Subaru Galaxy Surveys: HyperSuprime Cam

1
Subaru Galaxy Surveys Hyper-Suprime Cam
WFMOS(As an introduction of next talk by Shun
Saito)
Masahiro Takada (Tohoku Univ., Sendai, Japan)
Sep 11 07 _at_ Sendai
2
CMB Large-Scale Structure (LSS)

WMAP (z103)
SDSS (Tegmark etal03)
LSS (0ltzlt3)
3
Complementarity btw CMB and LSS
(e.g., Eisenstein, Hu Tegmark 98)

• CMB probes the statistical properties of
  fluctuations at z103
• All the fluctuations are well in the linear
  regime clean info
• Linear perturbation theory predictions, which are
  robust and secure, can be compared with the
  measurements
• A galaxy survey probes the density perturbations
  at low redshifts (0ltzlt3)
• The perturbation amplitudes significantly grow
  from z103, by a factor of 103 at least
• An uncertainty in the model predictions arises
  from non-linearities in structure formation
• Combining the two is very powerful (e.g., WMAP
  SDSS)
• Opens up a window to probe redshift evolution of
  the perturbations, which helps break parameter
  degeneracies
• Allow to constrain the neutrino mass
• Very complementary in redshift and wavenumbers
  probed

4
Linear growth rate (a case of CDM model)

• The density perturbation in the LSS, observable
  from a galaxy survey
• Linear growth describes the time-evolution of the
  density perturbations, form the CMB epoch
  (z103)
• In the matter-dominated regime, the CDM
  perturbations of different wavelengths grow at
  the same rate
• Combining the FRW eqns and the linearized
  GRBoltzmann eqns leads to the second-order
  differential equation
• Alternative, yet interesting ingredients
• The cosmic acceleration slows down the growth
• Adding massive neutrinos leads to suppression in
  the growth at low redshifts and on small scales

5
CDM Structure Formation Model P(k)
k3P(k,z)/2?2lt?2gtR1/k
Amplification in the density perturbation
amplitude by a factor of 1000, between z0 and
1000.
6
Massive neutrinos and LSS

• The experiments imply the total mass, m_totgt0.06
  eV
• Neutrinos became non-relativistic at redshift
  when T?,decm?
• Since then the neutrinos contribute to the energy
  density of matter, affecting the Hubble expansion
  rate
• The cosmological probes (CMB, SNe, BAO ) measure
  
• The massive neutrinos affect the CMB spectra,
  mainly through the effect on H(z) (see Ichikawa
  sans talk)
• The effect is generally small, also degenerate
  with other cosmo paras.

mtotgt0.11 eV
mtotgt0.06 eV
7
Suppression in growth of LSS

• Neutrinos are very light compared to CDM/baryon
  the free-streaming scale is 100Mpc (for
  m0.1eV), relevant for LSS
• At a redshift z
• The neutrinos slow down the growth of total
  matter pert.
• On large scales ?gt?fs, the neutrinos can grow
  together with CDM
• On small scales ?lt?fs, the neutrinos are smooth,
  ??0, therefore weaker gravitational force
  compared to a pure CDM case

? lt ?fs
? gt ?fs
CDM
CDM
Suppresses growth of total matter perturbations
Total matter perturbations can grow!
8
Suppression of growth rate (contd.)
9
Suppression of growth rate (contd.)
k_fs
The suppression is stronger at lower redshifts,
implying the usefulness of CMBLSS to probe the
neutrino effect E.g., the current limit on the
total neutrino mass, m_totlt0.9 eV (95) from WMAP
SDSS (Tegmark etal. 06)
10
Hyper Suprime-Cam (HSC)

• Replace the Subaru prime focus camera with the
  new one (HSC)
• PI S. Miyazaki (NAOJ)
• The grant (15M) to build the new camera was
  approved in 2006
• Construction 2006-2011
• FoV 1.5 - 2.0 degrees in diameter (10 ? the
  Suprime-Cams FoV)
• 4 - 5 broad band filters (BVRiz) available
• The first light in 2010 - 2011
• Plan to conduct a wide-field survey (primarily
  for WL) hopefully starting from 2011 for 3-5
  years

11
Suprime-Cam
From Y. Komiyama
2 degree FoV option
12
WFMOS (Wide Field Multi-Object Spectrograph)

• The project originally proposed by Gemini
  observatory (USEurope) to Japan (2005-)
• Now seriously considered as a next-generation
  Subaru instrument in the phase of the
  feasibility/design study
• Assume the HSC FoV
• 2000-4000 fibers
• If fully funded (gt50M) the first-light
  2015(?)-, after HSC

Echidna
Glazebrook et al. astro-ph/0507457
Proposed galaxy redshift survey

• Survey area 2000 deg2 _at_ 0.5ltzlt1.3
  (ng1000deg-2), 300deg2 _at_ 2.5ltzlt3.5 (ng2000
  deg-2) ? 300 nights
• Primary science cases dark energy, neutrinos

13
Advantage of high-redshift survey (I)

• For a fixed solid angle, a higher-redshift survey
  allows to cover a larger 3D comoving volume
• A more accurate measurement of P(k) is available
  with a larger surveyed volume
• A planned WFMOS (z1 survey with 2000 deg2 z3
  survey 300 deg2)
• 4 (z1) 1 (z3) 5 h-3 Gpc3
• For comparison, SDSS (z0.3) covers
• 1 h-3 Gpc3 with 4000 deg2 (Eisenstein etal
  05)
• V_wfmos 5 V_sdss

?_s
14
Advantage of high-redshift survey (II)

• At higher redshifts, weaker non-linearities in
  LSS
• A cleaner cosmological info is available up to
  kmax
• SDSS kmax0.1 h/Mpc
• WFMOS
• z1 kmax0.2 h/Mpc
• z3 kmax0.5 h/Mpc
• Surveyed volume in F.S.
• V_wfmos(k)30V_sdss(k)
• In total, accuracy of measuring P(k)
  ?2(lnP(k))1/V_s?V(k)

Springel etal. 2005, Nature
15
A measurement accuracy of P(k) for WFMOS
Neutrino suppress. 0.6 of ?_m ?4 effect on P(k)

• WFMOS allows a high-precision measurement of P(k)
• The characteristic scale-dependent suppression in
  the power of P(k) due to the neutrinos could be
  accurately measured (see Saito kuns talk)

16
The parameter degeneracy in P(k)

• Different paras affect P(k) in fairly different
  ways
• Combining galaxy survey with CMB is an efficient
  way to break degeneracies btw f_nu, n_s and alpha
  (MT, Komatsu Futamase 2005)

17
Different probes are complementary
From Tegmark04
18
Summary

• CMBLSS opens up a new window of constraining the
  neutrino mass, from the measured suppression in
  the growth of mass clustering
• A higher redshift survey, such as the survey of
  planned Subaru survey, allows a precise
  measurement of the galaxy power spectrum
• Need to develop more accurate theoretical
  predictions of P(k) for a mixed DM model that
  allow a secture comparison with the precise
  measurement (see Saito kuns talk!)

19
Suppression in P(k)

• Assume 3 flavors when relativistic
• Consistent with CMB and BBN
• Assume N? species become NR (or are massive) at
  low-z
• Suppression has scale-dependence
• P(k) amplitude is normalized by the primordial
  Pi(k)
• All P(k) have same amount suppression on
  sufficiently large k.

f?0.05 (??0.014)
WFMOS z3 slice
f?0.01 (??0.003)
20
Forecasted errors for neutrino paras

• 2D galaxy P(k) is very powerful to constrain mtot
• N.O. experiment neutrinos can be weighed at more
  than 1? ?(mtot)0.03eV
• Relatively difficult to constrain N? and mtot
  independently.
• If mtotgt0.45eV, models with ?N?1 can be
  discriminated at more than 1-sigma level

21
WFMOS Can Measure DE Clustering?
(MT 06 soon)

• Another important consequence of DE with w?-1 is
  its spatial clustering, ?de(x,t)
• A useful way explore fluid properties of DE
  (??de, ?pde, ?de, ), instead of modeling a form
  of DE Lagrangian
• Sound speed ce (??pde) defines the free-streaming
  scale of DE clustering (e.g., quintessence,
  c_e1)
• ?gt ?fs DE can cluster with DM
• ? lt ?fs DE perturbations are smooth (?de0)

22
Effect on P(k)
23
Sensitivity of WFMOS to DE perterbations

• If c_elt0.1, WFMOS can measure the DE
  perturbations at more than 1-sigma significance.
• The power is compatible with an all-sky imaging
  survey (CMB-galaxy cross-correlation, Hu
  Scranton 04).

24
Summary

• Hyper-Suprime/WFMOS survey will provide an
  ultimate, ideal dataset for performing BAO as
  well as WL tomography experiments.
• BAO and WL are complementary for DE constraints,
  and more important is the independent two methods
  from the same surveyed region will be very
  powerful to test various systematics.
• Issue for WL Need to study which type of
  galaxies gives a fair sample of WFMOS sample
• Current survey design (2000deg2, 4 or 5 colors,
  ng0.3/arcmin2 or ng5?10-4 h3 Mpc-3) seems
  optimal for joint BAO and WL experiments.
• Valued Sciences WFMOS can do
• Neutrino Mass sigma(mtot)0.03eV
• Dark energy clustering
• In this case, zlt1 survey is crucial for doing
  this (z3 survey cant do)
• Inflation parameters (ns and alpha) 10-3
• Having a well-defined survey geometry is crucial
  optimal survey strategy?