• Nenhum resultado encontrado

Braz. J. Phys. vol.32 número4

N/A
N/A
Protected

Academic year: 2018

Share "Braz. J. Phys. vol.32 número4"

Copied!
9
0
0

Texto

(1)

The Nature of Dark Energy

Lua Amendola,ClaudiaQuerellini, DomenioTohini-Valentini,

and Alessandro Pasqui

INAF-ObservatorioAstronomiodiRoma,

ViaFrasati33,00040MontePorzio Catone,Italia

Reeivedon23May,2001

Aording to a variety of osmologial observations at small and large redshifts, the universe is

omposedbyalargefrationofaweaklylusteredomponentwithnegativepressure, alleddark

energy. The natureof the dark energy, i.e. itsinterationand self-interation properties, is still

largely unknown. In this ontribution we review the properties of dark energyas inferred from

observations, withpartiular emphasisonthe osmimirowavebakground. We arguethat the

urrent dataset imposesstrong onstraintsonthe oupling ofdark energyto dark matter, while

itisstill insuÆient toonstraintheequationofstateorpotential. Futuredatawill dramatially

improvetheprospets.

I Rods, loks and andles

Fouryears after the rstobservationalhintsfrom the

supernovae Ia (SNIa)Hubble diagram about the

exis-tene of adominantomponentof unlustered matter

with negative pressure (Riess et al. 1998, Perlmutter

et al. 1999), theso-alleddarkenergyorquintessene

(Wetterih 1988; Ratra and Peebles 1988; Frieman et

al. 1995; Caldwell et al. 1998), there are still very

fewindiationsastoitsnature. Themainreason,

per-haps,isthatwelakanyspeitheoretialsuggestion

on the properties of the dark energy, i.e. on its

self-interation potentialand on howit interats with the

other osmologial omponents. At this stage, all we

andois toexplore awide rangeofphenomenologial

models speied by the potentialand theoupling to

the other elds in order to provide an overall best t

to theurrentdata. Fortunately,therearenowseveral

osmologial observations that are at leastpotentially

abletoput stringentonstraintsonthenatureofdark

energy.

As with anyother fundamental eld,the dark

en-ergy anbe haraterized byhow it interats with

it-self and with the other elds. In other words, it an

beharaterizedbyitspotentialand ouplings. While

there have been severalpapersthat tried to onstrain

thepotentialortheequationofstateofthedarkenergy

(Amendola 2001, Baigalupi et al. 2001, Corasaniti

and Copeland 2001, Bean and Melhiorri 2001), the

study of its oupling has been relatively sare,

al-thoughithasbeensuggestedseveraltimesthatwe

an-notexpetavanishingoupling(Carroll1998).

The observations that an reveal dark energy are

aount only the evolution of the homogeneous

bak-ground and those that investigate the growth of

per-turbations. The latter lass is so far still poor of

re-sults,duetothelargeerrorsin theobservationsofthe

growthof lustering, althoughtheprospetsare

inter-esting(Newmanetal. 2001). Moreover,the

perturba-tiongrowthisverysimilarforallmodelsofdarkenergy,

sinetheaelerationstopsthegravitationalinstability

forallbutonemodelofdarkenergy. Thesingleasein

whih thisisnottruehasbeendisussedin Amendola

andTohini-Valentini(2002).

Thebakground observations allredue essentially

tothe useof astandardmeasure,either andles,rods

orloks: any of this standard is in fat subjet to a

geometriapparentvariationwhenseenathighredshift

thatdependonosmology. Thestandardandleshave

beenemployedinthemethodbasedonthesupernovae

Ia, while standard rods are assumed in the

Alok-Pazinsky method (or rather, astandardisotropy, see

e.g. Calvaoet al. 2001), in the lensingstatistis (e.g.

Cooray and Huterer 1999, Giovi et al. 2001 ) and in

thesizeofthesoundhorizonatdeouplingasdeteted

from the CMB aousti peaks (see e.g. Doran et al.

2000). Finally, standard loks are employed in the

method based on the galaxy ages (Alaniz and Lima

2001, Jimenez and Loeb 2001). They all rely on the

fatthattheproperdistanetoanobjetatredshiftz

dependsonosmology:

S[r(z)℄= Z

z

0 dz

0

E(z 0

)

wherethefuntion S[x℄isj

k j

1=2

sin 1

(j

k j

1=2

x);x;

j

k j

1=2

sinh 1

(j

k j

1=2

(2)

ur-Friedmannequation

H 2

(z) = H 2

0 E(z)

2

E 2

(z) =

m (1+z)

3

+

de (1+z)

3wde

+

k (1+z)

2

:

The subsriptsrefer to matter(m) , darkenergy (de)

and urvature(k), andw

de

is thedarkenergy

param-eter of state, p

de = (w

de 1)

de

and by denition

m +

de +

k

=1. Itislearthereforethat all

quan-titiesthat depend ondistane, likeluminosity,angular

sizes, ages, also depend on osmology. For instane,

the method based on the SN Ia Hubble diagram

de-termines the distane to a supernova(whose absolute

magnitudeM hasbeeninferredandorretedbyloal

observations),viathedistane modulus

m M=5logd

L

(z)+25

(plus possibly aK-orretion)and thenomparesthis

value to the theoretial luminosity distane d

L (z) =

r(z)(1+z)to ndthe best t in terms ofthe

param-eters H

0 ;

m ;

de ;w

de

. The asew

de

=0orresponds

toapureosmologialonstantmodel.

The main advantages with this approah are that

thedependene ontheosmologialonstantis simple

andthatoneanobserver(z)atdierentredshifts,

de-pending on the objet. So far, SN Ia have been seen

up to z 1 (with a ouple of SN beyond this), but

future experiments an extendthe observations up to

z2andto amuh extendeddataset. Moreover,the

Alok-Pazyinsky method an be applied to quasars

atz5orbeyond. Thedistaneofthesoundhorizon

at deoupling, expressed as the angular diameter

dis-taned

A

(z)=r(z)=(1+z),inturn, reahesz 1000,

theredshift ofthe last satteringsurfae. Theruial

point isthat r(z)depends ontheparametersin a

dif-ferent way for dierent redshifts. In Fig. 1 we show

thelines ofonstantd

L

(z) forapure model

assum-ingthesouresareatz=0:7;4;10;1000: theontours

followadierentorientation,rotatingantilokwisefor

inreasingz,beingalmostvertialforz4. Therefore,

observationsatlowredshift,liketheSNIa,andathigh

redshift,likethesoundhorizonat deoupling,angive

orthogonalonstraints,thereby determiningwith high

preisiontheosmologialparameters.

The observationof the CMB utuation spetrum

fall in betweenthetwolasses,sine thespetrum

de-pends both on the bakground and on the

perturba-tions: in fat,at small angular salesthe spetrumis

mostlydeterminedbytheangular-diameterdistaneto

deoupling,whileatlargeangularsalesitdependson

the growth of the gravitational potentialthrough the

integratedSahs-Wolfeeetand ontheintrinsi

u-tuations ofthe darkenergyeld. In thisreview I will

fousmostlyontheostraintsfromthereenthigh

res-olution CMB data (Nettereld et al. 2001, Lee et al.

2001, Halverson et al. 2001) beause they appear to

rapid development. The reason is that they are not

subjet to the theoretial unertainties that still

en-shroud the SNIa method. In fat, in ontrast to the

SNIa,thephysisoftheCMButuationsisrelatively

well understood, and there are no problems of

evolu-tion. Moreover,experimentsareunderwaythatshould

expandthestatistisbyafatorofahundredormore

in afewyears.

0.2

0.4

0.6

0.8

M

0.2

0.4

0.6

0.8

z 10

0.2

0.4

0.6

0.8

M

0.2

0.4

0.6

0.8

z 1000

0.2

0.4

0.6

0.8

M

0.2

0.4

0.6

0.8

z 0.7

0.2

0.4

0.6

0.8

M

0.2

0.4

0.6

0.8

z 4

Figure 1. Contoursof onstant luminosity distane dL(z)

for various values of z. At small redshift the lines follow

approximately therelationm =onst, andatlarge

redshifttheorthogonal relation

m +

=onst.

Several works havetried to onstrain dark energy

models with the reent CMB data, leading in some

asestointerestingbutonitingresults. InAmendola

(2001) the dimensionless oupling parameter that

measuresthestrengthoftheinterationofdarkenergy

to darkmatterwithrespettothe gravitational

inter-ationwasonstrainedto besmallerthan 0.1roughly,

adoptinganexponentialpotential. Theslopeofthe

po-tential, ontheotherhand,wasfoundto beessentially

unonstrained by the data. Baigalupiet al. (2002)

foundthat, among power-lawpotentialsV

and

xing the Hubble onstant to h = 0:65 (in units of

100km/se/Mp),valuesofaroundunityarefavored

whiletheaseofpureosmologialonstant,=0,was

less likely by a fator of ve roughly. Corasanitiand

Copeland (2001),on theother hand,haveshown that

anequationofstatelosetothatofapureosmologial

onstantgivesthebest tto thedata,partiularlyfor

as onerns the position of the aoustipeaks. Bean

andMelhiorri(2002)alsoonludethatthepure

os-mologialonstantgivesthebestt to theCMBdata

whenthepriorontheHubbleonstantisbroadenedto

h=0:720:08.

Reently, we extended the previousstudies in two

(3)

en-teration withmatter. Seond,weput onlyveryweak

restritionsontheosmologialparametersh,

b; (the

density parameters of baryons and old dark matter,

respetively), and n

s

(the slope of the salar

pertur-bations). Theoniting resultsabovearein fat due

mostlytodierentpriors.

These generalizations allowed us to nd that the

present CMB data are apable to put a strong

on-straintonthe ouplingbut noton thesalareld

po-tential or equation of state. In fat, the degeneray

betweenh and w

, the darkenergy equation of state

(Huey etal. 1999)almost erasesthe sensitivityof the

CMB to thedarkenergypotential. Insharp ontrast,

theCMBspetraareverysensitivetothedarkenergy

oupling, sine the latter determines the equation of

state for a long stage after equivalene and there are

nostrongdegeneraieswiththeotherosmologial

pa-rameters.

II A Dark-dark oupling

Let us onsider two omponents, a salar eld

andCDM,desribedbytheenergy-momentumtensors

T

() andT

()

,respetively. Generalovariane

re-quirestheonservationoftheirsum,sothatitis

possi-bletoonsideraninterationbetweendarkenergyand

darkmattersuh that

T

();

= CT

()

; ;

T

();

= CT

()

;

: (1)

(2)

Suh a oupling an be obtained from the

onfor-mal transformation of a Brans-Dike gravity (see e.g.

Amendola 1999) and it has been onsidered several

timesinliteraturestartingfrom Wetterih(1988,1995)

and Wandset al. (1993). It hasbeen disussedin the

ontextofdarkenergymodelsinAmendola(1999,2000)

and, inalistwhihismeanttobeonlyrepresentative,

WandsandHolden(2000),Chimentoetal. (2000),

Bill-yard and Coley (2000), Chiba (2001), Albreht et al.

(2001),Esposito-FareseandPolarsky(2001). Inits

on-formallyrelatedBrans-Dikeformhasbeenstudied by

Uzan (1999), Chiba (1999), Chen and Kamionkowsky

(1999),Batista et al. (2000), Bertolami and Martins

(2000), Baigalupiet al. (2000), Faraoni (2000),Sen

andSen(2001). Theoretialmotivationsinsuperstring

modelsandinbraneosmologyhavebeenproposed

re-entlyin Gasperini, Piazza and Veneziano(2002)and

Pietroni(2002). Otherauthorsonsideredaouplingto

spei standardmodel elds ratherthanto ageneri

formofdarkmatter: Carroll(1998)to the

eletromag-netield, Horvat(2000)toneutrinos,Liet al. (2002)

to baryoni or leptoni urrent: in these ases, there

arestrong onstraintsfrom loal observationsorfrom

variationoffundamental onstants,seee.g. Damouret

al. (2002). Inontrast, a oupling to darkmatter is

observableonlywithosmologialexperiments,

involv-inggrowthofperturbationsorglobalgeometrieets

(properdistane).

IntheatonformalFRWmetrids 2

=a 2

( d 2

+

Æ

ij dx

i

dx j

)thesalareldanddarkmatteronservation

equationsare



+2H _

+a 2

U

;

= C

a

2

; (3)

_

+3H

= C

_

(4)

(dots refer to onformal time) where H = a=a._ We

supposenowthatthepotentialU()isspeiedbythe

followingrelation

U 0

=BU N

(5)

(theprime refersto derivationwithrespetto ): this

form inludes power laws, exponential potentials and

a pure osmologial onstant, i.e. the most ommon

forms of darkenergy potentials. In the aseof power

lawpotentialU =A

wehave

N =(1+)= (6)

andB = A ( 1=)

; whilefor theexponential

poten-tialN =1. We onsider only the rangeN 1sine

fornegativetherearenoasymptotiallyaelerating

models; for N ! 1wereoverthepure osmologial

onstant. Clearly,forA=0thepotentialvanishes,and

weredue to atheory whih is onformallyequivalent

topureBrans-Dikegravity.

Assumingthatthebaryonsarenotdiretlyoupled

tothedarkenergy(otherwiseloalgravityexperiment

wouldrevealafthfore, seeDamouret al. 1990)and

thatthe radiationaswellis unoupled (as itours if

the oupling is derived by a Brans-DikeLagrangian,

see e.g. Amendola 1999), the system of one Einstein

equationandfouronservationequations(forradiation,

,baryons,b,CDM,,andsalareld)anbe

onve-nientlywrittenintroduingthe followingvevariables

thatgeneralizeCopelandetal. (1997):

x=

H _

p

6 ; y=

a

H r

U

3 ; z=

a

H r

3 ; v=

a

H r

b

3

; w= H

a

(7)

where 2

=8G(notiethatw=dloga=dti.e. theusualHubbleonstant). Notiethat x 2

;y 2

;z 2

;et. orrespond

to thedensityparameterof eahomponent. Intermsoftheindependentvariablelogawehavethenthesystem:

x 0

=

z 0

1

x y

2N

w 2N 2

+(1 x 2

y 2

z 2

v 2

(4)

y 0 = xy 2N 1 w 2N 2 +y 2+ z 0 z ; z 0 = z 2 1 3x 2 +3y 2 z 2 ; v 0 = v 2 3x 2 +3y 2 z 2 w 0 = w 2

3+3x 2 3y 2 +z 2 (8) d where =C r 3 2 2

; =3 N 1 2N B p 6 : (9)

The dimensionless onstant 2

an be seen asthe

ra-tioofthedarkenergy-darkmatterinterationwith

re-spettogravity. Itanbeshownin fat(Damourand

Nordtvedt 1993, Wetterih 1995) that the fore

at-ing betweendarkmatterpartiles an bedesribed in

the Newtonian limit as arenormalized Newton's

on-stant ^

G=G(1+4 2

=3). Thisshowsthat,in termsof

the \post-Einsteinian" parameter (see Groom et al.

2001) whih haraterizesthe deviation from General

Relativity, we have = 4 2

=3. The eet of this

in-terationonstruturegrowthisdisussedinAmendola

andTohini-Valentini (2002).

Thesystem(8)inludesseveralqualitatively

dier-ent behaviors, already disussed in Amendola (2000),

Tohini-ValentiniandAmendola(2002). However,for

the range of values that are of osmologial interest,

the system passes through three distint phases after

equivalene.

MDE. Immediately after equivalene, the system

entersamatterdominatedepohwithanon-negligible

ontribution, that we denote MDE, in whih the

darkenergypotentialdensityisnegligiblewhilethe

ki-netienergydensityparameter

K =x

2

ofthesalar

eld givesaonstantontribution tothetotaldensity.

Aswillbeshowninthefollowing,theexisteneofsuh

an epoh is ruial forthe onstraintsthat we will be

able to put on the oupling. It is not diÆult to see

that,negletingradiationandbaryons,thepointy=0,

x = 2=3 is a saddlepointsolution of thesystem (8)

foranyNandforjj< p

3=2. Thesystemstaysonthe

MDEsolutionuntily startsgrowing. Alongthis

solu-tionthesalefatorexpandsslowerthanapureMDE,

i.e. as

at 4=(6+4

2

)

: (10)

Siney=0ontheMDE,itsexisteneisindependent

of the potential, although it has to veried for eah

potentialwhetheritisasaddle.

Traking trajetories. Let us now neglet baryons

and radiation and put = 0. Thetraking solutions

foundin Steinhardtet al. (1999)assumey=x=pand

y 2N 1

w 2N 2

=q where p;q are twomotion integrals.

Inthelimity 2

;x 2

1itiseasytoshowthatp 0 =q 0 =0 if p 2

= 4N 3 (11)

q =

3p

2(2N 1)

(12)

The sametraking behavior remains a good

approxi-mation for small . ForN = 1the traking solution

beomes atually a global attratorof the dynamial

system,see below. Intheother ases,thetraking

in-terpolatesbetweentheMDEandtheglobalattrator.

Globalattrators. Inthesystem(8)forN 6=1there

exists only the attrator x = 0;y = 1 on whih the

dark energy ompletely aounts for the matter

on-tent. For N = 1 the phase spae is muh riher,

and there are several possible global attrators, only

two of whih aelerated (Amendola 2000). One, for

>( + p

18+ 2

)=2,presentsaonstantnon-zero

: this attratoratually oinideswith thetraking

solutionsand in fat realizesthe onditionp= 1and

y = 3=(2) asrequestedbyEq. (11)for =0. This

\stationaryattrator"anbeaeleratedif >2and

ould solvetheoinidene problemsine

/

; it

hasbeendisussedindetailinAmendolaand

Tohini-Valentini (2001, 2002). The other aelerated

attra-tor ourswhen <( + p

18+ 2

)=2: in this ase

therearenotrakingsolutionsandtheglobalattrator

redues to x = 0;y = 1 as for N 6= 1. These

solu-tionshave alreadybeenompared to CMBfor N =1

inAmendola(2001). Theinlusionofthebaryons

mod-ies the onsiderationsabovebut, as longastheyare

muhsmaller thantheother omponents,the

qualita-tivebehaviorofthesystemremainsthesame. Itistobe

notied thatthenal attrator,onwhih thedark

en-ergydominatesompletelytheosmiuid,isyettobe

reahed, and therefore the existeneof an aelerated

epohatthepresentdepends mostlyonthetraking.

TheexisteneoftheMDEsaddleandofthe

trak-ing solutionsis ruial for ouranalysis. Infat, these

twoepohsguaranteethat theequationof stateof the

salareldispiee-wiseonstantthroughessentiallyall

thepost-equivaleneepoh. IntheMDEphasethe

ef-fetive parameter of state w

e =p

tot =

tot

+1 and the

eld equationofstatew

=p

=

+1are

w

e =1+

4 9 2 ; w

=2 (13)

whileduring thetrakingphase

(5)

(thelastrelationisapproximatedanditisatuallyonly

anupperbound tothepresentw

;itismorepreiseif

w

is identied with the average equation of state

af-ter MDE rather thanthe present equation of state).

Therefore,theosmievolutiondependsonalone

dur-ing the MDE, and on N alone during the traking.

Sine the position of the aousti peaks is related to

theequationofstatethroughtheangulardiameter

dis-tane, it appears that the CMB is ableto put diret

onstraints on both and N. Sine the sign of is

ininuent, we onne ourselveshereinafter to > 0:

InFig. 2weshowatypialtrajetorythatpresentsin

sequene the three epohs disussed above. InFig. 3

wepresentathree-dimensionalphasespaex;y;z (i.e.,

eld kinetienergy,potentialenergy,andradiation

en-ergy, respetively) for N =1, negletingbaryons(the

equation forw deouplesfor N =1). Notiethe

loa-tionoftheMDEandthenalattrator.

0

10

1

10

2

10

3

10

4

10

5

z

0

0.5

1

1.5

2

w

,w

e

w

w

e

MDE

T

A

0

10

1

10

2

10

3

10

4

10

5

0.001

0.01

0.1

1

rad

cdm

b

y

2

b

MDE

T

A

Figure 2. Numerial solutions of the system (8) for N =

2;=0:1;!

=0:1;!

b

=0:02;h=0:65 plottedagainstthe

redshift. Upper panel. Long dashedline: radiation; short

dashedline: CDM;unbrokenthinline: baryons;unbroken

thikline: salareld;dottedline: salareldpotential

en-ergy. Thehorizontalthinlinemarksthekinetienergy

den-sityofMDE,reahedjustafterequivalene. Lowerpanel.

Thikline: eetiveequationofstate;thinline: darkenergy

equationofstate. ThelabelsmarktheMDE,thetraking

N

1.,

Β

0.5,

Μ

3.

-1

-0.5

0

0.5

1

x

0

0.2

0.4

0.6

0.8

y

0

0.25

0.5

0.75

1

z

MDE

A

-1

-0.5

0

0 5

Figure3. Three-dimensionalphase spae for N =1. The

trajetories start at x = 1. They all end at the global

attrator,butmostpassthroughtheMDEsaddle.

There is however a ruial dierene between the

MDEandatrakingforwhatonernshere:whilethe

presentdarkenergyequationofstate,setbythe

trak-ing, is degenerated with h for as onerns the CMB

spetrum (see e.g. Huey et al. 1999; Bean and

Mel-hiorri2002), the equation of state during the MDE

isnot.Infat,theangulardiameterdistanetothelast

satteringsurfaed

A

isdegeneratealonglinesh(w

)for

whih

d

A

Z

1

ade da

!

a+(h

2

!

)a

4 3w

1=2

=onst

(15)

where!

h

2

(althoughthisisexatonlyfor=0

it remains a good approximation even for small

non-zerovalues). InFig. 4thedegeneraybetweenspetra

fordierent valuesof h and N (the other parameters

beingequal)appearslear. Assumingastrongprioron

hapeakemergesin thelikelihood forN but thenthe

resultislearlyprior-dependent. The sameholdstrue

formodelsinwhihtheequationofstateisslowly

vary-ing(seee.g. Huey etal. 1999,Doranetal. 2001). On

theotherhand, thefat that

6=0at deouplingin

oupledmodelsimpliesthattheeetsoftheoupling

ontheCMBarenotduesolelytotheangulardiameter

distane, and therefore the geometri degeneray an

be broken. This isshownin Fig. 5in whih C

`

spe-trafor variousvaluesof (allother parametersbeing

(6)

200

400

600

800

1000

{

1000

2000

3000

4000

5000

6000

7000

8000

{

{

1

C

{

2

Π

Μ

K

2

h

0.8 N

3.2

h

0.75 N

2.5

h

0.67 N

1.7

Figure4. Spetrafordegeneratedparameters(theother

pa-rametersare=0;!

b

=0:01;!

=0:1). Thespetrahave

beensaledupforlarity: theyoverlapalmostexatly.

200

400

600

800

1000

1000

2000

3000

4000

5000

6000

7000

8000

1

C

2

K

2

✁✂

0.

✁✂

0.1

✁✂

0.2

Figure5. CMBspetraforinreasingvaluesoftheoupling

onstant(theotherparametersareN=1:5;h=0:65;!b=

0:01;!=0:1).Notiethatthepeaksnotonlymovetothe

rightbutalsohangeinheight.

III Constraining past and

present equation of state

with CMB

Ourtheoretialmodeldependsonthreesalareld

pa-rametersandfourosmologialparameters:

;;N;n

s ;h;!

b ;!

(16)

where !

b

=

b h

2

and !

=

h

2

. In Amendola et

al. (2002)wealulated theC

`

spetrabyamodied

CMBFAST(SeljakandZaldarriaga1996)odethat

in-ludes the full set of oupled perturbation equations.

WeusedasdatasettheCOBEdataanalyzedinBondet

al. (2000),andthehigh resolutiondataofBoomerang

(Nettereld et al. 2002), Maxima (Lee et al. 2002),

andDASI(Halversonet al. 2002). Weadopted apure

log-normal likelihood; the overall amplitude and the

alibrationerrorsofBoomerang(10%)andofMaxima

and DASI(4%) havebeen integratedout analytially.

Thetheoretialspetrahavebeenbinnedasthe

exper-analyses we assume uniform priors with the

parame-tersonnedintherange 2(0;0:3); N 2(1;8:5);

n

s

2 (0:7;1:3); h2 (0:45;0:9); !

b

2(0:005;0:05);

!

2(0:01;0:3). Thesameageonstraints(>10Gyr)

usedin mostpreviousanalysesisadoptedhere.

0

0.2

0.4

0.6

0.8

w

0.5

0.6

0.7

0.8

0.9

h

1.1

1.5

2

5

N

Figure 6. Likelihood ontour plots in the spae N;h

marginalizing over the otherparameters at the 68,90 and

95% .l.. The white thik linesrefer to the oupled ase,

theblakurvestothe unoupledase. Thedottedline is

the likelihooddegeneray urve, thedashedline is the

ex-peteddegenerayurve. Notiethat onlyxinghsmaller

than0.65 itwouldbepossibletoexlude theosmologial

onstantat95%.l..

In Fig. 6 we show the likelihood urves for N;h

marginalizingovertheremainingparameters. The

on-tours ofthe likelihoodplotfollowtheexpeted

degen-eray of the angular diameter distane. The residual

deviation from the expeted degeneray line is due to

theageprior(>10Gyr)thatfavorssmallhvalues.

No-tiethatforh>0:75noupperlimittoN anbegiven

withtheurrentCMBdatanomatterhowpreisehis,

and that only assuming h < 0:65it beomes possible

to exludeat95%.l. thepureosmologialonstant.

In Fig. 7 we show the likelihood for all parameters,

marginalizingin turnovertheothers. Wendthe

fol-lowingonstraintsat95%.l.:

N >1:5; <0:16 (17)

(the limiton N orresponds to <2). The limit on

N is however only a formal one: the likelihood never

vanishes in the denition domain and even N = 1

(the exponentialpotential) is onlyafator of veless

likelythan thepeak and ertainlyannot be exluded

on this basis. Moreover, the lower bound on N is

prior-dependent: allowingsmallervaluesofhit would

weaken. Clearly,ifweadoptnarrowpriorsonh,wean

(7)

not depend sensibly on the prior on h and the

likeli-hoodfor doesvanishatlarge. InplaeofN and

weanuseaswelltheequationofstateduringtraking

andduringMDE,respetively,aslikelihoodvariables,

using Eqs. (13)and (14). Then weobtainat the95%

.l.

w

(trak ing)

<0:8; 1<w

e(MDE)

<1:01 (18)

This shows that the eetive equation of state

dur-ing MDE, i.e. between equivalene and traking, is

lose to unity (as in a pure matter dominated epoh)

towithinoneperent. Thestrikingdierenebetween

the level ofthe twoonstraintsin (18)wellillustrates

the main point ofthis paper: theCMB is muh more

sensitivetothedarkenergyouplingthantoits

poten-tial.

Figure 7. Marginalizedlikelihoodfortrakingtrajetories.

Theequationofstateinpanelaisw=1=(2N 1). In

pan-elsaandbweplotasadottedlinethelikelihoodassuming

h=0:650:05andasdashedlineh=0:750:05(gaussian

prior). In panel b the long-dashed line is the Plank-like

likelihood. Inpaneldthelong-dashed lineisfor=0.

Theotherparametersare(wegiveherefor

simpli-ity themean and theonesigma error,while the limit

isat the95%.l.)

n

s

=0:990:05; !

b

=0:0230:004;

!

=0:0920:02; h>0:62 (19)

Thetotaldarkenergydensityturnsouttobe

>0:60

(95% .l.). Itappearsthatthelimitsonthe

osmolog-ialparametersn

s; !

b ;!

arealmost independent of,

whileanon-zero favorshigherh(Fig. 7,paneld). It

is interesting to omparewith the urrentonstraints

from the Hubble diagramof the supernovaeIa, where

one obtainsquiteastrongerbound ontheequationof

state,w

<0:4orN >1:75atthesame.l..

Finally,inFig. 7(panelb)weplotthelikelihoodfor

that an experiment with no alibration unertainty

mission (de Zotti et al. 1999), an ahieve. We nd

< 0:05 (95% .l.) (using the speiations of the

MAPsatellitewend <0:1).

IV Conlusion: the CMB as a

gravity probe

Adarkmatter-darkenergyinterationwouldobviously

esapeanyloal gravityexperiment: osmologial

ob-servations liketheCMB are then theonly wayto

ob-servesuh aphenomenon. Sine observations require

thebaryonstobedeoupledfrom darkenergy(or

ou-pledmuh moreweaklythan darkmatter),thesearh

foranon-zero isinfatalsoatestoftheequivalene

priniple.WefoundthaturrentCMBdataareapable

toputaninterestingupperboundtothedarkmatter

-darkenergyoupling:

<0:16

(95%.l.) regardlessofthepotential(withinthe lass

weonsidered). This impliesthat thesalargravityis

at least1= 2

40 times weakerthan ordinarytensor

gravity. As shownin e.g. Amendola (1999),the limit

on anberestatedasalimitontheonstantofthe

non-minimally oupledgravity, dened bythe gravity

Lagrangiandensity

L= p

g

1 1

2

2

R+L

(20)

whereL

istheanonialsalareld Lagrangian(we

areassumingthat theLagrangianforthebaryonelds

ouplestoaonformallyrelatedmetri). Sineforsmall

onehas 2 2

=3,wehave<0:017. Similarly,one

an express the limit in terms of the original

Brans-Dikeparameter! denedbytheLagrangian

L= p

g( R+ !

;

;

) (21)

Inthisasewehave!>60.

An experiment like the Plank mission an lower

theupperbound to to 0.05(95% .l.): salar

grav-ity would be in this ase at least 400 times weaker

thanordinarytensorgravity. This limitisomparable

to thosethat loal gravityexperimentsimpose onthe

salargravityouplingtobaryons,expressedintermof

the\post-Einsteinian"(Groomet al. 2001)parameter

= 4 2

=3 = 0:00070:0010, i.e. 2

baryons < 10

3

(seee.g. Groometal. 2000).

In ontrast, CMB data, on their own, annot put

any rm limit to the dark energy potential, unless a

narrow prior on h is adopted: e.g. h = 0:650:05

gives w

= 0:550:2 while h = 0:750:05 gives

w

=0:350:2(assuminggaussianpriorsand

marginal-izingasusual overall theother parameters,inluding

theoupling).

(8)

ex-assumingthattheSNIadimmingisausedbysome

evo-lutionaryeet,wearestill onfrontedwith the

prob-lem of reonilinga atuniverse asseenby theCMB

data with the small fration of matter in lusters of

galaxies. Lakinganydenitesuggestionastoits

prop-erties,osmologistsannotdomuhbetterthantrying

to inferits nature fromobservations. Wemodeledthe

darkenergyasasalareld governedby ageneri

in-versepower-lawpotentialandaouplingto dark

mat-ter. The ouplingintroduesanewinterationthat is

unobservableonloalsalesbutgivesstrongsignatures

forasonernstheperturbationgrowthandtheproper

distane. Wereviewedtheurrentonstraintsfromthe

CMB, nding that theoupling of thedarkenergy to

darkmatterannotexeed0.16intermsofthestrength

ratio. Thefuture data from thePlankmission an

reduetheupperlimittoalevelomparabletothe

lim-its that loal gravityexperimentsput ontheoupling

tobaryons.

Referenes

[1℄ A.Albreht,C.P.Burgess,F.Ravndal,andC.Skordis,

astro-ph/0107573.

[2℄ J.S. Alaniz and J. A. Lima,Astrophys.J. 521, L87

(1999).

[3℄ L.Amendola,Phys.Rev.D60,043501(1999).

[4℄ L.AmendolaandD.Tohini-Valentini,Phys.Rev.D

64,043509(2001).

[5℄ L.Amendola,Phys.Rev.D62,043511(2000).

[6℄ L.Amendola,Phys.Rev.Lett.86,196(2001).

[7℄ L. Amendola and D. Tohini-Valentini(2002)

astro-ph/0111535.

[8℄ L. Amendola, C. Querellinni, D. Tohini-Valentini,

andA.Pasqui,(2002).

[9℄ C. Baigalupi, F.Perrotta, and S.Matarrese, Phys.

Rev.D61,023507(2000).

[10℄ C. Baigalupi, A. Balbi, S. Matarrese, F. Perrotta,

andN.Vittorio,(2001).

[11℄ A.B.Batista,J.C.Fabris,andR.deSaRibeiro,Gen.

Rel.Grav.33,1237 (2001).

[12℄ R.BeanandA.Melhiorri,astro-ph/0109097.

[13℄ O. Bertolami and P. J. Martins, Phys. Rev. D 61,

064007(2000).

[14℄ A.P.BillyardandA.A.Coley,Phys.Rev.D61,083503

(2000).

[15℄ J.R.Bondetal.,Ap.J.533,19(2000).

[16℄ R.R.Caldwell,R.Dave,andP.J.SteinhardtPhys.Rev.

Lett.80,1582 (1998).

[17℄ M.O. Calvao, J.R.T. De Mello Neto, and I. Waga,

astro-ph/0107029.

[18℄ X. Chen and M. Kamionkowsky, Phys. Rev. D60,

104036(1999).

[19℄ T.Chiba,Phys.Rev.D60,083508(1999).

[21℄ L.P.Chimento,A.S.Jakubi,andD.Pavon,Phys.Rev.

D62,063508(2000).

[22℄ A.R.CoorayandD.Huterer, Astrophys.J.513,L95

(1999).

[23℄ E.J.Copeland,A.R.Liddle,andD.Wands,Phys.Rev.

D57,4686(1997).

[24℄ P.S.CorasanitiandE.Copeland,astro-ph/0107378.

[25℄ T. Damour and K. Nordtvedt, Phys. Rev. Lett. 70,

2217(1993).

[26℄ T.Damour, G.W.Gibbons,and C.Gundlah,Phys.

Rev.Lett.64,123, (1990).

[27℄ T. Damour, F. Piazza, and G. Veneziano,

hep-th/0205111.

[28℄ G.deZottietal.,(1999),AIPConf.Pro.476: 3K

os-mology,204,eds.L.Maiani, F.Melhiorri,N.Vittorio

(Woodbury,N.Y.: AmerianInstituteofPhysis)

[29℄ M. Doran, M. Lilley, and C.Wetterih, (2001)

astro-ph/0105457.

[30℄ M.Doran,M.J.Lilley,J.Shwindt,andC.Wetterih,

Astrophys.J.559,501(2001).

[31℄ G.Esposito-Fareseand D.Polarsky, Phys.Rev.D63,

063504(2001).

[32℄ V.Faraoni,Phys.RevD62,023504(2000).

[33℄ J.Frieman,C.T.Hill,A.Stebbins,andI.Waga,Phys.

Rev.Lett.75,2077 (1995).

[34℄ M. Gasperini,F.Piazzaand G.Veneziano,(2001)

gr-q/0108016.

[35℄ F.Giovi,F.Ohionero,andL.Amendola,Mon.Not.

Roy.Astron.So.325,1097(2001).

[36℄ D.E.Groometal.Eur.Phys.J.C15,1(2000).

[37℄ N.W.Halversonetal.,(2001)astro-ph/0104489.

[38℄ D.J. Holden and D. Wands,Phys.Rev. D61, 043506

(2000).

[39℄ R.Horvat,astro-ph/0007168.

[40℄ G.Huey,L. Wang, R. Dave,R.R. Caldwell, andP.J.

Steinhardt,Phys.Rev.D59,063005(1999).

[41℄ R.JimenezandA.Loeb,arXiv:astro-ph/0106145.

[42℄ A.T.Leeetal.,Ap.J.561,L1(2001).

[43℄ M. Li, X. Wang, B. Feng, and X. Zhang

astro-ph/0112069.

[44℄ C.B.Nettereldetal.,(2001)astro-ph/0104460.

[45℄ J.Newman,C.Marinoni,A.CoilandM.Davis,

astro-ph/0109131.

[46℄ S.Perlmutteretal.Ap.J.517,565(1999).

[47℄ M.Pietroni,(2002)hep-ph/0203085.

[48℄ B.RatraB.andP.J.E.Peebles,Phys.Rev.D37,3406

(1988).

[49℄ A.G.Riessetal.A.J.,116,1009(1998).

[50℄ A.A.SenandS.Sen,MPLA,16,1303(2001).

[51℄ SeljakandM.Zaldarriaga,Ap.J.469,437(1996).

(9)

[53℄ D. Tohini-Valentini and L. Amendola, Phys. Rev.

D65,063508(2002).

[54℄ J.P.Uzan,Phys.Rev.D59,123510 (1999).

[55℄ D. Wands,E.S. Copeland, and A. Liddle, Ann.N.Y.

Aad.Si.688,647(1993).

[56℄ C.Wetterih,Nul.Phys.B302,668(1998).

Imagem

Figure 1. Contours of onstant luminosity distane dL(z)
Figure 2. Numerial solutions of the system (8) for N =
Figure 5. CMB spetra for inreasing values of the oupling
Figure 7. Marginalized likelihood for traking trajetories.

Referências

Documentos relacionados

Observando os dois cartogramas que representam a distribuição da população rural, em 2000 e em 2010, nota-se que de maneira geral houve uma redução da população rural no Brasil.

De acordo com vários autores (Stopar et al., 2003; Bregoli e Fabronni, 2007) a aplicação conjunta de mondadores à base de auxinas e citocininas são mais eficientes do que

Encontra-se que será difícil compreender ou gerir a identidade corporativa, sem conhecer a origem de tal fenómeno, justificando um estudo dos diferentes antecedentes ou origens

É neste sentido que a gestão de projetos pode servir como uma importante ferramenta para a melhoria dos serviços públicos e da forma de atuação das organizações públicas,

We investigated the magnitude of population replacement in Britain with qpAdm 2 by modelling the genome-wide ancestry of Neolithic, Copper and Bronze Age individuals,

The questionnaire delivered at the end of the second cycle to the two groups was prepared based on what had been presented in the zero cycle, but now had as its aim:

O Estatuto da Cr iança e do Adolescente (ECA), a Reforma Psiquiátr ica e o Sistema Único de Saúde (SUS) reforçaram as disposições da Constituição Federal de 1988 e a Rede

The LPM comprises different costs, such as: (i) system costs, which are related to energy production; (ii) network congestion, which is the cost for using other generation