Ben and Jerrys Dark Matter: Well Beyond Vanilla Cold Dark Matter and even beyond chocolate fudge

1
Ben and Jerrys Dark Matter(Well) Beyond
Vanilla Cold Dark Matter (and even beyond
chocolate fudge)

• Marc Kamionkowski
• Caltech
• Columbus, Ohio
• 7 January 2005

2
Plan

• Review of standard weakly interacting massive
  particle scenario
• (Possible) problems with CDM on small scales
• Self-interacting dark matter
• WIMPs from charged-particle decay (Sigurdson, MK
  Sigurdson, Caldwell, Doran, Kurylov, MK)
• How dark is dark? Dark-matter dipole moments

3
What do we know?

• Compelling cosmological evidence that nonbaryonic
  (non SM) dark matter exists.
• .
• Dark matter must be dark matter.
• But empirically, know little else.

4
Good news cosmologists don't need to "invent"
new particle

• Axions
• ma10-(3-6) eV
• arises in Peccei-Quinn
• solution to strong-CP
• problem

• Weakly Interacting Massive Particles (WIMPS).
  e.g.,neutralinos

(e.g., Raffelt 1990 Turner 1990)
(e.g., Jungman, MK, Griest 1996)
5
WIMPs

• Relic Density ??h2( 3x10-26 cm3/sec / ????sm)

Prospects for detection
direct
Neutrinos from sun/earth
Detection
indirect
anomalous cosmic rays
WIMP candidate motivated by SUSY
Lightest Neutralino, LSP in MSSM
6
Typical WIMP-WIMP elastic scattering cross
section 10-40 cm2 and mass 10-1000 GeV for halo
density GeV/cm3 and velocity 300 km/sec,
mean-free time for WIMP scattering is at least
1013/H0 thus, WIMPs act as collision-free dark
matter. Axion-axion cross section far
smaller, so also collisionless.
7
Problem 1 Halo cusps
N-body simulations show "cusp", ????r, for
small r for collisionless halos (Navarro,
Frenk, White 1996 Moore
et al. 1997) however, rotation curves for
(at least some, maybe most) galaxies show
dark-matter cores.
8
Problem 2 Halo substructure
N-body simulations show more than 10 times as
many dwarf galaxies in typical galactic halo
than are observed in Milky Way (Moore et al.
1999 Klypin et al. 1999)
9

Cluster
galactic halo
300 kpc
10
The self-interacting dark matter solution
(Spergel Steinhardt 1999)
Hypothesize that dark matter can elastically
scatter from itself Small self-interaction leads
to energy transport that reduces sharp
subgalactic features like cusp and
substructure.
11
Required cross section is 10-(21-24) cm2
(Mx/GeV), or self-opacity 0.5x(1-10) cm2/g
(Spergel Steinhardt 2000 Dave et al.
2000). Weaker interaction doesn't work
larger interaction leads to halo core collapse
on Hubble time (e.g., Moore et al. 2000, 2002
Yoshida et al. 2002 Burkert 2000 Kochanek
White 2000)
12

• "Particle physics" reaction
• Required elastic-scattering cross section at
  least 13 orders of magnitude than allowed for
  WIMP
• problem even worse with axions.
• More generally, for point particles, required
  cross section violates unitarity for MxGeV (MK
  Griest 1990)
• SIDM must be extended object . But what is it
  (e.g. Q-balls, Wimpzillas)? How does it get
  produced? Why does it have density
  comparable to critical density?

13
May be astrophysical solutions
to dwarf-galaxy dearth (e.g., Benson
et al. 2002 Bullock, Kravtsov, Weinberg
2000 Stoehr et al. 2002)
14
Another possible resolution Power suppression
on small scales from inflation with broken scale
invarianceMKLiddle, PRL 84, 4525
(2000)Yokoyama, PRD, 2000
I
V(?)
Inflaton ?
15
ad hoc
BSI
16
What about cusps? May be astrophysical solutions
to this problem As well (e.g., Binney, Dehnen,
Silk Katz, Weinberg Sellwood,
Milosavljevic)
17
Astrophysical Tests of SIDM
Need to account for existence of DM cores
in DM-dominated dwarf galaxies typical core
radii r2 kpc and velocity dispersions v50
km/sec. Proposed velocity dependences for
scattering cross section are either constant in
velocity or 1/v. If constant-velocity, then core
radii should scale as r?v3/2. So, for cluster
with v1000 km/sec, core radius should be 100
kpc.
18
Miralda-Escude 2000 Misalignment of lens arcs in
cluster MS2137-23 suggests elliptical (as opposed
to spherical expected for SIDM) core Constrains cross section arguments (based on size as well as shape of
cluster cores, as well as lensing statistics)
from Yoshida et al. 2000, Meneghetti et al.
2001, Dahle et al. 2002 find cm2/g. However, is consistent if cross section
???vk.
19
However, if cross section ???vk, then r ?
v(3-k)/2 and ??v(k-1)/2 (Miralda-Escude). X-r
ay observations of giant elliptical NGC 4636
find very dense dark halo with profile r-1.2,
with very stringent constraints to core.
Eliminates 1/v cross section
(Loewenstein Mushotzky 2002)
20
Lesson from SIDM
Clever observations and arguments can constrain
interactions of dark-matter particles
21
WIMPs from Charged-Particle Decay

• Charged particles couple to baryon-photon fluid,
  have pressure, so growth of structure
  suppressed.Growth of modes that enter horizon
  while dark matter is charged is suppressedIf
  charged particle has lifetime 3.5 yr, power on
  

22
Effect of Charged NLDP?
k 30 Mpc-1 3 Mpc-1
0.3 Mpc-1
Dark Matter (Standard Case) Dark Matter
(w/Charged NLDP) Charged Matter (BaryonsNLDP)
23

f? fraction of DM that is initially charged
If

• KS and Marc Kamionkowski
• Phys. Rev. Lett. 92, 171302 (2004)
  astro-ph/0311486

24
Small Scale Structure Problem

• Can solve this problem with charged-decay for
  lifetimes of order years.
• Long lifetime. Weak coupling?
• Measurements of small-scale P(k) can lead to
  cosmologically interesting lifetimes.

SuperWIMPS J. Feng et al. (2003)
25

Can charged-particle decay mimic Running of
spectral index?
(Profumo, Sigurdson, Ullio, MK, astro-ph/0410714)

Suppression by a factor in the
linear power spectrum.
26

• Simple suppression by a factor
• in the linear power spectrum for k-modes that
  enter the horizon prior to decay.
• But for small-scale (large k) modes we must
  consider the nonlinear gravitational evolution of
  the power spectrum. We can only measure the
  nonlinear power spectrum.

27
Nonlinear Evolution

Charged-particle decay can mimic running of
spectral index
28

Particle Theory Models?

• Are there any viable particle theory models that
  can give this behavior? Yes! Supersymmetric Dark
  Matter Models with the LDPneutralino and
  NLDPstau.
• The stau coannihilation region ( )
  of the MSSM can give the observed WMAP relic
  abundance.
• If The stau is long-lived because it
  decays via 4-body processes

29

21-cm Fluctuations

• Measurement of linear-regime P(k) with 21-cm
  spin-flip transition during the Cosmic Dark
  Ages, at redshifts z30-200 may discriminate
  between running of spectral index, and
  charged-particle decay

30
How Dark is Dark? How weak must coupling of DM
to photon be?

• Charge? No.
• A. Gould et al. (1990)
• Millicharge?
• S. L. Dubovsky et al. (2004)
• S. Davidson et al. (2000)
• What about a neutral particle with magnetic or
  electric dipole moments?
• Kris Sigurdson, Michael Doran, Andriy Kurylov,
  Robert R. Caldwell, Marc Kamionkowski
  Phys. Rev. D70 (2004) 083501 astro-ph/0406355

31
Effective Interaction

• The effective interaction Lagrangian
• In the nonrelativistic limit

32
Constraints From

• Cosmological Relic Abundance
• Direct Detection
• Cosmology (CMB and LSS)
• Precision Standard Model
• Production at Accelerators
• Gamma Rays

33
Relic Abundance

Standard Cosmological Freeze-out Calculation
34
Constraints
35
Direct Detection

CDMS (Soudan)
36
Constraints
37
Direct Detection

But if the dipole strength is too large dipolar
dark matter (DDM) will scatter in the
atmosphere and the rock above the detector and
arrive at the detector with an energy below the
detection threshold.
Strongest constraints from shallowest experiment
with a null result. Balloon and Rocket
experiments.
38
Constraints
39
Effects on the Matter Power Spectrum

40
Effects on the CMB

41
Constraints
42
Precision Standard Model

Muon g-2
Standard Model EDMs
Z-Pole
43
Precision Standard Model

Strongest Constraint
44
Constraints
45
Production at Accelerators

• B and K decays
• LEP, Tevatron? Tricky.

Look for missing energy
Need the full theory not the effective theory
46
Gamma Rays

• Annihilation at the Galactic center could produce
  a nearly monoenergetic line.
• EGRET Constraints
• Possible GLAST signal

47
Constraints
48
Dipolar Dark Matter?

• Dipolar Dark Matter A phenomenologically viable
  dark-matter candidate with a mass between an MeV
  and a GeV and predominantly dipole interactions.

49
Summary

• Self-interacting dark matter more tightly
  constrained than one might have thought
• Chunky Monkey dark matter may account for
  dwarf-galaxy dearth
• Dulce de Leche may mimic running of spectral
  index
• Frappucino may have tiny coupling to photon
• Plenty of other flavors

50
my favorite is still chocolate
51
Small Scale Structure Problem

• The theory of Cold Dark Matter predicts of order
  100 small structures for a halo in a galaxy the
  size of the Milky Way. Compare with the 11
  observed dwarf galaxies.
• Astrophysical mechanisms?
• A. J. Benson et al. (2002), R. S. Somerville
  (2002), L. Verde et al. (2002)
• Cutting off power on small scales reduces number
  of small structures.
• M. Kamionkowski and A. Liddle (2000)

52
Standard CDM
Power suppressed on small scales
Turnaround
Today
Moore et al, astro-ph/9903164
53
What do we not know? Disclaimer Not an
exhaustive list!

• Which theory describes dark matter and the rest
  of the dark sector (if it exists).
• The interactions of the LDP, NLDP, NNLDP
• In particular, the quantitative question
• How dark is dark?

How strongly does the dark sector couple to the
photon?
54
Other Phenomenological Constraints

• Larger effort to constrain interactions
• For Example
• Strongly-Interacting Dark Matter
• G. D. Starkman et al. (1990)
• Self-Interacting Dark Matter
• E. D. Carson et al. (1992)
• D. N. Spergel and P. J. Steinhardt (2000)

55
The Charged NLDP Case

• Decay shifts fraction of energy density from
  charged to neutral.
• Dark Matter Gravity, but overdensity now also
  sourced by decay.
• Photon-Baryon-NLDP Fluid Charged particles
  (including NLDP) coupled by Coulomb scattering.
  Photons coupled to electrons via Compton
  scattering. Gravity (from dark matter) vs.
  Radiation Pressure (from photons) yields acoustic
  oscillations.

56
Effect of Charged NLDP?

• Before Decay The NDLP couples to the
  photon-baryon fluid! NLDP perturbation modes
  that enter the horizon oscillate rather than
  grow. These modes source the Dark-Matter modes
  and thus suppress growth of dark-matter
  perturbations.
• After Decay Dark-Matter modes that enter the
  horizon grow under the influence of gravity, as
  in the standard case.

57
Small Scale Structure Problem

M. White and R. A. C. Croft (2000)
58
A Running Spectral index
Primordial Power Spectrum
Nearly Scale Invariant
Running Index
59
A Running Spectral index
60
A Running Spectral index

• Possible large running in analysis of WMAP
  Lyman- forest Data.
• Question If we detect a large running is that
  telling us something is wrong with simple models
  of inflation?
• Or Is there another effect that can mimic this
  behavior?

61
Mimic running?

Yes!
z4

• Stefano Profumo, KS, Piero Ullio and Marc
  Kamionkowski astro-ph/0410714

62

Particle Theory Models?
Only the stau can play the role of a quasistable
NLDP (with cosmologically interesting
lifetimes) in the MSSM.
63

The Charged Fraction
In the simplest case the charged fraction is
However, if there are other particles that are
quasi-degenerate that coannhilate the
stau-neutralino system which decouple from the
neutralino later than the stau then the late-time
charged fraction is
Where the sum is over all such all such
coannhilating particles.
64

Benchmark Scenarios
Bino-like neutralinos in mSUGRA Stau-neutralino
Higgsino-like neutralinos in mSUGRA Stau-neutralin
o w/ neutralino, chargino
Wino-like neutralinos in mAMSB Stau-neutralino w/
chargino
Bino-like neutralinos in mSUGRA Stau-neutralino
w/ selectron, smuon
65

Direct Detection?
Possible Direct Detection Signals! Indirect
Detection Unlikely.
66

LHC?
LHC? Yes. Bino-like may have m No. Higgisino-like, Wino-like
67

Stau Production

• If neutralinos are produced, will also produce
  staus!
• Might detect staus as highly penetrating particle
  in detectors.
• Could also trap in water tanks outside of the
  detector and observer stau decays. 10kton water
  tank might detect 10 staus/year.
• For example, Jonathan Feng and Brian Smith
  hep-ph/0409278

68

21-cm Fluctuations
From Loeb and Zaldarriaga 2004.