oleg ruchayskiy and alexey boyarsky- cosmological bounds on warm dark matter and astrophysical...

8/3/2019 Oleg Ruchayskiy and Alexey Boyarsky- Cosmological bounds on warm dark matter and astrophysical bounds on decaying dark matter

1/42

Cosmological bounds on warm dark matterand astrophysical bounds on decaying dark

matter

Oleg Ruchayskiy Ecole Politechnique F ed erale de Lausanne

Alexey BoyarskyETH Z urich & CERN

4th Patras workshop on Axions, WIMPs, and WISPsDESY. June 18, 2008


2/42

Dark Matter in the Universe

Extensive astrophysical evidence for the presence of the dark non-baryonic matter in the Universe

Rotation curves of stars in galaxies andof galaxies in clustersDistribution of (X-ray bright) intracluster gas Gravitational lensing data

Galaxy cluster CL0024+1654 (z = 0 .39 )Courtesy of ESA-NASA

Left: Galaxy cluster CL0024+1654 as agravitational lense

Courtesy of HST

Oleg Ruchayskiy D ARK MATTER BOUNDS 1


3/42

What is known about DM?

DM is not baryonic

DM is not a SM particle (neutrinos could be but . . . )

Any DM candidate must be

Produced in the early Universe and have correct relic abundance

Very weakly interacting with electromagnetic radiation (dark)

Be stable or cosmologically long-lived

What can be the mass of DM?



4/42

What can be the DM mass?

The model-independent lower limit on mass for fermionic DM: Tremaine,Gunn (1979)Dalcanton,Hogan (1990)The smaller is the DM mass the bigger is the number of particles

in a given DM-dominated object.

For fermions there is a maximal phase-space density (degenerateFermi gas). Hence, maximal number of fermions

Objects with highest phase-space density: dwarf spheroidalgalaxies Qobs = 10 4 105 M kpc

3 [km s 1 ] 3 .

Leads to the lower bound on the DM mass m 300 500 eV

Active neutrinos with m 300 eV have primordial phase-spacedensity Q Qobs . But DMh2 = m 94 eV Active neutrinos cannotconstitute 100% of DM



5/42

Phase-space density evolution

Collisionless disipationless dynamics can only lead to the decreaseof the coarse-grained phase-space density

N-body simulations show decrease of phase-space density duringthe collapse by O (102 103 )

WIMPs with M 100 GeV and decoupling temperature T d 10 MeV have primordial phase-space density Q 1021 Qobs

Can phase-space density decrease by 21 order during thegravitational collapse? Yes!

N-body simulations measure the change of phase-space densitybetween start of simulations z 102 and today z = 0 .

Start of the simulations: particles have peculiar (Zeldovich)velocities i 10 30 km/sec.

End of the simulations: particles have virial velocities f

100 km/sec. (f / i )3

102

103

.Oleg Ruchayskiy D ARK MATTER BOUNDS 4


6/42

What is known about DM?

DM is not baryonicDM is not a SM particle (neutrinos could be but . . . )

Any DM candidate must be

Produced in the early Universe and have correct relic abundance Very weakly interacting with electromagnetic radiation (dark)

Be stable or cosmologically long-lived

For fermionic DM mass 300 eVPossible interactions with SM matter?

Pessimistic scenario: DM interacts only in the early (very) hot

Universe (e.g. produced in inaton decays)Optimistic scenario: DM interact with ordinary matter Annihilation Decay possibility of indirect detection of DM



7/42

Searching for decayingdark matter



8/42

Decaying DM

DM with radiative signatures: DM + , + , e+ + e . . .

N s

e

W

W

(a)

p k

G p

k

R

(b)

p k

G

p

k

R

Appears in many models:

Sterile neutrino Dodelson & Widrow93;Asaka, Shaposhnikov et al.05

R Gravitino Takayama & Yamaguchi00Buchm uller et al.07

Volume Modulus Quevedo07

Decaying Majoron M. Lattanzi, J.W.F. Valle 07



9/42

Properties of decaying DM

WIMPs cannot decay. Their interaction strength with matter GF would lead to life-time of neutron in -decay: n p + e + e .

Decaying DM should interact superweakly GF and 1

Radiative decay channel : DM +

Photon energy E = m DM2Life-time = 1 / life-time of the Universe

Flux from DM decay:

F DM =E

m DM

M fovDM4D 2L

fov

8 line of sight

DM( r )dr (z 1 , fov 1)

DM(r )dr is roughly equal for a large class of objectsOleg Ruchayskiy D ARK MATTER BOUNDS 8


10/42

DM decay line search : advantages

Moore et al.2005

DM decay is an all sky signal



11/42


Moore et al.2005

DM annihilation signal is concentrated on GC



12/42


Decay signal DM(r )dr as compared to 2DM(r )dr forannihilation:Decay signal is not very sensitive to the precise form of DM prole(difference between cuspy and cored proles changes signal by afactor of 2 3)

For decay signal : freedom of choosing the observational targets(many targets of different nature have comparable signals). Do notneed to look at GC

If a DM decay line candidate is found can study its surface

brightness prole and its distribution over the sky



13/42

For decaying DM " indirect "

search becomes " direct " !



14/42

Where to look for DM decay line?

Extragalactic diffusebackground

Takayama & Yamaguchi, 2000

Dolgov & Hansen, 2000Boyarsky, Neronov, O.R., Shaposhnikov, 2005Buchm uller et al., 2007Bertone et al., 2007

Clusters of galaxies(Coma, Virgo)

Abazajian et al., 2001Boyarsky, Neronov, O.R., Shaposhnikov PRD 74, 2006

DM halo of the MilkyWay.

Signal increases as weincrease FoV!

Boyarsky, Neronov, O.R., Shaposhnikov, TkachevPRL 2006 [astro-ph/0603660]Riemer-Srense et al. ApJL 2006 [astro-ph/0603661]

Boyarsky, Nevalainen, O.R. A&A 2007Abazajian et al. [PRD 75, 2007]Boyarsky, Malyshev, Neronov, O.R., MNRAS 2008

Andromeda galaxy

(M31)

Watson et al. PRD 74, 2006 [astro-ph/0605424]Boyarsky, Iakubovsky, O.R., Savachenko 2007Bertone et al. 2007

Bullet cluster 1E 0657-56 Boyarsky, Markevitch, O.R. [ApJ 2008]

Soft XRB Boyarsky, den Herder, Neronov, O.R. [Astropart.Phys.07] S t r a t e g y

d e p e n d s o n

t h e o b j e c

t , e n e r g y r a n g e ,

i n s t r u m e n

t . . .



15/42

Restrictions on life-time of decaying DM MW (HEAO-1Boyarsky et al2005

Bullet clusterBoyarsky et al2006

LMC+MW(XMBoyarsky et al2006

MW(Chandra)Riemer-Srensen etal.; Abazajianet al.

MW (XMM)

Boyarsky et al2007

M31 Watsonet al. 2006;Boyarsky et al

2007

10 25

10 26

10 27

10 28

10 29

10 0 10 1 10 2 10 3

L i f e - t

i m e [ s e c

]

mDM [keV]

XMM, HEAO-1 SPI

= Universe life-time x 10 8

Chandra



16/42

What else do we know

about the DM?



17/42

Free-streaming DM and structure formation

DM particles erase primordial spectrum ofdensity perturbations on scales up to the DMparticle horizon free-streaming length

coFS = t0 v(t )dt a(t )Comoving free-streaming lengths peaks around t nr when p m

All DM models are thus divided into 3 groups:

CDM : free streaming is negligible WDM : free streaming at galaxy scales, tnr teq HDM : free streaming at cosmological scales tnr teq

HDM (e.g. active neutrinos with the mass 1 eV) is ruled out.Gives wrong large scale structure

CDM and WDM work equally well at large scales (CMB, SDSS,2dFG). CDM model could have been called WDM



18/42

Cold or warm ?

101

10010

6

107

108

109

r (kpc)

( M k p

c 3

)

Ursa MinorDracoCarina

Sextans1/r

CDM : success at large scalesAt galaxy scale predicts:

Cuspy proles. (cores observed?) Many satellites. (few detected?)

WDM : shares success of CDM atlarge scalesQualitatively explains small scalestructure



19/42

DM at small scales. Paris, Feb.13-15

Scientic organizers: A. Boyarsky, O. Ruchayskiy, J. Lesgourgues



20/42

CDM or WDM? A summary

1) Does CDM predictions contradict observations?

CDM simulations are pure DM . Pure N-body is not enoughAstronomers observe luminous matter.Baryonic feedback can be essentialExample: not all DM halos can acquire baryons

2) Any WDM simulations (N-body or hydrodynamical) should

properly include primordial velocities of the particlesuse correct power spectrum of initial density perturbations.

3) WDM is ruled out by Lyman- ?

No (discussion follows)

4) DM with keV mass still allowed?

Yes



21/42

Lyman- forest

To probe the DM properties at smallscales one can use Lyman- forestdata:

Red-shifted absorption Lyman- line in the spectra of distant QSOsNeutral hydrogen traces DMdistribution at red-shifts z 2 4.

Allows to measure one-dimensional non-linear powerspectrum:

P 1 D =

kP 3 D (k)kdk2



22/42

Lyman- forest : challenges

Need to compare measured non-linear 1D powerspectrum with thelinear 3D power spectrum, predicted by a cosmological model.

Simulations are needed! Each hydrodynamical simulation takesabout 36 hours (optimistic)

Need to t simultaneously 7+ cosmological parameters plus 20 astrophysical Lyman- parameters to the data (Lyman- pluspossible other experiments: CMB, 2dFG, SDSS, . . . ).

Try 2 values for each parameters. Have to explore 10 8 models,perform 108 simulations which would take 550 000 years

Honest processing of Lyman- data is computationallyprohibitive



23/42

Powerspectra of DM models

Solution? Perform numerical simulations only in a number of pointsin multi-D parameter space. Interpolate between them

But simulations depend on initial (linear) powerspectrum

In many WDM models: non-thermal momenta distribution powerspectra of complicated (non-universal) form

0.01

0.1

1

0.001 0.01 0.1 1 10 100

P W D M

( k ) / P

C D M

( k )

k [h/Mpc]

Example: mixture of colderand warmer components

Suppression startsearly, at F S of warmcomponent.But at smaler scales like CDM with smallernormalisation.This makes Ly- boundsweaker.



24/42

Bayesian approach to WDM bounds

To explore effectively a vast multi-D parameter space one usesBayesian approach

Frequentist approach (model rejection based on 2 ) can rule outa particular model

Bayesian approach can tell what model is the most likely one

Bayesian approach determines the 90% condence limit (CL) asa region into which 90% of the models would fall

This leads to the bounds, which are generically weaker than those

of frequentist approach



25/42

Lower DM mass bound in Bayesian approach

0.022 0.023 0.024 0.025 b h

20.1 0.11 0.12

c h2

0.94 0.96 0.98 1 1.02ns

0.7 0.75 0.8

13.2 13.4 13.6 13.8Age/GYr

0.2 0.25 0.3

m

0.75 0.8 0.85 0.9

8

8 10 12 14 16zre

70 75 80H0

Example: bayesiananalysis gives mDM 15.3 keV at 90% CL

Fixing the parameterof interest at 50%below this lowerbound (i.e. at 10 keV)

gives t with 2

5( 2.2)

In this case Bayesian

90% CL is weakerby 50% from theactual (frequentist)one



26/42

Summary: Lyman- bounds on WDM

Lyman- analysis of WDM is model-dependent (should be repeatedfor each new type of initial power spectrum )

Bayesian 2 bounds on the WDM parameters are narrower than theactual (frequentist) ones (i.e. lower bound can be lowered, upperbound can be raised)

Lyman- analysis suffers from a systematics which is hard toestimate (due to approximate ways of converting measured topredicted power spectra)

Upshot: Lyman- allows to probe mildly non-linear stage of

structure formation and can be very useful in determining the natureof DM. However, much more work should be done, until robustbounds are obtained



27/42

Sterile neutrinos:

decaying, (warm?) DMand much more



28/42

MSM: all masses below electroweak scale

Just add 3 right-handed (sterile) neutrinos N I to MSM:Asaka,ShaposhnikovPLB 620 , 17(2005)L MSM = L MSM + i N

I / N I L M DI N I + M I 2 N cI N I + h.c.

A very modest and simple modication of the SM which can explainwithin one consistent framework

. . . neutrino oscillations

. . . baryon asymmetry of the Universe

. . . provide a viable (warm or cold) dark matter candidate

. . . can incorporate ination

. . . can have a number of astrophysical applications



29/42

Choosing parameters of the MSM

Parameters of two sterile neutrinos are enough to explainbaryogenesis and t the oscillations data: If M 2 ,3 150 MeV 20 GeV and M 2 ,3 M 2 , 3 MSM explains

baryon asymmetry of the Universe.

Neutrino experiments can be explained within the same choiceof parameters. See-saw with masses below EW scale.The third (lightest) sterile neutrino can have cosmologically long life

time = 5 1026 sec keVM s5

10 8

2

2

Can be produced in the early Universe in the right amount: DodelsonWidrow93

Asaka, Laine,Shaposhnikov

(2006)

Via active-sterile neutrino oscillations Via resonant active-sterile neutrino oscillations in the presence

Shi, Fuller98

of lepton asymmetries . (Works well for sterile neutrinos in keV range. ) In inaton decays. ( Can produce sterile neutrinos up to the mass of few

Tkachev,Shaposhnikov(2006)

MeV)Can play the role of DM ( warm , cold or mixed )


d f l


30/42

Production of sterile neutrino DM

100

101

102

M1

/ keV

10-16

10-14

10-12

10-10

10-8

10-6

10-4

s i n

2 2

case 1

LMC

MW

MW

M3110

6n

e

/ s

2

4

8 1 2 1 6

0 .0

2 5 0 0

2 5

2 5 0

SPI

7 0

7 0 0

Sterile neutrino interactswith the rest of the SMmatter only via couplingwith active neutrinos,

parametrized by =m DM

Interaction is differentwith and without leptonasymmetry


Wi d f f il i DM


31/42

Window of parameters of sterile neutrino DM

S i n

2 ( 2

1 )

M1 [keV]

10 -16

10 -14

10-12

10 -10

10 -8

10 -6

0.3 1 10 100


Wi d f t f t il t i DM


32/42

Window of parameters of sterile neutrino DM Boyarsky, O.Ret al. 2008

S i n

2 ( 2

1 )

M1 [keV]

10 -16

10 -14

10-12

10 -10

10 -8

10 -6

0.3 1 10 100




33/42


S i n

2 ( 2

1 )

M1 [keV]

10 -16

10 -14

10-12

10 -10

10 -8

10 -6

0.3 1 10 100


Summary


34/42

Summary

A number of models provides decaying DM (sterile neutrino,gravitino, Majoron, volume modulus,. . . ). The DM candidates canbe light (keV MeV range)

Astrophysical search of decaying DM model is an experiment ofdirect detection type (if a line is detected, it can be distinguishedfrom the line of any other origin)

DM models can be warm (with various velocity distributionfunctions), this can be probed by the Lyman- data

So far the Lyman- data were obtained only for the models withthermally produced WDM particles. The 90% CL bounds should

be understood with at least 50% additional uncertainty

Improving on these results can rule out (or conrm) severalinteresting extensions of the SM ( MSM, volume moduli)



35/42

Thank you for yourattention



36/42

How to test this theory?


How to detect heavy sterile neutrinos


37/42

How to detect heavy sterile neutrinos

Missing energy signal in K, D and B decays ( 2 effect)

M N < M K : KLOE, NA48, E787 M N < 1 GeV : charm and factories

1 GeV < M N < M B : charm, and B-factories (plannedluminosity is not enough) Gorbunov &Shaposhnikov2007Decay processes N + , etc (nothing + ) (4 effect)

M N < M K : Any intense source of K-mesons (e.g. from protontargets of K2K, MiniBooNe or MINOS)

M N < M D : JPARC, MINOS, CNGS beams + very near detector

M N > M D : extremely difcult


Laboratory detection of sterile neutrino


38/42

Laboratory detection of sterile neutrino

Creation and detection in the lab: suppressed by 4

and hopeless.Creation somewhere and detection in the lab 2 effect. But the Bezrukov,

ShaposhnikovPRD 2007

only realistic possibility is to search for radiative decays of sterileneutrino in the DM clouds not a laboratory experiment.

Creation in the lab without subsequent detection the uniqueoption, 2 effect.

Possibilities:Forbidden decays, e.g.

0

N branching ratio is too small.

Hopeless.

-decay kinematics: 3 H 3 He + e + e is not the same as3 H 3 He + e + N !

Full kinematics event-by-event mass measurement: may work.COLTRIMS technology


MSM valid up to Planck scale?


39/42

MSM valid up to Planck scale?

For Higgs masses 129 GeV < M H < 189 GeV MSM can beconsistent quantum eld theory up to the Planck scale.

Thus MSM describes all particle physics experiments, explainsneutrino oscillations , provides the DM candidate and explains thebaryon asymmetry of the Universe without introduction of any newscale above the EW scale.

MSM (as well as MSM) suffers from the Landau pole in the Higgs self-coupling. Maiani et al.1978;Cabibbo et al.1979For Higgs mass M H < 189 GeV the Landau pole occurs above the Planck

Pirogov &

Zenin 1999;Hambye &Riesselmann1997

scale

For sufciently low Higgs masses the MSM vacuum is unstable. This does not

Altarelli &

Isidori 1994Casas et al.1995

happen (with the cut-off scale M Pl) for M H > 129 GeV.


MSM in LHC era


40/42

If LHC nds only Higgs boson the MSM can be an effectivetheory up to the Planck scale!

It introduces no new scale to explain physics there may be nohierachy problem.

If LHC nds Higgs + new physics it will not be possibleto e.g. to calculate the DM abundance. WDM/CDM/mixed,

decaying/annihilating and other properties of the DM still should betested.

other aspects of the MSM should be tested as well

If LHC does not nd Higgs the main predictions of the MSM arestill valid.


TOC


41/42

Dark Matter in the Universe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1What is known about DM? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2What can be the DM mass? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Phase-space density evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4What is known about DM? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6Decaying DM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7Properties of decaying DM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8DM decay line search : advantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

DM decay line search : advantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0DM decay line search : advantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Where to look for DM decay line? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3

Restrictions on life-time of decaying DM . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4Free-streaming DM and structure formation . . . . . . . . . . . . . . . . . . . . . . . 1 6Cold or warm ? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17DM at small scales. Paris, Feb.13-15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 8CDM or WDM? A summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19


TOC


42/42

Lyman- forest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 0

Lyman- forest : challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1Powerspectra of DM models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2Bayesian approach to WDM bounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 3Lower DM mass bound in Bayesian approach . . . . . . . . . . . . . . . . . . . . 24Summary: Lyman- bounds on WDM . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 MSM: all masses below electroweak scale . . . . . . . . . . . . . . . . . . . . . . 2 7Choosing parameters of the MSM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 8Production of sterile neutrino DM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 9Window of parameters of sterile neutrino DM . . . . . . . . . . . . . . . . . . . . . 3 0Window of parameters of sterile neutrino DM . . . . . . . . . . . . . . . . . . . . . 3 1Window of parameters of sterile neutrino DM . . . . . . . . . . . . . . . . . . . . . 3 2Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

How to detect heavy sterile neutrinos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 6Laboratory detection of sterile neutrino . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 7 MSM valid up to Planck scale? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 8 MSM in LHC era . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 9


oleg ruchayskiy and alexey boyarsky- cosmological bounds on warm dark matter and astrophysical...

Documents