R. Jimenez - Fundamental Physics from Astronomical Observations
1. Fundamental physics from astronomical observations
Raul Jimenez
ICREA
ICC University of Barcelona
icc.ub.edu/~jimenez
Courtesy of Planck and SKA teams
2. Ultimate Experiments
In cosmology one can actually perform ultimate experiments, i.e.
those which contain ALL information available for measurement in the
sky. The first one of its kind is be Planck (in Temperature) and in this
decade we will also have such experiments mapping the galaxy field.
Question is: how much can we learn about fundamental physics, if any, from
such experiments?
There are many examples:
1. Dark Energy
2. Inflation
3. Neutrino masses
4. Nature of initial conditions
3. Beyond the Standard Model Physics
3. Extremely successful model
State of the art of data then… (~1992)
~14 Gyr
(DMR)COBE
CMB
380000 yr
(a posteriori information)
5. Vacuum energy
(also known as dark energy or cosmological constant)
vacuum
1917
Λ
Negative
pressure!
V1
V2
E1= ρv V1
E2= ρv V2
E2>E1
6. Supernovae
Standard candles
L
DL =
4πF
Function of geometry and
Content of Universe
7. Dark Energy(The basics)
Simon, Verde, RJ PRD (2005)
Action describing the dynamics of the universe is:
3 m2
p g µν
S = ∫ dtd x − g − R+ ∂ µ q∂ν q − V (q) + Smatter }
16π
2
Consider quintessence a perfect fluid:
1 2
ρQ =
q + V (q)
2
1 2
pQ = q − V ( q )
2
Which has conservation law:
ρ q + 3 H ( ρ q + pq ) = 0
2
All left now is use Einstein eq:
a 8π
H = = 2 (ρ m + ρ q )
2
a 3m p
8. 2
All left now is use Einstein eq:
a 8π
H = = 2 (ρ m + ρ q )
2
a 3m p
And Klein-Gordon equation:
q + 3Hq + V ' = 0
What I want to know is shape of potential V
H
ε1
ε1 = − 2 ; ε 2 =
H Hε1
H2 1 1
V ( z ) = (3 − ε 1 ) − ∑ (1 − wi ) ρ i − ( ρ f − p f )
mp 2 i 2
But what I really need is V(q)
H2 1
K (q) = ε 1 − ( ρT + pT )
mp 2
9. We can measure dark energy because of its effects on the expansion
history of the universe: a(t)
a(t ) 1 dz
= H ( z) = −
a(t ) (1 + z ) dt
H 2 = H 2 0 [ ρ ( z ) / ρ (0)]
ρ Q = −3H ( z )(1 + w( z )) ρ Q
0
dt
d L = (1 + z ) ∫ (1 + z ' ) dz '
SN: measure dL z
dz '
CMB:θA and ISW a(t)
LSS or LENSING: g(z) or r(z) a(t)
z
−1 dz dz '
AGES: H(z) a(t)
5/ 2
H0 = −(1 + z ) {Ω m (0) + ΩQ (0) exp[3∫ wQ ]}1/ 2
dt 0
(1 + z ' )
10. Reconstruct w(z): use dz/dt
Non-parametric!
0
Note:
dt
d L = (1 + z ) ∫ (1 + z ' ) dz '
z
dz '
1 dz
H ( z) = −
(1 + z ) dt
w(z) in here
2
d 2 z dz 5 3
= (1 + z ) −1 + w( z )
dt 2 dt 2 2
3
− Ω m (1 + z ) 4 w( z )
2
z
(from Jimenez & Loeb 2002
21. Reconstruct w(z): CAN IT work?
At z=0 dz/dt gives Ho and we have SDSS galaxies:
1 dz
H ( z) = −
(1 + z ) dt
The edge for z<0.2
The value of H0
22. A good test, to determine H(z=0)
Moresco, RJ, Cimatti, Pozzetti JCAP (2010) H(0) = 72.3 ± 2.8
23. Stern, RJ et al. JCAP 2011
The data at z>0 Moresco et al. 2011
24.
25. However, one can go one step further and build an effective theory…
…of expansion. RJ, Talavera & Verde 2011 (arXiv:1107.2542)
The simplest theory of expansion involves, besides gravity, a single
canonically normalized expansion field described by the leading Lagrangian
density
I can copy Weinberg for QCD, BUT here I cannot do scattering… so how to do the
power counting?
mass gap
26. If I obtain the modifications to gravity from growth of structure and/or GW, then I
can obtain the lambdas from the expansion rate…
H
ε
ε1 = − 2 ; ε 2 = 1
H Hε 1
27. Multiple uses of H(z)
A factor 5 improvement on universe transparency (Avgoustidis, Verde, RJ
JCAP(2009))
Detection of aceleration/deceleration
(Avgoustidis, Verde, RJ JCAP
(2009))
28. Multiple uses of H(z)
Constraints on the mass and number of relativistic particles (de Bernardis et al.
JCAP0803:020,2008 Figueroa, Verde, RJ JCAP0810:038,2008) and on
the curvature (Stern et al. 2009)
29.
30. Summary
• Vast quantity of high quality cosmo data
fast approaching: CMB, BAOs,
Gravitational waves, 21cm,…
• Fruitful interplay between HEP/cosmo
theory and cosmological observation:
constraints on axions, neutrino masses,
neutrino hierarchy, nature of the initial
conditions…
• First determinations of the expansion
history of the Universe, H(z), already
available at ~ 10% level. They already
provide constraints on alternatives to the
LCDM model.