Extremely Light Dark Matter Particles:
ΨDM
Tzihong Chiueh, NTU
In collaboration with
Hsi-Yu Schive, Tom Broadhurst, Tak-Pong Woo, James Chan,
Shing-Kwang Wong, Ming-Hsuen Liao, Kuan-Wei Huang, Ue-han Chang, Su-Ron Chen
Contents
• CDM and its small scale problems
• WDM
• What is ψDM?
• Linear and nonlinear evolution of ψDM
• Soliton cores and fuzzy halos
• Determination of ψDM particle mass
• Core – halo relation
• Hints and Predictions for ψDM
• Conclusions
Cold Dark Matter (CDM)
• Thermally generated non-standard-model particles
• Heavy, e.g. 100 GeV due to WIMPS miracle
• Tiny thermal velocity because of its high mass
• Easy to implement to numerical simulations for investigations of nonlinear large-scale structures
• Fit observations, such as CMB, acoustic oscillations, Lyman-α forests, galaxy clusters, etc.
• The agreements are data of large scale and does not rule out other alternatives
• To really pin down CDM, look for singularity
• CDM is scale-free and forms halos within halos
• CDM simulation consistently shows Narravo-Frenk- White (NFW) radial profile in bound objects
ρ = a/ [r (r+ b) ²]
a & b are object-dependent
fitting parameters
Coma galaxy cluster (1 Mpc)
Fornax dwarf galaxy (1 kpc) Andromeda galaxy (30 kpc)
Three kinds of cosmic objects that have high dark matter concentration
Of the three, galaxies have too much baryon concentration at central 10 kpc
Not suitable for understanding the nature of dark matter
stacked lensing data of 4 rich galaxy clusters of 1015 solar mass
Galaxy Clusters
Observations of galaxy clusters agree with the NFW profile very well
Kleyna et al, 2003
N=10000
(7% of all stars) Ursa Minor dwarf galaxy
Dwarf Galaxies
Old & cold star clump
elliptical harmonic potential
elliptical NFW potential
1 kpc
• Another evidence against CDM is the lack of satellite galaxies near Milky Way, and this becomes more acute after the
recent Sloan Digital Sky Survey
• From CDM simulations, there should be hundreds of nearby satellite galaxies around Milky Way, but only few tens (< 10%) are detected.
• Both the galaxy core problem and the abundance problem are associated with dwarf galaxies on the scale < few kpc !!!
Dwarf galaxies
Scale symmetry of CDM is broken below few kpc ???
If so, what may be the cause?
Option 1: Warm Dark Matter
• Thermally generated WDM has residual thermal velocity that erases small scale structures due to free stream mixing
• It solves the problem of satellite galaxy deficiency
• Best observational constraint:
m > 2 keV (via Lyman-α forests)
• Major problem: for this range of mass
WDM can not produce a flat core as large as 1 kpc
CDM
2 keV 0.2 keV
Linear power spectrum
Option 2: Extremely light particles
• Non-thermally generated bosons, e.g. via axion mechanism, and cold, i.e., BEC condensate
• The universe is described by a single field obeying Klein-Gordon equation, and self-coupled via metric perturbations.
• Breaks the scale symmetry due to a finite mass, or a Compton wavelength λcomp
λcomp= few kpc ???
• Not really! One must take into consideration the evolution of a length scale in an expanding universe.
Linear Evolution of Compton Length
• λL(a) = (a/ain) λcomp
• ain?
• It turns out ain is determined by λcomp= c/H(ain) = Hubble radius in the radiation era
z=3000 k
primordial spectrum Harrison-Zeldovich
processed spectrum before z=3000
a0/a = 1+z a0: expansion factor now
(a0/ain)
• If λL(a0) were few kpc, then m ~ 10-20 eV
• Length scale greatly reduced during collapse of a galaxy, i.e., λNL = few kpc (observed dwarf galaxies size) << λL
Therefore, m << 10-20 eV
• One can define λNL(a) = F(a) λL(a) and ask what F(a) is.
This nonlinear evolution must be aided by computer simulations.
Nonlinear Evolution
Nonlinear Evolution in NR regime
•
Φ = Ψr cos(mt) – Ψi sin(mt) to identify a complex Ψ|Ψ|2 : normalized number density
• Simulations carried out by solving Poisson-Schroedinger equation
• This equation is almost scale free and has an approximate scaling relation
x’ = ξ-1 a-1/2 x, |Ψ’|2 = ξ4 a |Ψ|2 , V’ = ξ2 V, τ’ =ξ-2 a-1 τ In terms of bound objects, M’ = ξ a-1/2 M
background density
AdaptiveMeshRefinement calculations:
accurate but time-consuming, Δt ~Δx 2
Nature Physics 10, 496 (2014)
|Ψ|2
ψr2
Fuzzy halos & Soliton cores
z = 0 dwarf galaxies for any m
obey rc ~ ρc-4 scaling relation
(Schive, et al 2014,NaturePhys)
Single halo redshift evolution
(Schive at al. 2014, PRL)
Soliton core & NFW halo
Determination of boson mass
Fornax dwarf spheroidal galaxy
Data:
• Fornax dwarf spheroidal galaxy
• good spectral and star count data
• 3 populations of stars
Analysis:
• Choose the richest population
• Solve the Jeans equation for star density, treating stars as test particles in soliton potential
• Adjusting m to yield the least χ2, and similar approaches apply to opyimize Burkert and NFW and NF
• Check the result with other two populations
(Schive at al. 2014 Nature Phys)
m = 8x10-23 eV χ2 ≈ 1.2
1-σ ellipse
Core-Halo relat ion:
M
cdepends on conserved mass and energy?
M c ~ rc-1
rcσh ~ h/m
Non-local uncertainty principle
A dedicated numerical experiment:
soliton collision
= σ2 Mc ~ σ
Thermalization of the entire halo!
Core-Halo relation: M
c– M
halo-z scaling
M
c~ M
halo1/3(1+z)
1/2((Schive, et al.,PRL (2014))
We can predict the core for any redshift and any halo mass !
• For the earliest galaxies at z=12, Mc ~ 5 x 108 solar mass (determined by λcomp), and gravitational potential
(GMc/rc)z=12 ~ (1+z) ~ 13 (GMc/rc)z=0 for the same halo mass.
• The earliest quasar appeared at z=7, for which the age of universe is ~ 8 x 108 yr, requiring formation of supermassive black hole of MBH ~ 109 solar mass in a short time.
• 13 times deeper potential may help BH formation
IMPLICATIONS:
M
c~ M
halo1/3(1+z)
1/2, r
c ~M
halo-1/3(1+z)
-1/2Nonlinear Evolution: Summary
• Nonlinearity generates a mass hierarchy of different length scales.
Erases the feature of λcomp and replaced by a running length scale λJeans ~ ρlocal-1/4(a, xcore)
• λcomp only important for determining the size of the earliest galaxies at z=12
• After that, galaxies of different masses gradually appear, and λJeans of different ρlocal defines core sizes of the galaxies.
• At every epoch, there is a minimum halo with minimum ρlocal and maximum λJeans which defines few kpc scale at present epoch.
Hints for soliton core and fuzzy halo:
(1) velocity dispersion peaks at central 150 pc
Milky Way bulge Nearby elliptical galaxies
rotation dispersion
NGC3379
Mbh=1.4 x 107 solar mass
Zoccali et al., 2014
Mbh = 106 solar mass
Mbh = 1.4 x 108 solar mass
Shapiro et al. 2006
Krajnovi´c et al.2009
(2) Strong lens quad image flux anomaly
Possibility for detecting 7 images
7
7
5
3
Nonlinear Evolution: Summary
• Soliton is the ground state of this nonlinear Eq. and has a size of λ
Jeans~ (1+z)
1/4m
-1/2ρ
local-1/4• Fuzzy halo consists of excitations in the halo potential
• Excitations are random phased, and so granules in the fuzzy halo randomly appear and disappear
• Fuzzy halo & soliton core share the same temperature
or squared velocity dispersion ‹(|▽ln|Ψ|)
2›
Conclusions
• ΨDM predicts a soliton core and a fuzzy halo in every galaxy
• ΨDM halos resemble CDM halos by the large, except for minimum halos
• ΨDM is a falsifiable model, having various predictions
different from CDM. To name a few, strong lensing,
central anomalous velocity peaks, first galaxy forming
at z=12 which affects reionization, etc.
Why is ψDM boson mass 10 -22 eV?
Tilted wine bottle (axion) model + one general
assumption about the dark sector naturally yields m
a~ 10
-22eV
f
a~ 4 x 10
16GeV m
Λ~ 100 eV
Φ V(Φ)
arXiv:1409-0380 fa
Why is ψDM boson mass 10 -22 eV?
T = 200GeV
All particles in the same soup
Relativistic Λ dark particles + dark photon
Bright sector
Particle annihilation when non-
relativistic
Why is ψDM boson mass 10 -22 eV?
Present Universe Ω
m=0.26
Non-relativistic axion density ~ a-3
Non-relativistic Λ dark particles annihilate to create axion oscillation
3 species of neutrinos + Photons
Td ~ (1/3) Tb ~ a-4
0.2 MeV > mΛ > 0.3 eV
Matter-Radiation Equality