• Nenhum resultado encontrado

Gourmelon Adapted metrics for dominated splittings

N/A
N/A
Protected

Academic year: 2019

Share "Gourmelon Adapted metrics for dominated splittings"

Copied!
11
0
0

Texto

(1)

doi: 10.1017/S0143385707000272 Printed in the United Kingdom

Adapted metrics for dominated splittings

NIKOLAZ GOURMELON

I.M.B., UMR 5584 du CNRS, B.P. 47 870, 21078 Dijon Cedex, France (e-mail: [email protected])

(Received14January2005and accepted in revised form23September2005)

Abstract. A Riemannian metric isadaptedto a hyperbolic set of a diffeomorphism if, in this metric, the expansion/contraction of the unstable/stable directions is seen after only one iteration. Adominated splittingis a notion of weak hyperbolicity where the tangent bundle of the manifold splits in invariant subbundles such that the vector expansion on one bundle is uniformly smaller than that on the next bundle. The existence of anadapted metricfor a dominated splitting has been considered by Hirsch, Pugh and Shub (M. Hisch, C. Pugh and M. Shub.Invariant Manifolds (Lecture Notes in Mathematics, 583). Springer, Berlin, 1977). This paper gives a complete answer to this problem, building adapted metrics for dominated splittings and partially hyperbolic splittings in arbitrarily many subbundles of arbitrary dimensions. These results stand for diffeomorphisms and for flows.

1. Introduction

The best known and simplest examples of chaotic dynamical systems are uniformly hyperbolic systems, like Anosov diffeomorphisms. A diffeomorphism f on a compact Riemannian manifoldMis said to be anAnosov diffeomorphismif there exists a splitting of the tangent bundleT Minto two supplementary,d f-invariant subbundles, called thestable and theunstablebundles that are uniformly contracted and expanded, respectively, by an iterate of f. If the hyperbolic systems are now well understood, many dynamical systems are (robustly) not hyperbolic, so that several authors have tried to weaken the notion of hyperbolicity, in order to recover some of its properties on a larger class of systems.

(2)

Working on the stability conjecture†, Liao [Lia80], Ma˜n´e [Man82] and Pliss [Pli72] were led to the following general notion: adominated splittingfor f is a splitting ofT M into two supplementary invariant subbundles such that there exists an iterate ofd f that uniformly contracts more (or expands less) the first subbundle than the second one. This notion is a key tool for understanding non-hyperbolic systems.

• In dimension two, Pujals and Sambarino [PS00] proved that a diffeomorphism with a dominated splitting may beC1-approached by hyperbolic diffeomorphisms, and diffeomorphisms without dominated splitting may be approached by diffeomorphisms exhibiting a homoclinic tangency: as a consequence, any diffeomorphism of a compact surface can beC1-approximated either by hyperbolic (Axiom A) diffeomorphisms, or by diffeomorphisms that exhibit a homoclinic tangency (this was conjectured by Palis [PS70]).

• In any dimension, Bonattiet alshowed in [BDP00] that a robustly transitive generic diffeomorphism in Diff1(M)admits a non-trivial dominated splitting defined on the wholeM.

As recalled above, for a hyperbolic set K of a diffeomorphism f, the vectors in the stable and unstable bundles are uniformly contracted and expanded, respectively, by the derivatived fn, for somen>0. The hyperbolicity ofK does not depend on the metric on the manifold, but the smallest timenwhere the contraction/expansion phenomena are seen depends on the metric; a Riemannian metric is said to beadaptedto the hyperbolic setK if one can taken=1. Applying Holmes’ theorem (see [HPS77, p. 15]), we obtain that any hyperbolic set admits an adapted Riemannian metric. We will adapt this theorem to the case of dominated behaviours, to show Lemma 4.1. It was asked in [HPS77, p. 5] if there existed anadapted metric for a dominated splitting, that is a metric such thatd f uniformly contracts (or respectively expands) the first subbundle more (or respectively less) than the second one, at the first iteration.

The aim of this paper is to give a complete positive answer to this question, proving that such an adapted metric exists for any dominated splitting.

THEOREM1. Let f be a diffeomorphism of a Riemannian manifold M, and K a compact invariant subset of M, such that the restriction of f to K admits a dominated splitting T M|K=E1⊕E2⊕ · · · ⊕Ed, where the vectors in Ei are uniformly less expanded than

those in Ei+1by d fnfor some n>0. Then there exists a Riemannian metrick · kon M (necessarily equivalent to the first metric) and adapted to the dominated splitting: there exists a constant0< µ <1such that for any xK , any i∈ {1, . . . ,d−1}, and any unit vectors uEix,v∈Ei+1x , one haskd f(u)k< µ.kd f(v)k.

This result was already known by Hisch et al[HPS77] for a dominated splitting in two bundles,T M|K=E1⊕E2, such that dim(E1)=1 or dim(E2)=1. In addition, they

showed that any absolutely normally hyperbolic system admits an adapted metric, but it was not known whether it was true for a relatively normally hyperbolic system, which is, with our definitions, a partially hyperbolic system. We answer by showing (see Theorem 3)

(3)

that an adapted metric exists for any partially hyperbolic splitting, that is a metric adapted to the corresponding dominated splitting, and such that the stable/unstable bundles are uniformly expanded/contracted at the first iterate. Finally, we show in §5 how to transpose these results from diffeomorphisms to flows.

In order to present more clearly the idea of the proof, we will first focus, in §3, on dominated splittings into two subbundles, over an invariant compact set of a diffeomorphism. Then, we will show that there exists an adapted Finsler for any dominated splitting intod bundles for a Banach bundle automorphism (see Theorem 2).

2. Definition and notations

For a morphismAof normed vector spaces, define thenormand theminimum normofA: kAk = sup

kuk=1

kA(u)k, m(A)= inf

kuk=1kA(u)k.

When Ais invertible, m(A)= kA−1k−1. For a Banach bundle E, we denote by Ex the

fibre ofEabove a pointxof the base. IfAis an automorphism of a Banach bundleEwith compact baseK, then, for any pointxof K, we denote byAx the restriction ofAto the fibreEx. We refer the reader to [HPS77] for definitions.

We say that a sequence of functionsgn(x): K→Rconverges exponentially to zeroif

there exists positive constantsCandµ <1 such that, for allxandn, |gn(x)| ≤Cµn.

Given an automorphismAof a Banach bundleEwith compact baseK

E A //

π

E

π

K f //K

and a positive continuous functionr: K →R, we denote byRn(x)the product

Rn(x)= n−1

Y

i=0

r[fi(x)] =r(x)r[f(x)] · · ·r[fn−1(x)].

Definition 2.1. A positive continuous functionr: K→RdominatesA, if the sequence of

ratiosx→ kAn

xk/Rn(x)converges exponentially to zero asn→ ∞, whereAnx=An|Ex.

In this case we writed f|E≺r. Symmetrically, we say thatrisdominatedbyAand we

writer≺Aif and only if the ratioRn/m(An)goes exponentially to zero asn→ ∞. Note

thatr≺Ais equivalent toA−1≺1/r.

Definition 2.2. LetEbe a Banach bundle over a compact baseK, andE=E1⊕ · · · ⊕Ed be an invariant splitting for an automorphism A, where the Ei are vector subbundles with constant dimension. Then we say that it is adominated splittingif, for each integer 0<i<d, the ratiokAn

|Eik/m(A

n

(4)

We have written kAn

|Eik/m(An|Ei+1)for the function x7→ kAn|Ei x

k/m(An

|Exi+1). In this case, we say thatA|Ei isdominatedbyA|Ei+1, and we writeA|Ei ≺A|Ei+1. We recall that the subbundlesEiare necessarily continuous (see [BDV04, Appendix B] for a proof). Remark 2.3. Since the bundles and automorphisms are continuous, and the base K is compact, the definitions of domination and dominated splitting are independent of the Finsler. Thus we will be allowed to change to equivalent metrics; in finite dimension, all Finslers can be replaced by smooth Riemannian metrics.

A Finsler k · k∗ isadaptedto the dominated splitting if and only if, for alli<d, we

have

kA|Eik

m(A

|Ei+1) <1

where k · k∗ andm are the norm and the minimum norm, with respect to the Finsler

k · k∗. Equivalently, by compactness of the base, there exists a real number 0<C<1

such that, for any xK, for any non-zero unit vectors uEix, v∈Exi+1, we have kA(u)k∗<CkA(v)k∗.

3. Two-bundle splittings

Let M be a compact smooth manifold endowed with a Riemannian metrick · k, let f be a diffeomorphism of M and let K be an invariant compact set in M. We will show the following theorem.

THEOREM3.1. If TKM=EF is a dominated splitting for the diffeomorphism f on

the compact K , then there exists a smooth Riemannian metric on M that is adapted to that dominated splitting.

Proof. The proof consists in first building aseparatorfor the dominated splitting, that is, a positive functionr: K→Rsuch that we haved f|E≺rd f|F. Then, by a dominated

version of Holmes’ theorem, we will build two metricsk · kE andk · kFon the bundlesE

andF, such that, for anyxK, for any unit vectorsuEx andv∈Fx, we have

kd f(u)kE<r(x) <kd f(v)kF.

These metrics will induce, up to perturbation, an adapted Riemannian metric onM. LEMMA3.1. A two-bundle dominated splitting has a separator.

Proof. In the following, we fix a dominated splitting TKM=EF for the

diffeomorphism f. For simplicity, calld f|E=Aandd f|F=B. By hypothesis, the ratio

kAnk/m(Bn)tends exponentially to zero. In particular, for N large enough, the function

x7→ kANxk/m(BN

x )is smaller than 1/2. Therefore, fora>1>bclose enough to one, we

have for allxinK:

akAxNk1/N<bm(BN

x)1/N.

Hence, Lemma 3.1 comes from the following claim. ✷

CLAIM1. Any continuous function r: K→Rsuch that a.kAxNk1/Nr(x)b.m(BN

x )1/N

(5)

Proof. For any integern>N, and each 0≤kN−1, we can write the iterateAn xas the

composition

Al

fk+m N(x)◦A

m N fk(x)◦A

k x

for some integers 0≤lN−1 and m≥0. More precisely, take the integer part of (n−k)/N for m, andl=nn Mk. Denote byc the upper bound of the norms of theith forward or backward iterates of A, foriN,

c= sup

|i|≤N,y∈K

kAiyk. It is finite, asKis compact. We then have

kAnxk ≤ kAl

fk+nm(x)k m−1

Y

i=0

kAN

fk+i N(x)k

kAkxk kAnxk ≤c2 Y

j∈Jk kAN

fj(x)k

(1k)

where Jk is the set of integers {k+i N,i=0, . . . ,m−1}, that is, the set of integers

of the form k+i N and comprised between 0 and nN. Obviously, the sets Jk for

k=0, . . . ,N−1 are pairwise disjoint and their union is the interval{0, . . . ,nN}. Hence, taking the product of inequalities (1k), fork=0, . . . ,N−1 we obtain

kAnxkNc2N Y

j∈{0,...,n−N}

kAN

fj(x)k. SincekAN

fj(x)k

−1≤ kA−N

fj+N(x)k ≤c, we get kAnxkNc2NcN Y

j∈{0,...,n}

kAN

fj(x)k. Thus, asakAN

fj(x)k1/Nr[fj(x)], we get that, for anyxK, kAnxk ≤c3a−nRn(x),

which proves thatA≺r. Note that(1/b).kB−Nx k1/N≤1/r[FN(x)], for allx. Thus, we have in the same wayB−11/r and thenrB. This ends the proof of the claim, and

that of Lemma 3.1. ✷

We now show the following lemma (which can actually be seen as a particular case of Lemma 4.1 stated below).

LEMMA3.2. Let r: K→Rbe a positive function that separates the continuous splitting EF , that is, d f|E≺rd f|F. Then there exists a Riemannian metrick · k∗on M that

isadaptedto the domination; namely, for all xK , for all unit vectors uEx,v∈Fxwe

have

kd f(u)k∗<r(x) <kd f(v)k∗.

Proof. We define on Ea metrick · kEby

kuk2E=

X

n=0

(6)

for anyuEx, whereRn(x)=r(x)· · ·r[fn−1(x)]as above. By domination, this is a sum

of a normally convergent series of continuous functions; thereforek · kE is well defined

and continuous. As a sum of quadratic forms,k · k2E is a quadratic form, and thusk · kEis

a Hilbertian metric (it arises from an inner product). Moreover, we have kd f(u)k2E=

X

n=0

kd fn+1(u)k2 [Rn(x)]2

=

X

n=1

kd fn(u)k2 [Rn−1(x)]2

=r(x)2

X

n=1

kd fn(u)k2 [Rn(x)]2

since Rn−1(x)=Rn(x)/r(x). We obtainkd f(u)k2E=r(x)2[kuk2E− kuk2]wherekuk2

is the first term of the series defining kuk2E. Hence, for any non-zero u, kd f(u)kE<

r(x)kukE. Up to change of f into f−1andrinto 1/r, we find the same way a Hilbertian

metric k · kF on F such that, for all non-zerov in F,r(x)kvkF<kd f(v)kF. Consider

now the Hilbertian metrick · k∗onTKM that extendsk · kE andk · kF and that makesE

andForthogonal. It is continuous, sincek · kEandk · kFare continuous.

The inequality kd f(u)k∗<r(x) <kd f(v)k∗ holds for all unit vectorsuE, v∈F

above each pointxof the baseK. We extend the metrick · k∗to the wholeM, and smooth

it into a Riemannian metric by a small perturbation, so that, by compactness of K, the

inequality is preserved. ✷

This together with the existence of a separator (Lemma 3.1) ends the proof of

Theorem 3.1. ✷

4. Multiple bundles splittings

We will show the most general result of our paper in this section.

THEOREM2. Let E be a finite-dimensional Banach bundle on a compact base, and let

Abe an automorphism of E . If E=E1⊕ · · · ⊕Ed is a dominated splitting forA, then

there is a Finslerk · k∗on E adapted to the domination, that is, for each i=1. . .d−1,

for any xK , we have

kA|Ei xk∗<

m(A

|Ei+1x ).

Furthermore, if the original metric on E is Hilbertian, then the adapted metric can be chosen to be also Hilbertian.

LetFbe a Banach bundle with compact baseK, andBbe an automorphism

F B //

π

F

π

K f //K

Then we have this dominated version of Holmes’ theorem (see [HPS77, p. 15]).

LEMMA4.1. Let r,s: K→R be two positive continuous functions such that the domination r≺B≺s and the inequality r<s hold on K . Then there is a Finslerk · k∗on

F that is adapted to the domination, namely, for any xK , for any uFx\ {0}, we have

(7)

Proof. For alluinF, we define

kuk2=

X

n=1

Rn2[f−n(x)]kB−n(u)k2+

X

n=0

kBn(u)k2 [Sn(x)]2

wherek · kis the original metric onF, and whereRn[f−n(y)] =r(f−n(y))· · ·r(f−1(y)),

Sn(y)=s(y)s(f(y))· · ·s(fn−1(y))as before. Still by domination, the series normally

converges to a continuous function. Thusk · k∗is a well-defined Finsler. We have

kB(u)k2=

X

n=1

Rn2[f−n+1(x)]kB−n+1(u)k2+

X

n=0

kBn+1(u)k2 [Sn(f(x))]2

=

X

n=0

Rn+12 [f−n(x)]kB−n(u)k2+

X

n=1

kBn(u)k2

[Sn−1(f(x))]2

.

As we haveRn+1[f−n(x)] =Rn[f−n(x)].r(x)andSn−1[f(x)] =Sn(x)/s(x), we get

kB(u)k2= [r(x)]2

X

n=0

Rn2[f−n(x)]kB−n(u)k2+ [s(x)]2

X

n=1

kBn(u)k2

[Sn(x)]2

.

On the other hand, we haveR0(x)kB−0(v)k = kvk = kB0(v)k/S0(x)sinceR0andS0are

empty products equal to one. So, from the definition ofkuk2, we have

kuk2=

X

n=0

Rn2[f−n(x)]kB−n(u)k2+

X

n=1

kBn(u)k2

[Sn(x)]2

.

Finally, sincer<s, we obtain, for any non-zero vectoru, the inequality [r(x)]2kuk2<kB(u)k2<[s(x)]2kuk2

the square root of which concludes the proof. ✷

Remark 4.2. Having chosen this quadratic construction, ifk · kis a Hilbertian metric, then k · k∗is still a Hilbertian metric.

Proof of Theorem 2. By definition, the ratioskAn

|Eik/m(An|Ei+1)converge exponentially to zero, for eachi, asngoes to infinity. Thus we can find an integerN such that, for eachi, the ratiokAN

|Eik/m(A|ENi+1)is smaller than 1/4. The proof of Lemma 3.1 still works when T M|K=EF is replaced by the Banach bundleEiEi+1, and whend f|K is replaced

by the automorphismA|Ei⊕Ei+1.

Then choose a family(ri)0<i<d of continuous functions such that 21/NkA|ENik1/N< ri <2−1/Nm(AN

|Ei+1)

1/N. We then haver

1(x) <· · ·<rd−1(x), for all xK;

further-more, by Claim 1, we have

E1≺r1≺E2≺ · · · ≺rd−1Ed.

(8)

a Finslerk · ki on eachEi that is adapted to the dominationriEi+1ri+1. Define the

new metric

kuk∗=

s X

i=1,...,d

kpi(u)k2i,

for alluE, where pi is the projection onEi alongE1⊕ · · · ⊕Ei−1Ei+1· · · ⊕Ed.

It is a Finsler that is clearly adapted to the dominated splitting: for any unit vectorsuEix, v∈Exi+1we havekuk∗= kuki<ri(x) <kvki+1= kvk∗.

Remark 4.3. Obviously by the previous remark, if the original metrick · kwas a Hilbertian metric, then the metricsk · ki are so, and the metrick · k∗we have built is still a Hilbertian

metric.

After smoothing the adapted Hilbertian metric, we obtain the following, which is a reformulation of Theorem 1.

COROLLARY4.4. Let M be a Riemannian manifold, K a compact invariant set for a diffeomorphism f , and TKM=E1⊕ · · · ⊕Eda dominated splitting for f above K . Then

there exists a smooth Riemannian metric on M that is adapted to it.

The existence of an adapted metric was shown forabsolute-and notrelative-normally hyperbolic systems (see [HPS77] for proofs and definitions). The bases of all Banach bundles are still compact. A dominated splittingE=E1⊕ · · · ⊕Edfor an automorphism

A is partially hyperbolic if and only if, for some 1≤k<k+1<ld, the bundles Es=E1⊕ · · · ⊕Ek andEu=El⊕ · · · ⊕Ed are respectively stable and unstable, that

iskAn

|Esk andkA−n|Eukconverge exponentially to zero as n goes to infinity. We say that a metric k · k∗ isadapted to such a partially hyperbolic splitting if it is adapted to the

dominated splitting, and ifkA|Esk<1 andm(A|Eu) >1.

THEOREM3. A partially hyperbolic splitting has an adapted metric.

Proof. We show it in the three-bundle case (it is the same idea for the general case). Consider a partially hyperbolic splitting E=EsEcEu with compact base for an automorphismA. ThenkAn

|EskandkA−n|Euktend exponentially to zero asngoes to infinity. From the construction we gave in the proof of Lemma 3.1, we can find two functions 0<r<1<ssuch that

A|Esr≺A|Ecs≺A|Eu.

With respect to this domination, the metric k · k∗ produced in the proof of Theorem 2

is adapted to the dominated splitting and satisfies kA|Esk<1<m(A|Eu). Hence, it is

adapted to the partially hyperbolic splitting. ✷

5. Dominated splittings for flows

In this section, we briefly show that the same results apply for flows. In the following,φ is a flow on a compact subset K of a Riemannian manifold M. A Finslerk · k∗ onM is

adaptedto a dominated splittingTKM=E1⊕ · · · ⊕Edforφ, if and only if, for any point

(9)

wheredφt is the derivative of the time-t map ofφ. The existence of adapted metrics for flows is not a straightforward consequence of our results on diffeomorphisms. At best, applying the former results would provide, for eachǫ >0, a metrick · k∗ such that the

inequality above holds for all t> ǫ. To get adapted metrics, we have to transpose the notion of separator to the flow case.

Let E be a subbundle of TKM, invariant byφ. Fix two strictly positive, continuous

functionsr,s: K→R. Then, for anyxK, for allt∈R, define Rt(x)=exp

Z t

0

ln[r(φu(x))].du

,

St(x)=exp

Z t

0

ln[s(φu(x))].du

.

For any fixedx, the functionst7→Rt(x)andt7→St(x)areC1, and for all real numbers

t,k,

Rt+k(x)=Rt(x).Rkt(x)], (1)

St+k(x)=Rt(x).Skt(x)]. (2)

Assume that r<s, and that we have the domination relation rdEφ≺s, that is, for

allxK, for any vectorv∈Ex, the quantitieskdφt(v)k/St(x)andkdφt(v)k/Rt(x)go

exponentially to zero, respectively, ast goes to+∞, and ast goes to−∞. Then, define the Finslerk · k∗onEby

kvk2= Z 0

−∞

kdφt(v)k2 Rt(x)2

.dt+ Z ∞

0

kdφt(v)k2 St(x)2

.dt

for anyxK, for allv∈Ex.

CLAIM2. For any non-zero vectorv∈E above any point xK , the metrick · k∗satisfies

for all k>0, Rk(x).kvk∗<kdφk(v)k∗<Sk(x).kvk∗.

That is, the metrick · k∗isadaptedto the dominationrdEφ≺s.

Remark 5.1. This is merely Lemma 4.1 for flows. Proof. After a change of variable, we get

kdφk(v)k2= Z k

−∞

kdφt(v)k2 R2t−k◦φk(x).dt+

Z +∞

k

kdφt(v)k2 St−k2 ◦φk(x).dt

=Rk2(x) Z k

−∞

kdφt(v)k2 Rt2(x) .dt+S

2 k(x)

Z +∞

k

kdφt(v)k2 St2(x) .dt, by formulae (1) and (2). Letθ (k)be the quotientSk2(x)/Rk2(x), and define the function

f: k7→ kdφ

k(v)k2 ∗

Rk2(x) = Z k

−∞

kdφt(v)k2

R2t(x) .dt+θ (k) Z +∞

k

kdφt(v)k2 St2(x) .dt.

(10)

f′(k)=θ′(k) Z +∞

k

kdφt(v)k2 St2(x) .dt.

Hence f′ is strictly positive, and f is strictly increasing. As we have f(0)= kvk2/R02(x)= kvk2, the inequality f(k) > f(0)leads to Rk(x)2.kvk2∗<kdφk(v)k2∗, for

allk>0. The inequalitykdφk(v)k2<Sk2(x).kvk2comes the same way, considering this time the functiong: k7→ kdφk(v)k2/Sk2(x)

. This concludes the proof of the claim. ✷

On the other hand, any dominated splitting for a flow has a separator. Let TKM=

EFbe a dominated splitting. For simplicity,8will denote the restriction ofdφto the bundle E. For instance, we will writek8txkfor the maximum norm of the restriction of dφtoEx. More precisely, we assert the following claim.

CLAIM3. The function r= x7→a.k8Txk1/T

is a separator ErF , for some a>1, and some T large enough.

Remark 5.2. This is Lemma 3.1 for flows; the proof is comparable step by step to that of Claim 1.

Proof. Fix a realt>T. Letmbe the largest integer such thatT +mTt. For any real 0≤κ≤T, we decompose8tto obtain, for allxK,

k8txk ≤ k8λ

φκ+mT(x)k · k8Tφκ+(m−1)T(x)k · · · k8φTκ+T(x)k · k8φTκ(x)k · k8κxk,

whereλ >0 satisfiest=κ+mT+λ. Denote bycthe upper boundc=sup|τ|≤2T,y∈K k8τyk, and take the logarithm of the inequality,

lnk8txk ≤2 ln(c)+lnk8T

φκ+(m−1)T(x)k + · · · +lnk8

T

φκ(x)k.

This stands for all 0≤κ≤T. Therefore, we have Z T

0

lnk8txk.dκ ≤ Z T

0

2 ln(c).dκ+ Z T

0

lnk8T

φκ+(m−1)T(x)k.dκ + · · · +

Z T

0

lnk8Tφκ(x)k.dκ,

T lnk8txk ≤2T ln(c)+ Z mT

0

lnk8Tφu(x)k.du,

T lnk8txk ≤4T ln(c)+ Z t

0

lnk8Tφu(x)k.du,

since 0≤tmT ≤2T, and−lnk8Tyk ≤lnk8−Ty k ≤ln(c). Then, dividing by T, we obtain

lnk8txk ≤ln(c4)+ Z t

0

ln[a−1.r(φu(x))].du.

Writing the exponential form, we get k8txk ≤c4a−tRt(x), which means that8=dEφ

r. We are left to check thatrdFφfor someT >0 big enough, and somea>1. This

is done the same way as in Lemma 3.1. ✷

Clearly, referring the reader to the proof of Theorem 2, flow versions of Lemmas 3.1 and 4.1 are all the ingredients we need, to transpose the results of §4 to flows.

(11)

Acknowledgements. I would like to thank Christian Bonatti and Sylvain Crovisier for second reading and corrections. Special thanks also go to the referee for his great help in writing this paper.

REFERENCES

[BDP00] C. Bonatti, L. J. Diaz and E. R. Pujals. AC1-generic dichotomy for diffeomorphisms: weak forms of hyperbolicity or infinitely many sinks or sources.Ann. of Math.158(2) (2000), 355–418. [BDV04] C. Bonatti, L. J. Diaz and M. Viana.Dynamics Beyond Uniform Hyperbolicity. Springer, Berlin,

2004.

[BP74] M. Brin and Ya. Pesin. Partially hyperbolic dynamical systems.Izv. Akad. Nauk. SSSR38(1974), 177–212.

[Hay97] R. S. Hayashi. Connecting invariant manifolds and the solution of theC1-stability and-stability conjectures for flows.Ann. of Math.145(1997), 81–137.

[HPS77] M. Hisch, C. Pugh and M. Shub. Invariant Manifolds (Lecture Notes in Mathematics, 583). Springer, Berlin, 1977.

[Lia80] S. D. Liao. On the stability conjecture.Chinese Ann. Math.1(1980), 9–30. [Man82] R. Man´e. An ergodic closing lemma.Ann. of Math.116(1982), 503–540. [Pli72] V. Pliss. On a conjecture due to smale.Differ. Uravn.8(1972), 262–268.

[PS70] J. Palis and S. Smale.Structural Stability Theorems (Proceedings of the Institute on Global Analysis, XIV). American Mathematical Society, Providence, RI, 1970, pp. 223–232.

Referências

Documentos relacionados

A infestação da praga foi medida mediante a contagem de castanhas com orificio de saída do adulto, aberto pela larva no final do seu desenvolvimento, na parte distal da castanha,

b) Organizações do setor voluntário – Aplicação de formulário com questões abertas e fechadas, conforme mostra o Apêndice A, a 28 pessoas sendo: uma

Felizmente a tendência é de abolição deste tipo de conceções, uma vez que, são cada vez mais escassas as decisões em se exige uma violência desta natureza para o preenchimento

As to disclose the rate determining step of the extraction process, six extraction curves were measured for different flow rates and ethanol content, and three simplified

No entanto, ao fim de 60 minutos de oxidação a α- amilase de Bacillus amyloliquefaciens possui uma maior percentagem de resíduos oxidados (64%) do que a α-amilase salivar

Stepanov Institute of Physics, National Academy of Sciences of Belarus, Minsk, Republic of Belarus 92 National Scientific and Educational Centre for Particle and High Energy

Portanto, conclui-se que existem dois posicionamentos em torno da aplicabilidade ou não do princípio da anterioridade da lei tributária à norma que revoga ou diminui uma isenção

During the course ofinspecting vegetation at four aluminum smelters in Brasil in 1982, 1983 and 1986, more than 230 species of plants from tropical, semi-tropical, and temperate