Dark Energy and Dark Matter Interaction f (R) A Worked Example Wayne Hu Florence, February 2009
Why Study f(r)? Cosmic acceleration, like the cosmological constant, can either be viewed as arising from Missing, or dark energy, with w p/ ρ < 1/3 Modification of gravity on large scales G µν = 8πG ( ) Tµν M + Tµν DE F (g µν ) + G µν = 8πGT M µν Proof of principle models for both exist: quintessence, k-essence; DGP braneworld acceleration, f(r) modified action Compelling models for either explanation lacking Study models as illustrative toy models whose features can be generalized
Parameterized Post-Friedmann Description Smooth dark energy parameterized description on small scales: w(z) that completely defines expansion history, sound speed defines structure formation Parameterized description of modified gravity acceleration Many ad-hoc attempts violate energy-momentum conservation, Bianchi identities, gauge invariance; others incomplete Parameterize the degrees of freedom in the effective dark energy F (g µν ) = 8πGT DE µν retaining the metric structure of general relativity but with component non-minimally dependent on metric, coupled to matter Non-linear mechanism returns general relativity on small scales
Parameterized Post-Friedmann Description Smooth dark energy parameterized description on small scales: w(z) that completely defines expansion history, sound speed defines structure formation Parameterized description of modified gravity acceleration Many ad-hoc attempts violate energy-momentum conservation, Bianchi identities, gauge invariance; others incomplete Parameterize the degrees of freedom in the effective dark energy F (g µν ) = 8πGT DE µν retaining the metric structure of general relativity but with component non-minimally dependent on metric, coupled to matter Non-linear mechanism returns general relativity on small scales
Three Regimes Three regimes defined by γ= Φ/Ψ BUT with different dynamics Examples f(r) and DGP braneworld acceleration Parameterized Post-Friedmann description Non-linear regime follows a halo paradigm but a full parameterization still lacking and theoretical, examples few: f(r) now fully worked General Relativistic Non-Linear Regime Scalar-Tensor Regime Conserved-Curvature Regime r r * c halos, galaxy large scale structure CMB r
Outline f(r) Basics and Background Linear Theory Predictions N-body Simulations and the Chameleon Collaborators: Marcos Lima Hiro Oyaizu Michael Mortonson Hiranya Peiris Iggy Sawicki Fabian Schmidt Yong-Seon Song
f (R) Basics
Cast of f(r) Characters R: Ricci scalar or curvature f(r): modified action (Starobinsky 1980; Carroll et al 2004) S = d 4 x [ ] R + f(r) g 16πG + L m
Cast of f(r) Characters R: Ricci scalar or curvature f(r): modified action (Starobinsky 1980; Carroll et al 2004) S = d 4 x [ ] R + f(r) g 16πG + L m f R df/dr: additional propagating scalar degree of freedom (metric variation) f RR d 2 f/dr 2 : Compton wavelength of f R squared, inverse mass squared B: Compton wavelength of f R squared in units of the Hubble length B f RR 1 + f R R H H d/d ln a: scale factor as time coordinate
Modified Einstein Equation In the Jordan frame, gravity becomes 4th order but matter remains minimally coupled and separately conserved G αβ + f R R αβ ( ) f 2 f R g αβ α β f R = 8πGT αβ Trace can be interpreted as a scalar field equation for f R with a density-dependent effective potential (p = 0) 3 f R + f R R 2f = R 8πGρ
Modified Einstein Equation In the Jordan frame, gravity becomes 4th order but matter remains minimally coupled and separately conserved G αβ + f R R αβ ( ) f 2 f R g αβ α β f R = 8πGT αβ Trace can be interpreted as a scalar field equation for f R with a density-dependent effective potential (p = 0) 3 f R + f R R 2f = R 8πGρ For small deviations, f R 1 and f/r 1, f R 1 (R 8πGρ) 3 the field is sourced by the deviation from GR relation between curvature and density and has a mass m 2 f R 1 R = 1 3 f R 3f RR
Interacting Dark Sector Conformal transformation to Einstein frame g E µν = exp( ψ)g µν S = d 4 x g E { 1 16πG [ R E 3 ] 2 ( Eψ) 2 V (ψ) } + exp(2ψ)l m [gµν E exp(ψ)] Gravity normal, but matter and scalar field ψ coupled For f R 1, f R = ψ and two frames coincide to lowest order Calculate in Jordan frame so that matter falls on geodesics of the metric Given metric usual kinematical tools apply: CAMB, PM N-body codes etc. For dark matter only simulations, represents both universal coupling and dark matter only coupling
f (R) Linear Theory
Three Regimes Three regimes defined by γ= Φ/Ψ BUT with different dynamics Examples f(r) and DGP braneworld acceleration Parameterized Post-Friedmann description Non-linear regime follows a halo paradigm but a full parameterization still lacking and theoretical, examples few: f(r) now fully worked General Relativistic Non-Linear Regime Scalar-Tensor Regime Conserved-Curvature Regime r r * c halos, galaxy large scale structure CMB r
PPF f(r) Description Metric and matter evolution well-matched by PPF description Standard GR tools apply (CAMB), self-consistent, gauge invar. 1.1 100 Φ /Φ i 1.0 0.9 k/h 0 =0.1 30 0.8 f(r) PPF Hu, Hu Huterer & Sawicki & Smith (2007); (2006) Hu (2008) 0.1 1 a
Galaxy-ISW Anti-Correlation Large Compton wavelength B 1/2 creates potential growth which can anti-correlate galaxies and the CMB In tension with detections of positive correlations across a range of redshifts B 0 =0 B 0 =5 Hu, Song, Huterer Peiris & Smith Hu (2007) (2006)
Linear Power Spectrum Linear real space power spectrum enhanced on scales below Compton scale in the background Scale-dependent growth rate and potentially large deviations on small scales P L (k) (Mpc/h) 3 10 4 10 3 B 0 1 0.1 0.01 0.001 0 (ΛCDM) 0.001 0.01 0.1 k (h/mpc)
Redshift Space Distortion Relationship between velocity and density field given by continuity with modified growth rate (f v = dlnd/dlna) Redshift space power spectrum further distorted by Kaiser effect 0.55 B 0 =1 f v 0.5 0.1 0.45 0.01 0.001 0 (ΛCDM) 0.001 0.01 0.1 k (h/mpc)
f (R) Non-Linear Evolution
Three Regimes Three regimes defined by γ= Φ/Ψ BUT with different dynamics Examples f(r) and DGP braneworld acceleration Parameterized Post-Friedmann description Non-linear regime follows a halo paradigm but a full parameterization still lacking and theoretical, examples few: f(r) now fully worked General Relativistic Non-Linear Regime Scalar-Tensor Regime Conserved-Curvature Regime r r * c halos, galaxy large scale structure CMB r
Non-Linear Chameleon For f(r) the field equation 2 f R 1 3 (δr(f R) 8πGδρ) is the non-linear equation that returns general relativity High curvature implies short Compton wavelength and suppressed deviations but requires a change in the field from the background value δr(f R ) Change in field is generated by density perturbations just like gravitational potential so that the chameleon appears only if f R 2 3 Φ, else required field gradients too large despite δr = 8πGδρ being the local minimum of effective potential
Non-Linear Dynamics Supplement that with the modified Poisson equation 2 Ψ = 16πG δρ 1 3 6 δr(f R) Matter evolution given metric unchanged: usual motion of matter in a gravitational potential Ψ Prescription for N-body code Particle Mesh (PM) for the Poisson equation Field equation is a non-linear Poisson equation: relaxation method for f R Initial conditions set to GR at high redshift
Environment Dependent Force Chameleon suppresses extra force (scalar field) in high density, deep potential regions Oyaizu, Hu, Lima, Huterer Hu &(2008) Smith (2006)
Environment Dependent Force For large background field, gradients in the scalar prevent the chameleon from appearing Oyaizu, Hu, Lima, Huterer Hu &(2008) Smith (2006)
N-body Power Spectrum 512 3 PM-relaxation code resolves the chameleon transition to GR: greatly reduced non-linear effect 0.8 0.6 P(k)/P GR (k)-1 0.4 0.2 0 Oyaizu, Hu, Lima, Huterer Hu &(2008) Smith (2006) 0.1 1 k (h/mpc)
N-body Power Spectrum Artificially turning off the chameleon mechanism restores much of enhancement 0.8 0.6 P(k)/P GR (k)-1 0.4 0.2 0 Oyaizu, Hu, Lima, Huterer Hu &(2008) Smith (2006) 0.1 1 k (h/mpc)
N-body Power Spectrum Models where the chameleon absent today (large field models) show residual effects from a high redshift chameleon 0.8 0.6 P(k)/P GR (k)-1 0.4 0.2 0 Oyaizu, Hu, Lima, Huterer Hu &(2008) Smith (2006) 0.1 1 k (h/mpc)
Distance Predicts Growth With smooth dark energy, distance predicts scale-invariant growth to a few percent - a falsifiable prediction Mortonson, Hu, Huterer (2008)
Scaling Relations Fitting functions based on normal gravity fail to capture chameleon and effect of extra forces on dark matter halos 0.6 P(k)/P GR (k)-1 0.4 0.2 0 Oyaizu, Hu, Lima, Huterer Hu &(2008) Smith (2006) 0.1 1 k (h/mpc)
Mass Function Enhanced abundance of rare dark matter halos (clusters) with extra force Rel. deviation dn/dlogm Schmidt, Hu, Lima, L, Huterer Oyaizu, & Smith Hu (2008) (2006) M 300 (h -1 M )
Halo Bias Halos at a fixed mass less rare and less highly biased Schmidt, Lima, Oyaizu, Hu (2008)
Halo Mass Correlation Enhanced forces vs lower bias Schmidt, Lima, Oyaizu, Hu (2008)
Halo Model Power spectrum trends also consistent with halos and modified collapse 1.2 1 f R0 =10-4 P(k)/P GR (k)-1 0.8 0.6 0.4 halo model + modified collapse 0.2 0 0.1 1 10 k (h/mpc) Schmidt, Hu, Lima, Huterer Oyaizu, & Smith Hu (2006) (2008)
Summary General lessons from f(r) example 3 regimes: large scales: conservation determined intermediate scales: scalar-tensor small scales: GR in high density regions, modified in low Given fixed expansion history f(r) has additional continuous parameter: Compton wavelength Enhanced gravitational forces below environment-dependent Compton scale affect growth of structure Enhancement hidden by non-linear chameleon mechanism at high curvature high density) N-body (PM-relaxation) simulations show potentially observable differences in the power spectrum and mass function