Dark Matter Phenomenology
Alejandro Ibarra
Technische Universität München
Cinvestav
Ciudad de Mexico
November 2018
Outline
Lecture 1: Evidence for dark matter.
Lecture 2: Dark matter production.
Lecture 3: Indirect detection
Lecture 4: Indirect detection (cont.)
Lecture 5: Direct detection, collider signals
Indirect dark matter searches
General idea:
1) Dark matter particles annihilate or decay producing a flux of
stable particles: photons, electrons, protons, positrons, antiprotons or
(anti-)neutrinos.
DM
DM
DM
Indirect dark matter searches
General idea:
1) Dark matter particles annihilate or decay producing a flux of
stable particles: photons, electrons, protons, positrons, antiprotons or
(anti-)neutrinos.
2) These particles propagate through the galaxy and through the
Solar System. Some of them will reach the Earth.
Indirect dark matter searches
General idea:
1) Dark matter particles annihilate or decay producing a flux of
stable particles: photons, electrons, protons, positrons, antiprotons or
(anti-)neutrinos.
2) These particles propagate through the galaxy and through the
Solar System. Some of them will reach the Earth.
3) The products of the dark matter annihilations or decays are detected
together with other particles produced in astrophysical processes
(for example, cosmic ray collisions with nuclei in the interstellar medium).
The existence of dark matter can then be inferred if there is a significant
excess in the fluxes compared to the expected astrophysical backgrounds.
Cosmic antimatter: state of the art
Antiprotons
Positrons
Cosmic antimatter: state of the art
Antiprotons
Positrons
Gamma-rays
Production of gamma-rays
The gamma ray flux from dark matter annihilations/decays has two components:
 Inverse Compton Scattering
radiation of electrons/positrons
produced in the annihilation/decay.
 Always smooth spectrum.
 Prompt radiation of gamma rays
produced in the annihilation/decay
(final state radiation, pion decay...)
 May contain spectral features.
Inverse Compton Scattering radiation
The inverse Compton scattering of electrons/positrons from
dark matter annihilation/decay with the interstellar and
extragalactic radiation fields produces gamma rays.
e from dark matter
annihilation/decay
Ee  1 TeV
Upscattered photon
Interstellar radiation field (starlight,
thermal radiation of dust, CMB)
Porter et al.
This produces
Eg*  100 GeV
Prompt radiation
Annihilation
Source term
(particle physics)
Decay
Line-of-sight integral
(astrophysics)
Propagation
Where to look for annihilating dark matter
Baltz et al.
arXiv:0806.2911
Kuhlen, Diemand, Madau
Where to look for annihilating dark matter
Baltz et al.
arXiv:0806.2911
Galactic center
Kuhlen, Diemand, Madau
Where to look for annihilating dark matter
Baltz et al.
arXiv:0806.2911
Satellites
Kuhlen, Diemand, Madau
Where to look for annihilating dark matter
Baltz et al.
arXiv:0806.2911
Galactic halo
Kuhlen, Diemand, Madau
Where to look for annihilating dark matter
Baltz et al.
arXiv:0806.2911
Extragalactic
background
Kuhlen, Diemand, Madau
Where to look for annihilating dark matter
Baltz et al.
arXiv:0806.2911
Galaxy clusters
Kuhlen, Diemand, Madau
Where to look for annihilating dark matter
Baltz et al.
arXiv:0806.2911
Features in the
energy spectrum
Kuhlen, Diemand, Madau
Problem for discovery: for typical channels and typical masses,
the expected flux lies well below the background.
E 2  GeV cm2 s 1 sr1
105
Galactic center region, for thermally produced DM
106
107
bb
mDM=500 GeV
108
109
20
30
50
70
100
E GeV
150
200
300
Discovery requires a very good understanding of the backgrounds,
or to select targets and search strategies which optimize the
signal-to-background ratio.
- Dwarf spheroidal galaxies
- Diffuse galactic emission.
- Sharp spectral features.
Dwarf spheroidal galaxies
Segue 1: Optical image
Dwarf spheroidal galaxies
Segue 1: Optical image
Dwarf spheroidal galaxies
Segue 1: Optical image
Dwarf spheroidal galaxies
Mass-to-light ratio
~ 3400 M/L
Most DM-dominated
object known so far!
Segue 1: Optical image
Dwarf spheroidal galaxies
Segue 1: Gamma-ray image
(simulated!)
Dwarf spheroidal galaxies
Gamma-ray image taken with the MAGIC telescopes
Dwarf spheroidal galaxies
MAGIC coll.
arXiv:1312.1535
Dwarf spheroidal galaxies
Relatively close
High mass-to-light ratio:
dwarf galaxies contain large
amounts of dark matter
Assume a Navarro-Frenk-White dark matter halo profile inside
the tidal radius:
Constraints on WIMP dark matter models
Fermi coll.
arXiv:1503.02641
Diffuse Galactic emission
Divide the sky in different regions:
3°  3°
Diffuse Galactic emission
Divide the sky in different regions:
5°  30°
Diffuse Galactic emission
Divide the sky in different regions:
10° - 20° galactic latitude
Diffuse Galactic emission
Divide the sky in different regions:
Galactic poles
But beware of backgrounds when searching for dark matter...
Background I: sources
Background II: modelling of the diffuse emission
Inverse Compton
Bremsstrahlung
p0-decay
Conservative approach: demand that the flux from dark matter annihilation
does not exceed the measured flux
Cirelli, Panci, Serpico
Cirelli, Panci, Serpico
… but one can do better: backgrounds
can be modelled (to some extent)
Region |b|<5°, |l|<5°
Daylan et al. '14
Goodenough, Hooper '09, '10
Hooper, Linden '11
Huang et al '13
Abazajian et al '14
…
Region |b|<5°, |l|<5°
Daylan et al. '14
Goodenough, Hooper '09, '10
Hooper, Linden '11
Huang et al '13
Abazajian et al '14
…
Subtract
- Sources from 2FGL
Region |b|<5°, |l|<5°
Daylan et al. '14
Goodenough, Hooper '09, '10
Hooper, Linden '11
Huang et al '13
Abazajian et al '14
…
Subtract
- Sources from 2FGL
- Spatial template for diffuse
galactic emission
 p0 component
 Bremsstrahlung component
 ICS component
Region |b|<5°, |l|<5°
Daylan et al. '14
Goodenough, Hooper '09, '10
Hooper, Linden '11
Huang et al '13
Abazajian et al '14
…
Subtract
- Sources from 2FGL
- Spatial template for diffuse
galactic emission
 p0 component
 Bremsstrahlung component
 ICS component
- Fermi bubbles
Region |b|<5°, |l|<5°
Daylan et al. '14
Goodenough, Hooper '09, '10
Hooper, Linden '11
Huang et al '13
Abazajian et al '14
…
Subtract
- Sources from 2FGL
- Spatial template for diffuse
galactic emission
 p0 component
 Bremsstrahlung component
 ICS component
- Fermi bubbles
- Isotropic (extragalactic) component
Region |b|<5°, |l|<5°
Daylan et al. '14
Goodenough, Hooper '09, '10
Hooper, Linden '11
Huang et al '13
Abazajian et al '14
…
Subtract
- Sources from 2FGL
- Spatial template for diffuse
galactic emission
 p0 component
 Bremsstrahlung component
 ICS component
- Fermi bubbles
- Isotropic (extragalactic) component
Energy spectrum of the excess
Daylan et al.
Energy spectrum of the excess
Daylan et al.
DM DM → b b
mDM = 35.25 GeV
sv = 1.010-26 cm3/s
Region 1°<|b|<20°, |l|<20°
Daylan et al.
Region 1°<|b|<20°, |l|<20°
Daylan et al.
DM DM → b b
mDM = 35.25 GeV
sv = 1.010-26 cm3/s
DM DM → b b
mDM = 35.25 GeV
sv = 1.710-26 cm3/s
This analysis assumes one concrete spatial template for the
Galactic diffuse model.
The excess could simply be the result of a wrong modelling for the
foreground emission …
This analysis assumes one concrete spatial template for the
Galactic diffuse model.
The excess could simply be the result of a wrong modelling for the
foreground emission …
… or may be not.
Sytematic uncertainties in the GC excess
Calore et al.
Sytematic uncertainties in the GC excess
Calore et al.
Morphology of the GC excess
Calore et al.
Morphology of the GC excess
Calore et al.
The excess does look like a WIMP signal!
 The excess is extended: it is also observed in the region
1°<|b|<20°, |l|<20°, with roughly the same spectrum.
 The excess looks spherically symmetric.
 The excess is present for any diffuse background model.
 The excess is “theoretically sound”: simple models can reproduce
the observations.
Testing the DM interpretation of the GC excess
 The required mass and cross section could be large enough to produce
an observable signal from dwarf galaxies, however no signal is observed
Keeley et al. '17
Testing the DM interpretation of the GC excess
 The required mass and cross section could be large enough to produce
an observable signal from dwarf galaxies, however no signal is observed.
 Also in antiprotons. Antiproton data inconclusive.
Cuoco et al.
1610.03071
Reinert, Winkler.
1712.00002
Testing the DM interpretation of the GC excess
 The required mass and cross section could be large enough to produce
an observable signal from dwarf galaxies, however no signal is observed.
 Also in antiprotons. Antiproton data inconclusive.
 And in radio, however no signal is observed.
Bringmann, Vollmann, Weniger
Testing the DM interpretation of the GC excess
 The required mass and cross section could be large enough to produce
an observable signal from dwarf galaxies.
 Also in antiprotons.
 And in radio.
 Besides, some (less understood) astrophysical backgrounds were not
included in the templates.
 Millisecond pulsars?
Wang et al.'05, Abazajian '11, Gordon, Macias '13
Hooper et al. '13, Cholis et al. '14
 Active past of the galactic center?
- Hadronic interactions of protons injected in a burst event.
- ICS of electrons injected in a burst event. Petrovic et al.
Carlson et al.
Testing the DM interpretation of the GC excess
 The required mass and cross section could be large enough to produce
an observable signal from dwarf galaxies.
 Also in antiprotons.
 And in radio.
 Besides, some (less understood) astrophysical backgrounds were not
included in the templates.
 Millisecond pulsars?
Wang et al.'05, Abazajian '11, Gordon, Macias '13
Hooper et al. '13, Cholis et al. '14
 Active past of the galactic center?
- Hadronic interactions of protons injected in a burst event.
- ICS of electrons injected in a burst event. Petrovic et al.
Carlson et al.
Millisecond pulsar contribution to the bulge emission?
 The energy spectrum of the excess is similar to the energy spectrum
of known millisecond pulsars
Daylan et al. '14
Millisecond pulsar contribution to the bulge emission?
 The energy spectrum of the excess is similar to the energy spectrum
of known millisecond pulsars
 The morphology and intensity of the excess is similar to the one
predicted from a MSP population generated by tidally disrupted
globular clusters
Brandt, Kocsis '15
Millisecond pulsar contribution to the bulge emission?
 The energy spectrum of the excess is similar to the energy spectrum
of known millisecond pulsars
 The morphology and intensity of the excess is similar to the one
predicted from a MSP population generated by tidally disrupted
globular clusters
 The excess seems to be “clumpy”, as if originated by point sources
Maps of simulated photon counts in a 20°20° region
Millisecond pulsar contribution to the bulge emission?
 The energy spectrum of the excess is similar to the energy spectrum
of known millisecond pulsars
 The morphology and intensity of the excess is similar to the one
predicted from a MSP population generated by tidally disrupted
globular clusters
 The excess seems to be “clumpy”, as if originated by point sources
Maps of simulated photon counts in a 20°20° region
ed
r
o
v
Fa ata
d
by
Lee et al. '14'15
Bartels et al '15
Limits on dark matter parameters
Fermi coll.'17
Gamma-ray features
Idea: Search for a gamma-ray excess with an energy spectrum
qualitatively different from the background.
E 2  GeV cm2 s 1 sr1
105
106
107
bb
mDM=500 GeV
108
109
20
30
50
70
100
E GeV
150
200
300
Gamma-ray features
Idea: Search for a gamma-ray excess with an energy spectrum
qualitatively different from the background.
E 2  GeV cm2 s 1 sr1
105
106
107
bb
mDM=500 GeV
108
109
gg
20
30
50
mDM=100 GeV
70
100
E GeV
150
200
300
“Smoking gun” for dark matter: no (known) astrophysical process can
produce a sharp feature in the gamma-ray energy spectrum
Gamma-ray features: strategy for searches
dN/dE
Monochromatic
signal at E=100 GeV
10
15
20
30
50
70
100
150
200
E
Gamma-ray features: strategy for searches
dN/dE
Assume power-law
background
10
15
20
30
50
70
100
150
200
E
Gamma-ray features: strategy for searches
dN/dE
Total spectrum
10
15
Fit data to
20
30
50
70
100
150
200
E
Gamma-ray features: strategy for searches
Data don't really look like a power law...
Signal concentrated in
a narrow energy range
10
15
20
30
50
70
100
150
200
Gamma-ray features: strategy for searches
Data don't really look like a power law...
Signal concentrated in
a narrow energy range
10
15
20
30
50
70
100
150
200
Gamma-ray features: strategy for searches
Data don't really look like a power law...
In a narrow energy interval, the
background resembles a power-law
(Taylor's theorem)
Signal concentrated in
a narrow energy range
10
15
20
30
50
70
100
150
200
Gamma-ray features: strategy for searches
Data don't really look like a power law...
Repeat the search with different
“windows” postulating a signal
at different DM masses.
Signal concentrated in
a narrow energy range
10
15
20
30
50
70
100
150
200
Gamma-ray features: strategy for searches
Fermi collaboration
arXiv:1205.2739
Gamma-ray features: strategy for searches
One can do better searching for gamma-ray spectral features
in regions where it is most likely to find a signal.
Former approach: select a geometrically simple region of the sky
and search for features.
e.g region |b|>10° plus a 20°20°
square centered at the Galactic Center
(Fermi coll.)
Gamma-ray features: strategy for searches
One can do better searching for gamma-ray spectral features
in regions where it is most likely to find a signal.
Former approach: select a geometrically simple region of the sky
and search for features.
e.g region |b|>10° plus a 20°20°
square centered at the Galactic Center
(Fermi coll.)
Disadvantage: in the chosen region the background could be too large
and bury the signal
Instead, choose regions where, for a
given dark matter profile, the
signal-to-background ratio is maximized
Gamma-ray features: strategy for searches
One can do better searching for gamma-ray spectral features
in regions where it is most likely to find a signal.
Bringmann, Huang, AI, Vogl, Weniger , arXiv:1203.1312
Weniger, arXiv:1204.2797
Gamma-ray features: strategy for searches
One can do better searching for gamma-ray spectral features
in regions where it is most likely to find a signal.
Bringmann, Huang, AI, Vogl, Weniger , arXiv:1203.1312
Weniger, arXiv:1204.2797
Gamma-ray features: strategy for searches
One can do better searching for gamma-ray spectral features
in regions where it is most likely to find a signal.
Bringmann, Huang, AI, Vogl, Weniger , arXiv:1203.1312
Weniger, arXiv:1204.2797
Gamma-ray features: strategy for searches
One can do better searching for gamma-ray spectral features
in regions where it is most likely to find a signal.
Fermi collaboration
arXiv:1503.02641
Gamma-ray features: strategy for searches
One can do better searching for gamma-ray spectral features
in regions where it is most likely to find a signal.
H.E.S.S. collaboration
arXiv:1301.1173
Gamma-ray spectral features in Particle Physics
Three gamma-ray spectral features have been identified:
Gamma ray line
Srednicki, Theisen, Silk '86
Rudaz '86
Bergstrom, Snellman '88
Internal bremsstrahlung
Bergstrom '89
Flores, Olive, Rudaz '89
Bringmann, Bergstrom, Edsjo '08
Gamma ray box
AI, Lopez Gehler, Pato '12
Gamma-ray lines
The dark matter particle is electrically neutral.
The annihilation DM DM → g g arises at the one loop level
DM
SM
DM
DM
SM
DM
Monochromatic signals are quite generic in WIMP models
Gamma-ray lines
The dark matter particle is electrically neutral.
The annihilation DM DM → g g arises at the one loop level
DM
SM
DM
DM
SM
DM
Monochromatic signals are quite generic in WIMP models
However, with rates generically very suppressed:
Fermi-LAT collaboration
arXiv:1503.02641
“Canonical value of sv”
Fermi-LAT collaboration
arXiv:1503.02641
“Canonical value of sv”
Fermi-LAT collaboration
arXiv:1503.02641
Generically expected cross section
Sommerfeld enhancement
Hisano et al '03,'04
In models where the dark matter interacts with a lighter mediator, the one-loop
calculation cannot be trusted  Necessary to resum diagrams at all loops.
 enhancement of the cross section
“Inert doublet Model”
Expected
including
Sommerfeld
Garcia-Cely,
Gustafsson, AI '15
Expected neglecting
Sommerfeld
Internal bremsstrahlung
DM
SM
med
DM
SM
Assume a model where the dark matter particle
annihilates into Standard Model light particles
via the interaction with a mediator in the t-channel
Internal bremsstrahlung
Diagrams contributing to the process DM DM → SM SM g:
DM
SM
med
DM
g
SM
Internal bremsstrahlung
Diagrams contributing to the process DM DM → SM SM g:
DM
SM
med
DM
g
SM
Internal bremsstrahlung
Diagrams contributing to the process DM DM → SM SM g:
DM
SM
med
DM
g
SM
In the case mDM  mmed the scalar propagator gets enhanced
when ESM is small (which corresponds to Eg large).
Enhancement of the amplitude (and the rate)
when Eg is close to the kinematic end-point.
Internal bremsstrahlung
Bringmann, Huang,
AI, Vogl, Weniger
arXiv:1203.1312
Internal bremsstrahlung
Expected annihilation cross section for the 2  3 process.
DM
SM
DM
SM
+ ...
Limits on the annihilation cross section from the Fermi-LAT data
NFW
NFW
Bringmann, Huang,
AI, Vogl, Weniger
arXiv:1203.1312
Limits on the annihilation cross section from the Fermi-LAT data
“Canonical value of sv”
“Canonical value of sv”
NFW
NFW
Bringmann, Huang,
AI, Vogl, Weniger
arXiv:1203.1312
Gamma-ray box
DM
DM
Assume on shell production of an
intermediate scalar, f.
Gamma-ray box
DM
DM
Assume on shell production of an
intermediate scalar, f.
Assume that the scalar decays into two photons
Photon spectrum in the rest frame of the scalar
Gamma-ray box
DM
DM
Assume on shell production of an
intermediate scalar, f.
Assume that the scalar decays into two photons
Photon spectrum in the rest frame of the scalar
Photon spectrum in the galactic frame
“box-shaped spectrum”
mDM=100 GeV
mf=90 GeV
mf=70 GeV
0
20
40
60
80
100
New aspect: the spectral feature arises from a tree level 22 annihilation.
The strength of the signal could be unsuppressed, depending on the cross
section DM DM → ff and BR(fgg).
AI, Lopez Gehler, Pato
arXiv:1205.0007
Expected if BR=1
AI, Lamperstorfer, Lopez-Gehler, Pato, Bertone
Gamma-ray spectral features in Particle Physics
Recapitulation
“Smoking gun” for dark matter: no (known) astrophysical process can
produce a sharp feature in the gamma-ray energy spectrum
Three gamma-ray spectral features have been identified:
Gamma ray line
Internal bremsstrahlung
Gamma ray box
Gamma-ray spectral features in Particle Physics
Recapitulation
“Smoking gun” for dark matter: no (known) astrophysical process can
produce a sharp feature in the gamma-ray energy spectrum
Three gamma-ray spectral features have been identified:
Gamma ray line
Internal bremsstrahlung
Gamma ray box
Bright future for dark matter searches using gamma-rays!
H.E.S.S. II – in operation
CTA – Construction started in 2017
HAWC
in operation
DAMPE – Launched in 2015
GAMMA 400
Launch in 2021