1511 05962 PDF
1511 05962 PDF
1511 05962 PDF
Department of Physics and Astronomy, University of British Columbia, Vancouver, BC, V6T 1Z1, Canada
2
Department of Physics, Simon Fraser University, Burnaby, BC, V5A 1S6, Canada
3
Institut de Physique Theorique, CEA, IPhT, CNRS, URA 2306, F-91191Gif/Yvette Cedex, France
4
DAMTP, Centre for Mathematical Sciences, University of Cambridge, Wilberforce Road, Cambridge CB3 0WA, U.K.
5
National Astronomy Observatories, Chinese Academy of Science,
Beijing, 100012, Peoples Republic of China and
6
Institute of Cosmology and Gravitation, University of Portsmouth, Portsmouth, PO1 3FX, United Kingdom
Modified gravity theories often contain a scalar field of gravitational strength which interacts with
matter. We examine constraints on the range and the coupling strength of a scalar gravitational
degree of freedom using a subset of current data that can be safely analyzed within the linear
perturbation theory. Using a model-independent implementation of scalar-tensor theories in MGCAMB
in terms of two functions of the scale factor describing the mass and the coupling of the scalar degree
of freedom, we derive constraints on the f (R), generalized chameleon, Symmetron and Dilaton
models. Since most of the large scale structure data available today is from relatively low redshifts,
only a limited range of observed scales is in the linear regime, leading to relatively weak constraints.
We then perform a forecast for a future large scale structure survey, such as Large Synoptic Survey
Telescope (LSST), which will map a significant volume at higher redshifts, and show that it will
produce much stronger constraints on scalar interactions in specific models. We also perform a
principal component analysis and find that future surveys should be able to provide tight constraints
on several eigenmodes of the scalar mass evolution.
I.
INTRODUCTION
A non-vanishing cosmological constant, , is the simplest and the most common explanation of the observed
cosmic acceleration [1, 2]. Because, gravitationally, is
equivalent to the large vacuum energy predicted in particle physics, its value requires a technically unnatural
fine-tuning [3, 4] in order to be consistent with observations. The cosmological constant could be embedded
in a larger class of dark energy models, where dynamics
dictate the value of the vacuum energy. Because of the
absence of apparent violation of Lorentz invariance in the
Universe, dark energy is commonly described by the field
theory of a scalar. Usually, some degree of fine-tuning of
the parameters of the model must be introduced.
Another explanation could be provided by a modification of the laws of gravity on large scales. Such modifications generically involve a scalar degree of freedom which
can lead to dynamical dark energy when the range of
the scalar interaction is cosmological. As a result, scalartensor models with couplings to matter represent a wellmotivated and versatile class of dark energy. Theories
describing the behaviour of the scalar field involve conformal [5] and disformal couplings to matter [6, 7]. It turns
out that the disformal coupling is severely constrained by
local experiments and cosmological observations [8, 9].
On the other hand, the conformal couplings, albeit large
on cosmological scales, can be screened in the local environment where none of their effects, such as deviations
from Newtons law, have been uncovered.
In this paper, we will focus on scalar-tensor models
with screening mechanisms that are broadly classified
to be of chameleon type [10, 11], i.e. where either the
2
constrained by observations both cosmological and astrophysical. The strongest bound on the range of the scalar
interaction, expressed in terms of the parameter fR0 , is
at the level of 107 and comes from astrophysical tests
of modified gravity using the period of cepheids or the
gas dynamics of dwarf galaxies [4547]. The cosmological bounds are less effective, at the level of 105 [43, 48].
On the other hand, dilatons and symmetrons have not
been constrained as systematically as f (R) on cosmological scales. Only a few tests have been performed using the [m(a), (a)] parameterisation [42]. The strongest
bounds on dilatons and symmetrons still spring from local gravitational tests [49]. Local tests of gravity for the
chameleon-type models of modified gravity imply that
the range of the scalar interaction cannot exceed 1 Mpc,
implying that linear analyses are limited to probing only
some of the features of the chameleon screening mechanisms. On the other hand, studying effects of modified
gravity on shorter scales requires the use of either semianalytical methods suited to the quasi-linear regime of
cosmological perturbations or N-body simulations, both
of which are model-specific. Here, in order to keep our
analysis as model-independent as possible, we shall restrict ourselves to observables which are sensitive to the
linear regime only.
The range of scales that are safely in the linear regime
at low redshifts is quite limited. Most of the large scale
structure data available today is from relatively low redshifts and provides only weak constraints on scalar-tensor
models unless one considers information from non-linear
scales. The only way to do so is to run N-body simulations
for specific models. On the other hand, future surveys,
such as LSST [50] and Euclid [51], will provide a high volume of data from higher redshifts at which the range of
linear scales is significantly larger, allowing one to deduce
stronger constraints on scalar interactions not only for
specific models but in a more general model-independent
way. In this paper, we start by deriving constraints on
f (R), Symmetron and Dilatons models from the subset
of todays data that can be safely considered to be in
the linear regime. Then we perform a Fisher forecast for
the same models assuming data from a future LSST-like
survey in combination with other types of data expected
over the next 5-10 years to show that they will be significantly tighter. Finally, we perform a principal component
analysis (PCA) forecast of m(a) for the same future data,
assuming that (a) is a slowly varying function that can
be taken to be a O(1) constant over the range of redshifts
relevant to LSST.
II.
THE MODEL
()2
V () .
2
(3)
(4)
For some forms of V () and A(), the effective potential can have a density dependent minimum, (). The
scalar force will be screened if either the mass of the
field happens to be extremely large or the coupling happens to be negligibly small at the minimum of Veff ().
Such models can be broadly classified as Generalized
Chameleons (GC), and include the original chameleon
model [10], f (R), dilatons [14] and symmetrons [15].
We note that the GC scalar-tensor theories considered
in this work are viable only if the field stays at the minimum of the effective potential Veff () [18]. In this case,
the effective dark energy equation of state is indistinguishable from 1 and the expansion history practically
the same as in the CDM model. Furthermore, as long
as the scalar field is at its density dependent minimum,
(), the theory can be described parametrically from
the sole knowledge of the mass function m() and the
coupling () at the minimum of the potential [17, 18]
Z c
() c
1
()
= 2
d 2 ,
(5)
mPl
mPl
m ()
where we have identified the mass as the second derivative
m2 () =
(2)
d2 Veff
|=()
d2
(6)
d ln A
|=() .
d
(7)
3
It is often simpler to characterize the functions m() and
() using the time evolution of the matter density of the
Universe
0
(8)
(a) = 3
a
where a is the scale factor whose value now is a0 = 1.
This allows one to describe characteristic models in a
simple way and the full dynamics can be recovered from
the time evolution of the mass and coupling functions,
m(a), (a).
A.
= (k, a)
(9)
(10)
(11)
In the quasi-static approximation, the equation governing the evolution of matter density contrast reads
3
00 + H 0 m H2 (k, a) = 0
2
(14)
(15)
= A2 ()
(16)
then
(12)
B.
where
(k, a) =
2 2 (a)
.
1 + m2 (a)a2 /k 2
(13)
1
2
6000
104
4000
P ( k)
(l +1)ClTT (2)1
5000
3000
2000
103
102
0
0.000
0.001
0.002
0.003
0.004
0.005
0.006
101
102
103
101
P(k)/PCDM(k)
ClTT /ClTTCDM
1000
101
102
CDM
103
Symmetron
2.0
10-3
10-2
10-1
100
10-3
10-2
10-1
100
1.5
1.0
0.5
0.0
Dilaton
k [h/Mpc]
Hu-Sawicki
FIG. 1. Plots of CMB temperature anisotropy ClT T (left) and the matter power spectrum P (k) (right) for the models studied
in this paper. The parameters used for the Symmetron model are: a? = 0.25, ? = 1 and ? = 103 . The parameters used for
the Dilatons model are: 0 = 3 and 0 = 6 103 . The parameters used for Hu-Sawicki f (R) model are fR0 = 104 and n = 1.
The yellow shaded region shows the scales that are not taken into account in the data analysis.
fR0 R0n+1
,
n Rn
(18)
4 + m a3
4 + m
(n+2)/2
(19)
where and m are the dark energy and matter density fractions today, and m0 is a mass scale that can be
expressed in terms of fR0 as [18]
s
4 + m
m 0 = H0
.
(20)
(n + 1)fR0
6
Local tests of gravity require fR0 <
10 [49], while astrophysical constraints from dwarf galaxies imply that
7
fR0 <
10 [47]. These bounds depend on accurate modelling of non-linear physics. In what follows, we will derive
the constraint on fR0 from current cosmological data using only information from linear scales, and also forecast
constraints expected from future surveys like LSST.
5
C.
Another relevant example is the environmentally dependent Dilaton [14], where the screening mechanism is
of the Damour-Polyakov type [16]. This model, inspired
by string theory in the large string coupling limit, has an
exponentially runaway potential
V () = V0 e/mPl ,
(21)
with the value of V0 set to generate the current acceleration of the Universe, while the coupling function is
A() = 1 +
A2
( ? )2 .
2m2Pl
(22)
In dense environments, the minimum of the effective potential approaches = ? , and the coupling function
(a) vanishes. The coefficient A2 has to be large to satisfy
6
local tests of gravity; typically A2 >
10 . These models
can be described by a mass function given by
m2 (a) = 3A2 H 2 (a)
? =
and
r
(a) = ?
? =
where 0 = /m 2.7 is related to V0 , and is determined by requiring that plays the role of dark energy.
We will present our constraints on the mass in terms of
a scalar-force range parameter 0 , defined as
H0
1
0 =
=
,
c m0
3A2
(25)
a 3
?
H0 1
.
c m?
(30)
(31)
E.
In our forecasts, we will also consider generalized models of chameleon type [61] defined by
m(a) = m0 ar , (a) = 0 as .
D.
(23)
(24)
(28)
m
mpl m2? 2?
=
,
3
a?
2?
(32)
Another example of a GBD model with the DamourPolyakov screening mechanims is the Symmetron [15],
where the scalar field has a quartic potential,
"
2
4 #
m2? 2?
1
1
V () = V0 +
+
(26)
2
2 2?
4 ?
and a coupling function,
A() = 1 +
? 2
.
2?
(27)
F.
Binned Model
As discussed so far, for any of the aforementioned models, each with its own theoretical motivation, one can determine the functional forms of m(a) and (a). This effectively reduces the two free functions m(a) and (a) to a
handful of parameters. However, one might be interested
in knowing how well the two functions are constrained in
general, without regard for any specific model. One can
then proceed by discretizing either of the two functions
in bins of redshift space and treating the amplitude in
each bin as a free parameter to be constrained.
Varying both, the coupling and the mass functions,
simultaneously would be redundant, since their effect is
largely degenerate. Since it is the mass parameter that affects the shape of the matter power spectrum, we fix (a)
6
to a constant value of order unity and bin m(a) in redshift. If a non-zero m1 (a) were detected, it would signal
the presence of a scalar interaction and further investigation would be required to determine if the variation
occurs in (a), m(a) or both.
While a binning scheme gives a model independent
(rather a far less model dependent) treatment of m(a),
the larger number of parameters (values of m in each bin)
results in weaker constraints on the individual parameters. To extract useful information, we apply the Principal Components Analysis (PCA) technique (reviewed
in Section IV G). The resulting Principal Components
(PCs) are linear combinations of the original bin values
and the propagated uncertainty (from original errors on
the bins) in their values can inform us about those PCs
that are best constrained by data and the number of degrees of freedom the can potentially be constrained.
III.
A.
P
P
Fixed
m
Varying
m
P
Data sets
f R0
fR0
m (eV)
PLC+BAO
0.05 (0.14)
0.08 (0.23)
0.24(0.35)
+CMBLens 3 (8) 103 0.6 (1.6) 102 0.22(0.31)
+MPK
0.6(1.6) 104 0.7 (1.7) 104 0.24(0.34)
+WL
3 (7) 105
4 (9) 105
0.23(0.33)
TABLE I. The 68% (95%) CL upper limits of fR0 and the
sum of neutrino masses using different combinations of data
sets shown in the table.
B.
Constraints on f (R)
7
P
P
f (R)+AL , fixed
m
f (R)+AL , varying
m
P
fR0
AL
fR0
AL
m
+0.07
(0.12)
+0.10
(0.16)
5
4
3 (8) 10
1.080.05 (0.13) 0.4 (1.0) 10
1.110.06 (0.15) 0.30 (0.38)
TABLE II. 68% (95%) CL bounds on fR0 , AL and
1.0
1
2
log10fR0
P/Pmax
0.8
0.6
0.4
4
5
6
0.2
0.0 -7
10
7
0.0
10-6
PLC+BAO
10-5
10-4
fR0
10-3
+CMBLens
10-2
10-1
100
+CMBLens+MPK+WL
0.1
PLC+BAO
0.2
0.3
0.4
m (eV)
+CMBLens
0.5
0.6
0.7
+CMBLens+MPK+WL
P
m in the Hu-Sawicki
FIG. 3. Joint contours for fR0 and
model (n = 1) after marginalizing over all other cosmological parameters. The darker and lighter shades correspond
respectively to the 68% C.L. and the 95% C.L. . Datasets are
described in text and also in the caption of Fig. 2
PLC+BAO
+CMBLens+MPK+WL
0.8
P/Pmax
log10fR0
4
5
6
7
fixed m
X
varying m
1.0
0.6
0.4
0.2
0.6
0.8
1.0
AL
1.2
1.4
0.0 -6
10
10-5
10-4
10-3
10-2
10-1
100
solid
P line) as well as after marginalizing over a varying
m (blue dashed line). We find an upper bound of
? < 1.5 103 at 95 % C.L, which corresponds to a
Compton wavelength of a few Mpc. Our bounds are
summarized in Table III.
As mentioned above, current data is unable to simultaneously constrain all the model parameters because they
are highly correlated. We also note that one cannot derive meaningful constraints for smaller values of coupling
constant ? as the modification of growth is relatively
small for the scales and redshifts currently probed. Further, since a? sets the onset of modified growth, we would
see tighter constraints on ? for smaller a? values. Nevertheless, as we will show in Sec. IV, future surveys with
larger sky and deeper redshift coverage will be able to
constrain ? along with the other two parameters.
(33)
D.
The results of the analysis with varying AL are summarized in Table II.
C.
P
P
Fixed
m
Varying
m
P
?
?
m
Symmetron
0.8 (1.5) 103 0.9 (1.8) 103 0.16 (0.27)
P
0
0
m
Dilaton
2.1 (3) 103
2.3 (3) 103 0.15 (0.25)
TABLE III. Summary of the 95% CL upper limits of the MG
parameters and the sum of neutrino masses (in unit of eV)
derived from current observations described in Sec. III A.
9
X
fixed m
X
varying m
1.0
P/Pmax
0.8
0.6
0.4
0.2
0.0 -5
10
10-4
10-3
10-2
10-1
FIG. 6. Marginalized posterior distribution for 0 in the Dilatons model with 0 = 5. The datasets used in the analysis are PLC+BAO+lensing+MPK+WL as described in section III A. The red solid line shows
the case with massive neuP
trinos with a fixed mass of
m = 0.06eV, while the blue
solid
P lines shows the PDF after marginalizing over a varying
m .
IV.
A.
B.
Fisher analysis
FORECASTS
Fab = fsky
`X
max
`=`min
2` + 1
Tr
2
C` 1 C` 1
C
C
pa ` pb `
, (34)
` is the
where pa is the ath parameter of our model and C
observed covariance matrix with elements C`XY that
include contributions from noise:
C`XY = C`XY + N`XY .
(35)
10
1E-4
GC
+WL
Full
6E-5
fR0
fR0
8E-5
LSST
DES
LSST, fixed n
6E-5
4E-5
4E-5
2E-5
2E-5
1.0
1.5
2.0
FIG. 7. Expected 1 bounds on the parameters of the HuSawicki model. The assumed fiducial model is marked with
a star. The importance of using theFull set of observables
(WL, GC and their cross-correlation) is clearly demonstrated.
The Planck CMB data is included in all cases and is important
for constraining the standard cosmological parameters.
C.
Fig. 9 shows the bounds on the parameters of the Symmetron parameters expected from LSST+. As a fiducial
model, we assume ? = 1 and a mass scale of ? = 103 ,
which corresponds to a range of a few Mpc. Current data
is unable to constrain ? if a? = 0.5 or larger. For this
reason, the bound on ? in Sec. III was derived for a
0.05
0.10
m [eV]
0.15
D.
2.5
a =0.5
a =0.25
1.05
0.5
1.00
0.95
0.504
6E-4
8E-4
1E-3
1.2E-3
1.4E-3
6E-4
8E-4
1E-3
1.2E-3
1.4E-3
0.502
0.500
0.498
0.496
FIG. 9. Expected 1 constrains from LSST+ on the parameters of the Symmetron model. The assumed fiducial models
are marked with a star. Unlike current data, LSST+ can simultaneously constrain ? and a? to a few percent level, and
will improve the current bounds on ? . See Table IV for a
quantitative comparison.
fixed a? = 0.25. We perform a forecast using two different fiducial values: a? = 0.25 and 0.5. In the former
case, LSST+ clearly improves on the bound in Sec. III,
even after marginalizing over a? and ? . It will also be
able to provide a non-trivial bound on ? for a? = 0.5,
which is the value assumed in much of the previous literature. The current and expected bounds are summarized
in Table IV.
It is interesting to examine the possible degeneracy be-
11
2E-03
1E-03
1.1
1E-03
1.2
a Marg.
a Fixed
a , Fixed
1E-03
0.9
8E-04
0.8
0.05
0.1
0.15 0.2
m[eV]
E.
0.08
0.25
m[eV]
6E-04
0
F.
1.0
0.07
0.06
0.05
9.6E-4
9.8E-4
1E-3
1.02E-3
1.04E-3
G.
12
9.6E-4
1E-3
1.04E-3 0.97
1.03
0.506
0.504
0.5
0.497
0.494
0.8
0.7
0.6
0.5
0.4
0.3
0.2
(36)
(37)
V.
m(a) m(a)
N
X
i ei (a)
(38)
i=1
(39)
Modifications of gravity on cosmological scales can potentially explain the origin of cosmic acceleration. The
Generalized Brans-Dicke theory, in which there is an additional scalar degree of freedom that mediates a fifth
force, is one of the viable MG models that are able to fit
13
101
10-1
ei (z)
i (z)
100
Marg.
Fixed
10-2
10-3
6
PC#
10
0.8
0.6
0.4
0.2
0.0
0.2
0.4
PC #1
PC #2
PC #3
PC #4
FIG. 13. Left: The uncertainties (square roots of eigenvalues) associated with the eigenmodes of m(a) for the case when the
coupling is fixed at = 0.4 (solid line), and when it is marginalized over (dashed line). Right: The first four best constrained
eigenmodes of m(a) after marginalizing over .
Hu-Sawicki f (R)
Symmetron
Dilaton
Parameters
f R0
n=1
?
? = 1 a? = 0.25 (0.5)
0
0 = 1 (5)
Current 1 4 105
103
unconstrained (2.3 103 )
LSST+ 1 6 106 0.3
2 (2.9) 104 0.05 (0.07) 0.001 (0.005)
2.7 105
2.3 101
TABLE IV. The current 68 % C.L. uncertainties and those expected from LSST+. The blocks with mean the parameter
was fixed at its fiducial value. The values in parenthesis indicate those obtained for an alternative fiducial value.
14
those from a photometric survey like LSST [84].
ACKNOWLEDGMENTS
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
at SFU. GBZ is supported by the Strategic Priority Research Program The Emergence of Cosmological Structures of the Chinese Academy of Sciences Grant No.
XDB09000000. This research was enabled in part by support provided by WestGrid [85] and Compute Canada
[86]. P.B. acknowledges partial support from the European Union FP7 ITN INVISIBLES (Marie Curie Actions,
PITN- GA-2011- 289442) and from the Agence Nationale
de la Recherche under contract ANR 2010 BLANC 0413
01. ACD acknowledges partial support from STFC under
grants ST/L000385/1 and ST/L000636/1.
15
ph.CO]].
[43] J. Dossett, B. Hu and D. Parkinson, JCAP 1403 (2014)
046 [arXiv:1401.3980 [astro-ph.CO]].
[44] P. A. R. Ade et al. [Planck Collaboration], Planck
2015 results. XIV. Dark energy and modified gravity,
arXiv:1502.01590 [astro-ph.CO].
[45] B. Jain, V. Vikram and J. Sakstein, Astrophys. J. 779
(2013) 39 [arXiv:1204.6044 [astro-ph.CO]].
[46] V. Vikram, A. Cabr, B. Jain and J. T. VanderPlas, JCAP
1308 (2013) 020 [arXiv:1303.0295 [astro-ph.CO]].
[47] J. Sakstein, B. Jain and V. Vikram, Int. J. Mod. Phys. D
23 (2014) 12, 1442002 [arXiv:1409.3708 [astro-ph.CO]].
[48] J. Bel, P. Brax, C. Marinoni and P. Valageas, Phys. Rev.
D 91 (2015) 10, 103503 [arXiv:1406.3347 [astro-ph.CO]].
[49] P. Brax, Phys. Rev. D 90 (2014) 2, 023505
[arXiv:1310.2562 [astro-ph.CO]].
[50] http://www.lsst.org
[51] http://www.euclid-ec.org
[52] C. Schimd, J. P. Uzan and A. Riazuelo, Phys. Rev. D 71,
083512 (2005) [astro-ph/0412120].
[53] http://www.camb.info
[54] Http://www.sfu.ca/~aha25/MGCAMB.html
[55] P. Brax and P. Valageas, Phys. Rev. D 88, no. 2, 023527
(2013) [arXiv:1305.5647 [astro-ph.CO]].
[56] C. Llinares and D. Mota, Phys. Rev. Lett. 110, no. 16,
161101 (2013) [arXiv:1302.1774 [astro-ph.CO]].
[57] S. Capozziello, S. Carloni and A. Troisi, Recent Res. Dev.
Astron. Astrophys. 1, 625 (2003) [astro-ph/0303041].
[58] S. M. Carroll, V. Duvvuri, M. Trodden and M. S. Turner,
Phys. Rev. D 70, 043528 (2004) [astro-ph/0306438].
[59] S. A. Appleby and R. A. Battye, Phys. Lett. B 654, 7
(2007) [arXiv:0705.3199 [astro-ph]].
[60] A. A. Starobinsky, JETP Lett. 86, 157 (2007)
[arXiv:0706.2041 [astro-ph]].
[61] P. Brax, A. C. Davis, B. Li, H. A. Winther and
G. B. Zhao, JCAP 1304 (2013) 029 [arXiv:1303.0007
[astro-ph.CO]].
[62] http://cosmologist.info/cosmomc/
[63] P. A. R. Ade et al. [Planck Collaboration], Planck
2015
results.
XIII.
Cosmological
parameters,
arXiv:1502.01589 [astro-ph.CO].
[64] P. A. R. Ade et al. [Planck Collaboration],
arXiv:1502.01591 [astro-ph.CO].
[65] P. A. R. Ade et al. [Planck Collaboration], Astron. Astrophys. 571, A16 (2014) [arXiv:1303.5076 [astro-ph.CO]].