Dark Energy, Modified Gravity and The Accelerating Universe

Dragan Huterer
Kavli Institute for Cosmological Physics University of Chicago

Makeup of universe today

(stars 0.4%, gas 3.6%) 4%
(suspected since 1930s established since 1970s)

Baryonic Matter

(suspected since 1980s established since 1998)

Dark Energy

Dark Matter

22%

74%
Also: radiation (0.01%)

Some of the early history of the Universe is actually understood better!

Physics quite well understood

95% of contents only phenomenologically described

DE status ~8 years after discovery

High-Z SN Search Team Supernova Cosmology Project

m-M (mag)

Measurements much better, LCDM still a good fit

=0.3, =0.3, =1.0, =0.7 =0.0 =0.0

Strong indirect (non-SNa Ia) evidence for DE from CMB+LSS Physical mechanism responsible completely unknown

Vacuum energy dominated

(m-M) (mag)

Matter dominated
0.01 0.10 z 1.00

A lot of work on modified gravity proposals and observational signatures

Riess et al 1998; Perlmutter et al 1999

Current constraints

DE 0.7

What if gravity deviates from GR?

For example:
8G H F (H) = , 3
2

or

8G H = 3
2

3F (H) + 8G

Modified gravity

Dark energy

Modified gravity proposals

Introduce modifications to GR (typically

(can be hard)

near horizon scale) to explain the observed acceleration of the universe

Make sure Solar System tests are passed Constrain the MG theory using the
cosmological data (can be hard!)

Try to distinguish MG vs. standard DE

Example: f(R) gravity

1 S= 16G d x g [R + f (R)]
4

Einstein equations are now 4th order Two classes

fRR<0 (never Matter Dominated, long range forces) fRR>0 (MD in the past, can evade Solar system tests)

Carroll, Duvvuri, Trodden, Turner 2005; Mena, Santiago & Weller 2006; Navarro & van Acoleyen 2006; Song, Hu & Sawicki 2006; many others....

Example: DGP braneworld theory

1 extra dimension
brane

(bulk) in which only gravity propagates

matter lives on the weakening of gravity

at large distances = appearance of DE

Credit: Iggy Sawicki

Dvali, Gabadadze & Porrati 2000; Deffayet 2001

The structure of DGP

H 8G 2 H = rc 3
rc is a free parameter
(to be consistent with observation, rc ~ 1/H0)
2GM=rg Scalar-Tensor

5D GR

4D GR

r*

rc

New scale r =

2 1/3 rg rc

5D GR
Credit: Iggy Sawicki

Dvali, Gabadadze & Porrati 2000; Deffayet 2001

DGP linear growth

1

Growth relative to EdS

0.95

LCDM LCDM

0.9

0.85

g(a) y
0.8 0.75

DGP4D

dark energy
0.7

0.65 0

DE Mimicking DGP expansion

0.2 0.4

DGP DGP

0.6

0.8

Scale factor
Lue, Scoccimarro & Starkman; Koyama & Maartens; Sawicki, Song & Hu

a x

ISW in DGP
LCDM Phi DGP

Song, Sawicki, & Hu 2007

So DGP is (almost) ruled out

Disfavored at a few sigma from distances (SNe etc) Disfavored at a few more sigma from CMB ISW Decisive rule-out will come from ISW cross-correlation at high z:

50

100

50

100

50

100

50

100

Song, Sawicki, & Hu 2007

Dark Energy or Modied Gravity?

A given DE and modied gravity models may both t the expansion history data very well But they will predict different structure formation history, i.e. deviation from + 2H 4M = 0

 In standard GR, H(z) determines distances and growth of

structure

+ 2H 4M = 0

So check if this is true by measuring separately

Distances
(a.k.a. kinematic probes) (a.k.a. 0th order cosmology)

Growth
(a.k.a. dynamical probes) (a.k.a. 1st order cosmology)

Price of ignorance of MG
0.8 0.6 0.4 0.2

allows for modified gravity

wa

-0.2 -0.4 -0.6 -0.8 -1.2

-1.15 -1.1 -1.05

neglects modified gravity having !"=0.1

w0

-1

-0.95

-0.9

-0.85

-0.8

Huterer & Linder, astro-ph/0608681

Cosmological Probes of Dark Energy (and Modified Gravity)

Kinematic probes: SNe Ia

High-Z SN Search Team Supernova Cosmology Project

m-M (mag)

=0.3, =0.3, =1.0,

=0.7 =0.0 =0.0

Get pure (luminosity) distances

(m-M) (mag)

0.01

0.10 z

1.00

Kinematic probes: CMB and BAO

T = 2.726 K
T 105 T
Credit: WMAP team

2-pt correlation

Sound horizon

Distance
Bennett et al 2003 (WMAP collaboration)

to recom

bination

Structure formation probes: Galaxy cluster counts

d2 N r(z)2 = n(z) d dz H(z)
10
5

!M=1, !DE=0 w=-0.8

dN/dz (4000 deg )

10

10

!M=0.3, !DE=0.7, w=-1

10
Credit: Quinn, Barnes, Babul, Gibson

0.5

1.5

Essentially fully in the nonlinear regime (scales ~1 Mpc)

Structure formation probes: Weak Gravitational Lensing

Pshear
0

W (r)Pmatter (r)dr
10
-2

True, nonlinear
10
-4

l(l+1)Pl /(2")

10

-6

Linear theory
10
-8

10

-10

Credit: Colombi & Mellier

10

Multipole l

100

1000

10000

Mostly in the nonlinear regime (scales ~10 arcmin, or ~1 Mpc)

More general approach

Measure the DE parameters from distances and growth separately

0.726

OmegaDE

0.724 0.722 0.72 0.718 0.716 0.714 0.712 0.71 0.708 0.8 0.6 0.4 0.7 w_1 0.2 0 1.1 0.9 1 w_0 0.6

Omega_de

wa

w0

0.8

Ishak, Upadhye and Spergel 2006; others...

Still more general approach:

measure functions r(z) and g(z) see if they are consistent

Knox, Song & Tyson 2005

Minimalist Modified Gravity vs. DE

Describe deviations from GR via a single new parameter
g(a) = exp a
a

d ln a[M (a) 1]
0

Excellent fit to standard DE growth function with

= 0.55 + 0.05[1 + w(z = 1)]

Also fits the DGP braneworld theory with = 0.13
Huterer & Linder, astro-ph/0608681 see also Linder & Cahn, astro-ph/0701317

4000

3000

(dN/dz) !z

2000

"=0.55 !"=0.1 !w=0.05 !m#=0.3 eV

Cluster counts

1000

0.5

1.5

2
-4

10

# = 0.55 # = 0.65

44 33 22

Weak lensing tomography

l(l+1) P ii / (2")

10

-5

{i,i}=11
10
-6

100

1000

10000

Constraints on the growth index

sig(w0) WL +SNE +Planck +Clusters 0.33 0.06 0.06 0.05 sig(wa) 1.16 0.28 0.21 0.16
sig(gamma)

0.23 0.10 0.044 0.037

Recall, for DGP = 0.13

Huterer & Linder, astro-ph/0608681

Discarding the small-scale info in weak lensing

30

(error or bias in !) / "!, fid

Degradation factor in ! error

10

0.3 0.1

10

100

kcut (h / Mpc)
Using the Nulling Tomography of weak lensing (Huterer & White 2005)

South Pole Telescope

Planck

Supernova/Acceleration Probe

LSST

Conclusions

distinguishing dark energy from modied gravity is becoming one of the key goals of cosmology in years to come assuming nonlinear clustering that follows the usual prescription even with MG, we nd that future probes can achieve very interesting constraints on this parameter restriction to linear scales severely degrades the errors, but well worth pursuing ambitious, general approach: measure functions r(z) and g(z), check if they are consistent minimalistic approach: measure a single parameter that describes departures between DE and MG bright future with upcoming powerful surveys

Physically motivated MG parametrization

ds2 = a2 ( ) (1 + 2)d 2 + (1 2)dx2

= (1 +
80

) and assume

DE 0 M

[l(l+1)Cl/2!]

1/2

60

LCDM !0.1 !0.3 +0.1

(K)

40

20

10

100

1000

l: multipole moment

Caldwell, Cooray & Melchiorri, astro-ph/0703375

Physically motivated MG parametrization

Weak lensing power spectrum CMB-galaxy cross-correlation

10
!3

LCDM 0.1 !0.1 10 l(l+1)Cl /2! 10

!5 !4

10

!6

10

10

10 l

10

Caldwell, Cooray & Melchiorri, astro-ph/0703375