point at p.<br />
Note that Proposition 2.3.1 states that we can choose such that none of its critical<br />
points on the boundary are degenerate and such that critical points do not accumulate<br />
on the boundary.<br />
PROPOSITION 5.1<br />
If the Cauchy data spaces agree, that is, if C Ɣ 1 = C Ɣ 2 ,thenV1(p) = V2(p).<br />
Proof<br />
Let a be a holomorphic function on M0 which is purely imaginary on Ɣ0 with a(p) = 0<br />
and a(p ′ ) = 0 to large order for all other critical points p ′ of . The existence of a is
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 107<br />
ensured by Lemma 2.2.4.Letu1 and u2 be H 2 solutions on M0 to<br />
(g + Vj)uj = 0<br />
constructed in Section 4 with = φ + iψ for Carleman weight for u1 and − for<br />
u2, and thus of the form<br />
u1 = e /h (a + ha0 + r1) + e /h (a + ha0 + r1) + e ϕ/h r2,<br />
u2 = e −/h (a + hb0 + s1) + e −/h (a + hb0 + s1) + e −ϕ/h s2,<br />
and with boundary value uj|∂M0 = fj, where fj vanishes on Ɣ0. By Green’s formula,<br />
we can write<br />
<br />
M0<br />
<br />
u1(V1 − V2)u2 dvg =−<br />
<br />
=−<br />
M0<br />
∂M0<br />
(gu1.u2 − u1.gu2)dvg<br />
(∂νu1.f2 − f1.∂νu2)dvg.<br />
Since the Cauchy data for g + V1 agrees on Ɣ with that of g + V2, there exists a<br />
solution v of the boundary-value problem<br />
(g + V2)v = 0, v|∂M0 = f1,<br />
satisfying ∂νv = ∂νu1 on Ɣ. Sincefj = 0 on Ɣ0, thisimpliesthat<br />
<br />
<br />
u1(V1 − V2)u2 dvg =− (gu1.u2 − u1.gu2)dvg<br />
M0<br />
<br />
=−<br />
<br />
=−<br />
<br />
=−<br />
M0<br />
∂M0<br />
∂M0<br />
M0<br />
(∂νu1.f2 − f1.∂νu2)dvg<br />
(∂νv.f2 − v.∂νu2)dvg<br />
(gv.u2 − v.gu2)dvg = 0 (17)<br />
since g + V2 annihilates both v and u2. We substitute in the full expansion for u1<br />
and u2, and setting V := V1 − V2 and using the estimates in Lemmas 4.1.2, 4.3.1,and<br />
4.2.1, we obtain<br />
0 = I1 + I2 + o(h), (18)
108 GUILLARMOU and TZOU<br />
where<br />
I1 =<br />
<br />
M0<br />
<br />
I2 = 2h<br />
V (a 2 + a 2 <br />
)dvg + 2<br />
M0<br />
<br />
V Re ae 2iψ/h<br />
<br />
s1<br />
h<br />
+ e −2iψ/h<br />
<br />
a0 + r1<br />
<br />
h<br />
M0<br />
V |a| 2 Re(e 2iψ/h )dvg, (19)<br />
+ b0<br />
<br />
+ b0 + a0 + s1 + r1<br />
h<br />
<br />
dvg. (20)<br />
Remark 5.2<br />
We observe from the last identity in Lemma 4.1.2 that r1/h in the expression I2 can<br />
be replaced by the term r12 satisfying 2ir12∂ψ = b up to an error which can go in<br />
the o(h) in (18), and similarly for the term s1/h, which can be replaced by a terms12<br />
independent of h.<br />
We apply the stationary phase to these two terms in Lemmas 5.3 and 5.4.<br />
LEMMA 5.3<br />
The estimate<br />
<br />
I2 = 2h<br />
M0<br />
V Re <br />
a(b0 + a0 +s12 +r12)dvg + o(h)<br />
holds true where r12, s12,r12 ands12 are independent of h.<br />
Proof<br />
We start with the following.<br />
LEMMA 5.4<br />
Let f ∈ L 1 (M0). Thenash → 0,<br />
<br />
M0<br />
e 2iψ/h f dvg = o(1).<br />
Proof<br />
Since C k (M0) is dense in L 1 (M0) for all k ∈ N, it suffices to prove the lemma for<br />
f ∈ C k (M0).Letɛ>0 be small, and choose a cutoff function χ which is identically<br />
equal to 1 on the boundary such that<br />
<br />
M0<br />
χ|f | dvg ≤ ɛ.
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 109<br />
Then, splitting the integral and using stationary phase for the 1 − χ term, we obtain<br />
<br />
<br />
e 2iψ/h <br />
<br />
f dvg<br />
≤ (1 − χ)e 2iψ/h <br />
<br />
f dvg<br />
+ χe 2iψ/h <br />
<br />
f dvg<br />
≤ ɛ + Oɛ(h),<br />
M0<br />
M0<br />
which concludes the proof by taking h small enough depending on ɛ. <br />
The proof of Lemma 5.3 is a direct consequence of Lemma 5.4 and Remark<br />
5.2.1. <br />
The second lemma will be proved at the end of this section.<br />
LEMMA 5.5<br />
The estimate<br />
I1 =<br />
<br />
M0<br />
V (a 2 + a 2 )dvg + hCpV (p)|a(p)| 2 Re(e 2iψ(p)/h ) + o(h)<br />
holds true with Cp = 0 and is independent of h.<br />
With Lemmas 5.3 and 5.4, we can write (18) as<br />
<br />
0 = V (a 2 + a 2 )dvg + O(h)<br />
M0<br />
and thus we can conclude that<br />
<br />
0 =<br />
M0<br />
M0<br />
V (a 2 + a 2 )dvg.<br />
Therefore, (18) becomes, with Lemma 5.3,<br />
0 = CpV (p)|a(p)| 2 Re(e 2iψ(p)/h <br />
) + 2 V Re a(b0 + ˜s12 + a0 + ˜r12) dvg + o(1).<br />
M0<br />
Since ψ(p) = 0, we may choose a sequence of hj → 0 such that Re(e 2iψ(p)/hj ) = 1<br />
and another sequence ˜hj → 0 such that Re(e 2iψ(p)/˜hj ) =−1 for all j. Adding the<br />
expansion with h = hj and h = ˜hj, we deduce that<br />
0 = 2CpV (p)|a(p)| 2 + o(1)<br />
as j →∞, and since Cp = 0, a(p) = 0, we conclude that V (p) = 0. The set<br />
of p ∈ M0 for which we can conclude this is dense in M0, by Proposition 2.3.1.<br />
Therefore, we can conclude that V (p) = 0 for all p ∈ M0. <br />
We now prove Lemma 5.5.
110 GUILLARMOU and TZOU<br />
Proof of Lemma 5.5<br />
Let χ be a smooth cutoff function on M0 which is identically 1 everywhere except<br />
outside a small ball containing p and no other critical point of , andletχ = 0 near<br />
p. We split the oscillatory integral into two parts:<br />
<br />
(e 2iψ/h + e −2iψ/h )V |a| 2 <br />
dvg = χ(e 2iψ/h + e −2iψ/h )V |a| 2 dvg<br />
M0<br />
M0<br />
<br />
+<br />
M0<br />
(1 − χ)(e 2iψ/h + e −2iψ/h )V |a| 2 dvg.<br />
The phase ψ has nondegenerate critical points. Therefore, a standard application of<br />
the stationary phase at p gives<br />
<br />
(1 − χ)(e 2iψ/h + e −2iψ/h )V (p)|a| 2 dvg = hCp|a(p)| 2 V (p)Re(e 2iψ(p)/h ) + o(h),<br />
M0<br />
where Cp is a nonzero number which depends on the Hessian of ψ at the point p.<br />
Define the potential V (·) := V (·) − V (p) ∈ C1,α (M0). Then we show that<br />
<br />
(1 − χ)(e 2iψ/h + e −2iψ/h )V |a| 2 dvg = o(h). (21)<br />
M0<br />
Indeed, first by integration by parts and using gψ = 0, one has<br />
<br />
(1 − χ)(e 2iψ/h + e −2iψ/h )V |a| 2 dvg<br />
M0<br />
= h<br />
2i<br />
= h<br />
2i<br />
<br />
<br />
M0<br />
M0<br />
〈d(e 2iψ/h − e −2iψ/h (1 − χ)|a|2<br />
),dψ〉V<br />
|dψ| 2<br />
dvg<br />
(e 2iψ/h − e −2iψ/h <br />
(1 − χ)|a| 2V<br />
) d<br />
|dψ| 2<br />
<br />
,dψ dvg.<br />
But we can see that 〈d((1 − χ)|a| 2V/| dψ| 2 ),dψ〉∈L1 (M0): this follows directly<br />
from the fact that V is in the Hölder space C1,α (M0) and V (p) = 0, and from the<br />
nondegeneracy of Hess(ψ). It then suffices to use Lemma 5.4 to conclude that (21)<br />
holds. Using a similar argument, we now show that<br />
<br />
χ(e 2iψ/h + e −2iψ/h )V |a| 2 dvg = o(h).<br />
M0
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 111<br />
Indeed, since a vanishes to large order at all boundary critical points of ψ, we may<br />
write<br />
<br />
χ(e 2iψ/h + e −2iψ/h )V |a| 2 dvg<br />
M0<br />
= h<br />
<br />
〈d(e<br />
2i M0<br />
2iψ/h − e −2iψ/h ),dψ〉V χ|a|2<br />
dvg<br />
|dψ| 2<br />
=− h<br />
<br />
(e<br />
2i M0<br />
2iψ/h − e −2iψ/h <br />
)divg V χ|a|2<br />
|dψ| 2 ∇g <br />
ψ dvg<br />
+ h<br />
<br />
(e<br />
2i<br />
2iψ/h − e −2iψ/h )V |a|2<br />
|dψ| 2 ∂νψ dvg.<br />
∂M0<br />
For the interior integral, we use Lemma 5.4 to conclude that<br />
− h<br />
<br />
(e<br />
2i M0<br />
2iψ/h − e −2iψ/h <br />
)divg V χ|a|2<br />
|dψ| 2 ∇g <br />
ψ dvg = o(h),<br />
and for the boundary integral we write ∂M0 = Ɣ0 ∪ Ɣ and observe that on Ɣ0, ψ = 0,<br />
so (e2iψ/h−e−2iψ/h ) = 0, while on Ɣ we have V = 0 from the boundary determination<br />
proved in Proposition A.1 of the appendix. Therefore,<br />
<br />
χ(e 2iψ/h + e −2iψ/h )V |a| 2 dvg = o(h),<br />
M0<br />
and the proof is complete. <br />
6. Inverse scattering<br />
We first obtain, as a trivial consequence of Theorem 1.1, a result about inverse<br />
scattering for asymptotically hyperbolic (AH) surfaces. Recall that an AH surface<br />
is an open complete Riemannian surface (X, g) such that X is the interior of a<br />
smooth compact surface with boundary ¯X, and for any smooth boundary-defining<br />
function x of ∂ ¯X, ¯g := x 2 g extends as a smooth metric to ¯X, with curvature tending<br />
to −1 at ∂ ¯X. IfV ∈ C ∞ ( ¯X) and if V = O(x 2 ), then we can define a scattering<br />
map as follows (see [20], [13] or[14]). First, the L 2 kernel kerL 2(g + V ) is<br />
a finite-dimensional subspace of xC ∞ ( ¯X) and in one-to-one correspondence with<br />
E := {(∂xψ)| ∂ ¯X; ψ ∈ kerL 2(g + V )}, where ∂x := ∇ ¯g x is the normal vector field to<br />
∂ ¯X for ¯g. Then, for f ∈ C ∞ (∂ ¯X), there exists a function u ∈ C ∞ ( ¯X), unique modulo<br />
kerL 2(g + V ), such that (g + V )u = 0 and u| ∂ ¯X = f . Then one can see that the<br />
scattering map S : C ∞ (∂ ¯X) → C ∞ (∂ ¯X)/E is defined by Sf := ∂xu| ∂ ¯X. We thus<br />
obtain the following.
112 GUILLARMOU and TZOU<br />
COROLLARY 6.1<br />
Let (X, g) be an asymptotically hyperbolic manifold, and let V1,V2 ∈ x 2 C ∞ ( ¯X) be<br />
two potentials and Ɣ ⊂ ∂ ¯X an open subset of the conformal boundary. Assume that<br />
∂xu ∂ ¯X ; u ∈ kerL 2(g + V1) = ∂xu ∂ ¯X ; u ∈ kerL 2(g + V2) ,<br />
and let Sj be the scattering map for the operator g + Vj for j = 1, 2.IfS1f = S2f<br />
on Ɣ for all f ∈ C ∞ 0 (Ɣ), thenV1 = V2.<br />
Proof<br />
Let x be a smooth boundary-defining function of ∂ ¯X,let¯g = x 2 g be the compactified<br />
metric, and define ¯Vj := Vj/x 2 ∈ C ∞ ( ¯X). By conformal invariance of the Laplacian<br />
in dimension 2, one has<br />
g + Vj = x 2 (¯g + ¯Vj),<br />
and so if kerL 2(g + V1) = kerL 2(g + V2) and if S1 = S2 on Ɣ, then the Cauchy<br />
data spaces C Ɣ i for the operator ¯g + ¯Vj are the same. Then it suffices to apply the<br />
result in Theorem 1.1. <br />
Next we consider the asymptotically Euclidean scattering at zero frequency. An<br />
asymptotically Euclidean surface is a noncompact Riemann surface (X, g) which<br />
compactifies into a smooth surface ¯X with boundary such that the metric, in a collar<br />
neighborhood (0,ɛ)x × ∂ ¯X of the boundary, is of the form<br />
g = dx2 h(x)<br />
+<br />
x4 x<br />
where h(x) is a smooth 1-parameter family of metrics on ∂ ¯X such that h(0) = dθ2 S1 is<br />
the metric. Notice that using the coordinates r := 1/x, g is asymptotic to dr2 +r 2dθ2 S1 near r →∞. A particular case is given by the surfaces with Euclidean ends, that<br />
is, ends isometric to R2 \ B(0,R), where B(0,R) ={z ∈ R2 ; |z| ≥R}. Note that<br />
g is conformal to an asymptotically cylindrical metric (or “b-metric” in the sense of<br />
Melrose [27])<br />
2 ,<br />
gb := x 2 g = dx2<br />
+ h(x)<br />
x2 and that the Laplacian satisfies g = x2gb . Each end of X is of the form (0,ɛ)x ×S 1 θ ,<br />
and the operator gb has the expression in the ends<br />
gb =−(x∂x) 2 + ∂ ¯X + xP(x,θ; x∂x,∂θ)
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 113<br />
for some smooth differential operator P (x,θ; x∂x,∂θ) in the vector fields x∂x,∂θ<br />
down to x = 0. Let us define Vb := x −2 V , which is compactly supported, and<br />
H 2m<br />
b := u ∈ L 2 (X, dvolgb); m<br />
gb u ∈ L2 (X, dvolgb) , m ∈ N0.<br />
We also define the following spaces for α ∈ R:<br />
Fα := ker x α H 2 b (gb + Vb).<br />
Since the eigenvalues of S 1 are {j 2 ; j ∈ N0}, the relative index theorem of Melrose<br />
[27, Section 6.2] shows that gb + Vb is Fredholm from xαH 2 b to xαH 0 b if α /∈ Z.<br />
Moreover, from [27, Section 2.2.4], we have that any solution of (gb + Vb)u = 0 in<br />
xαH 2 b has an asymptotic expansion of the form<br />
u ∼ <br />
ℓj <br />
j>α,j∈Z ℓ=0<br />
x j (log x) ℓ uj,ℓ(θ) as x → 0<br />
for some sequence (ℓj)j of nonnegative integers and some smooth function uj,ℓ on<br />
S 1 . In particular, it is easy to check that kerL 2 (X,dvolg)(g + V ) = F1+ɛ for ɛ ∈ (0, 1).<br />
<strong>THEOREM</strong> 6.2<br />
Let (X, g) be an asymptotically Euclidean surface, let V1,V2 be two compactly supported<br />
smooth potentials, and let x be a boundary-defining function. Let ɛ ∈ (0, 1),<br />
and assume that, for any j ∈ Z and any function ψ ∈ ker x j−ɛ H 2 b (g + V1), thereisa<br />
ϕ ∈ ker x j−ɛ H 2 b (g + V2) such that ψ − ϕ = O(x ∞ ), and conversely. Then V1 = V2.<br />
Proof<br />
The idea is to reduce the problem to the compact case. First, we notice that by unique<br />
continuation, ψ = ϕ where V1 = V2 = 0. Now it remains to prove that, if Rη denotes<br />
the restriction of smooth functions on X to {x ≥ η} and if V is a smooth compactly<br />
supported potential in {x ≥ η}, then the set ∞ j=0 Rη(F−j−ɛ) is dense in the set NV<br />
of H 2 ({x ≥ η}) solutions of (g + V )u = 0. The proof is well known for positive<br />
frequency scattering (see, e.g., [28, Lemma 3.2]). Here it is very similar, so we do not<br />
give details. The main argument is to show that it converges in the L2 sense and then<br />
uses elliptic regularity; the L2 convergence can be shown as follows. Let f ∈ NV<br />
such that<br />
<br />
∞<br />
fψdvolg = 0, ∀ ψ ∈ F−j−ɛ.<br />
x≥η<br />
Then we want to show that f = 0. By[27, Proposition 5.64], there exists k ∈ N<br />
and a generalized right inverse Gb for Pb = gb + Vb (here, as before, x 2 Vb = V )<br />
j=0
114 GUILLARMOU and TZOU<br />
in x−k−ɛ H 2 b such that PbGb = Id. This holds in x−k−ɛ H 2 b for k large enough since<br />
the cokernel of Pb on this space becomes zero for k large. Let ω = Gbf so that<br />
(gb + Vb)ω = f ; in particular, this function is zero in {x n,<br />
while Brown [5] studied the case of Lipschitz domains with a continuous conductivity.<br />
Since the result in our setting is not explicitly written down but is certainly known<br />
from specialists, we provide a short proof with few details using the approach of [5].<br />
We prove the following.
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 115<br />
PROPOSITION A.1<br />
Let Ɣ ⊂ ∂M0 be a nonempty open subset of the boundary. If V1,V2 ∈ C0,α (M0) for<br />
some α>0 and if their associated Cauchy data spaces C Ɣ 1 , C Ɣ 2 defined in (1) are<br />
equal, then V1|Ɣ = V2|Ɣ.<br />
The key to proving this proposition is the existence of solutions to (g + Vi)u = 0,<br />
which concentrate near a point p ∈ Ɣ. First, we need a solvability result for the<br />
equation (g + Vi)u = f , which is an easy consequence of the Carleman estimate<br />
of Proposition 3.1, and follows the method of Salo and Tzou [31, Section 6]. If we<br />
fix h>0 small and take ϕ = 1 in the Carleman estimate of Proposition 3.1, we<br />
easily obtain that there is a constant C such that for all functions in H 2 (M0) satisfying<br />
u|∂M0 = 0,<br />
u 2<br />
2<br />
H 2 +∂νuL2 (Ɣ0)<br />
2<br />
2<br />
≤ C(( + Vi)uL2 +∂νuL2 (Ɣ) ). (A.1)<br />
As a consequence, we deduce the following solvability result. Let<br />
B := w ∈ H 2 (M0) ∩ H 1<br />
0 (M0) | ∂νw|Ɣ = 0 <br />
be the closed subspace of H 2 (M0) under the H 2 norm, and let B ∗ be its dual space.<br />
Then we have the following.<br />
COROLLARY A.2<br />
Let i = 1, 2. Then for all f ∈ L 2 (M0) there exists u ∈ H 2 (M0) solving the equation<br />
(g + Vi)u = f<br />
with boundary condition u|Ɣ0 = 0, and uL2 ≤ Cf B∗. Proof<br />
Set A := {w ∈ H 1 0 (M0) | (g + Vi)w ∈ L2 ,∂νw|Ɣ = 0} equipped with the inner<br />
product<br />
<br />
(v, w)A := (g + Vi)v(g + Vi)w dvg.<br />
M0<br />
Thanks to (A.1), A is a Hilbert space and A = B. For each f ∈ L 2 (M0), let us define<br />
the linear functional on B:<br />
By (A.1), we have, for all w ∈ B,<br />
<br />
Lf : w ↦→<br />
M0<br />
wf dvg.<br />
|Lf (w)| ≤f B ∗wB ≤f B ∗wA.
116 GUILLARMOU and TZOU<br />
Therefore, by the Riesz theorem, there exists vf ∈ A such that<br />
<br />
<br />
(g + V )vf (g + Vi)w dvg = wf dvg<br />
M0<br />
for all w ∈ B. Furthermore, (g + Vi)vf L2 ≤fB∗. Setting u := (g + Vi)vf ,<br />
we have (g + Vi)u = f and uL2 ≤fB∗. To obtain the boundary condition for<br />
u, observe that since<br />
<br />
M0<br />
<br />
u(g + Vi)wdvg =<br />
for all w ∈ B, then by Green’s theorem,<br />
<br />
<br />
u∂νw dvg = 0 =<br />
M0<br />
Ɣ0<br />
M0<br />
M0<br />
fwdvg<br />
u∂νw dvg<br />
for all w ∈ B. Thisimpliesthatu = 0 on Ɣ0. <br />
Clearly, it suffices to assume that Ɣ is a small piece of the boundary which is contained<br />
in a single coordinate chart with complex coordinates z = x + iy, where |z| ≤1,<br />
Im(z) > 0, and the boundary is given by {y = 0}. Moreover, the metric is of the<br />
form e 2ρ |dz| 2 for some smooth function ρ.Letp ∈ Ɣ, and possibly by translating the<br />
coordinates, we can assume that p ={z = 0}. Letη ∈ C ∞ (M0) be a cutoff function<br />
supported in a small neighborhood of p. Forh>0 small, we define the smooth<br />
function vh ∈ C ∞ (M0) supported near p via the coordinate chart Z = (x,y) ∈ R 2 by<br />
vh(Z) := η(Z/ √ h)e (1/h)α·Z , (A.2)<br />
where α := (i, −1) ∈ C 2 is chosen such that α · α = 0 and the dot product · means<br />
the canonical symmetric C-bilinear form on C 2 . Thus we get (∂ 2 x + ∂2 y )eα·Z = 0 and<br />
thus ge α·Z = 0 by conformal covariance of the Laplacian. Therefore, we have in<br />
local coordinates<br />
gvh(Z) = 1<br />
h e(1/h)α·Z (gη)<br />
<br />
Z√h<br />
<br />
+ 2<br />
<br />
Z√h<br />
<br />
e(1/h)α·Z dη ,α· dZ . (A.3)<br />
h3/2 g<br />
LEMMA A.3<br />
If V ∈ C 0,α (M0) for q>2, then there exists a solution uh ∈ H 2 to (g + V )u = 0<br />
of the form<br />
uh = vh + Rh,<br />
with vh defined in (A.2) and RhL 2 ≤ Ch5/4 , satisfying supp(Rh|∂M0) ⊂ Ɣ.
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 117<br />
Proof of Lemma A.3<br />
We need to find Rh satisfying RhL 2 ≤ Ch5/4 and solving<br />
(g + V )Rh =−(g + V )vh =: Mh.<br />
Thanks to Corollary A.2, it suffices to show that MhB ∗ ≤ Ch5/4 . Thus, let w ∈ B;<br />
then we have, by (A.3),<br />
<br />
wMh dvg = I1 + I2 + I3,<br />
where<br />
<br />
I1 : =<br />
I2 : = 1<br />
h<br />
M0<br />
|Z|≤ √ h<br />
<br />
I3 : = 2<br />
h 3/2<br />
wVivhe 2ρ dZ,<br />
|Z|≤ √ we<br />
h<br />
(1/h)α·Z χ1<br />
<br />
|Z|≤ √ wχ2<br />
h<br />
<br />
Z√h<br />
<br />
e 2ρ dZ,<br />
<br />
Z√h<br />
<br />
e (1/h)α·Z dZ,<br />
and χ1 = gη, χ2 = i∂xη − ∂yη. In the above equation, the term I3 has the worst<br />
growth when h → 0. We analyze its behavior, and the preceding terms can be treated<br />
in similar fashion. One has<br />
I3 =−h 1/2<br />
=−h 1/2<br />
<br />
<br />
|Z|≤ √ wχ2<br />
h<br />
|Z|≤ √ (∂x − i∂y)<br />
h<br />
2<br />
<br />
Z√h<br />
<br />
(∂x − i∂y) 2 e (1/h)α·Z dZ<br />
<br />
Z√h<br />
<br />
wχ2 e (1/h)α·x dZ.<br />
Notice that the boundary term in the integration by parts vanishes because w ∈ H 1 0<br />
and ∂νw|∂M0 vanishes on the support of η. The term (∂x − i∂y) 2 (wχ2(Z/ √ h)) has<br />
derivatives hitting both χ2(Z/ √ h) and w. The worst growth in h would occur when<br />
both derivatives hit χ2(Z/ √ h), in which case a h −1 factor would come out. Combined<br />
with the h 1/2 term in front of the integral this gives a total of a h −1/2 in front. By this<br />
observation we have improved the growth from h −3/2 to h −1/2 . Repeating this line of argument<br />
and using the Cauchy-Schwarz inequality, we can see that |I3| ≤Ch 5/4 wH 2<br />
(an elementary computation shows that functions of the form χ(Z/ √ h)e (1/h)α·Z have<br />
L 2 norm bounded by Ch 3/4 ). Therefore, (1/h 3/2 )χ2(Z/ √ h)e (1/h)α·Z B ′ ≤ Ch5/4 ,<br />
and we are done.
118 GUILLARMOU and TZOU<br />
Proof of Proposition A.1<br />
It suffices to plug the solutions u1 h ,u2 h from Proposition A.3 into the boundary integral<br />
identity (17). A simple calculation using the fact that V1 − V2 is in C0,γ (M0) yields<br />
<br />
0 = u 1<br />
h (V1 − V2)u 2<br />
h dvg = Ch 3/2 V1(p) − V2(p) + o(h 3/2 ),<br />
M0<br />
and we are done. <br />
Acknowledgments. This work started during a summer evening in Pisa thanks to the<br />
hospitality of M. Mazzucchelli and A. G. Lecuona. We thank Sam Lisi, Rafe Mazzeo,<br />
Mikko Salo, and Eleny Ionel for pointing out very helpful references. The first author<br />
thanks the Mathematical Sciences Research Institute and the organizers of the 2008<br />
program “Analysis on Singular Spaces” for support during part of this project. Part of<br />
this work was also done while the first author was at the Institute for Advanced Study<br />
in Princeton, and he acknowledges the hospitality extended to him.<br />
References<br />
[1] G. ALESSANDR<strong>IN</strong>I, Singular solutions of elliptic equations and the determination of<br />
conductivity by boundary measurements, J. Differential Equations 84 (1990),<br />
252 – 272. MR 1047569 114<br />
[2] K. ASTALA and L. PÄIVÄR<strong>IN</strong>TA, Calderón’s inverse conductivity problem in the plane,<br />
Ann. of Math. (2) 163 (2006), 265 – 299. MR 2195135 84<br />
[3] K. ASTALA, L. PÄIVÄR<strong>IN</strong>TA, and M. LASSAS, Calderón’s inverse problem for<br />
anisotropic conductivity in the plane, Comm. Partial Differential Equations 30<br />
(2005), 207 – 224. MR 2131051 84<br />
[4] M. I. BELISHEV, The Calderon problem for two-dimensional manifolds by the<br />
BC-method, SIAM J. Math. Anal. 35 (2003), 172 – 182. MR 2001471 84<br />
[5] R. M. BROWN, Recovering the conductivity at the boundary from the Dirichlet to<br />
Neumann map: A pointwise result, J. Inverse Ill-Posed Probl. 9 (2001), 567 – 574.<br />
MR 1881563 114<br />
[6] A. L. BUKHGEIM, Recovering a potential from Cauchy data in the two-dimensional<br />
case, J. Inverse Ill-Posed Probl. 16 (2008), 19 – 33. MR 2387648 84, 85<br />
[7] A. L BUKHGEIM and G. UHLMANN, Recovering a potential from partial Cauchy data,<br />
Comm. Partial Differential Equations 27 (2002), 653 – 668. MR 1900557 84<br />
[8] A. P. CALDERÓN, “On an inverse boundary value problem” in Seminar on Numerical<br />
Analysis and its Applications to Continuum Physics, (Rio de Janeiro, 1980), Soc.<br />
Brasileira de Matemática 12, Soc. Brasileira de Matemática, Rio de Janeiro, 1980,<br />
65 – 73. MR 0590275 83<br />
[9] D. DOS SANTOS FERREIRA, C. E. KENIG, M. SALO,andG. UHLMANN, Limiting<br />
Carleman weights and anisotropic inverse problems, Invent. Math. 178 (2009),<br />
119 – 171. MR 2534094 84
CALDERÓN <strong>IN</strong>VERSE PROBLEM ON RIEMANN SURFACES 119<br />
[10] L. D. FADDEEV, Increasing solutions of the Schrödinger equation (in Russian), Dokl.<br />
Akad. Nauk SSSR 165 (1965), 514 – 517; English translation in Sov. Phys. Dokl.<br />
10 (1966), 1033 – 1035. 85, 100<br />
[11] H. M FARKAS and I. KRA, Riemann Surfaces, 2nd ed., Grad. Texts in Math. 71,<br />
Springer, New York, 1992. MR 1139765 86<br />
[12] I. M. GELFAND, “Some aspects of functional analysis and algebra” in Proceedings of<br />
the International Congress of Mathematicians (Amsterdam, 1954),<br />
North-Holland, Amsterdam, 253 – 276. MR 0095423 83<br />
[13] C. R. GRAHAM and M. ZWORSKI, Scattering matrix in conformal geometry, Invent.<br />
Math. 152 (2003), 89 – 118. MR 1965361 111<br />
[14] C. GUILLARMOU and L. GUILLOPÉ, The determinant of the Dirichlet-to-Neumann map<br />
for surfaces with boundary, Int. Math. Res. Not. IMRN 2007, no. 22, art. ID<br />
rnm099. MR 2376211 111<br />
[15] C. GUILLARMOU and A. SÁ BARRETO, Inverse problems for Einstein manifolds,<br />
Inverse Probl. Imaging 3 (2009), 1 – 15. MR 2558301 84<br />
[16] C. GUILLARMOU and L. TZOU,“Calderón inverse problem for the Schrödinger<br />
operator on Riemann surfaces” in The AMSI-ANU Workshop on Spectral Theory<br />
and Harmonic Analysis (Canberra, 2009), Proc. Centre Math. Appl. Austral. Nat.<br />
Univ. 44, Austral. Nat. Univ., Canberra, 2010, 129 – 141. MR 2655386 84<br />
[17] G. HENK<strong>IN</strong> and V. MICHEL, Inverse conductivity problem on Riemann surfaces, J.<br />
Geom. Anal. 18 (2008), 1033 – 1052. MR 2438910 84, 85<br />
[18] G. HENK<strong>IN</strong> and R. G. NOVIKOV, On the reconstruction of conductivity of bordered<br />
two-dimensional surface in R 3 from electrical currents measurements on its<br />
boundary, to appear in J. Geom. Anal., preprint, arXiv:1003.4897v1 [math.AP]<br />
85<br />
[19] O. Y. IMANUVILOV, G. UHLMANN, andM. YAMAMOTO, The Calderón problem with<br />
partial data in two dimensions, J.Amer.Math.Soc.23 (2010), 655 – 691. 84, 85,<br />
86, 95, 96, 103, 105<br />
[20] M. S. JOSHI and A. SÁ BARRETO, Inverse scattering on asymptotically hyperbolic<br />
manifolds, Acta Math. 184 (2000), 41 – 86. MR 1756569<br />
[21] C. E. KENIG, J. SJÖSTRAND,andG. UHLMANN, The Calderón problem with partial<br />
data, Ann. of Math. (2) 165 (2007), 567 – 591. MR 2299741 111 84<br />
[22] R. KOHN and M. VOGELIUS, Determining conductivity by boundary measurements,<br />
Comm. Pure Appl. Math. 37 (1984), 289 – 298. MR 0739921 114<br />
[23] M. LASSAS, M. TAYLOR,andG. UHLMANN, The Dirichlet-to-Neumann map for<br />
complete Riemannian manifolds with boundary, Comm. Anal. Geom. 11 (2003),<br />
207 – 222. MR 2014876 84<br />
[24] M. LASSAS and G. UHLMANN, On determining a Riemannian manifold from the<br />
Dirichlet-to-Neumann map. Ann. Sci. Éc. Norm. Supér. (4) 34 (2001), 771 – 787.<br />
MR 1862026 84<br />
[25] R. MAZZEO and M. TAYLOR, Curvature and uniformization, Israel J. Math. 130 (2002),<br />
323 – 346. MR 1919383 95<br />
[26] D. MCDUFF and D. SALAMON, J -Holomorphic curves and symplectic topology, Amer.<br />
Math. Soc. Colloq. Publ. 52, Amer. Math. Soc., Providence, 2004. MR 2045629<br />
87, 88
120 GUILLARMOU and TZOU<br />
[27] R. B. MELROSE, The Atiyah-Patodi-Singer Index Theorem, Res. Notes Math. 4,A.K.<br />
Peters, Wellesley, 1993. MR 1348401 112, 113, 114<br />
[28] ———, Geometric Scattering Theory, Cambridge Univ. Press, Cambridge, 1995.<br />
MR 1350074 113<br />
[29] A. NACHMAN, Reconstructions from boundary measurements, Ann. of Math. (2) 128<br />
(1988), 531 – 576. MR 0970610 84<br />
[30] R. G. NOVIKOV, A multidimensional inverse spectral problem for the equation<br />
−ψ + (v(x) − Eu(x))ψ = 0 (in Russian), Funktsional. Anal. i Prilozhen. 22<br />
(1988), no. 4, 11 – 22, 96; English translation in Funct. Anal. Appl. 22 (1988), no.<br />
4, 263 – 272 (1989). MR 0976992 84<br />
[31] M. SALO and L. TZOU, Carleman estimates and inverse problems for Dirac operators,<br />
Math. Ann. 344 (2009), 161 – 184. MR 2481057 115<br />
[32] J. SYLVESTER and G. UHLMANN, A global uniqueness theorem for an inverse boundary<br />
value problem, Ann. of Math. (2) 125 (1987), 153 – 169. MR 0873380 84<br />
[33] ———, Inverse boundary value problems at the boundary—continuous dependence,<br />
Comm. Pure Appl. Math. 41 (1988), 197 – 219. MR 0924684 114<br />
[34] K. UHLENBECK, Generic properties of eigenfunctions, Amer. J. Math. 98 (1976),<br />
1059 – 1078. MR 0464332 91, 92<br />
Guillarmou<br />
École Normale Supérieure, UMR CNRS 8553, F 75230 Paris CEDEX 05, France;<br />
cguillar@dma.ens.fr<br />
Tzou<br />
Department of Mathematics, Stanford University, Stanford, California 94305, USA;<br />
ltzou@math.stanford.edu
QUANTITATIVE VERSION OF THE OPPENHEIM<br />
CONJECTURE FOR <strong>IN</strong>HOMOGENEOUS<br />
QUADRATIC FORMS<br />
GREGORY MARGULIS and AMIR MOHAMMADI<br />
Abstract<br />
We prove a quantitative version of the Oppenheim conjecture for inhomogeneous<br />
quadratic forms. We also give an application to eigenvalue spacing on flat 2-tori with<br />
Aharonov-Bohm flux.<br />
Contents<br />
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121<br />
2. Passage to space of inhomogeneous lattices . . . . . . . . . . . . . . . 128<br />
3. The case where p ≥ 3 . . . . . . . . . . . . . . . . . . . . . . . . . . 130<br />
4. The case of signature (2, 2) . . . . . . . . . . . . . . . . . . . . . . . . 132<br />
5. Contribution from quasi-null subspaces . . . . . . . . . . . . . . . . . 136<br />
6. Proof of Theorem 4.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . 146<br />
7. Proof of Theorem 1.10 . . . . . . . . . . . . . . . . . . . . . . . . . . 149<br />
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153<br />
A. Equidistribution of spherical averages . . . . . . . . . . . . . . . . . . . 153<br />
B. Number of quasi-null subspaces . . . . . . . . . . . . . . . . . . . . . . 157<br />
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159<br />
1. Introduction<br />
Let Q be a nondegenerate indefinite quadratic form on R n . Let ξ ∈ R n beavector,<br />
and define the (inhomogeneous) quadratic form Qξ by<br />
Qξ(x) = Q(x + ξ) for all x ∈ R n . (1)<br />
We will refer to Q = Q0 as the homogeneous part of Qξ. We say that Qξ has signature<br />
(p, q) if Q does. Recall that a quadratic form Qξ is called irrational if it is not a scalar<br />
DUKE MATHEMATICAL JOURNAL<br />
Vol. 158, No. 1, c○ 2011 DOI 10.1215/00127094-1276319<br />
Received 4 March 2010. Revision received 22 July 2010.<br />
2010 Mathematics Subject Classification. Primary 11E20; Secondary 58J50.<br />
Margulis’s work partially supported by National Science Foundation grant DMS-0801195.<br />
121
122 MARGULIS and MOHAMMADI<br />
multiple of a form with rational coefficients. In other words, Qξ is irrational if either Q<br />
is irrational as a homogeneous form, or, if Q is a rational form, then ξ is an irrational<br />
vector.<br />
Let ν be a continuous function on the sphere {v ∈ R n : v =1}. Define<br />
={v ∈ R n : v
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 123<br />
Recall from [EMM2]thatifQ is irrational of signature (2, 2), then it has at most four<br />
rational null subspaces. Let<br />
ÑQ,(a,b,T ) = #<br />
<br />
x ∈ Z n :<br />
Eskin, Margulis, and Mozes proved the following.<br />
x is not in a null subspace of Q<br />
<br />
. (6)<br />
x ∈ T and a
124 MARGULIS and MOHAMMADI<br />
As in [EMM1], we also have the following uniform version of Theorem 1.5. Let<br />
I(p, q) denote the space of inhomogeneous quadratic forms whose homogeneous<br />
parts are quadratic forms of signature (p, q) and discriminant ±1.<br />
<strong>THEOREM</strong> 1.6<br />
Let D be a compact subset of I(p, q) with p ≥ 3 and q ≥ 1, and let n = p + q.<br />
Then for every interval (a,b) and every θ>0, there exists a finite subset P of D<br />
such that each Qξ ∈ P is a rational form and, for every compact subset F ⊂ D \ P ,<br />
there exists T0 such that, for all Qξ ∈ F and T ≥ T0, we have<br />
(1 − θ)λQ,(b − a)T n−2 ≤ NQ,ξ,(a,b,T ) ≤ (1 + θ)λQ,(b − a)T n−2 , (10)<br />
where λQ, is as in (3).<br />
The proofs of the above theorems are straightforward inhomogeneous versions of the<br />
arguments and ideas developed in [DM] and[EMM1].<br />
As we mentioned before, Theorem 1.5 fails in signature (2, 2) and (2, 1). In this,<br />
paper, we prove an inhomogeneous version of Theorem 1.3. Indeed,asin[EMM2],<br />
one needs to assume certain Diophantine conditions on the quadratic form. Using<br />
similar ideas, we also give a partial result in the (2, 1) case (see Theorem 1.10 below).<br />
We start with the following definitions.<br />
Definition 1.7<br />
Let κ>0. A vector ξ = (ξ1,...,ξn) ∈ R n is called κ-Diophantine if there exist C =<br />
C(ξ) > 0 such that, for all 0
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 125<br />
then there are at most four subspaces L for which the above can hold. Let<br />
ÑQ,ξ,(a,b,T ) = #<br />
<br />
x ∈ Z n :<br />
x is not in an exceptional subspace of Qξ<br />
x ∈ T and a
126 MARGULIS and MOHAMMADI<br />
planes” interfere with each other. This fact prevents us from successfully using the<br />
ideas in [EMM2] when the signature is (2, 1). The absence of such an asymptotic<br />
in the homogeneous case with explicit Diophantine conditions is responsible for the<br />
“extra” assumption (i) in Theorem 1.10. Indeed, using this assumption, we are able to<br />
reduce the study to an investigation of the contribution from null vectors. We are then<br />
able to impose and utilize an explicit Diophantine condition on ξ (see Section 7 for<br />
details).<br />
Let us also mention that the Diophantine conditions in Theorems 1.10 and 1.9<br />
are necessary. Similar to [EMM1, Theorem 2.2], [Mark, Theorem 1.3], and [Sar,<br />
Theorem 2], it is possible to construct sets of second Baire category of quadratic<br />
forms for which the above fail. To be more explicit, it is shown in [EMM1], using<br />
category arguments, that there exists a dense set of second Baire category γ such<br />
that NQ,(1/8, 2,Ti) >Ti(log Ti) 1−ε for an infinite sequence Ti, where Qγ (x) =<br />
x2 1 + x2 2 − γ 2 (x2 3 + x2 4 ). Nowifwetakeanyξ∈ Z4 , then the same holds for Q γ<br />
ξ .<br />
This already gives the examples we wanted. However, these examples are in a sense<br />
“degenerate" since ξ ∈ Z4 . We will reproduce the argument in [EMM1] to get more<br />
“generic" counterexamples. First, note that an argument like that in [Mark, Appendix<br />
10] shows that, for any (α, ς) ∈ Q × Q4 ,wehave<br />
# v ∈ Z 4 : Q α<br />
ς (v) = 0 and v ≤T ∼ Cς,αT 2 log T. (15)<br />
Now the same argument as in [EMM1, Lemma 3.15] (see also [Mark, Section 9])<br />
gives: for every ε>0 and any S>0, the set (β,ζ) ∈ R × R 4 for which there exists<br />
T>Ssuch that NQ β ,ζ,(1/4, 1,T) ≥ T 2 (log T ) 1−ε is dense.<br />
Given that S>0, letUS be the set of ordered pairs (γ,ξ) ∈ R × R 4 for which<br />
there exists (β,ζ) ∈ R × R 4 and T>Ssuch that<br />
NQ β ,ζ,(1/4, 1,T) ≥ T 2 (log T ) 1−ε , |β − γ |
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 127<br />
Let the Hamiltonian be the geodesic flow for flat 2-tori. Sarnak [Sar] proved that,<br />
for almost all 2-dimensional flat tori (with respect to Lebesgue measure on the moduli<br />
space of 2-dimensional flat tori), the pair correlation density function converges to the<br />
pair correlation density of a Poisson process. He used averaging arguments to reduce<br />
the pair correlation problem to one of spacing between the values at integers of binary<br />
quadratic forms. This is related to the quantitative Oppenheim problem in the case of<br />
signature (2, 2). One corollary of Theorem 1.3 is that the Berry-Tabor conjecture holds<br />
for the pair correlation of 2-dimensional flat tori under certain explicit Diophantine<br />
conditions.<br />
Similarly, Theorem 1.9 has a corollary in this direction. Let h be a lattice in R 2 .<br />
Also, let α = (α1,α2) ∈ R 2 . Now the eigenvalues of the Laplacian<br />
− =− ∂2 ∂2<br />
−<br />
∂y2 ∂y2 with quasi-periodicity conditions<br />
(17)<br />
φ(x + v) = e 2πi〈α,v〉 φ(x) for all x ∈ R 2 and all v ∈ h (18)<br />
are of the form 4π 2 w + α 2 , where w ∈ h ∗ and where h ∗ is the dual lattice to h. Let<br />
0 ≤ λ0
128 MARGULIS and MOHAMMADI<br />
α ∈ R 2 be given, and define β as above. Suppose that at least one of the following<br />
holds.<br />
(i) The vector β = (β1,β2) is Diophantine.<br />
(ii) There exists N,C > 0 such that, for all triples of integers (p1,p2,q) with<br />
q ≥ 2, we have<br />
<br />
<br />
max Ai −<br />
i=1,2<br />
pi<br />
<br />
<br />
><br />
q<br />
C<br />
q<br />
Then for any interval (a,b) with 0 /∈ (a,b), we have<br />
lim<br />
T →∞ Rh,α(a,b,T ) = c 2<br />
h (b − a). (22)<br />
Hence the spectrum satisfies the Berry-Tabor conjecture for the pair correlation<br />
function.<br />
The case of h = Z 2 was proved by Marklof [Mark]. His approach utilizes results from<br />
the theory of unipotent flows combined with an application of theta sums. We also use<br />
the theory of unipotent flows in our proof; however, our strategy to control the integral<br />
of unbounded functions over certain orbits is dynamical and rests heavily on [EMM1]<br />
and [EMM2].<br />
Outline of the proof. Let Qξ be a quadratic form of signature (p, q),andletn = p+q.<br />
Fix an interval (a,b), andletU⊂Rnbe a “suitably chosen" compact set such that<br />
a
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 129<br />
Quadratic forms. Let n ≥ 3, andletn = p + q, where p ≥ 2. Let {e1,...,en} be<br />
the standard basis for R n . If p ≥ 3, let B be the “standard" form<br />
i<br />
B<br />
i=1<br />
xiei<br />
<br />
= 2x1xn +<br />
p<br />
i=2<br />
x 2<br />
i −<br />
n−1<br />
i=p+1<br />
x 2<br />
i . (23)<br />
Let H = SO(B),andlet{at} be the 1-parameter subgroup of H given by ate1 = e −t e1,<br />
atei = ei for 2 ≤ i ≤ n − 1, andaten = e t en. And let K = H ∩ ˆK, where ˆK is the<br />
group of orthogonal matrices with determinant 1. We let dk denote the Haar measure<br />
on K normalized so that K is a probability space.<br />
If (p, q) = (2, 2), welet<br />
B(x1,x2,x3,x4) = x1x4 − x2x3<br />
be the standard form on R4 . This is the determinant on M2(R) if we identify R4 with M2(R). Note that this identification shows that SO(2, 2) is locally isomorphic<br />
to SL2(R) × SL2(R) with the action v → g1vg −1<br />
2 , which leaves the determinant<br />
invariant. We let H = SL2(R) × SL2(R), K = SO(2) × SO(2), andat = (bt,bt),<br />
where bt = diag(e −t/2 ,e t/2 ). We let dk denote the Haar measure on K normalized so<br />
that K is a probability space. We often work with the standard lattice Z 4 and the form<br />
Q, in which case we continue to denote by {at} and K the corresponding 1-parameter<br />
and maximal compact subgroup of SO(Q).<br />
If (p, q) = (2, 1), welet<br />
B(x1,x2,x3) = x1x3 − x 2<br />
2<br />
be the standard form on R3 . This is the determinant on Sym2(R), the space of 2 × 2<br />
symmetric matrices, if we identify R3 with Sym2(R). This identification shows that<br />
SO(2, 1) is locally isomorphic to SL2(R) with the action v → gvtg, where tg is<br />
the transpose matrix. We let H = SL2(R). We let at = diag(e−t/2 ,et/2 ),andwe<br />
let K = SO(2) be the maximal compact subgroup of H. As before, dk denotes the<br />
normalized Haar measure on K.<br />
Let f be a continuous function with compact support on Rn . We define the theta<br />
transform of f by<br />
f ˆ(<br />
+ ξ) = <br />
f (v), (26)<br />
v∈+ξ<br />
where +ξ is any unimodular inhomogeneous lattice in R n . Note that ˆ<br />
f is a function<br />
on the space of inhomogeneous lattices.<br />
(24)<br />
(25)
130 MARGULIS and MOHAMMADI<br />
We fix some more notations. Let n = p + q, andletG = SLn(R) ⋉R n . Let<br />
Ɣ = SLn(Z) ⋉Z n , which is a lattice in G. We have the following, which is similar to<br />
Siegel’s integral formula.<br />
LEMMA 2.1<br />
Let f and fˆ be as above. Let µ be a probability measure on G/ Ɣ which is invariant<br />
under Rn . Then<br />
<br />
<br />
f ˆ(g)<br />
dµ(g) =<br />
G/ Ɣ<br />
Rn f (x) dx. (27)<br />
Proof<br />
Note that Rn is the unipotent radical of G. Now the lemma follows from Fubini’s<br />
theorem and the fact that µ is Rn-invariant. <br />
We end this section by recalling the definition of the α functions defined on the space<br />
of lattices. Let be a lattice in R n . A subspace L of R n is called -rational if L ∩ <br />
is a lattice in L. For a -rational subspace L, letd(L) be the volume of L/(L ∩ ).<br />
For 0 ≤ i ≤ n,define<br />
<br />
1<br />
αi() = sup<br />
d(L)<br />
<br />
: L is a -rational subspace of dimension i , (28)<br />
and let α() = maxi αi(). Now if ξ = + ξ is an inhomogeneous lattice, let<br />
α(ξ) = α(). There is a constant c = c(f ) depending on f such that, for any<br />
inhomogeneous lattice ξ = + ξ, wehave<br />
ˆ<br />
f (ξ)
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 131<br />
Theorem 1.6 is proved using the following, which is a result of combining Theorems<br />
3.1, A.3, andA.4. We have the following.<br />
<strong>THEOREM</strong> 3.2 ([EMM1, Theorem 3.5])<br />
Suppose that p ≥ 3 and q ≥ 1. Let f and fˆ be as above. Let ν be any continuous<br />
function on K. Then, for every compact subset D of G/ Ɣ, there exist finitely many<br />
points x1,...,xℓ ∈ G/ Ɣ such that<br />
(i)<br />
(ii)<br />
the orbit Hxi is closed and has finite H -invariant measure for all i;<br />
for any compact set F ⊂ D \ <br />
i Hxi, there exists t0 > 0 such that, for all<br />
x ∈ F and t>t0, we have<br />
<br />
<br />
<br />
<br />
f ˆ(atkx)<br />
ν(k) dk −<br />
<br />
fdµ ˆ<br />
<br />
<br />
νdk<br />
≤ ε, (31)<br />
K<br />
G/ Ɣ K<br />
where either µ is the G-invariant measure on G/ Ɣ or H ⋉R n x is closed and has<br />
H ⋉R n -invariant probability measure and µ is this measure.<br />
Proof<br />
We may and will assume that φ is nonnegative. Now define<br />
A(r) = ∈ G/ Ɣ : α1() >r . (32)<br />
Let gr be a continuous function on G/ Ɣ such that gr() = 0 if /∈ A(r),gr() = 1<br />
for all ∈ A(r + 1), and0≤ gr() ≤ 1 if r ≤ α1() ≤ r + 1. We have<br />
f ˆ = ( fˆ − fgr) ˆ + fgr. ˆ Note that fˆ − fgr ˆ is a continuous function with compact<br />
support on G/ Ɣ.<br />
Note that H 0 ⋉Rn is a maximal connected subgroup of G. Hence for every<br />
δ>0, there exists r0 such that if H ⋉Rny is a closed orbit of H ⋉Rn in G/ Ɣ with<br />
an H ⋉Rn-invariant probability measure σ ,thenσ (A(r) ∩ H ⋉Rny) r0. Hence for r sufficiently large, we get<br />
<br />
<br />
<br />
fdµ− ˆ ( fˆ − fgr) ˆ <br />
dµ
132 MARGULIS and MOHAMMADI<br />
Hence, if we apply Theorem 3.1, then there is a constant B depending on B1 such that<br />
we get<br />
<br />
( fgr)(atkx) ˆ ν(k) dk ≤ B sup |ν(k)| r −β/2<br />
(35)<br />
K<br />
for all x ∈ D.<br />
Now choose r>r0sufficiently large so that B(supk∈K |ν(k)|)r−β/2
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 133<br />
above. Define<br />
f ˜(g<br />
: qξ) = <br />
v∈X(qξ )<br />
f (gv). (36)<br />
The following is an analogue of [EMM2, Theorem 2.3] and will provide us with<br />
the upper bound required for the proof of Theorem 1.9. The proof of this theorem is<br />
the main technical part of this paper and will occupy the rest of this paper.<br />
<strong>THEOREM</strong> 4.1<br />
Let G, H, K, and {at} be as in Section 2 for the signature (2, 2) case. Let Qξ be a<br />
quadratic form of signature (2, 2) which is Diophantine. Let q ∈ SL4(R) and qξ be<br />
as above. Let ν be a continuous function on K. Then we have<br />
<br />
<br />
<br />
lim sup<br />
t→∞<br />
˜f (atk : qξ)ν(k) dk ≤ f ˆ(g)<br />
dµ(g)<br />
K<br />
G/ Ɣ<br />
K<br />
ν(k) dk, (37)<br />
where µ is the G-invariant probability measure on G/ Ɣ if the homogenous part, Q,<br />
is irrational and where the H ⋉R 4 -invariant probability measure on the closed orbit<br />
is H ⋉R 4 · qξ if Q is a rational form.<br />
Proof of Theorem 1.9<br />
Suppose that Qξ is as in the statement of Theorem 1.9. An argument like that<br />
of [EMM1, Sections 3.4, 3.5] combined with Theorem 4.1 gives: if 0 /∈ (a,b),<br />
then<br />
lim sup<br />
T →∞<br />
NQ,ξ,(a,b,T ) = lim sup ÑQ,ξ,(a,b,T ) ≤ λQ,(b − a)T<br />
T →∞<br />
2 . (38)<br />
This upper bound, combined with the lower bound obtained by Theorem 1.4, proves<br />
Theorem 1.9. <br />
The proof of Theorem 4.1 extensively utilizes results and ideas from [EMM1]<br />
and [EMM2], and we will try to use their terminology and notation for the convenience<br />
of the reader. We recall these theorems and terminology when we need them. Let us<br />
start with the following.<br />
<strong>THEOREM</strong> 4.2<br />
Let {at} and K be as in Theorem 4.1. Let be any lattice in R 4 . Then for i = 1, 3<br />
and any ε>0, we have<br />
<br />
sup αi(atk)<br />
t>0 K<br />
2−ε dk < ∞. (39)
134 MARGULIS and MOHAMMADI<br />
Hence there exists a constant c depending on ε and such that, for all t>0 and<br />
0
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 135<br />
If L is a µ1-quasi-null subspace and if T/2 ≤v L ≤T, then there exists C>0<br />
depending on µ1 such that either π2(v L )
136 MARGULIS and MOHAMMADI<br />
Indeed, an estimate like (43) would actually hold for f ˆ (by virtue of [Sch, Lemma<br />
2]) if A was defined by taking the union over all quasi-null subspaces. But we have<br />
replaced fˆ by f ˜,<br />
and this implies that we can take the union over Q instead. To see<br />
this, consider one of these subspaces, say L1. Assume that ξ ∈ L1 + w1ξ, where<br />
w1ξ ∈ Z4 . Now if there is k ∈ K such that (44) is satisfied with L = L1 and v ∈ Z4 ,<br />
then there is a constant cξ ≥ 1 depending on ξ such that d(atkH) 1/cξδ}. Hence there is c such that (43)<br />
holds.<br />
Now the proof of Theorem 4.1 will be completed if we can show that there is<br />
some η>0 depending on Qξ such that |AL t (δ)|
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 137<br />
Furthermore, if L is a null subspace (of the first type), then<br />
k ∈ K : d(atkL) 0 be small, and let<br />
A L,1<br />
t (δ, ε)<br />
<br />
=<br />
k ∈ K : δ1+ε
138 MARGULIS and MOHAMMADI<br />
For simplicity, let ζ = qξ, where q ∈ SL4(R) was chosen such that Q(v) =<br />
B(qv) for all v ∈ R 4 . Fix t > 0 and 0
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 139<br />
(i) there exists λL ∈ such that, if for some λ ∈ there is kφ ∈ EL t (δ, kθ,ε) for<br />
which the plane (L + λ + ζ )φ intersects B(r), then L + λ = L + λL;<br />
(ii) for all λ ∈ , we have maxkφ (at(kθ,kφ)[λ + ζ ]) Lφ ⊥ >c/(δ(1−ε)/2 ).<br />
Furthermore, if (ii) holds, then there exists a computable constant η1 > 0 such that,<br />
for 0 0 such that the function hλ is (C1,β1) good. It follows from the<br />
definition of (C, α)-good functions (see [KM]) that fλ(φ) =(at(kθ,kφ)(v L ∧ (λ +<br />
ζ )) is (C2,β2)-good for some C2,β2 > 0.<br />
Recall now that<br />
c<br />
maxkφ∈E (at(kθ,kφ)[λ + ζ ]) ⊥ Lφ > δ (1−ε)/2 ,<br />
δ1+ε
140 MARGULIS and MOHAMMADI<br />
Note that, for φ ∈ Eλ,wehavefλ(φ)
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 141<br />
let aφ =(at(kθ,kφ)[λ0 + ζ ]) Lφ ⊥. Recall that T/2 ≤vL ≤T, hence aT/2 ≤<br />
w ≤aT. As was mentioned above, K2 acts on the plane spanned by {e123,e124}<br />
by rotation on R 2 , and e123 is the contracting direction of {at}. Also note that we are<br />
assuming that |E L t (λ0)| >δ η2 . Hence, there exist kφ ∈ E L t (λ0) such that<br />
aT δ η2<br />
2 ≤ at(kθ,kφ)[v L ∧ (λ0 + ζ )] = aφat(kθ,kφ)v L . (57)<br />
Note now that we have aφ ≤ r and at(kθ,kφ)v L ≤δ. So we get a ≤ (2rδ 1−η2 )/T,<br />
as we wanted to show. <br />
Before proceeding, let us draw the following corollary. This will be used in the proof<br />
of Theorem 4.1 to control the contribution of “small” subspaces.<br />
COROLLARY 5.5<br />
Let η1 be as in Proposition 5.3, and let η2 1, there<br />
is a δ0 = δ0(M,) such that if δ2, one of the following holds.<br />
(i) The number of quasi-null subspaces of Q of norm between T/2 and T is<br />
O(T 1−τ2 ).
142 MARGULIS and MOHAMMADI<br />
(ii) There exists a split integral form Q ′ with coefficients bounded by a fixed power<br />
of T τ2 and 1 ≤ λ ∈ R satisfying Q − (1/λ)Q ′ ≤T −ρ such that the number<br />
of quasi-null subspaces of Q with norm between T/2 and T which are not null<br />
subspaces of Q ′ is O(T 1−τ2 ).<br />
We refer to [EMM2, Section 10] and also Section 7 of this paper for a more careful<br />
analysis of this theorem. The main ingredient in the proof is the system of inequalities<br />
from [EMM1]. The main difference is that these inequalities are applied to a certain<br />
dilated lattice in 2 R 4 = R 6 .<br />
Let us now outline the rest of the proof. Let Qξ be the inhomogeneous quadratic<br />
form as in the statement of Theorem 1.9. We apply the above theorem with appropriate<br />
parameters ρ,τ2 to be determined later. Let T ≥ 2 be given. Now if (i) above holds<br />
(which is always the case if Q is not EWAS), then we already have a good control on<br />
the number of quasi-null subspaces in question, and we will get the desired control<br />
on the measure of the set <br />
L AL t (δ, ε). Hence, we may assume that (ii) above holds.<br />
Again, if L is not a null subspace of the appropriate approximation of Q, we proceed<br />
as in case (i). So we need to consider the contribution from quasi-null subspaces which<br />
are null subspaces of some rational approximation. In this case, using Lemma 5.3,we<br />
are reduced to the case where only one translate of L has contribution. We then use<br />
the Diophantine property of ξ, and we get a control on the number of such subspaces.<br />
This completes the proof.<br />
We need to fix some more notations before proceeding with the above outline.<br />
If Q is a rational form, we may choose µ1 small enough such that all µ1-quasi-null<br />
subspaces are null subspaces. Also in this case, by replacing Q with a scalar multiple,<br />
we may and will assume that Q is a primitive integral form.<br />
Let T ≥ 2 be a fixed number. Recall from Theorem 5.6 that there are two<br />
possibilities for quasi-null subspaces. The case which requires more study is case<br />
(ii), so let us assume that we are in this case. Let QT = Q ′ (resp., QT = Q) ifQ<br />
is an irrational form (resp., if Q is split integral form), where Q ′ is given as in (ii)<br />
of Theorem 5.6. LetQt(δ, ε, τ2) be the set of all such quasi-null subspaces (resp.,<br />
null subspaces if Q is an rational) which are not exceptional subspaces and such that<br />
A2 δ (•,ε) is nonempty for them. Let L ∈ Qt(δ, ε, τ2) be such a subspace. We further<br />
assume that L is of the first type and that T/2 ≤vL≤T. Since QT is a split integral form, after possibly multiplying by a scalar bounded<br />
by a fixed power of T τ2 , there exists a nonsingular integral matrix such that QT (v) =<br />
B(pv). Furthermore, Theorem 5.6 guarantees that we may choose p such that its<br />
entries are bounded by a fixed power of T τ2 . Recall that the null subspaces of B are<br />
of two types based on the fact that the corresponding vector vL is in V1 or V2. From<br />
now on by a null space of first kind for B, we mean the following.
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 143<br />
Null subspaces of the first type. These are subspaces which are orbits of x11<br />
0<br />
x12 under<br />
0<br />
SL2 × SL2.<br />
Throughout the rest of the section, by a null subspace we mean a rational null<br />
subspace of the first type. Now let M be a rational null subspace of the first type for<br />
B. Such subspaces are characterized as annihilators of primitive row vectors. Hence<br />
an integral basis for M is m 0 0 m<br />
, , where gcd(m, n) = 1. We refer to this basis<br />
n 0 0 n<br />
as the standard integral basis for M.<br />
Recall that L ∈ Qt(δ, ε, τ2) with T/2 ≤vL≤T.Since L is a null subspace of<br />
QT , the subspace M = pL is a null subspace of B. Now let {v1,v2} be the standard<br />
basis for M, and let wi be the unique primitive integral multiple of p−1vi. Then<br />
{w1,w2} is a basis for L. Furthermore, since p is an integral matrix whose entries are<br />
bounded by a fixed power of T τ2 1/2−τ3 , we have T ≤wi ≤T 1/2+τ3 , where τ3 is a<br />
fixed multiple of τ2. We refer to the basis constructed in this way as a τ3-round basis<br />
for L.<br />
Fix 0 δ η2 ,<br />
in particular, that (i) of Proposition 5.3 holds for some λ ∈ . Fix{w1,w2} as a<br />
τ3-round. Note that τ3 is fixed when τ2 is chosen.<br />
Let q −1 λ = v ∈ Z 4 . Now, using Lemma 5.4, there is a constant c = c(q) such<br />
that (v + ξ)L ⊥ < (c(q)δ1−η2 )/T. This and the fact that L is a null subspace of QT<br />
give<br />
|〈wi ,v+ ξ〉QT |≤T 1/2+τ3 · cδ 1−η2<br />
T<br />
cδ1−η2<br />
=<br />
T 1/2−τ3<br />
, (58)<br />
where c is an absolute constant depending on Q. So {〈wi ,ξ〉QT }≤(cδ 1−η2 )/(T 1/2−τ3 ),<br />
where {}denotes the distance to the closest integer. Let us now collect the result of<br />
the above discussion in the following.<br />
LEMMA 5.7<br />
Let ε and τ2 be small, and let L ∈ Qt(δ, ε, τ2). Letkθ ∈ p1(A qL<br />
t (δ, ε)), and suppose<br />
that |E qL<br />
t (δ, kθ,ε,ξ)| >δ η2 with η2 as in Lemma 5.4. In particular, (i) in Proposition<br />
5.3 holds for qL and some λ ∈ . Then {〈wi ,ξ〉QT }≤(cδ 1−η2 )/(T 1/2−τ3 ),<br />
where τ3 is a fixed multiple of τ2 as above. Furthermore, since {w1,w2} is the image of<br />
standard basis {v1,v2} of M = pL, then we have {〈vi , pξ〉B} ≤(cδ 1−η2 )/(T 1/2−τ3 ).<br />
This lemma brings us to the situation where we can now use the Diophantine property<br />
of the vector ξ.<br />
As we mentioned in the brief outline following it, Theorem 5.6 deals with the<br />
Diophantine properties of Q. If Q fails to have the desired Diophantine condition, ξ
144 MARGULIS and MOHAMMADI<br />
needs to be a Diophantine vector thanks to our assumption on Qξ. In what follows,<br />
we make use of this assumption, and we control the number of quasi-null subspaces<br />
for which Lemma 5.7 can hold.<br />
The following is a simple consequence of Definition 1.7 (i.e., the Diophantine<br />
condition). The proof is easy, and we include it for the sake of completeness (and also<br />
for later reference).<br />
LEMMA 5.8<br />
Let x = (x1,...,xn) ∈ R n be a κ-Diophantine vector. Then for any κ ′ > (n+1)κ +1,<br />
we have that, for all β > 0 and all 0
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 145<br />
Proof<br />
We may and will assume that L is of the first type. We continue to use the notations<br />
as in Lemma 5.7. In particular, let M = pL, andlet{v1,v2} be the standard basis<br />
for M. Then Lemma 5.7 implies that {〈vi, pξ〉B} ≤(cδ 1−η2 )/(T 1/2−τ3 ). Recall that<br />
v1 = (m, 0,n,0) and v2 = (0,m,0,n), where gcd(m, n) = 1, and we have<br />
T 1−τ3 ≤v M =m 2 + n 2 ≤ T 1+τ3 , (60)<br />
where τ3 is a fixed multiple of the constant τ2 appearing in Theorem 5.6. We also note<br />
that 〈v1 , pξ〉B = m(pξ)4 − n(pξ)2 and that 〈v2 , pξ〉B(pξ)1 − m(pξ)3.<br />
Let τ ′ 1 ≪ 1/32κ be chosen. Then choose τ2 in Theorem 5.6 such that τ2 T τ ′ 1. Taking τ2 even smaller, we may and will assume that<br />
τ3
146 MARGULIS and MOHAMMADI<br />
Proof<br />
Let η1 be as in Proposition 5.3, andletη2 δ η2 for some kθ ∈ p1(A qL<br />
t (δ, ε)) is O(T 1−τ1 ). On the<br />
other hand, using Theorem 5.6, we see that the number of quasi-null subspaces with<br />
T/2 ≤v L ≤T and A L t (δ, ε) = ∅which are not in Qt(δ, ε, τ2) is O(T 1−τ2 ). The<br />
corollary follows with η = η1/2 and τ = min{τ1,τ2}. <br />
6. Proof of Theorem 4.1<br />
We complete the proof, of Theorem 4.1 in this section. Before proceeding to the proof,<br />
we need the following two statements.<br />
LEMMA 6.1<br />
If Qξ is an irrational (2, 2) form, then the number of 2-dimensional null subspaces,<br />
say L,ofQ such that ξ ∈ v + L for some v ∈ Z 4 is at most four.<br />
Proof<br />
First, note that using [EMM2, Lemma 10.3], we may and will assume that Q is a<br />
rational form. In this case, we actually show that there are at most two such subspaces.<br />
In order to see this, suppose that there are two null rational subspaces of the same<br />
type, say Li, for i = 1, 2 such that ξ ∈ vi + Li, where vi ∈ Z 4 . Now since the Li are<br />
of the same type, they are transversal. Let {w i 1 ,wi 2 } be an integral basis for Li. Then<br />
〈ξ,w i j 〉B ∈ Z for i, j = 1, 2. This (thanks to the transversality of the Li) implies that<br />
ξ is a rational vector, which is a contradiction. <br />
The following is essential to the proof of Theorem 4.1 and proved in Section 7.<br />
PROPOSITION 6.2<br />
The number of quasi-null subspaces with norm between T/2 and T is O(T ).<br />
As we mentioned, this is proved in Section 7. However, for the time being, let us<br />
remark that in the case where Q is a split integral form, this is immediate. Indeed,<br />
in that case we are dealing with null subspaces, and hence we need to show this for<br />
Q = B. As we observed, however, the null subspaces of B are classified by primitive<br />
integral vectors (m, n). Now if L is a null subspace of either type which corresponds<br />
to (m, n), then v L =m 2 + n 2 . Thus the result is obvious.<br />
We now turn to the proof of Theorem 4.1.
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 147<br />
Proof of Theorem 4.1<br />
We may and will assume that ˜f is nonnegative. Now define<br />
A(r) = ∈ X : α1() >r . (63)<br />
Let gr be a continuous function on X such that gr() = 0 if /∈ A(r), gr() = 1<br />
for all ∈ A(r + 1), and0 ≤ gr() ≤ 1 if r ≤ α1() ≤ r + 1. Let the constant c<br />
(depending on f and ξ) beasin(43), and let s ∈ N be a large number. Define<br />
( fgr) ˜ ≤<br />
s = ( fgr)χ ˜<br />
{ : fgr ˜ ()≤c2s } and ( fgr) ˜ ><br />
s = ( fgr)χ ˜<br />
{ : fgr ˜ ()>c2s }. (64)<br />
Now let µ be as in the statement of Theorem 4.1.Sincef˜−fgr ˜ is bounded and<br />
continuous, Theorems A.3 and A.4 imply that<br />
<br />
<br />
lim sup ( f˜ − fgr)(atkqξ)ν(k) ˜<br />
dk ≤ fdµ ˆ νdk. (65)<br />
t→∞ K<br />
G/ Ɣ K<br />
Hence Theorem 4.1 will be proved if we show that: given ɛ>0, we can choose r0<br />
<br />
such that, if r>r0, then lim supt K ˜ fgr dk < ɛ.<br />
Now let ɛ>0 be an arbitrary small number. Let us first control the contribution<br />
of ( fgr) ˜ > s when s is large enough. Let M>0 beanumberfixed(fornow)andlarge<br />
enough such that 1/M < ɛ/6C,where C is a universal constant appearing in (70) and,<br />
in particular, is independent of µ1 in the definition of quasi-null subspaces. Further<br />
assume that M is large enough such that all exceptional subspaces have norm less<br />
than M. Let µ1 in the definition of quasi-null subspace be small enough such that,<br />
for all L ∈ Qt(δ, ε) with vL 0<br />
be large enough such that the conclusions of Theorems 4.2 and 4.5, for this µ1 which<br />
we chose, as well as of Corollaries 5.5 and 5.11, hold for δ = 1/2j whenever j>s.<br />
Recall now that we have<br />
<br />
k ∈ K : ˜f (atkξ) >c2 j ⊂ k ∈ K : α13(atkξ) > 2 j <br />
1<br />
∪ Bt<br />
2j <br />
1<br />
∪ At<br />
2j <br />
,<br />
where At(1/2 j ) and Bt(1/2 j ) are as in (43). Let Ct(1/2 j ) ={k ∈ K : α13(atkξ) ><br />
2 j }.Wehave<br />
<br />
K<br />
h ><br />
s (atKqξ) dk ≤ <br />
s
148 MARGULIS and MOHAMMADI<br />
all j>s,we have<br />
<br />
1<br />
Ct<br />
2j <br />
<br />
+<br />
<br />
Bt<br />
<br />
1<br />
2j <br />
<br />
+<br />
<br />
At<br />
1<br />
2 j<br />
<br />
\ <br />
L∈Qt ( 1<br />
2j ,ε)<br />
AL <br />
1<br />
<br />
<br />
t ,ε <<br />
2j 1<br />
. (68)<br />
2 (1+ε/4)j<br />
Let η1 < 1/4 be as in Proposition 5.3, andletη = η1/2. The Conclusions of<br />
Corollaries 5.5 and 5.11 hold with this η and the corresponding τ.Now let Q < t (1/2j ,ε)<br />
(resp., Q ≥ t (1/2 j ,ε)) be the set of quasi-null subspaces in Qt(1/2 j ,ε) with norm less<br />
than M (resp., greater than or equal to M.)Wehave<br />
| <br />
L∈Q < t (1/2j ,ε) AL 1<br />
t (<br />
| <br />
L∈Q ≥ t (1/2 j ,ε) AL t<br />
( 1<br />
2 j ,ε)|≤ <br />
i<br />
t i+j e /(2 )<br />
i et /(2i+j+1 )<br />
t i+j e /(2 )<br />
et /(2i+j+1 )<br />
2 j ,ε)| ≤ <br />
<br />
L |It (δ)| ≤C1 i<br />
L |It (δ)|≤C2<br />
<br />
<br />
i<br />
1<br />
2 (1+η)j+i/2 +<br />
1<br />
2 (1+η)j+i/2 ,<br />
e −τt<br />
2 (1/2−τ)i+(1−τ)j<br />
where C1 and C2 are absolute constants independent of µ1. The inequality in the<br />
first line above follows from Corollary 5.5 and from the fact that the definition of<br />
Qt(1/2 j ,ε) excludes exceptional subspaces. The inequalities in the second line follow<br />
from Corollary 5.11. Wenowhave<br />
<br />
2 j<br />
<br />
<br />
<br />
j>s<br />
L∈Qt<br />
<br />
1<br />
2j ,ε<br />
<br />
,<br />
(69)<br />
AL <br />
1<br />
<br />
1<br />
t ,ε ≤ C<br />
2j 2s 1<br />
<br />
+ . (70)<br />
η M<br />
Here C ′ is an absolute constant which depends on C in (70). This inequality together<br />
with (68) gives<br />
<br />
K<br />
( fgr) ˜ ><br />
s (atKqξ) dk ≤ C′<br />
+ C<br />
2εs/4 <br />
1 1<br />
+<br />
2ηs M<br />
<br />
. (71)<br />
Recall that M was chosen such that C/M < ɛ/6. Now we choose s large enough<br />
such that the right-hand side of (71)islessthanɛ/2. The above estimate holds for all<br />
r.<br />
As we mentioned in the proof of Theorem 3.2, There exists r0 = r0(s, ɛ) such<br />
that if r>r0, then µ(A(r)) r0(s, ɛ). We have<br />
lim sup<br />
t→∞<br />
<br />
K<br />
( fgr) ˜ ≤<br />
s (atkqξ)ν(k) dk ≤ 2 s lim sup<br />
t→∞<br />
<br />
K<br />
gr(atkqξ)ν(k) dk ≤ ɛ/2.<br />
(72)<br />
Thus (72) and(71) give:ifr>r0(s, ε), then<br />
<br />
lim sup<br />
t→∞<br />
fgr(atkqξ) ˜ dk < ɛ. (73)<br />
K
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 149<br />
This finishes the proof of Theorem 4.1. <br />
7. Proof of Theorem 1.10<br />
The proof of Theorem 1.10 is very similar to that of Theorem 1.9. Indeed, our study<br />
in this case is simpler as we are dealing with the case where the homogeneous part Q<br />
is rational. Hence, we only need to consider the contribution of null subspaces to the<br />
counting function.<br />
As before, let q ∈ SL3(R) be such that Q(v) = B(qv) for all v ∈ R 3 . Since Q is<br />
rational, we may assume that q is in PGL3(Q). Let p ∈ GL3(Q) be a representative<br />
for q. Let = qZ 3 , and define qξ = q(Z 3 + ξ). As in Section 4, letX(qξ)<br />
be the set of vectors in qξ not contained in qL, where L ⊂ Z 3 is an exceptional<br />
(1-dimensional) subspace for Q. An argument like that in Lemma 6.1 shows that there<br />
are at most three such subspaces when Q is rational and that ξ is an irrational vector.<br />
For any continuous compactly supported function f on R 3 , define<br />
f ˜(g<br />
: qξ) = <br />
v∈X(qξ )<br />
f (gv). (74)<br />
As in previous discussions, we reduce the proof of Theorem 1.10 to the following<br />
theorem.<br />
<strong>THEOREM</strong> 7.1<br />
Let G, H, K, and {at} be as in Section 2 for the signature (2, 1) case. Let Qξ be a<br />
quadratic form of signature (2, 1) as in the statement of Theorem 1.10.Letq∈ SL3(R)<br />
and qξ be as above. Let ν be a continuous function on K. Then we have<br />
lim sup<br />
t→∞<br />
<br />
K<br />
<br />
˜f (atk : qξ)ν(k) dk ≤<br />
G/ Ɣ<br />
<br />
f ˆ(g)<br />
dµ(g)<br />
K<br />
ν(k) dk, (75)<br />
where µ is the H ⋉R 3 -invariant probability measure on the closed orbit H ⋉R 3 ·qξ.<br />
The proof of this theorem is very similar to that of Theorem 4.1. We will use the<br />
same notations as those in the previous sections for the sake of simplicity. The main<br />
notational difference to bear in mind is that in previous sections L denotes a 2dimensional<br />
subspace, where in this section L is a 1-dimensional (null) subspace.<br />
As before, we need to study the subsets of K where the function ˜f is “large”.<br />
We start by recalling the following. There is c>0 such that, for all large t and small
150 MARGULIS and MOHAMMADI<br />
0 c<br />
<br />
⊂ k ∈ K : α1(atkqξ)<br />
δ<br />
> 1<br />
<br />
∪ k ∈ K : α2(atkqξ) ><br />
δ<br />
1<br />
<br />
. (76)<br />
δ<br />
We fix t and δ as above. Let us make two important remarks before we continue.<br />
Remarks 7.2<br />
(i) Using reduction theory of the orthogonal group, we see that if α2(atkqξ) ><br />
1/δ, then actually α1(atkqξ) > 1/δ. Hence we only need to study the contribution<br />
coming from α1.<br />
(ii) The fact that Q is a rational form implies that there exists δ0 depending only on<br />
Q such that if 0
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 151<br />
For any null subspace L of Q, the subspace pL is a (rational) null subspace of<br />
B. We let vL denote the standard basis for pL. For any such L, let AL t (δ) be the<br />
corresponding set in (77). We have the following.<br />
LEMMA 7.3<br />
There exists 0
152 MARGULIS and MOHAMMADI<br />
Some important properties of these intervals were proved in [EMM2] and recalled in<br />
Lemma 5.1. What is important for us in Section 7 is property (ii) in Lemma 5.1.This<br />
property, tailored to our current assumptions, gives<br />
|A L<br />
t<br />
−t<br />
L e δ<br />
1/2 (δ)| ≤|It (δ)| ≈ . (81)<br />
T<br />
Proof of Theorem 7.1<br />
Let M be a large number which is fixed for now. Let t>0 be a large number. Assume<br />
also that j is a large number. We assume throughout that η = η1, where η1 is as in<br />
Lemma 7.3. LetL0,L1 be nonexceptional null subspaces such that {〈v Lk , pξ〉B} <<br />
1/2 j for k = 0, 1 and such that v L0 ≤v L1 are minimal with these properties.<br />
Indeed, we fix any two such subspaces if there are more than two subspaces satisfying<br />
these conditions. Assume that e t /2 j+i0+1 ≤v L 2 jη2 , where η2 is as in Lemma 7.3. Now using this<br />
lemma, we see that, for all i ≤ i0, the number of null subspaces L with e t /2 j+i+1 ≤<br />
v L i0, the subspaces of norm at most e t /2 i+j have<br />
no contribution to the set At(1/2 j ), that is, A L t (1/2j ) =∅for all such subspaces L.<br />
Hence if we use this fact and (81), we have: if j is such that Case 1 holds, then<br />
we have<br />
<br />
<br />
<br />
<br />
Lnull<br />
AL <br />
1<br />
t<br />
2j <br />
<br />
<br />
≤<br />
i<br />
≤ 1<br />
2 j(1+η)<br />
<br />
<br />
<br />
et /(2i+j+1 )≤vL
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 153<br />
Now, if we argue just as in Case 1, we get<br />
<br />
<br />
<br />
L null<br />
AL <br />
1<br />
t<br />
2j <br />
<br />
≤<br />
i<br />
<br />
<br />
<br />
et /(2i+j+1 )≤vL
154 MARGULIS and MOHAMMADI<br />
of G, and let K be the maximal compact subgroup of H . The question which is<br />
addressed in this section is that of the equidistribution of sets of the form atKx,<br />
where x ∈ G/ Ɣ and where A ={as : s ∈ R} is a suitable 1-parameter subgroup<br />
of H. This is a well-studied question (see [Sh2]). The main tools in the analysis are<br />
indeed Ratner’s theorem on classification of unipotent flow-invariant measures on<br />
G/ Ɣ, and linearization techniques for the action of unipotent groups on G/ Ɣ which<br />
were developed by Dani and Margulis [DM].<br />
Let H and W be closed subgroups of G. Following Dani and Margulis [DM],<br />
define X(H,W) ={g ∈ G : Wg ⊂ gH}. We recall the following.<br />
<strong>THEOREM</strong> A.1 ([DM, Theorem 3])<br />
Let G be a connected Lie group, and let Ɣ be a lattice in G. Let U ={ut} be<br />
an Ad-unipotent 1-parameter subgroup of G. Let φ be a bounded continuous function<br />
on G/ Ɣ. Let D be a compact subset of G/ Ɣ, and let ε>0 be given. Then<br />
there exist finitely many proper closed subgroups H1 = H1(φ,D,ε),...,Hk =<br />
Hk(φ,D,ε) such that Hi ∩ Ɣ is a lattice in Hi for all i, and compact subsets<br />
C1 = C1(φ,D,ε),...,Ck = Ck(φ,D,ε) of X(H1,U),...,X(Hk,U), respectively,<br />
for which the following holds. For any compact subset F of D \ <br />
i CiƔ/Ɣ<br />
there exists T0 > 0 such that, for all x ∈ F and T>T0, we have<br />
<br />
<br />
1<br />
T<br />
<br />
<br />
φ(utx) dt − φdg<br />
0.<br />
Note that the standard representation of H 0 on Rn is irreducible, hence H 0 is a<br />
maximal connected subgroup of H ⋉Rn . Note also that as H 0 is a maximal connected<br />
subgroup of SLn(R), we have that H 0 ⋉Rn is a maximal connected subgroup of G.
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 155<br />
Now let Ɣ be a lattice in G. Then Ɣ ∩ R n is a lattice in R n . We let = ϑ(Ɣ) =<br />
Ɣ/ Ɣ ∩ R n (this is a lattice in SLn(R)). We abuse the notation and let ϑ also denote<br />
the projection from G/ Ɣ onto SLn(R)/. We have the following.<br />
LEMMA A.2<br />
Let x ∈ G/ Ɣ. Then the orbit Hϑ(x) is closed in SLn(R)/ if and only if H ⋉R n x<br />
is closed in G/ Ɣ.<br />
Proof<br />
Note that closed orbits of H ⋉R n (resp. H ) have a finite H ⋉R n -invariant (resp.,<br />
H -invariant) measure by [Mar86, Section 3]. Let x = (g, v)Ɣ. Note that H ⋉R n x =<br />
ϑ −1 (Hϑ(x)), hence if the Hϑ(x) is closed, then so is H ⋉R n x. Suppose now that<br />
H ⋉R n x is closed. Then Ɣ1 = H ⋉R n ∩ (g, v)Ɣ(g, v) −1 is a lattice in H ⋉R n ,<br />
and hence Ɣ1 ∩ R n is a lattice in R n and Ɣ1/Ɣ1 ∩ R n is a lattice in H. Hence Hϑ(x)<br />
is closed, as we wanted. <br />
The following is special case of [EMM1, Theorem 4.4].<br />
<strong>THEOREM</strong> A.3<br />
Let the notation be as above. Further assume that is a lattice in H 0 ⋉R n . Let φ<br />
be a compactly supported continuous function on H 0 ⋉R n /. Then for every ε>0<br />
and any bounded measurable function ν on K, and for every compact subset D of<br />
H 0 ⋉R n /, there exist finitely many points x1,...,xℓ ∈ H 0 ⋉R n / such that<br />
(i) the orbit H 0 xi is closed and has finite H 0 -invariant measure for all i;<br />
(ii) for any compact set F ⊂ D \ <br />
i Hxi, there exists s0 > 0 such that, for all<br />
x ∈ F and s>s0, we have<br />
<br />
<br />
<br />
φ(askx) ν(k) dk −<br />
K<br />
H 0 ⋉R n /<br />
<br />
φdg<br />
K<br />
<br />
<br />
νdk<br />
≤ ε. (89)<br />
We also need a slight variant of [EMM1, Theorem 4.4]. This is the content of the<br />
following.<br />
<strong>THEOREM</strong> A.4<br />
Let G, H, K, and Ɣ be as above. Let A ={as : s ∈ R} also be as above. Let φ be<br />
a compactly supported continuous function on G/ Ɣ. Then for every ε>0 and any<br />
bounded measurable function ν on K, and for every compact subset D of G/ Ɣ,there<br />
exists finitely many points x1,...,xℓ ∈ G/ Ɣ such that<br />
(i) the orbit H ⋉R n xi is closed and has finite H ⋉R n -invariant measure for all<br />
i;
156 MARGULIS and MOHAMMADI<br />
(ii) for any compact set F ⊂ D \ <br />
i H ⋉Rn xi, there exists s0 > 0 such that, for<br />
all x ∈ F and s>s0, we have<br />
<br />
<br />
<br />
φ(askx) ν(k) dk −<br />
K<br />
G/ Ɣ<br />
<br />
φdg<br />
K<br />
<br />
<br />
νdk<br />
≤ ε. (90)<br />
Proof<br />
The proof of this theorem goes along the same lines as in [EMM1, Section 4]. Let U be<br />
as defined above. Let Hi = Hi(φ,KD,ε) and Ci = Ci(φ,KD,ε) for i = 1,...,k<br />
be given as in Theorem A.1 corresponding to U.For1 ≤ i ≤ k, define<br />
Gi = g ∈ G : Kg ⊂ X(Hi,U) . (91)<br />
Note that the group generated by <br />
k∈K k−1Uk is H 0 , as U is not contained in any<br />
proper normal subgroup of H and K is the maximal compact subgroup of H. Now let<br />
g ∈ Gi. Then k−1Uk ⊂ gHig−1 for all k ∈ K. Hence, H 0 ⊂ gH 0<br />
i g−1 . Now as H 0<br />
acts irreducibly on Rn , the only possibilities for gH 0<br />
i g−1 are H 0 and H 0 ⋉Rn . Thus<br />
we have<br />
gH 0<br />
i g−1 ⊂ H 0 ⋉R n , for any g ∈ Gi, 1 ≤ i ≤ k. (92)<br />
We also note that if g ∈ G is such that gH 0 g −1 ⊂ H 0 ⋉R n , then g ∈ NSLn(R)(H 0 ) ⋉<br />
R n . This fact and (92) say that if g1,g2 ∈ Gi for some i, then g −1<br />
1 g2 ∈ NSLn(R)(H 0 ) ⋉<br />
R n . Since H 0 ⋉R n is of finite index in NSLn(R)(H ) ⋉R n , we get that Gi can be<br />
covered by finitely many cosets of H 0 ⋉R n . Hence there are finitely many points<br />
x1,...,xℓ ∈ G/ Ɣ such that the H ⋉R n xi are closed and have a finite H ⋉R n -<br />
invariant measure for all 1 ≤ i ≤ ℓ and such that<br />
<br />
GiƔ/ Ɣ ⊂ <br />
H ⋉Rn xi. (93)<br />
i<br />
1≤i≤ℓ<br />
Note that since the X(Hi,U) are analytic submanifolds of G and since K is connected,<br />
we have that, for any g ∈ G \ <br />
i GiƔ/ Ɣ,<br />
<br />
<br />
k ∈ K : kg ∈ <br />
<br />
X(Hi,U)Ɣ = 0. (94)<br />
i<br />
Now (93), (94), and the fact that Ci ⊂ X(Hi,U), give<br />
|{k ∈ K : kx ∈ <br />
CiƔ/ Ɣ}| = 0 (95)<br />
i
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 157<br />
for any x ∈ F ⊂ D \ <br />
i H ⋉Rnxi. Now if we apply [EMM1, Lemma 4.2], then<br />
there exists an open subset W ⊂ G/ Ɣ such that <br />
i CiƔ/ Ɣ ⊂ W and such that<br />
|{k ∈ K : kx ∈ W }| 0 such that<br />
<br />
<br />
1<br />
T <br />
<br />
φ(uty) − φdg<br />
0 be any sufficiently small number, and let T > 2. There exists a rational<br />
3-dimensional subspace U of 2 R 4 with a reduced integral basis of norm at most<br />
T τ whose projections into V2 have norm less than T τ−1 such that one of the following<br />
holds.<br />
(a) The restriction of Q (6) to U is anisotropic over Q, in which case (i) in Theorem<br />
5.6 holds.<br />
(b) The restriction of Q (6) to U splits over Q, in which case (ii) in Theorem 5.6<br />
holds. Furthermore, in this case Q ′ (as in loc. cit.) is proportional to f (U,U ⊥ ),<br />
where U ⊥ is the orthogonal complement of U with respect to Q (6) and where f<br />
is as in Lemma 4.3. Moreover, the number of quasi-null subspaces with norm<br />
between T/2 and T which are not in U or U ⊥ is O(T 1−τ ).<br />
Let us denote by Q (3) (v) the restriction of Q (6) to U. As we mentioned before, this is<br />
a form with signature (2, 1) and U is a rational subspace. If L is a quasi-null subspace<br />
with T/2 ≤v L ≤T, then we have Q (6) (v L ) = 0. Hence if L is a quasi-null<br />
subspace in U, then Q (3) (v L ) = 0. In other words, Proposition 6.2 follows from the<br />
following.
158 MARGULIS and MOHAMMADI<br />
PROPOSITION B.2<br />
Let Q0(x,y,z) = 2xz − y 2 be the standard quadratic form of signature (2, 1) on<br />
R 3 . Let = gZ 3 ,whereg ∈ GL3(Q). Let T > 0 be a large parameter, and let τ<br />
be a sufficiently small parameter. Assume that the entries of g are rational numbers<br />
whose nominator and denominators are bounded by a fixed power of T τ , and also<br />
suppose that |det g| ≤T τ . Further assume that the length of the shortest vector in <br />
is c = O(1). Then if T is large enough, we have<br />
# w ∈ P () : Q0(w) = 0 and w ≤T 0. Note that C is dilation-invariant. Let λ = disc() 1/3 , and define<br />
1 = 1/λ. The lattice 1 is unimodular, and the length of the shortest vector in<br />
1 is at least c/λ. Let¯g ∈ PGL3(Q) denote the image of g. Indeed, 1 = ¯gZ 3 . The<br />
counting problem in (98) follows if we show that<br />
<br />
# w ∈ P (1) : Q0(w) = 0 and w ≤ T<br />
<br />
<strong>IN</strong>HOMOGENEOUS QUADRATIC FORMS 159<br />
Hence we see that C ∩P (1) ⊂ Ɣmv0. As is a finite set, we only need to consider<br />
Ɣmv0.<br />
Let e1 = (1, 0, 0), and let N0 ⊂ H be the stabilizer of e1. Let BR ={h ∈ H :<br />
h · e1 ∈ CR}, and let BR = BR/N0. Now let χR denote the characteristic function of<br />
CR. Define the following:<br />
FR(h) = <br />
χR(hγ v), for h ∈ H. (100)<br />
Ɣm/Ɣm∩N<br />
This is a function on H/Ɣ.Now using the well-developed machinery for counting<br />
problems using the mixing property (see in particular [BO] and[MS]), there exists<br />
ε>0 depending only on the spectral gap of H and τ such that<br />
FR(e) = |N0/N0 ∩ Ɣ|<br />
vol(BR)<br />
|H/Ɣ|<br />
1 + O(vol(BR) −ε ) , (101)<br />
where vol denotes the H -invariant Haar measure on H/N. Hence we have<br />
#(Ɣmv0 ∩ CR) ≤ FR/v0(e) = O(R/v0). (102)<br />
Since v0 ≥c/λ, this finishes the proof. <br />
Acknowledgments. We would like to thank J. Marklof for reading the first draft and<br />
for many helpful comments. We also thank the anonymous referees for their remarks<br />
and suggestions. The second author would like to thank A. Eskin for conversations<br />
regarding these remarks.<br />
References<br />
[BO] Y. BENOIST and H. OH, Effective equidistribution of S-integral points on<br />
symmetric varieties, preprint, arXiv:0706.1621v1 [math.NT] 159<br />
[BT] M. V. BERRY and M. TABOR, Level clustering in the regular spectrum, Proc.Royal<br />
Soc. A 356 (1977), 375 – 394. 126<br />
[DM] S. G. DANI and G. A. MARGULIS, “Limit distributions of orbits of unipotent flows<br />
and values of quadratic forms” in I. M. Gelfand Seminar, Adv.SovietMath.<br />
16, Amer. Math. Soc., Providence, 1993, 91 – 137. MR 1237827 123, 124,<br />
128, 154<br />
[EMM1] A. ESK<strong>IN</strong>, G. A. MARGULIS,andS. MOZES, Upper bounds and asymptotics in a<br />
quantitative version of the Oppenheim conjecture, Ann. of Math. (2) 147<br />
(1998), 93 – 141. MR 1609447 122, 123, 124, 125, 126, 128, 130, 131, 132,<br />
133, 134, 142, 155, 156, 157
160 MARGULIS and MOHAMMADI<br />
[EMM2] ———, Quadratic forms of signature (2,2) and eigenvalue spacings on<br />
rectangular 2-tori, Ann.of Math. (2) 161 (2005), 679 – 725. MR 2153398<br />
122, 123, 124, 125, 126, 128, 133, 134, 135, 136, 137, 141, 142, 146, 152, 157<br />
[GT] B. GREEN and T. TAO, The quantitative behavior of polynomial orbits on<br />
nilmanifolds, preprint. 151<br />
[KM] D. Y. KLE<strong>IN</strong>BOCK and G. A. MARGULIS, Flows on homogeneous spaces and<br />
Diophantine approximation on manifolds, Ann. of Math. (2) 148 (1998),<br />
339 – 360. MR 1652916 139<br />
[Mar] G. A. MARGULIS, “Lie groups and ergodic theory” in Algebra—Some Current<br />
Trends (Varna, 1986), Lecture Notes in Math. 1352, Springer, Berlin, 1988,<br />
130 – 146. MR 0981823<br />
[Mark] J. MARKLOF, Pair correlation densities of inhomogeneous quadratic forms, Ann.<br />
of Math. (2) 158 (2003), 419 – 471. MR 2018926 126, 128<br />
[MS] A. MOHAMMADI and A. S. GOLSEFIDY, Translates of horospherical measures and<br />
counting problems, preprint. 159<br />
[Sar] P. SARNAK, “Values at integers of binary quadratic forms” in Harmonic Analysis<br />
and Number Theory (Montreal, 1996),CMSConf.Proc.21,<br />
Amer. Math. Soc., Providence, 1997, 181 – 203. MR 1472786 126, 127<br />
[Sch] W. M. SCHMIDT, Asymptotic formulae for point lattices of bounded determinant<br />
and subspaces of bounded height, Duke Math. J. 35 (1968), 327 – 339.<br />
MR 0224562 130, 136<br />
[Sh2] N. SHAH, Limit distributions of expanding translates of certain orbits on<br />
homogeneous spaces, Proc. Indian Acad. Sci. Math. Sci. 106 (1996),<br />
105 – 125. MR 1403756 154<br />
Margulis<br />
Department of Mathematics, Yale University, New Haven, Connecticut 06520, USA;<br />
margulis@math.yale.edu<br />
Mohammadi<br />
Department of Mathematics, University of Chicago, Chicago, Illinois 60637, USA;<br />
amirmo@math.uchicago.edu