Dark Matter:
What do we really know?
ICISE inauguration, Quy Nhon, August 11-17, 2013
Charling TAO tao@cppm.in2p3.fr
Centre de Physique des Particules de Marseille (CPPM), IN2P3,
Marseille, France
Tsinghua Center for Astrophysics (THCA), Tsinghua University,
Beijing, China
Dark Matter:
What do we really know?
DM: - particle that does not emit observable radiation
- interacts gravitationally…
- non baryonic
DM: we know it exists in the Universe!
Assuming standard Big bang Cosmology with GR
Wealth of astrophysical evidence for DM

Galaxy rotation curves (V. Rubin)

Dynamics of galaxy clusters (Zwicky)

X-ray clusters

Bullet cluster (Clowe+,2006)

Gravitational lensing mass reconstruction
Evidence for dark matter:
clusters
Zwicky ApJ 86 ,217 (1937): Coma Cluster velocities
Clusters velocity dispersion 
4
masses ~ 100 x visible mass
Galactic level: (!930’s) Oort discrepancy in the Milky Way
April 18, 20015
disk: factor 2 now disappeared
Evidence for dark matter:
rotation curves of spiral galaxies
V. Rubin
1970’s
A.Bosma
Some numbers ...
A galaxy like the Milky Way or Andromeda has a total visible
mass of about 61010 Msun.
- rotation velocity is ~220 km/sec
- radius about ~30 kpc
Newton:
2
vrot
R
GM
vrot 
M 
R
G
 total mass: 3.31011 Msun
 ~5 times more mass than visible
 Local density 0.3- 0.4 GeV/cm3
Evidence for dark matter:
Gravitational Lensing:
Gravitational Lensing:
a property of General Relativity
Perfectly Aligned
Slightly Misaligned
GR: light trajectory bent by a gravitational field.
Evidence for dark matter:
Bullet Cluster
Clowe+
2006
X-ray vs gravitational lensing:
Gaz clearly separated from mass potential peaks
Dark Matter:
What do we really know?
DM: - particle that does not emit observable radiation
- interacts gravitationally…
- non baryonic
DM common paradigm:
it exists!
- Contributes to energy density in the Universe,
- Measured in clusters and galaxies
The Universe energy density content
after Planck
% precision
Cf Y. Giraud-Heraud‘s talk
Wikipedia
Matter today ~ 31.7%
energy density of the Universe
84.5% of the matter is dark matter
What do we know about DM
nature ?
Particle : stable?
mass?
interaction cross-sections?
charge?
spin ?
Constraints from non-observation in direct/indirect/LHC searches
AND
Observations in Astrophysics / Cosmology
Very different DM candidates
Modified
Gravity
dust
Cold Molecular
Hydrogen
MACHOs
Black holes
SIMPs
1Neutrino
3. Light axions
Exotica
2. WIMPs
Weakly interacting
massive particles
10-1000GeV
Snowmass
2013
Why WIMPs?
“WIMP”= “Weakly Interacting” Massive Particles
G. Altarelli: « still most optimal candidates !»
Arguments in the 1980’s:
•
Need for Cold Dark Matter from Large Scale
Structures
• Very good Particle physics candidate: SUSY LSP
• Weak neutrino size cross sections expected which our
detectors Ge, NaI were sensitive to…
Why WIMPs?
“WIMP”= “Weakly Interacting” Massive Particles
Assumption: DM= Relic Particles from Big Bang
If DM survives today
rate annihilation < rate expansion
If (rate annihilation << rate expansion),
too much DM today
At « freeze out »
< s v > ~ 10-26/ W h2 cm3/s
Scale of weak interactions !
Coincidence with W, Z physics?
Argument in 80’s, now weaker?
Searches for massive neutrinos cross sections exclude cross-section
< 0.1 sn
Particle physics preferred DM:
SUSY Neutralinos ?
• A natural particle physics solution
• Stable linear combination gauginos and higgsinos (LSP)
0
0
˜
˜
˜
˜  Z  H1  H2
  
• SUSY > 7 parameters MSSM  no predictive power
• Experimental Constraints LEP, pp, b-->s,LHC ...
Look everywhere possible !
Direct and Indirect
Detections
WIMP searches
Direct detection
M
Indirect detection

MN
Ge, Si, NaI, LXe, …
Accelerator particle
production,
eg, LHC
+ Galactic, cluster, Universe scales…
n,,p, e+
Indirect Detection: Principle

SMMG
Sun, Earth, Accumulation
Galactic center,
+
clumps?
n Annihilation
Possible final states: t+t-, lepton pairs,
qq, WH, ZH, WW, ZZ ; Hadronisation
and decay
Astroparticle detectors:
positrons, antiprotons, antideutons
gammas, neutrinos
Astrophysical
of observed
Non dedicatedorigin
experiments
signals,eg, AMS, are hard to
exclude (cf Lee SC’s talk)
Need discovery at accelerators!
Still hope at LHC ?
WIMPs Indirect Detection

- e+
n,,p,
Shore station
Light Sources
K40
Bioluminescence
Optical M odules
track
hydrophone
Cerenkov Light
~60m
float
Compass,
tilt meter
2500m
Electro-optical
underwater cable
~40km
300m
Electronics Containers
Readout Cables
50m
anchor
Junction box
acoustic detector
Present limits
Neutrino limit: Billard+ 2013
Snowmass 2013
WIMP search: direct detection
Cf. B. Sadoulet’s talk
Usual assumptions of DM distribution in our
Galaxy
Usual h y p othesis
DM= 0.3 GeV/cm 3 , =10 -3 ,
Maxwellian distribution of
velocities, vrms =270 km/s
Rotation curves
 c (r )
(r ) 
(r / a)  (1  (r / a)  ) (  ) / 
?
a = halo core radius
  
Isothermal profile
2 2 0
 =0 without cusp
vSun=220 km /s
Navarro-Frenk-White 1 3 1
« Simplified Model »of
Matter in our Galaxy:
SMMG
Mo Moore +
1.5 3 1.5)
Used for most comparisons…
But is it the reality? Clumps? Corotation?
Galactic scale N-body simulations
with Baryons
Ling+ 2009 Dark Matter
Direct Detection Signals
inferred from a
Cosmological N-body
Simulation with Baryons
Fin
2 DM populations :
halo DM +disk DM
 only measurements can tell
LCDM simulation at small
scales might have problems
DM properties from Large Scale Structures
LSS
Cf beautiful movies of G. Smoot
Planck CMB map
Primordial perturbation seeds for structure formation
DM potential wells
Structure formation: Bottom-up Scenario
!Density perturbations collapse into DM haloes.
Small Haloes merge into bigger haloes.
Gas in DM haloes collapse in galactic disks.
Shapes of galaxies change over time.
Due to merging of haloes
Hubble Tuning
Fork Diagram
Before 2000: Nature of DM
Hot or Cold?
CDM is non-relativistic
at decoupling,
Form structures in a
Hierarchical
bottom-up scenario.
HDM relativistic at decoupling
Mean free path large
 Large structures form first
Comparisons of observations with
pre-2000 N-body Simulations prefer CDM
Z=3
Z=1
Z=0
OMEGA = 0.3
LAMBDA = 0.7
H0 = 70 km/(Mpc sec)
Sigma8 = 0.9
LCDM
OMEGA = 1
LAMBDASCDM
=0
H0 = 50
km/(Mpc sec)
Sigma8 =tCDM
0.51
OMEGA = 0.3
LAMBDAOCDM
=0
H0 = 50 km/(Mpc
sec)
Sigma8 = 0.51
OMEGA = 0.3
LAMBDA = 0
H0 = 70 km/(Mpc sec)
Sigma8 = 0.85
Collaboration VIRGO 1996
http://www.mpa-garching.mpg.de/~virgo/virgo/
N-Body simulations: CDM
Preferred paradigm:
Most N-Body simulations use stable CDM halos as seed for
structures:
structures evolve, merge and cluster
- DM halos
- cuspy density profiles,
- Triaxial halos
- central density depends on the mass of the halo.
Dark matter distribution—Density profiles
Cusp
Universal Density Profile
from N-body simulations
NFW
Navarro, Frenk, White 1996
Cluster central density profile X-ray
~2000 : CDM crisis at small scales
Comparing data with N-body
Simulations
•cusp/core at GC
•Missing galactic satellites
Galaxy profiles prefer core at center
CDM Simulations  cusps
(Navarro, Frenk, White 1996):
Problems at smaller scales?
Observations favour
Core profile
rotation curves
Galaxy core vs cusp
Salucci & Frigerio Martins, 2009
Data prefer
Burkert Core Profile
Too low number of visible Satellite galaxies
Satellite galaxies are seen in Milky Way, e.g. Saggittarius, MCs
Predicted number
Observed number
of luminous
satellite galaxies
10km/s
20km/s
100km/s
Alternatives to CDM
• Self-Interacting Dark Matter (Spergel & Steinhardt 2000)
• Strongly Interacting Massive Particle
• Annihilating DM
• Decaying DM (eg. Zhang XM+, Nguyen Quynh Lan in // session)
• …
• WDM: reduce the small scale power
Norma G.Sanchez, Hector J. de Vega+… Chalonge series
DM Self-interaction constraints
DM particles might interact with themselves or other new
particles, mediated by new, dark gauge bosons.
Interactions affect the structures of DM halos:
DM scatters  energy and angular momentum transfers
For hard-sphere elastic scattering,
observations of the structure of galaxy clusters
constraints σ /m
4.5 E-7 (t/E10 yr)-2 < s/m <~ 1 cm2/g Bullet cluster
Williams & Saha 2011
SL cluster analysis
< 0.02
elliptical core MS2137-23
Miralda-Escude 2002
Non neutral DM/CHAMPs
Strong constraints :
Charged (CHAMPS) or small electric or magnetic
dipole moment
coupling to the photon-baryon fluid before
recombination,
alter the sub-degree-scale of CMB and matter power
spectrum.
Cf
- Sigurdson+ , Dark-matter electric and magnetic dipole moments,
(2004);
- McDermott, H.-B. Yu, & K. M. Zurek, Turning off the lights: How dark is
dark matter? (2011)
“Evidence” for WDM ?
•"missing satellite problem'',
•''cusp-core problem'',
• mini-voids The sizes of mini-voids in the local universe: an argument
in favor of a warm dark matter model? Tikhonov et al.
•HI determinations of velocity function profiles
N-Body simulation Comparisons with Virgo results by Arecibo Legacy
(ALFALFA)
N-Body simulations: WDM
Stable WDM looks like stable CDM on scales> 10 Mpc,
- WDM create a cutoff in the matter power spectrum
At late times, the evolution of the matter power spectrum is
more subtle as halos form.
Large WDM halos are virtually indistinguishable from stable
CDM halos
somewhat less concentrated,
smaller halos, fluffier and less cuspy than CDM halos.
The subhalo mass function drops significantly on mass scales
corresponding to that cutoff scale.
Does not solve everything
Nature of DM
Hot or Cold, or Warm?
CDM is non-relativistic
at decoupling, forms
structures in a hierarchical,
bottom-up scenario.
HDM is tightly bound by
observations
and LSS formation
WDM
10 h/Mpc, keV
WDM?
CLUES simulations, Yepes, 2010
WDM vs CDM
Velocity function
Density profile
From Jing 2000
CDM vs WDM: HI velocity functions
Virgo and Anti Virgo directions
arXiv:1005.2687: Constrained Local
UniversE Simulations (CLUES)
Gottloeber, Hoffman , Yepes
No simple feedback
mechanism to explain
the factor 10 depletion
from CDM?
Velocity widths in Galaxies
Velocity widths in galaxies from 21 cm HI surveys
Papastergis et al, 2011; Zavala et al., 2009
NB: The red curve is for 1 keV WDM
Limits on mass of eventual WDM particles
• Stellar dynamics in MW satellites (Boyanovsky, de Vega, Sanchez 2008;
de Vega and Sanchez 2009)
• High-z QSO LF (e.g. Song and Lee 2009)
• Ly-alpha forest to constrain P(k) at small scales and different z’s
(Most popular method: Narayanan et al 2000; Viel et al 2005;2008)
• Ly-a + SDSS results (Boyarsky et al 2009)
• QSO lensing ( Miranda & Maccio 2007 )
• Abundance of dwarf satellites of MW (Maccio & Fontanot 2010;
Polysensky & Ricotti, 2010)
 Mass WDM ~ 1- 5 keV
A fashionable (?) candidate
Sterile neutrinos
Constraints on sterile neutrinos
~2000 :Problems with CDM at small scales
Comparing data with N-body
Simulations
• Galactic satellites
• cusp/core at GC
Problems can perhaps be solved with better
resolution and additional physics in N-Body
simulations (SN, AGN feedback, stellar winds…)
Einasto vs NFW
CDM Simulations
 cusps
rather Einasto
profiles than NFW
Ma Chung Pei,
Chang, P.,
Zhang, 2009
Missing satellites: CDM way out
• satellites do exist, but star formation suppressed (after reionization?)
• satellites orbit do not bring them to close interaction with disk, so
they will not heat up the disk.
• Local Group dwarf velocity dispersion underestimated
• Galaxies may not follow dwarves
Halo substructures may be probed by
- Lensing
- local Milky Way structures
More faint or dark galaxies discovered
Eg, Belokurov et al, 2010
Nature of dark matter or
astrophysics process?
What we know:
Comparisons of observations with
N-body Simulations today
prefer
Non-Hot DM
Probing DM Particle properties
CL0024
Mandelbaum et al. (2006)
Tyson, Kochanski, &
Dell’Antonio (1998)
Stacked galaxy—galaxy weak lensing
signal fit with various profiles.
12/16/2009 70
Progress in Gravitational Lensing
• Strong lensing arclets
• Weak lensing
• Flexion
“Weak Lensing”
Distorsion of galaxy shapes by foreground matter
without lensing
Lensing effect
Weak Lensing mass reconstruction
Image ellipticity -> shear->
invert the equation
RXJ1347.5-1145
(Bradac et al 2005)
Galaxy-scale DM density profile
Generalized NFW model → Dark Matter mass
Sensitivity of detection scales by lensing
Weak lensing:
Flexion:
Strong lensing:
< 100
kpc
10-100 kpc
1-10 kpc
Surface density profile measurements
obtained from galaxy groups
in the COSMOS survey
Leauthaud et al. 2010
Dec
2012
Galaxy-galaxy lensing
Measure the correlation of shear of the background galaxies
with mass of the foreground galaxies
To achieve the galaxy-galaxy lensing signal, we need two important
ingredients that we can extract from the data
1)redshift distribution of the lensed background galaxies
2)shape of the lensed background galaxies
Future Measurements of
DM properties with lensing
From 100 sq deg scale at CFHT
to 5000 – 20000 sq deg sky surveys
KDUST?
WFIRST?
BigBoss-like/MS-DESI can provide 3D
Euclid slide + new logo
Cosmic shear power spectra
Markovic et al. 2010 Euclid-like DE space survey +Planck:
Integral effects → better than matter power spectrum
Sensitive to
m_WDM < 2.5 keV
keV WDM effect around k=10 h/Mpc
Issues
•
Galaxy evolution alters DM halos and the matter power
spectrum .
Rudd, Zentner & Kravtsov, Effects of Baryons and Dissipation on the Matter
Power Spectrum (2008);
Pedrosa,Tissera, & Scannapieco, The joint evolution of baryons and dark matter
halos, (2010);
Scannapieco +, The Aquila Comparison Project: The Effects of Feedback and
Numerical Methods on Simulations of Galaxy Formation, arXiv:1112.0315.
•
Most of the simulations
(even today) are DM-only
- DM halos extremely sensitive to the implementation of the
galaxy physics in the codes.
- DM halo morphologies and galaxy properties need
resolutions: giant molecular cloud (GMC) sized regions .
But a lot of concern/work
in the last 3 years.
N-Body simulations with baryons
Jing Y. (2005)
More recent comparisons of WDM and CDM simulations.
eg Gao+, Jing+ , Yepes+ ,
- Non-linear collapse of WDM structures
Caveat: Strong Reliance on N-body
simulations
might be misleading!
Baryon physics (eg.,AGN feedback)
affects Matter Power Spectrum
Semboloni+ (2011)
Van Daalen+(2011)
Shale + :OWLS simulation
 Consequences on WL
cosmological parameters fits
Baryon effects
different from neutrino effects
Semboloni et al. 2011
Dark Matter:
What do we really know?
DM: - particles that does not emit observable radiation
- interacts gravitationally…
- non baryonic
DM: we know it exists!
Or Do We Really?
Alternatives to DM?
Not so many models any more, but still… some are still doubting:
eg http://www.astro.uni-bonn.de/~pavel/kroupa_SciLogs.html
Famaey & Mc Gaugh Living Reviews in Relativity, vol. 15, no. 10 2012
- MOND- Milgrom /TEVES-Beckenstein needs neutrinos to explain
Bullet Cluster…
- MOG : Moffat and collaborators
Scalar-Tensor-Vector Model of gravity : “few parameters can explain
away DE and DM”.
Main observational argument for
alternative to DM
Local Universe:
- Velocity analysis
 local density ~ 0.07-0.08
- Dwarf galaxies: observed ones seem to be
Tidal dwarf galaxies not expected to be dominant with DM
models but seem to be observationnally in disk of Milky Way
and Andromeda
How representative is it?
Universe with Torsion
- Extension to GR:
in simplest CARTAN model :
(eg, Schucker and Tilquin)
Lambda/DE still needed but… DM reduced (to zero?)
- Difficulties with many extensions
eg Gauss theorem not valid, pathologies…
Summary: What do we know about DM?
• Astrophysical observations
 existence of non baryonic Dark Matter
• N-Body simulations and Observations of LSS
 existence of not-hot DM?
. Many problems with CDM simulations can be solved with
O(1keV) WDM or Baryon physics ?
• More work on baryonic N-body simulations needed!
We love CDM
but need to find CDM in
accelerators and DD/ID experiments!
A mysterious Dark Universe !
What we know is only
4-5 %
of the energy density of
the Universe
We now measure
with precision
the extent of
our ignorance !
Graph source: Wikipedia
cảm ơn bạn
Thank you
谢谢
Towards a large South Pole Dome A
Kunlun Dark Universe Survey Telescope (KDUST)
 Multiprobe measurements (SNIa, Weak Lensing, BAO, Clusters) for
cosmology and ancillary science
 First stage 2011-2015: 3 x 67 cms telescopes
(AST3)
- one AST3 installed in Dome A in fall 2011,
 THCA contributes to one AST3 and take responsibility for SN
search (need computing capability)
- Collaboration with Australia, US and France
2.5 m KPATH (Kunlun Pathfinder): 2013(?)-2017
 Larger (> 4m) KDUST:
Timescale too early to define!
Antarctica Schmidt Telescopes (AST3)
Aperture：67 cm；
FOV：4.2°；
Wave Band：400nm-900nm ( i,g, r, or IR? filter for 3 telescopes );
Scale：1 arcsec/pixel;
Image quality：80％energy encircled in one pixel；
CCD: 9micron /pixel, 10580x10560 (95.22mm x 95.05mm image area)；
Type: STA1600；
 Working mode: frame transfer readout
 Focal length: 1867mm
 Distorsion in the whole field: 0.012% (less than 1 pixel)
 Total optical length: 2.2m
First AST3 in Dome A,
some data in 2012
Summer 2011 in Xuyu
Dec 2011 in Dome A
The Kunlun Dark Universe Survey Telescope
5000 sq deg down to mag 29