Alexander Kusenko (UCLA) Q2C Discovering dark matter in space · Alexander Kusenko (UCLA) Q2C...

Alexander Kusenko (UCLA) Q2C

Discovering dark matter in space

• The evidence for dark matter

• The candidates, and the detection strategies

• The opportunities in space

• Sterile neutrinos as dark matter:

1


Dark matter

The evidence for dark matter is very strong:

• galactic rotation curves cannot be explained by the disk alone

• cosmic microwave background radiation

• gravitational lensing of background galaxies by clusters is so strong thatit requires a significant dark matter component.

• clusters are filled with hot X-ray emitting intergalactic gas; some(merging) clusters show displacement of dark and baryonic matter

2


Galactic rotation curves

3


Cosmic microwave background radiation (CMBR)

At redshift zdec = 1089 ± 1, the atoms formed and the universe becametransparent to radiation. Radiation emitted at that time, tdec = (379 ±8) kyr, has been red-shifted into the microwave range. Fluctuations havebeen measured first by COBE, and later by BOOMERANG, MAXIMA, ...,WMAP:

4


Cosmic microwave background radiation (CMBR)

At redshift zdec = 1089 ± 1, the atoms formed and the universe becametransparent to radiation. Radiation emitted at that time, tdec = (379 ±8) kyr, has been red-shifted into the microwave range. Fluctuations havebeen measured first by COBE, and later by BOOMERANG, MAXIMA, ...,WMAP:

5


These fluctuations can be represented in the form of a power spectrum Cl.First, one expands in spherical harmonics:

∆T

T=∑

l,m

almYlm(θ, φ)

And then one plots

Cl ≡1

2l + 1

l∑

m=−l

|alm|2

6


Power spectrum measured by WMAP

7


Power spectrum measured by WMAP

shows angular sizeof the horizon at the time of recombination

densities

sensitive to the ratio ofordinary to dark matter

WMAP: Ωmatter = 0.234 ± 0.035; Ωb = 0.0446 ± 0.0016

8


Gravitational lensing: seeing the invisible

9


Foreground cluster CL0024+1654 produces multiple images of a bluebackground galaxy in the HST image (left). Mass reconstruction (right).

10


None of the known particles can be dark matter

11


None of the known particles can be dark matter

12


Dark matter ⇒ new physics

13


Dark matter: what is it?

Can take guesses based on...

• ...compelling theoretical ideas

• ...simplicity

• ...observational clues

Dark matter properties determine the detection strategy

14


Dark matter: compelling theoretical ideas

SUSY is an appealing theoretical idea

Dark matter comes as part of the package as one of the following:

• Lightest supersymmetric particle, stable because of R-parity

– neutralino

– gravitino

– axino

• SUSY Q-balls, stable for gauge-mediated SUSY breaking, thanks to the baryon

number and energy conservation.

Theoretically motivated! Mass vs cross section OK, even ”natural”.

By no means minimal. No experimental evidence so far, but the search isunder way.

15


SUSY

WIMP

Direct detection

Weakly interacting particles (WIMP)can penetrate through rock; they can bedetected in underground detectors, such asZeplin and others, screened by the rockfrom cosmic rays (unwanted background)

16


Indirect detection of WIMPs

Annihilations of WIMPs in the center of MW galaxy can be detected bygamma ray and x-ray telescopes in space

17


Axion

�

�

�

�Strong CP problem: QCD vacuum is a superposition |θ〉 =

∑

n exp{−inθ}|n〉of topologically distinct vacuum states |n〉.

LQCD = Lpert + θg2

32π2F F̃

θ̄ = θ + arg det M

Experiment: |θ̄| ≪ 10−10 !

18


Possible solution: Peccei-Quinn symmetry

An elegant solution: θ̄ relaxes to zero dynamically, by a VEV of the axion .

Additional U(1) symmetry, Peccei-Quinn symmetry is spontaneouslybroken by instantons ⇒ axion has small mass.

Axion is a weakly interacting particle ⇒ dark matter

19


Axion dark matter

20


Axion dark matter

Various searches:

• axion-photon conversion in magnetic field

• detection of solar axions

• long-range forces; possible advantage in space: drag-free environment

21


Sterile dark matter: a simple (minimalist) solution

Need one particle ⇒ add just one particle

If a fermion, must be gauge singlet (anomalies)

Interactions only through mixing with neutrinos

⇒ sterile neutrino

Small mass and, therefore, stability! No symmetries required.

22


Sterile neutrinosThe name ”sterile” was coined by BrunoPontecorvo in a paper [JETP, 53, 1717 (1967)],which also discussed

• lepton number violation

• neutrinoless double beta decay

• rare processes (e.g. µ → eγ)

• vacuum neutrino oscillations

• detection of neutrino oscillations

• astrophysical neutrino oscillations

23


Pontecorvo: neutrino oscillations can”convert potentially active particles into

particles that are, from the point of view

of ordinary weak interactions, sterile, i.e.practically unobservable, since they have

the ”incorrect” helicity” [JETP, 53, 1717(1967)]

24


Sterile neutrinos with a small mixing to active neutrinos

{

|ν1〉 = cos θ|νe〉 − sin θ|νs〉|ν2〉 = sin θ|νe〉 + cos θ|νs〉 (1)

The almost-sterile neutrino, |ν2〉 was never in equilibrium. Production ofν2 could take place through oscillations.

The coupling of ν2 to weak currents is also suppressed, and σ ∝ sin2 θ.The probability of νe → νs conversion in presence of matter is

〈Pm〉 =1

2

[

1 +

(

λosc

2λs

)2]−1

sin2 2θm, (2)

where λosc is the oscillation length, and λs is the scattering length.

25


Sterile neutrinos in the early universe

Sterile neutrinos are produced in primordial plasma through

• off-resonance oscillations. [Dodelson, Widrow; Abazajian, Fuller; Dolgov,Hansen; Shaposhnikov et al.]

• oscillations on resonance, if the lepton asymmetry is non-negligible[Fuller, Shi]

26


Mixing is suppressed at high temperature [Dolgov, Barbiieri; Stodolsky]

sin2 2θm =(∆m2/2p)2 sin2 2θ

(∆m2/2p)2 sin2 2θ + (∆m2/2p cos 2θ − V (T ))2, (3)

For small angles,

sin 2θm ≈sin 2θ

1 + 0.79 × 10−13(T/MeV)6(keV2/∆m2)(4)

Production of sterile neutrinos peaks at temperature

Tmax = 130 MeV

(

∆m2

keV2

)1/6

27


The resulting density of relicsterile neutrinos in conventionalcosmology, in the absence of alarge lepton asymmetry:

Ων2 ∼ 0.3(

sin2 2θ

10−8

)

(

ms

keV

)2

1e-11 1e-10 1e-09 1e-08 1e-07

sin2θ

1

10

m s [

keV

]

Ω > 0.3

Ω = 0.3

s

s

dark matter

[Dodelson, Widrow; Dolgov, Hansen; Fuller, Shi; Abazajian, Fuller, Patel]

28


Lyman-α forest: a look at the small-scale structure

29


The resulting density of relicsterile neutrinos in conventionalcosmology, in the absence of alarge lepton asymmetry:

Ων2 ∼ 0.3(

sin2 2θ

10−8

)

(

ms

keV

)2

Lyman-α forest clouds showsignificant structure on smallscales. Dark matter must be coldenough to preserve this structure.

1e-11 1e-10 1e-09 1e-08 1e-07

sin2θ

1

10

m s [

keV

]

Ω > 0.3

Ω = 0.3

s

s

dark matter

too warm

Seljak et al.

Viel et al.

30


Cold or warm dark matter?

CDM works well, but...

There are problem problems with cold dark matter on small scales

31


Some CDM problems eliminated by WDM• overproduction (by an order of magnitude!) of the satellite halos for galaxies of the

size of Milky Way.

• WDM can reduce the number of halos in low-density voids. [Peebles]

• observed densities of the galactic cores (from the rotation curves) are lower than whatis predicted based on the ΛCDM power spectrum. [Dalcanton et al.; van den Bosch etal.; Moore; Abazajian]

• The “angular-momentum problem”: in CDM halos, gas should cool at very early timesinto small halos and lead to massive low-angular-momentum gas cores in galaxies.[Dolgov]

• disk-dominated (pure-disk) galaxies are observed, but not produced in CDM becauseof high merger rate. [Governato et al.; Kormendy et al.]

• observations of dwarf spheroidal galaxies ⇒ m ∼ keV [Gilmore et al.]

32


Radiative decay

Sterile neutrino in the mass range of interest have lifetimes longer thanthe age of the universe, but they do decay:

ν2 W+ ν1

l -l -

γ

ν2 l - ν1

W+W+

γ

Photons have energies m/2: X-rays. Large lumps of dark matter emit someX-rays. [Abazajian, Fuller, Tucker; Dolgov, Hansen; Shaposhnikov et al.]

33


X-ray observations

Virgo cluster image from XMM-Newton

34


Chandra, XMM-Newton can see photons: νs → νeγ

[Abazajian et al; Hansen et al.; Boyarsky et al.; Watson et al.]

35


Chandra, XMM-Newton can see photons: νs → νeγ

dSphs (favored)

[Abazajian et al; Hansen et al.; Boyarsky et al.; Watson et al.]

36


Emission of sterile neutrinos from a supernova

• Sterile neutrino emission from a supernova is anisotropic

• Sterile neutrinos with masses and mixing angles consistent with darkmatter can explain the pulsar velocities

[AK, Segrè; Fuller, AK, Mocioiu, Pascoli]

37


The pulsar velocities.

Pulsars have large velocities, 〈v〉 ≈ 250 − 450 km/s.[Cordes et al.; Hansen, Phinney; Kulkarni et al.; Lyne et al. ]

A significant population with v > 700 km/s,about 15 % have v > 1000 km/s, up to 1600 km/s.[Arzoumanian et al.; Thorsett et al. ]

38


A very fast pulsar in Guitar Nebula

HST, December 2001HST, December 1994

39


Map of pulsar velocities

40


Proposed explanations:

• asymmetric collapse [Shklovskii] (small kick)

• evolution of close binaries [Gott, Gunn, Ostriker] (not enough)

• acceleration by EM radiation [Harrison, Tademaru] (kick small, predictedpolarization not observed)

• asymmetry in EW processes that produce neutrinos [Chugai; Dorofeev,Rodinov, Ternov] (asymmetry washed out)

• “cumulative” parity violation [Lai, Qian; Janka] (it’s not cumulative )

41


Asymmetric collapse

“...the most extreme asymmetric collapsesdo not produce final neutron star velocities above 200km/s” [Fryer ’03]

42


Supernova neutrinos

Nuclear reactions in stars lead to a formation of a heavy iron core. Whenit reaches M ≈ 1.4M⊙, the pressure can no longer support gravity. ⇒collapse.

Energy released:

∆E ∼ GNM2Fe core

R∼ 1053erg

99% of this energy is emitted in neutrinos

43


Pulsar kicks from neutrino emission?

Pulsar with v ∼ 500 km/s has momentum

M⊙v ∼ 1041 g cm/sSN energy released: 1053 erg ⇒ in neutrinos. Thus, the total neutrinomomentum is

Pν; total ∼ 1043 g cm/s�

�

�

�a 1% asymmetry in the distribution of neutrinos

is sufficient to explain the pulsar kick velocities

But what can cause the asymmetry??

44


Magnetic field?

Neutron stars have large magnetic fields. A typical pulsar has surfacemagnetic field B ∼ 1012 − 1013 G.Recent discovery of soft gamma repeaters and their identification asmagnetars

⇒ some neutron stars have surface magnetic fields as high as1015 − 1016 G.

⇒ magnetic fields inside can be 1015 − 1016 G.Neutrino magnetic moments are negligible, but the scattering of neutrinosoff polarized electrons and nucleons is affected by the magnetic field.

45


Core collapse supernova

Onset of the collapse: t = 0

Fe

core

46



Shock formation and “neutronization burst”: t = 1 − 10 ms

PNSburst

shock

ν

Protoneutron star formed. Neutrinos are trapped. The shock wave breaksup nuclei, and the initial neutrino come out (a few %).

47



Thermal cooling: t = 10 − 15 s

PNSthermal

ν

Most of the neutrinos emitted during the cooling stage.

48


Electroweak processes producing neutrinos (urca),

p + e− ⇀↽ n + νe and n + e+ ⇀↽ p + ν̄e

have an asymmetry in the production cross section, depending on the spinorientation.

σ(↑ e−, ↑ ν) 6= σ(↑ e−, ↓ ν)The asymmetry:

ǫ̃ =g2

V− g2

A

g2V

+ 3g2A

k0 ≈ 0.4 k0,

where k0 is the fraction of electrons in the lowest Landau level.

49


In a strong magnetic field,

20 30 40 50 600

0.1

0.2

0.3

0.4

0.5

0.6

0.7

B=10 G16

B=3x10 G15

0K

20 30 40 50 60

µ, MeV

B=3x10 G16

k0 is the fraction of electrons in the lowest Landau level.

Pulsar kicks from the asymmetric production of neutrinos?[Chugai; Dorofeev, Rodionov, Ternov]

50


Can the weak interactions asymmetry cause an

anisotropy in the flux of neutrinos due to a large

magnetic field?

No

eνeνeν

eνeν

eν

eν

eν

eν

eνeν

eν

eνeν e

ν

Neutrinos are trapped at high density.

51


Can the weak interactions asymmetry cause an

anisotropy in the flux of neutrinos due to a large

magnetic field?

NoRescattering washes out the asymmetry

In approximate thermal equilibrium the asymmetries in scattering amplitudesdo not lead to an anisotropic emission. Only the outer regions, nearneutrinospheres, contribute (a negligible amount). [Vilenkin,AK, Segrè]

However, if a weaker-interacting sterile neutrino was produced inthese processes, the asymmetry would, indeed, result in a pulsarkick!

52


Sterile neutrinos leave the star without scattering. Hence, they give thepulsar a kick.

eνeν

νsνs

νs

eν

νs

eν

νs

eν

B

53


Allowed range of parameters (time scales, fraction of total energy emitted):

1e-11 1e-10 1e-09 1e-08 1e-07

sin2θ

1

10

m s [

keV

]

��

��

Ω = 0.3

Ω > 0.3

ν

ν

dark matter

pulsar kick

oscillations)(off−resonance

[Fuller, AK, Mocioiu, Pascoli]

54


Resonant active-sterile neutrino conversions in matter

Matter potential:

V (νs) = 0

V (νe) = −V (ν̄e) = V0 (3Ye − 1 + 4 Yνe)

V (νµ,τ) = −V (ν̄µ,τ) = V0 (Ye − 1 + 2 Yνe) + cZ

L

~k · ~Bk

cZ

L=

eGF√2

(

3Ne

π4

)1/3

[D’Olivo, Nieves, Pal]

55


The magnetic field shifts the position of the resonance because of the~k· ~B

kterm in the potential:

µν

νs

νs

νs

νs

νµµν µν

µν

µν

In the absence of magnetic field, νs escape isotropically

56


The magnetic field shifts the position of the resonance because of the~k· ~B

kterm in the potential:

µν

νsνs

νs

νs

νsνs

µνµν

µν eν

µν

µνµν

B

Down going neutrinos have higher energies

of the

57


The range of parameters (resonance, adiabaticity, weak damping):

1e-11 1e-10 1e-09 1e-08 1e-07

sin2θ

1

10

m s [

keV

]

��

��

Ω = 0.3

Ω > 0.3

ν

ν

pulsar kick(resonant oscillations)

dark matter

2

58


Resonance & off-resonance oscillations

1e-11 1e-10 1e-09 1e-08 1e-07

sin2θ

1

10

m s [

keV

]

��

��

2

Ω > 0.3s

1

3

��

��pulsar kick

pulsar kick and dark matter (L=0)

too warm

[ A.K., Segrè, PL B396, 197 (1997); Fuller, A.K.,Mocioiu,Pascoli, PR D 68, 103002 (2003)]

59


Other predictions of the pulsar kick mechanism

• Stronger supernova shock [Fryer, AK, ApJ, in press; astro-ph/0512033]

iii

iii

ConvectionRegion

60




61




• No B − v correlation is expected because

– the magnetic field inside a hot neutron star during the first ten seconds is verydifferent from the surface magnetic field of a cold pulsar

– rotation washes out the x, y components

• Directional ~Ω − ~v correlation is expected, because

– the direction of rotation remains unchanged– only the z-component survives

B

62


Reionization

WMAP, three years of data, reionization redshift: zr = 10.9+2.7−2.3.

(This improves the one-year WMAP result, zr = 17 ± 5.)

Observations of distant quasars: reionization must be completed by z = 6.

First stars can ionize gas, but can they form so early?

WMAP 3 yrs ⇒ new challenge: can one end reionization by z = 6 withoutexceeding the optical depth τ

WMAP= 0.10 ± 0.03?

Small halos collapse first and start ionizing gas. If reionization is to becompleted by z = 6, small halos shine to early, too bright, and exceedτ

WMAP.

63


Need suppression of star formation in small halos by an order ofmagnitude: [Haiman, Bryan, astro-ph/0603541]

Warm dark matter (gravitinos) could suppress small structure and butthey would also delay the star formation.

What about sterile neutrinos?

• they are warm ⇒ small halos suppressed

• they decay and produce x-rays, and x-rays can ionize gas!

64


Photons from radiative decays

Sterile neutrino in the mass range of interest have lifetimes longer thanthe age of the universe, but they do decay:

ν2 W+ ν1

l -l -

γ

ν2 l - ν1

W+W+

γ

Photons have energies m/2: X-rays. X-rays can ionize gas.

65


Sterile neutrino decays: an increase in ionization fraction

10 20 50 100 200 500 1000

0.001

0.01

0.1

1

xe

z

The ions too few to explain the WMAP results [Ferrara, Mapelli]...

...but it’s a much higher fraction than in the absence of sterileneutrinos. Ionization catalyzes formation of molecular hydrogen [AK;P.L. Biermann]...

production of molecular hydrogen speeds up gas cooling, halocollapse and star formation

66


[Tegmark, et al., ApJ 474, 1 (1997) ]

67


Molecular hydrogen

H + H → H2 + γ − very slow!In the presence of ions the following reactions are faster:

H+ + H → H+2 + γ,H+2 + H → H2 + H+.

H+ catalyze the formation of molecular hydrogen![Biermann, AK, PRL 96, 091301 (2006)]

68


The fraction of molecularhydrogen f

ḟ ≈ km(t) nH(t) xe(t),

where km is the rate shown

0 200 400 600 800 10000

0.5

1

1.5

2

z

mk , 1

0 cm

/s

−16

3

End result: H2 production is enhanced at z ∼ 100Sterile neutrino decays can precipitate the early star formation. Stars canreionize the universe by redshif zr = 11 ± 3 (WMAP) while avoiding theminihalo problem

69


Clues of sterile neutrinos

The sterile neutrinos can explain:

• the light neutrino masses

• dark matter

• baryon asymmetry via leptogenesis [ Asaka, Shaposhnikov; Akhmedovet al. ]

• pulsar kicks

• reionization

70


Clues of sterile neutrinos

Depending, of course, on how far down it goes.This could be the greatest discovery of the century.

71


Summary

• Dark matter is out there and needs to be discovered!

• NASA can make the discovery in space using x-ray and gamma-ray telescopes

• A sterile neutrino with keV mass is a viable dark matter candidate, free of CDMproblems, consistent with dSphs observations.

• The same neutrino is emitted from a supernova with a sufficient anisotropy to explainthe pulsar velocities

• The same neutrino can boost the production of molecular hydrogen and precipitate arapid early star formation.

• A rather minimal extension of the Standard Model, the addition of three sterileneutrinos explains all the present data, including dark matter, the baryonasymmetry of the universe, the pulsar velocities, and reionization

72

Alexander Kusenko (UCLA) Q2C Discovering dark matter in space · Alexander Kusenko (UCLA) Q2C...

Documents

Transcript of Alexander Kusenko (UCLA) Q2C Discovering dark matter in space · Alexander Kusenko (UCLA) Q2C...