ebook img

Separating Dark Physics from Physical Darkness: Minimalist Modified Gravity vs. Dark Energy PDF

0.27 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Separating Dark Physics from Physical Darkness: Minimalist Modified Gravity vs. Dark Energy

Separating Dark Physics from Physical Darkness: Minimalist Modified Gravity vs. Dark Energy Dragan Huterer1 and Eric V. Linder2 1Kavli Institute for Cosmological Physics and Astronomy and Astrophysics Department, University of Chicago, Chicago, IL 60637 2Berkeley Lab, University of California, Berkeley, CA 94720 (Dated: February 5, 2008) The acceleration of the cosmic expansion may be due to a new component of physical energy density or a modification of physics itself. Mapping the expansion of cosmic scales and the growth oflargescalestructureintandemcanprovideinsightstodistinguishbetweenthetwoorigins. Using Minimal Modified Gravity (MMG) – a single parameter gravitational growth index formalism to parameterize modified gravity theories – we examine the constraints that cosmological data can placeonthenatureofthenewphysics. Fornextgenerationmeasurementscombiningweaklensing, 7 supernovaedistances, and thecosmic microwave background we can extend thereach of physicsto 0 0 allow for fitting gravity simultaneously with the expansion equation of state, diluting the equation 2 ofstateestimationbylessthan25%relativetowhengeneralrelativityisassumed,anddetermining the growth index to 8%. For weak lensing we examine the level of understanding needed of quasi- n and nonlinear structure formation in modified gravity theories, and the trade off between stronger a precision but greater susceptibility tobias as progressively more nonlinear information is used. J 1 3 I. INTRODUCTION approach and its range of validity. §III lays out the cos- mologicalprobes, fiducial models, and survey data char- 2 acteristics used in our analysis. We examine in §IV the v Whether the acceleration of the cosmic expansion is ability of next generation cosmological probes to reveal 1 due to a new physical component or a modification of 8 gravitation, the answer will involve groundbreaking new new gravity simultaneous with fitting cosmological ex- 6 pansion,andthe cosmologicalbiasincurredifweneglect physics beyond the currentstandardmodels for high en- 8 toallowthepossibilityofbeyondEinsteingravity. Lever- ergyphysicsandcosmology. Toobtaintheclearest,unbi- 0 age and systematics from the nonlinear regime are dis- asedpictureofthe fundamentalphysicsweneedtoallow 6 cussed in §V. 0 forthe possibility ofgravitationbeyondEinsteinandsee / where the data lead. This article presents simultaneous h fitting ofthe cosmicexpansionandthe theoryofgravity. p - Direct measurements of the expansion history can be II. GRAVITATION AND GROWTH o interpreted generally as the equation of state of the uni- r t verse;thismayormaynotcorrespondtoaphysicalcom- Attempting to invent a generalprescription for taking s ponent. Measurementsofthegrowthhistoryofmassfluc- into account the effects of modified gravity is like seek- a : tuations combine information on the expansion and the ing a general treatment of non-Gaussianity: there are so v theoryofgravity. Thetwoincomplementaritythusallow many ways in which a theory can be “not”, in which a i X separation of the possible origins for cosmic acceleration symmetry can be broken, that it seems a hopeless task. – a physicaldark component, e.g. a new field in high en- However,wedonotseekanall-encompassingdescription r a ergy physics, or dark (new) physics, e.g. a modification of modified gravityin all its aspects, but rather its gross of Einstein gravity. effectsoncosmologicalobservablesbeyondtheexpansion Within a specific theory of modified gravity one can rate. attempt to calculate the cosmological observables and As a beginning step, we take a fairly conservative ap- determine the goodness of fit with data. This model de- proach we call Minimal Modified Gravity (MMG). The pendent approach must proceed on a case by case basis modifications we consider are small, since general rela- and moreover suffers from difficulties in computation of tivity givesa predominantlysuccessfuldescriptionofthe many quantities due to the complexity of the theories. universe and its large scale structure, and homogeneous, Nor are the modifications necessarily well motivated or i.e.notdependentonenvironment`alachameleonscenar- completelyfreefrompathologies. Thealternateapproach ios [1]. We assume that structure formationcontinuesto taken here is phenomenological, using a model indepen- be described adequately by growth from Gaussian den- dent yet physically reasonable and broad parameteriza- sity perturbations. For weak gravitational lensing, we tionofthegravitymodificationtogaininsightintotheef- do not have to get too deeply involved in the nonlinear fect ofgeneralizingEinsteingravity. This is closelyanal- density regime and can concentrate on the growth law; ogoustothewidelysuccessfulequationofstateapproach weassumechangestothegravitationaldeflectionlaware to modifications of the expansion history, giving model negligible. independent constraints and understanding. This approach seeks to build understanding of modi- In §II we discuss the development of the parametrized fied gravity by taking a modest step away from general 2 relativity. By examining an alteration in the linear per- A key aspect to note is that much of the growth is turbationgrowth,preservingthestandardmappingfrom determined by the expansion history, even in modified linear to nonlinear fluctuations, we obtain a clear pic- gravity,soweshouldnotthrowawaythatknowledge. By ture of a specific effect of modified gravity (see §V for following the effects, treating the expansion in terms of relaxation of the assumption of standard mapping and the well-developed equation of state formalism (whether discussion of scale dependence). This serves as a proxy arising from a physical dark energy or a modified Fried- for presumably more complicated and model dependent mann equation), i.e. effects. The form of the growth equation can be written so as H2(z)/H02 = ΩM(1+z)3+δH2(z) (5) to directly show the influence of the cosmic expansion 1 dlnδH2 and the additional effect of the gravity theory. For a w(z) = −1+ , (6) 3dln(1+z) matter density perturbation δ ≡δρ/ρ, the linear growth factor g = δ/a (scaling out the matter dominated uni- andaddinganewparametertoincorporatetheeffectsof verse behavior δ ∼a) evolves in general relativity as modified gravity specifically on the growth source term, we render the physics appropriately. This was the mo- 1dlnH2 g′′ + 5+ a−1g′ tivation behind the gravitational growth index formal- (cid:20) 2 dlna (cid:21) ism of [2], which we follow, calling the ansatz MMG. An 1dlnH2 3 alternate approach is to define wholly separate growth + 3+ − Ω (a) a−2g =0, (1) (cid:20) 2 dlna 2 M (cid:21) variables (see, e.g., [8, 9]). The gravitational growth index serves as a proxy for whereaprimedenotesderivativewithrespecttothescale thefullmodifiedgravitytheory. Thelineargrowthfactor factor a, H = a˙/a is the Hubble parameter, and ΩM(a) is approximated by isthedimensionlessmatterdensity. Sincetheglobalcos- mology parameters H and ΩM(a) are, essentially, the g(a)=e 0adlna[ΩM(a)γ−1], (7) expansion history, we see that the cosmic expansion de- R termines the structure growth. where γ is the growth index. This was shown to be ac- To make the relation between growth probes and ex- curate to 0.2% compared to the exact solution within pansion probes such as the luminosity distance-redshift general relativity for a wide variety of physical dark en- relationship r (z) even more explicit, we can start from l ergy equation of state ratios. We verify explicitly that the growth equation as for dynamical dark energy where the equation of state δ¨+2Hδ˙−4πρδ =0, (2) ratio is parametrized as w(a)=w0+wa(1−a), the for- mulas for γ(w) given in [2] recover g(a) to better than where a dot denotes time derivative, and use 0.3% when w0+wa < −0.1. That is, the growth index parametrization of linear growth is extremely robust as r =a−1 dt/a=a−1 da/(a2H) (3) long as the early matter dominated epoch is not upset. l Z Z The gravitational growth index formalism has also beentestedandfoundaccurateto0.2%forasinglemod- (for a flat universe to keep the notation simple) to write ified gravity scenario [2], DGP [10, 11] braneworld grav- d2δ dδ ity, giving γ =0.68. We conjecture that it may work for dr2(a−2−Hrl)2− dr (Ha−2+a¨/a)−4πρδ =0. (4) modifiedgravitytheorieswith monotonic,smooth(Hub- l l ble timescale) evolution in the source term, so long as Thusingeneralrelativitythegrowthclearlycontainsthe the matterdominatedepochis notdisrupted. Forexam- same cosmologicalinformation as the distance relation. ple,inDGPthesourcetermreceivesasmoothcorrection As discussed in [2], we can alter the growth equation 1−(1/3)(1−Ω2 (a))/(1+Ω2 (a)) [11–14]. Preliminary M M in modified gravity by changing the matter source term resultsinscalar-tensortheoryalsoindicatesuccessfulap- (proportional to Ω (a) in Eq. (1) or ρδ in Eq. (2)) or proximation [15]. Future work includes testing this for M adding a new source through a nonzero right hand side. other specific models; here we use the growth index as Greenfunctionsolutionsforsuchmodificationsaregiven an indicator of possible effects of modified gravity, with in [2]. The matter source term can be written as Qδ, the advantage of knowing at least it is robust for many and Q = ∇2Φ/δ arises from the equivalent of the Pois- cases beyond ΛCDM. son equation relating the metric potential Φ to the mat- ter perturbation δ. One possibility is to make a phe- nomenological modification to the Poisson equation and III. FIDUCIAL MODEL AND COSMOLOGICAL investigate its effects on structure growth; this has been DATA investigatedby [3–6] and we revisitit in §V. An intrigu- ingapproachofgeneralquadraturerelationsbetweenthe To assess the leverage of cosmological observations matter perturbations and metric potentials is discussed to reveal dark physics vs. physical darkness, we si- by [7]. multaneously fit nine cosmological parameters: A, the 3 normalization of the primordial power spectrum at is [20] k = 0.05hMpc−1; physical matter and baryon den- fid hγ2 i nsietuietsrinΩoMmh2assaensdmΩνB,hm2a,ttseprecetnraerlgiynddeexnsnit,ystuomdayofretlhae- Ciκj(ℓ)=Piκj(ℓ)+δij n¯init , (10) tive to critical ΩM, and parameters describing the ef- where hγ2 i1/2 is the rms intrinsic shear in each compo- fective dark energy equation of state w and w , where int 0 a nent, taken to be 0.22, and n¯ is the average number of i w(a) = w + (1 − a)w . The mass power spectrum 0 a galaxiesintheithredshiftbinpersteradian. Thecosmo- ∆2(k,a)≡k3P(k,a)/(2π2) is written as logicalconstraintscanthenbecomputedfromtheFisher matrix ∆2(k,a)= 254ΩA2M (cid:18)kkfid(cid:19)n−1(cid:18)Hk0(cid:19)4g2(a)T2(k)Tnl(k,a) FiWj L =Xℓ ∂∂Cpi Cov−1 ∂∂pCj, (11) (8) where p are the cosmological parameters and Cov−1 is where T(k) is the transfer function, and T (k,a) is the i nl theinverseofthecovariancematrixbetweentheobserved prescription for the nonlinear power. Modified gravity power spectra whose elements are given by enters through the ninth parameter, the gravitational growth index γ in the linear growth function g(a) from Cov Cκ(ℓ),Cκ(ℓ′) = δℓℓ′ × (12) Eq. (7), with the growth index γ = 0.55 for the fiducial ij kl (2ℓ+1)f ∆ℓ (cid:2) (cid:3) sky cosmologycorrespondingto aflatuniversewith Einstein Cκ(ℓ)Cκ(ℓ)+Cκ(ℓ)Cκ (ℓ) . gravity and a cosmological constant. The other fiducial ik jl il jk (cid:2) (cid:3) values adopted correspond roughly to the current con- ThefiducialWLsurveyassumes1000squaredegreeswith cordance cosmology, with Ω = 0.3, w = −1, w = 0, tomographicmeasurementsin10uniformlywideredshift M 0 a Ω h2 = 0.023, Ω h2 = 0.14, n = 0.97, m = 0.2 eV bins extending out to z =3. The effective source galaxy B M ν (one massive species), and A=2×10−9 (corresponding density is 100 per square arcminute. to σ ≃ 0.9). While the exact values of some of these WewillalsosometimesconsideraSouthPoleTelescope 8 parameters,especially σ , are still a subject of much de- (SPT [21]) type cluster survey with sky coverageof5000 8 bate, we do not expect that different values allowed by deg2 andatotalofabout25000clusters(fortheassumed current data will change any of our conclusions on the cosmology with σ = 0.9), giving a number-redshift test 8 detectability of non-standard growth. involving the geometric volume and the number density The linear power spectrum uses the fitting formulae fromgrowthofstructure. For simplicity, we neither con- of [16]. We always use the linear growth function from sider the additional information provided by masses of Eq. (7) and account for its dependence on cosmological the clusters, nor degradation of constraints due to the parameters Ω , w , w and γ when taking the deriva- imperfectlycalibratedmass-observablerelation(herethe M 0 a tives for the Fisher matrix. To complete the calcula- observableis the Sunyaev-Zel’dovichflux). Ourprevious tionofthefullnonlinearpowerspectrumweusethehalo tests have shown that these two effects roughly cancel model fitting formulae of [17]. out in the final cosmological constraints [22]. Then the For the cosmological probes, we assume future weak cluster Fisher matrix is lensing (WL) and Type Ia supernova (SN) data as pro- 1 ∂N(z )∂N(z ) vided by the SNAP experiment [18] as well as cosmic Ficjlus = N(z ) ∂p k ∂p k (13) microwavebackgroundanisotropy (CMB) data provided Xk k i j by the upcoming Planck satellite [19]. where N(z ) is number of clusters in kth redshift bin. k In this work, for weak lensing we only consider the The SN surveyprovides a luminosity distance-redshift two point correlation function. The weak lensing shear test,with2800SNedistributedinredshiftouttoz =1.7 power spectrum measures cosmology through both the asgivenby[18],andcombinedwith300localsupernovae mass power spectrum and distance factors, uniformly distributed in the z = 0.03−0.08 range. We add systematic errors in quadrature with intrinsic ran- ∞ dz ℓ Pκ(ℓ)= W (z)W (z)∆2 ,z , dom Gaussian errors of 0.15 mag per SNe. The system- ij Z r(z)2H(z) i j (cid:18)r(z) (cid:19) aticerrorscreateaneffectiveerrorfloorof0.02(1+z )/2.7 0 i (9) mag per bin of ∆z =0.1 centered at redshift z . i where r(z) is the comoving distance and H(z) is the For the CMB we use the full Fisher matrix predicted Hubble parameter. The weights Wi are given by for the Planck experiment with polarization informa- Wi(χ)∞= (3/2)ΩMH02fi(χ)(1 + z) where fi(χ) = tion (W. Hu, private communication). Note that most, r(χ) χ dχsni(χs)r(χs − χ)/r(χs), χ is the radial co- though not all, information about dark energy is cap- ordinRate and n is the comoving density of galaxies if χ tured in the distance to the last scattering surface from i s fallsinthedistancerangeboundedbytheithredshiftbin theacousticpeaksofthepowerspectrum(e.g.[23]). The and zero otherwise. We employ the redshift distribution effective precision of the angular diameter distance to of galaxies of the form n(z)∝ z2exp(−z/z ) that peaks z =1089 from Planck is 0.4% with temperature and po- 0 at 2z =1.0. The observed convergencepower spectrum larization information [24]. 0 4 IV. FITTING GRAVITY The combinationof probes of the expansionhistoryof 0.8 the universe and the growth history of large scale struc- 0.6 allows for ture tests the nature of the acceleration physics. The modified gravity expansion history is described by the effective equation 0.4 of state parameters w and w , and the deviation of the 0 a 0.2 growth history from that given by Einstein gravity un- der that expansion is measured by the growth index γ. wa 0 Almost all cosmological analyses to date, however, have assumed Einstein gravity or worked within a specific al- -0.2 ternate theory of gravity,rather than fitting for gravity. -0.4 Ignoringthe possibility ofmodified gravitycreatesthe neglects modified risk of the biasing our cosmological conclusions. This -0.6 gravity having ∆γ=0.1 holds not only for “gravitational”parameters but all in- -0.8 formation. Neglectingpossiblemodificationisequivalent -1.2 -1.15 -1.1 -1.05 -1 -0.95 -0.9 -0.85 -0.8 w to fixing the gravitationalgrowthindex γ to its Einstein 0 value(e.g.γ =0.55forgeneralrelativityandacosmolog- icalconstantmodel); howeverthiswillbiasthe otherpa- FIG. 1: 68% and 95% CL constraints on the expansion his- rameters due to their covariances with γ (this “gravity’s tory equation of state parameters w0 and wa, marginalized bias” was illustrated for the linear growth factor alone, overallotherparameters. Thetwoblue(filled)contoursgive rather than the weak lensing shear power spectrum, in constraintsfromcombiningweaklensing,supernovae,cluster, Fig. 5 of [25]). and CMB data, while simultaneously fitting for beyond Ein- Suppose the true value of the growth index differs by stein gravity. The black dot shows the fiducial model. The ∆γ fromits assumedgeneralrelativityvalue. Thisprop- two red (empty) contours show the biased constraints if the agates through into the weak lensing shear cross power modified gravity growth rate, here with ∆γ =0.1, is ignored spectrumCκ(ℓ)atmultipoleℓforthepairofredshiftbins (i.e. fixing γ to thegeneral relativity value). α α≡{i,j},changing it from the assumed generalrelativ- ity value of C¯κ(ℓ). Using the Fisher matrix formalism, α the bias on any of the P cosmologicalparameters is must attempt to fit for beyond Einstein deviations. Including the gravitational growth index as an addi- δp = F˜−1 Cκ(ℓ)−C¯κ(ℓ) tional parameter, and marginalizing over it to estimate i ij α α Xℓ (cid:2) (cid:3) the effective equation of state parameters describing the ∂C¯κ(ℓ) expansion,removesthe biasbutnecessarilydegradespa- ×Cov−1 C¯κ(ℓ),C¯κ(ℓ) β (14) rameter determination. (This would of course become α β ∂p (cid:2) (cid:3) j more severe if we required more growth variables than ≈ (∆γ)F˜−1F (15) just γ.) ij jg Thedegradationonthe weaklensingshearconstraints where the last line follows if the finite difference is re- from marginalizing over γ causes a factor ∼ 2 increase placed by a derivative. Here F˜ is the (P −1)×(P −1) in the contour area. While adding CMB or supernovae Fisher matrix that specifically does not include the data does not directly constrain the growth, they prove growthindexγ,F isthefullP×P Fishermatrix,summa- valuable in breaking degeneracies between parameters. tionsoverj andtheredshiftbinindicesα,β areimplied, AddingCMBtoWLimprovesconstraintsby30-35%,but and g is the index corresponding to the γ parameter in still suffers the factor of 2 weakeningrelative to fixing γ. the matrix F. The inclusion of SN is potent in reducing uncertainties. Biasinthecosmologyfromneglectingthepossibilityof WL+SN+CMB allows for fitting modified gravity, im- modified gravity can be significant. The red (open) con- proving parameter estimation by 5-7 times and the area toursinFig.1demonstratethatassumingEinsteingrav- constraint by 40 times over WL alone. Conversely, it ity in a universe with γ actually higher by 0.1 can shift only dilutes estimation of w0 by 23%, wa by 14%, and the expansion characteristics (effective equation of state the contour area by 35% relative to the “fixed to Ein- parameters) by ∼ 2σ. Recall that the DGP braneworld stein” case. This seems a fairly modest price to pay for model has ∆γ = 0.13 with respect to the standard cos- extending the physics reach to beyond Einstein gravity. mologicalconstantcase. Notetoothatwithinthissimple The marginalized uncertainties on the key parameters treatment of modified gravity a shift in γ moves the w - describing the nature of the effective dark physics are 0 w contour along the degeneracy direction, else the bias shown in Table I. Since SN and CMB do not probe a wouldbe evenlarger. Given the price ofclosing oureyes growth, we start with WL measurements and add other to the issue of possible gravitational modifications, we probes in sequence. 5 TABLE I: Fiducial constraints on the gravitational growth index γ (last column), as well as the two expansion history parametersw0andwa. Startingwiththeweaklensingsurvey, we then add the supernova, CMB, and cluster information consecutively. Note that both WL and cluster surveys are partially sensitive to nonlinear physics; constraints that only rely on the linear regime are discussed in §V. Probe σ(w0) σ(wa) σ(γ) Weak lensing 0.33 1.16 0.23 + SNeIa 0.06 0.28 0.10 + Planck 0.06 0.21 0.044 + Clusters 0.05 0.16 0.037 Whilemarginalizingoverthegrowthindexallowsusto account for modified gravity rather than ignoring it, we also wantto measurethe modificationitself. That is, we want to extract information quantifying the deviation, to guide us towardany “darkphysics”, not just say that there is some inconsistency with general relativity. Fig- ure 2 show the constraints on the growthindex from the cosmological probes, in the γ-Ω plane (marginalizing M FIG.2: 68%CLconstraintsonthegravitationalgrowthindex over the equation of state and other parameters). γ and the matter density ΩM, marginalizing over the effec- Weseeanenormousdifferenceintheleverageonmod- tive equation of state (and all other parameters). To fit for ified gravity as we combine probes, again due to the beyond Einstein gravity as well as the expansion history re- breakingofdegeneracies. Thedeterminationofγ reaches quiresacombinationofprobes. Othereffectsongrowth,such 0.044 for the combination WL+SN+CMB, representing asneutrinomasses, should betakenintoaccount aswell; the inner contour of each pair shows the effect of holding this 8% precision with respect to the Einstein value. If we fixed,most severe for weak lensing in isolation. do not include supernova data, then the uncertainty in- creases by a factor 2 (5 with only WL), and furthermore the overall contour area in the γ-Ω plane increases by M 11 times (44 for WL only). So complementaritybetween per ∆z = 0.1 as a function of redshift for our fiducial WL and SN is quite important for answering the ques- survey (motivated by the South Pole Telescope), for two tion of whether we are facing dark physics or physical values of the growth index (γ = 0.55 and 0.65). Since darkness. we normalized the power spectrum at high redshift, the Complementarityofprobesisimportantforseparation number density of clusters is independent of the growth of different physical effects on growth. While inclusion index at high redshift. As z → 0, the volume element, of neutrino mass, which can suppress growth (as does which is independent of the growth index by definition, increasing γ), broadens the uncertainty area of WL by dominatesoverthenumberdensityandmakesthecounts 74%, adding other methods immunizes against such a go to zero for either model. Therefore the biggest differ- “theory systematic”, with the effect on WL+SN+CMB ence between the two models is around the peak of the limited to 10%. redshift distribution at z ∼ 0.6, as illustrated in the left panelofFig.3. Alsoshownarenumbercountpredictions when the equation of state and neutrino mass have been V. STRUCTURE IN THE NONLINEAR perturbed from their fiducial values by 0.05 and 0.3 eV REGIME respectively. While the strong degeneracy of the growth index with other cosmological parameters is apparent, A. Beyond the linear regime it is worth noting that clusters can in principle provide several other observables to break this degeneracy, chief The gravitational growth index modifies the linear among them being the mass information. However, in growth factor, which then propagates into the full non- the nonlinear regime we are at increased risk of the sim- linear growthof structure. Here we examine some issues pleMMGmodelbreakingdown,possiblyrequiringmodel relatedtothe nonlinearregime. Inparticular,clustersof dependent simulations for a specific theory of gravity. galaxies involve nonlinear growth and we might wonder Weak lensing also involves scales in the quasi-linear whether special sensitivity to modified gravity arises in and nonlinear regimes. The sensitivity of a weak lens- this regime. TheleftpanelofFig.3showsclustercounts ing survey is shown in the right panel of Fig. 3. For 6 4000 44 γ = 0.55 γ=0.55 -4 γ = 0.65 10 3000 ∆γ=0.1 ) 33 ∆dz) z2000 ∆∆wm=ν=00.0.35 eV κπ / (2Pii 22 N/ 1) 10-5 d + ( (l l 1000 {i,i}=11 0 10-6 0 0.5 1 1.5 2 100 1000 10000 z l FIG. 3: Left panel: Cluster counts per ∆z =0.1 as a function of redshift for our fiducial survey, for two values of the growth index(γ =0.55 and0.65). Alsoshownarecaseswhentheequationofstateandneutrinomasshavebeenperturbedfrom their fiducialvaluesby0.05 and 0.3 eVrespectively. Rightpanel: Auto-correlation powerspectra of4-binweak lensing tomography for the two values of thegrowth index,with statistical errors shown for the γ =0.55 model. simplicity,weconsiderthe samefiducialsurveyasbefore however, very directly depends on the metric potentials but with 4-bin (instead of 10-bin) tomography, with di- which themselves are likely to be altered in the modi- visions z = [0,0.5],[0.5,1],[1,1.5],[1.5,3]. We show the fied gravity theory). By including nonlinear scales we four auto-correlation power spectra with corresponding increasethestatisticaldiscriminationpowerwithrespect statistical errors for γ =0.55, and the same but without togrowthbutmaybiasthe results asaresultofemploy- errors for γ = 0.65. The raw signal-to-noise for distin- ing an improper nonlinear prescription for the modified guishingthetwovaluesofthegrowthindexincreaseswith gravity growth. We now consider the trade off between redshift,givinganadvantagetoadeepersurvey. Thean- these two trends. gularscaleatwhichthetwovaluesarebestdistinguished Since weak lensing probes a range of scales, we can decreaseswith redshift, so the multipole increases,going consider limiting the weak lensing information to scales fromℓ∼500forthe firstbintoℓ∼5000forthe lastbin, k ≤ k and investigate how the information on the cut making advantageous a higher resolution survey. growth index is degraded. This “k-cutting” can remove Wehavealsocheckedthatthesensitivitytothegrowth thosephysicalscaleswherewelackdependableestimates index increases as the density of resolvedsource galaxies ofmodifiedgravityeffects. Themoststraightforwardway ng and their mean distance, parametrized by the mean to implement the cutting is to use weak lensing power of the redshift distribution, are increased. All these fac- spectrum nulling tomography (see §5 of [26]) where, for tors indicate that a space-basedweak lensing survey has a given multipole ℓ, only the lens planes with distances certain definite advantages for testing beyond Einstein r(z) > ℓ/k are allowed to contribute information. cut gravity. At this time, however, we have not further pur- Higher ℓ then contribute increasingly less information, suedthesesensitivitytestsnortriedtodeviseanoptimal and exhaust all information below k . For a fixed k , cut cut strategy to determine the growth index. The reason is we compute the ratio of the error in the growth index that by far the dominant uncertainty is a theory uncer- relative to the error with k = ∞. The results are cut tainty — our ability to predict the nonlinear clustering shownwiththe solidline inFig.4. Restricting the infor- statistics and associated observables — in a given modi- mation to purely linear scales (k <0.2hMpc−1) leads cut fied gravity theory. to degradations in the marginalized error in γ of more than a factor of 10 relative to the full nonlinear fidu- cialcase. However,when the quasi-linearscalesare used B. Uncertainty and bias (k < 1hMpc−1), the degradation is kept to a factor cut of a few, and the resulting constraints on the growth in- So far we have included nonlinear scales when obtain- dex are still interesting. Therefore, even if the nonlinear ing constraints from measurements involving the growth prediction out to k ≈ 10hMpc−1 (ℓ ≈ 10000) in modi- of structure. Indeed, none of the cosmological growth fied gravity theories is unfeasible, efforts to understand probes is solely a linear theory probe, with the possi- predictions on quasi-linear scales are well worthwhile. bleexceptionoftheIntegratedSachs-Wolfeeffect(which, Since we desire to retain values of k beyond the lin- 7 30 k 3 error bias δ(∆2(k,z))=c(cid:18)k∗(z)(cid:19) (17) d σγ, fi 10 c=0.05 where c is a dimensionless constant which represents the ) / fractional error in the 1-halo term. We then use the γn Fisher formalism to propagate this offset into biases on s i 3 cosmologicalparameters. (The expansionparametersw0 bia c=0.01 and wa are not appreciably affected by nulling, so we r focus on γ.) o or 1 The less of the nonlinear regime we use, the less ef- r fect the misestimated 1-halo,or c, term has. The results r e ( are shown in Figure 4 where the dashed curves give the biasinthegrowthindexdividedbythefiducialstatistical 0.3 error (i.e. the statistical error for k = ∞). It is clear 0.1 1 10 100 cut k (h / Mpc) that,asweperformincreasinglymoredrasticnulling,de- cut creasing the value of k , the bias/error ratio decreases cut significantly. For c = 0.05 (0.01), cutting information FIG. 4: The solid (black) line shows the degradation in con- beyond k = 1 (6) hMpc−1 gives bias in γ that is be- straintson thegrowth index γ from aSNAP-typeweak lens- cut low the statistical error, while increasing the error by a ing survey (plus SN+CMB) as small-scale weak lensing in- factor of three (25%). One might expect the bias model formation is increasingly dropped. For each valueof kcut, we drop all information from k > kcut following [26] and com- of Eq. (17) to be cut off in the stable clustering regime, putetheratiooftheconstraintonγ relativetothecasewhen causing bias in γ to level off or decline at k greater than all information is used, that is, when kcut =∞. The dashed afewtimesknl,e.g.k >1hMpc−1. Thiswouldallowuse curves (dark (red) and light (orange)) lines show the bias in of more of the nonlinear regime. thegrowthindexdividedbythekcut =∞statisticalerrorfor the bias model given in Eq. (17) with c= 0.05 or 0.01. The optimum choice of kcut, with the least risk, occurs near the C. Scale dependent growth intersection of theerror and bias curves. MMG, withthe gravitationalgrowthindex formalism, has been adopted as the simplest reasonable method of earregime,fortheir leverage,it is instructiveto make at accounting for the effects of beyond Einstein gravity on least an approximate estimate of the bias that might be cosmological probes involving the growth of structure. induced by keeping such data. Within the halo model, One part of its simplicity is that γ acts in a scale inde- the linear regime corresponds to the 2-halo term of clus- pendentfashion;thisshouldreproduceglobaleffectssuch tering between dark matter halos, while the nonlinear as time varying gravitational coupling. However many regimeinvolvesthe1-halotermoftheprofileandconcen- modifications to gravity will introduce scale dependence trationwithinahalo. Fork≈0.2−1hMpc−1,the1-halo in the growth. For example, in the DGP braneworld we contributionmaybeapproximatedbyawhitenoiseterm mightexpectchangestothegrowthequationbeyondthe in the power spectrum [27], due to Poisson fluctuations varying coupling on both small and large scales, due to in the number of halos. In other words, the Vainshtein radius (where the scalar degrees of free- dombecomerelevant)andrelativisticeffectsrespectively (R. Scoccimarro, private communication). 3 k Scale dependence on small scales affects the nonlin- ∆2(k,z)≃∆2 (k,z)+ . (16) lin (cid:18)k∗(z)(cid:19) ear regime, and this is just what we looked at with the bias calculations above. On large scales, we may treat We find a good fit to the full power spectrum in the the modification of the source term mode by mode in quasi-linear regime for k∗(z) = (5/3)knl(z), where the the linear growth equation. From a harmonic analysis nonlinear scale k (z) is defined via ∆2(k (z),z)=1. of metric perturbations, [28, 29] found long wavelength nl nl Since neither the presence of a 1-halo/2-halo split nor corrections to the Poisson equation; recall from §II this the Poisson fluctuations in number should rely on the altersthefactorQinthesourcetermQδ. SuchJLWcor- specific gravity theory, we adopt this approach and con- rectionswouldmultiply Ω (a)by Q∼1+αe−(k/kJLW)2. M sider what happens if we improperly estimate the 1-halo We can then either solve the modified differential equa- effects. This wouldshift the white noiseterm to a differ- tion for the linear growth, or we can retain the growth ent value (equivalent to changing the scale at which the index approach,but similarly multiply Ω (a) in Eq. (7) M 2-haloand1-haloterms are comparable),i.e.biasing the by that same factor (this can also be viewed as making power spectrum by the growth index γ a function of k). The characteristic 8 scale is the horizon scale, k ∼H, and the amplitude morestatisticalpower,beingmoresuitabletonear-future JLW α ∼ O(1). We find that effects on the growth are then data, while we think the loss in generality is minimal. negligible for k >10−3hMpc−1. We also emphasize the advantage in retaining the maxi- Thus horizonsizescale dependence, asshouldholdfor malphysicsinformationbytreatingtheexpansioneffects theDGPcaseandscalar-tensortheoriesaswell,willhave on growththrough the effective equation of state, giving essentially no effect on the weak lensing probe, as the clear separation from deviations in the gravity theory. WL power spectrum errors on the near-horizon scales HowaccuratelysuchaprogramcanrevealbeyondEin- are very large due to sample variance; see e.g. Fig. (10) steingravity–darkphysicsvs.physicaldarkness–isthe in [30]. Weak lensing can be reasonably treated by the main topic of this paper. We have shown that measure- scaleindependentγ factorover10−3 <∼k/hMpc−1 <∼10. ments from weak gravitational lensing, Type Ia super- Such scale dependence may however need to be taken novae, and the CMB combined can measure the growth into accountfor attempts to use the ISW effect to probe index to about 8%, or to ±0.04 around its ΛCDM value cosmology, as mentioned previously. Only if the gravi- (and galaxy cluster data could potentially reduce this tation theory possesses a substantially smaller scale, ap- evenfurther). Atthesametime,theconstraintsonother proachingthenonlinearscale,asinthephenomenological cosmological parameters are not appreciably degraded, alterationsofthePoissonequationin[3–6],areweforced essentially because the surveys probe a range of scales to more elaborate parameterizations than γ. and thus their complementarity breaks the degeneracies Thespecificoptimumofthetradeoffbetweenleverage between parameters. on cosmological parameter constraints from the added One particular concern in the program of distinguish- information of smaller scales and increasing risk of bias inggeneralrelativityfrommodifiedgravityisthenonlin- from gaps in our understanding will depend on the spe- ear density regime of structure formation. Even for the cific gravity theory. Since this is what we are trying to limitednumber ofwell-definedmodifiedgravitytheories, obtain insight into, a rational, model independent ap- details of nonlinear clustering are currently unknown. proach might be to carry out both a wide area survey While in principle the nonlinear structure formation is to squeeze the most statistical power out of the rela- calculable from N-body simulations of modified gravity, tively weakly discriminating low ℓ (linear) regime and creating these simulations in practice is extremely diffi- a deep, high resolution and high number density survey cultexceptforsomeverysimplecases. Thestructureand to probe the richer high ℓ (quasi- and nonlinear)regime. evolutionofgalaxyclusters,whicharenonlinearobjects, This indicates possible strong complementarity between isfairlystronglydependentonthenonlinearphysics,and a ground-based weak lensing survey such as LSST [31] is consequently problematic. and a space-based survey like SNAP. Weak lensing, on the other hand, probes a range of scales,and we studied how our results behave if we drop small-scale (that is, nonlinear) information. Using the VI. DISCUSSION nulling tomography approachand a reasonably well mo- tivatedtoymodelforbiasduetouncertaintyinnonlinear This article presents an approachfor simultaneous fit- structure,wefoundthatcuttingoutthesmallscaleinfor- ting of the cosmic expansion and the theory of gravity. mation (k >∼ 1hMpc−1) can lead to significant decrease We have advocated a “minimalist” strategy of distin- in the resulting bias in the growth index, at the expense guishing modified gravity from dark energy, which con- of increasing the statistical error in it by a factor of a sists in measuring a single parameter, the growth index few. γ. In addition to reproducing the linear growthfunction Other cosmological observables exist with sensitivity for essentially all standard gravity, dark energy models to the growth of fluctuations, and hence can be used to (parametrized with w and w ), the growth index also constrain MMG, but we have not discussed them in any 0 a fits the linear growth of a single known modified grav- detail. Forexample,thebispectrumofweakgravitational ity theory, the DGP braneworld scenario. Therefore, it lensing is a potentially powerful probe [32]; however, it is reasonable to expect that the growth index can be hasprovedtobeatoughtasktocalibratethebispectrum used to measure deviationsfrom standardgravity: given even in standard general relativity. The same holds for themeasurementsofthebackgroundexpansionrate(pa- Lyman-alpha forest observations. The Integrated Sachs- rameterized, say, by Ω , w and w ), standard gravity Wolfe effect is a potentially strong discriminant of mod- M 0 a predicts the value of γ, and a statistically significant de- ified gravity models (see in particular recent predictions viationfrom this value can in principle be interpretedas oftheISWinDGPmodels[14]);however,themetricpo- evidence for – and characterizationof – beyond Einstein tentials are particularly sensitive to the structure of the gravity. modified gravity theory. Neither we nor anyone else has This is a first step, hence our emphasis in calling it yetsucceededinfinding a genericparametrizationofthe Minimal Modified Gravity. One could examine more deviations from general relativity for the ISW effect. complex schemes but these may be more model depen- Ourworkoutlinesafirststepintreatingmodifiedgrav- dent or employ more parameters and are therefore likely ity models. At the time of this writing, there is hardly togiveweakerresults. TheMMGapproachthereforehas a single well-defined modified gravity theory that does 9 not look like standard general relativity in terms of ob- By continuing forward with advances in measurements, servables, but is not already ruled out or disfavored by theory, and computation we can lift the darkness on the data. Considerable effort is underway to construct such new physics. theories (e.g. [33–47]; for a review see [48–50]) and test them experimentally. Itishearteningthatthestrongcomplementarityincos- Acknowledgments mologicalprobessuchasthecombinationofweaklensing, supernovae, and the CMB does provide important infor- mationonthequestionofdarkphysicsvs.physicaldark- This work has been supported in part by the Di- ness. Such data from next generation experiments, here rector, Office of Science, Department of Energy under calculatedspecificallyforSNAPandPlanck,doesfurnish grantDE-AC02-05CH11231andbyNSFAstronomyand arealtestnotjustofindividualmodelsbutofthephysics Astrophysics Postdoctoral Fellowship under Grant No. framework beyond Einstein. We will be able to measure 0401066. We gratefully acknowledge discussions with the effective equation of state describing the cosmic ex- Wayne Hu (who supplied the full CMB Fisher matrix) pansion and simultaneously reveal the theory of gravity. and Martin White. [1] P. Brax, C. van de Bruck, A-C. Davis, J. Khoury, [24] D. Eisenstein, W. Hu, and M. Tegmark, Astrophys. J. A. Weltman, Phys. Rev. D 70, 123518 (2004) [astro- 518, 2 (1999) [astro-ph/9807130] ph/0408415] [25] E.V.Linder,Astropart.Phys.(2006) [astro-ph/0604280] [2] E.V. Linder, Phys. Rev. D 72, 043529 (2005) [astro- [26] D. Huterer and M. White, Phys. Rev. D, 72, 043002 ph/0507263] (2005) [astro-ph/0501451] [3] M. White and C.S. Kochanek, ApJ 560, 539 (2001) [27] A.E.Schulz&M.White,Astropart.Phys.25,172(2006) [astro-ph/0105227] [astro-ph/0510100] [4] C. Sealfon, L. Verde and R. Jimenez, Phys. Rev. D, 71, [28] M.W.Jacobs, E.V.Linder,R.V.Wagoner,Phys.Rev.D 083004 (2005) [astro-ph/0506707] 45, R3292 (1992) [5] A. Shirata, T. Shiromizu, N. Yoshida, Y. Suto, Phys. [29] M.W.Jacobs, E.V.Linder,R.V.Wagoner,Phys.Rev.D Rev.D 71, 064030 (2005) [astro-ph/0501366] 48, 4623 (1993) [6] H.StabenauandB.Jain,Phys.Rev.D74,084007(2006) [30] D. Huterer, Phys. Rev. D, 65, 063001 (2002) [astro- [astro-ph/0604038] ph/0106399] [7] E. Bertschinger, ApJ, 648, 797 (2006) [astro- [31] LSST – http://www.lsst.org ph/0604485] [32] M.Takada.andB.Jain,MNRAS,348,897(2004)[astro- [8] L. Knox, Y.-S. Song, J.A. Tyson, Phys. Rev. D 74, ph/0310125] 023512 (2006) [astro-ph/0503644] [33] A.Lue,R.ScoccimarroandG.Starkman,Phys.Rev.D, [9] M. Ishak, A. Upadhyeand D. Spergel, Phys. Rev. D 74, 69, 044005 (2004) [astro-ph/0307034] 043513 (2006) [astro-ph/0507184] [34] S.M.Carrolletal.,Phys.Rev.D70,043528(2004)[astro- [10] G. Dvali,G. Gabadadze, M. Porrati, Phys.Lett. B 485, ph/0306438]; Phys. Rev. D 71, 063513 (2005) [astro- 208 (2000) [hep-th/0005016] ph/0410031] [11] A.Lue,R.ScoccimarroandG.Starkman,Phys.Rev.D, [35] T.Chiba,Phys.Lett.B575,1(2003)[astro-ph/0307338] 69, 124015 (2004) [astro-ph/0401515] [36] O.Mena,J.SantiagoandJ.Weller,Phys.Rev.Lett.96, [12] K. Koyama and R. Maartens, JCAP, 0601, 016, (2006) 041103 (2006) [astro-ph/0510453] [astro-ph/0511634] [37] S. Nojiri and S.D. Odintsov, Gen. Rel. Grav. 36, 1765 [13] I.Sawicki, Y.-S.Song and W. Hu,astro-ph/0606285 (2004)[hep-th/0308176];Phys.Rev.D68,123512(2003) [14] Y.-S.Song, I.Sawicki and W. Hu,astro-ph/0606286 [hep-th/0307288] [15] C.Baccigalupi,S.Garbari,E.V.Linder,S.Matarrese,F. [38] I. Navarro and K. Van Acoleyen, Phys. Lett. B 622, Perrotta, in draft 1 (2005) [gr-qc/0506096]; JCAP 0603, 008 (2006) [gr- [16] D.J. Eisenstein, and W. Hu, ApJ, 511, 5 (1999) [astro- qc/0511045] ph/9710252] [39] G. Gabadadze and A. Iglesias, Phys. Rev. D 72, 084024 [17] R.E. Smith et al., MNRAS, 341, 1311 (2003) [astro- (2005) [hep-th/0407049] ph/0207664] [40] Y.-S. Song, Phys. Rev. D 71, 024026 (2005) [astro- [18] SNAP– http://snap.lbl.gov ; ph/0407489]; astro-ph/0602598 G. Aldering et al. 2004, astro-ph/0405232 [41] P. Zhang, Phys. Rev. D, 73, 123504 (2006) [astro- [19] Planck Surveyor– http://planck.esa.int ph/0512422] [20] W. Hu, Astrophys. J., 522, L21 (1999) [astro- [42] L. Amendola, D. Polarski and S. Tsujikawa, astro- ph/9904153] ph/0603703 [21] South Pole Telescope – http://astro.uchicago.edu/spt [43] A.deFelice,M.HindmarshandM.Trodden,JCAP0608, [22] D.Huterer, A.Kim, L.M. Krauss and T. Broderick, As- 005 (2006) [astro-ph/0604154] trophys.J. 615, 595 (2004) [astro-ph/0402002] [44] S.M. Carroll, I.Sawicki, A.Silvestri, M.Trodden, astro- [23] J.Frieman,D.Huterer,E.LinderandM.S.Turner,Phys. ph/0607458 Rev.D,67, 083505 (2003) [astro-ph/0208100] [45] S. Nojiri, S.D. Odintsov, M. Sasaki, Phys. Rev. D 71, 10 123509 (2005) [hep-th/0504052] [48] A. Lue, Phys.Rept., 423, 1 (2006) [astro-ph/0510068] [46] T. Koivisto and D.F. Mota, astro-ph/0606078 [49] J.-P. Uzan,astro-ph/0605313 [47] G. Calcagni, B. de Carlos, A. De Felice, Nucl. Phys. B [50] M. Trodden,astro-ph/0607510 752, 404 (2006) [hep-th/0604201]

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.