Realistic estimation for the detectability of dark matter subhalos - - PowerPoint PPT Presentation

Valentina De Romeri - IFIC UV/CSIC Valencia

Realistic estimation for the detectability of dark matter subhalos using Fermi-LAT catalogs

Valentina De Romeri

(IFIC Valencia - UV/CSIC)

Halo Substructure and Dark Matter Searches 27-29 June 2018, IFT (Madrid)

1

Based on Phys.Rev. D96 (2017) no.6, 063009 with F. Calore, M. Di Mauro, F. Donato and F. Marinacci

Valentina De Romeri - IFIC UV/CSIC Valencia

There is overwhelming evidence for the existence of dark matter:

Cosmological and astrophysical

• bservations

2 CMB anisotropies, Clusters (X-rays, lensing), Large Scale Structures, Galaxies (rotation curves, fits…)

26.8% 68.3%

The content of the Universe in terms of paellas (after Planck)

credit: R.A. Lineros

Valentina De Romeri - IFIC UV/CSIC Valencia

…....not included in the Standard Model Many candidates in Particle Physics

3

→ WIMPs, axions …

Additional assumptions for this talk:

• dark matter is a WIMP (GeV - TeV mass scale)
• WIMPs cluster in galaxies as dark halos (a main smooth halo and many subhalos)
• can pair annihilate or decay to produce SM particles
• accounts for the measured relic density

What do we know about DM?

• Non-baryonic (BBN, CMB)
• Collisionless (bullet cluster)
• Stable on cosmological scales (or lifetime >> tU

~13.8 Gyr)

• Neutral
• Massive
• Cold or Warm (structure formation)
• Not in conflict/excluded by DM experiments and

cosmological data

Park, E.-K. DMSAG Report on the Direct Detection and Study of Dark Matter (2007)

Valentina De Romeri - IFIC UV/CSIC Valencia

2.1 Antimatter in the cosmic rays (antiprotons, antideuterons, positrons…) 2.2 Neutrinos (DM annihilation inside celestial bodies) 2.3 Photons (DM annihilation in the galactic halo(s))

4

• 1. DIRECT DETECTION (looks for energy

deposited within a detector by the DM- nuclei scattering)

• 2. INDIRECT DETECTION (looks for

WIMP annihilation (or decay) products)

+ complementary searches at colliders

If DM is made of particles that interact among themselves and with SM particles we may hope to detect it. Two strategies:

{

Valentina De Romeri - IFIC UV/CSIC Valencia

• 1. Prompt photons from DM annihilation:
• Two-body annihilation into photons (gamma-ray lines)
• Photon production in hard process (bremsstrahlung of charged particles)
• Two-photon decay of neutral pions π0 → γγ dumped by the hadronization

chain of strongly interacting annihilation products (continuum)

• 2. Secondary photons from radiative processes associated with stable, charged

particles produced by DM annihilation or decay (electrons and positrons): e.g. inverse-Compton and synchrotron emission.

5

Gamma-rays from WIMPs? - Annihilation processes

DM γ DM γ DM γ DM f f π0 π0 π0 DM γ DM

W-/Z/q W+/Z/q

π0 π0 π0 γ γ γγ γ

Valentina De Romeri - IFIC UV/CSIC Valencia

6

dΦγ dEγ (Eγ, ψ, θ, ∆Ω) = dΦP P

γ

dEγ (Eγ) × J(ψ, θ, ∆Ω)

The γ-ray flux from DM annihilation is defined as the number of photons collected by a detector per unit of time, area, energy and solid angle:

Valentina De Romeri - IFIC UV/CSIC Valencia

7

dΦγ dEγ (Eγ, ψ, θ, ∆Ω) = dΦP P

γ

dEγ (Eγ) × J(ψ, θ, ∆Ω)

The γ-ray flux from DM annihilation is defined as the number of photons collected by a detector per unit of time, area, energy and solid angle:

dΦP P

γ

dEγ = 1 4π hσvi 2m2

DM

X

i

dN i

γ

dEγ Bi

PARTICLE PHYSICS factor:

• bb, μ⁺μ⁻, τ⁺τ⁻ final states
• Bi = 1 (representative of larger class of models)

Velocity averaged annihilation cross-section Photon energy spectrum per annihilation

Characteristic Energy Spectrum Important to:

• identify a DM signal
• determine the DM mass
• determine the annihilation process

Valentina De Romeri - IFIC UV/CSIC Valencia

8

dΦγ dEγ (Eγ, ψ, θ, ∆Ω) = dΦP P

γ

dEγ (Eγ) × J(ψ, θ, ∆Ω)

The γ-ray flux from DM annihilation is defined as the number of photons collected by a detector per unit of time, area, energy and solid angle:

dΦP P

γ

dEγ = 1 4π hσvi 2m2

DM

X

i

dN i

γ

dEγ Bi

J(ψ, θ, ∆Ω) = Z ∆Ω dΩ Z

los

ρ2(r(s, ψ, θ))ds

ASTROPHYSICAL factor:

• Sensitivity to different DM

halo profiles Integration of the squared DM density at a distance s from the Earth in the direction along the l.o.s and in the observational cone of solid angle ΔΩ PARTICLE PHYSICS factor:

• bb, μ⁺μ⁻, τ⁺τ⁻ final states
• Bi = 1 (representative of larger class of models)

Valentina De Romeri - IFIC UV/CSIC Valencia

γ-ray experiments relevant for DM searches

Space based: Fermi-LAT (2008)

(Pair conversion detector)

9

Ground based: MAGIC (2003), VERITAS (2006), H.E.S.S. (2002)

(Atmospheric Cherenkov Telescopes)

Effective area: O(1m2) Observation times: O(yr) Energies: 20 MeV - 1 TeV Effective area: O(105 m2) Observation times: O(100hr) Energies: ~50 GeV - 10 TeV (GeV to TeV)

HAWC (2015)

(Water Cherenkov Telescope)

Effective area: O(104 m2) Observation times: O(100hr) Energies: 100 GeV - 10 TeV

Valentina De Romeri - IFIC UV/CSIC Valencia

Detectability of dark matter subhalos with Fermi-LAT

10

• Two Fermi-LAT catalogs:
• 3FGL (Acero et al. 2015): 4 years, 0.1 – 300 GeV, Pass 7 data, 3000 sources (at a

latitude |b| > 20◦ mainly AGN)

• 2FHL (Ackermann et al. 2015): 80 months, 50 – 2000 GeV, Pass 8 data, 360 sources
• In both catalogues, a large fraction of sources remain unassociated: about 15% in the

2FHL and 30% in the 3FGL.

• Unassociated sources are point-like gamma-ray emitters detected as such by the LAT,

but lacking association with astrophysical objects known in other wavelengths.

• The sample of unassociated sources in the Fermi-LAT catalogues might already contain

gamma-ray emitting DM SHs.

Valentina De Romeri - IFIC UV/CSIC Valencia

Modelling the DM distribution in the Galaxy

11

Springel et al. (2008) Marinacci et al. (MNRAS 2013)

• Q. Zhu et al. (MNRAS, 2016)
• We want to predict the detectability of galactic DM subhalos by the Fermi-LAT.
• Numerical simulations predict a large amount of subhalos in a galaxy-size Milky Way halo

(of which dSphs are a manifestation).

• Estimates for the numbers of SHs

detectable by Fermi-LAT strongly depend on the assumptions made about the local distribution of DM SHs and the shapes of the density profiles.

• For modelling the SH population in the

Galaxy, we use one of the most recent cosmological numerical simulations that includes baryonic physics, Hydro Aquarius (Marinacci et al. 2015)

• We use two runs, DMO and Hydro

Valentina De Romeri - IFIC UV/CSIC Valencia

Cosmological simulations: DM-only vs HYDRO

12

Effects of baryons:

• increasing the density in the center of the Galaxy
• removing both DM and luminous matter and redistribute them in the SHs
• evaporating the gas and preventing gas accretion from the intergalactic medium

Differences:

• Fewer SHs in the Hydro simulation
• Low-mass SHs depleted in the Hydro

simulation

• Depletion mostly near the center.

Springel et al. (2008) Marinacci et al. (MNRAS 2013)

• Q. Zhu et al. (MNRAS, 2016)

Valentina De Romeri - IFIC UV/CSIC Valencia

13

• SH spatial distribution: Einasto profile
• n−2 = 0.66±0.06 (0.50±0.03)
• α = 1.17 ± 0.15 (2.20 ± 0.29)
• r−2 = 0.64 ± 0.02 (0.65 ± 0.02) Rvir
• SH mass distribution: dN/dM ∼ M−1.9

Cosmological simulations: DM-only vs HYDRO

• Radial abundance lower for Hydro simulation, mostly in the central region.
• Vmax (MSH) dependence of radial distribution — stronger for Hydro simulation.

Calore, VDR et al. Phys.Rev. D96 (2017) no.6, 063009

The gamma-ray emissivity from DM annihilation in SHs is determined by the internal spatial profile of the DM subhalos.

Valentina De Romeri - IFIC UV/CSIC Valencia

DM distribution and density profile of the SH

14

• DM distribution and density profile of the SHs: Einasto α = 0.16

J(ψ, θ, ∆Ω) = Z ∆Ω dΩ Z

los

ρ2(r(s, ψ, θ))ds

rmax = rs x 2.189

Calore, VDR et al. Phys.Rev. D96 (2017) no.6, 063009

Valentina De Romeri - IFIC UV/CSIC Valencia

Realistic estimation of the flux sensitivity

15

• We assume that gamma rays are produced from DM annihilations via the prompt mechanism
• We take into account two channels: b bar and τ+τ- which give the largest fluxes
• We take the gamma-ray spectra from DM annihilation from Cirelli et al. 2011 (Pythia 8).

bb

MDM = 10, 100, 800, 5000 GeV

F[

Spectral energy distribution of the signal: Super-exponential cutoff parametrisation Epeak = mDM/20 (bbar)

Integrated photon flux:

Calore, VDR et al. Phys.Rev. D96 (2017) no.6, 063009

Valentina De Romeri - IFIC UV/CSIC Valencia

Estimation of the flux sensitivity (3FGL)

16

• The sensitivity flux is the flux for which TS=25.
• For each DM mass we simulate DM SHs with different channels and at different latitudes.
• We simulate also the IEM and isotropic components taking the reference models for the 3FGL

and 2FHL catalogs.

• The data analysis details (exposure time, energy range,...) are the same as in the 3FGL and

2FHL catalogs.

MDM = 8 GeV MDM = 30 GeV MDM = 80 GeV MDM = 300 GeV MDM = 600 GeV MDM = 1200 GeV

bb

3FGL

Sensitivity flux threshold as a function of DM mass

Calore, VDR et al. Phys.Rev. D96 (2017) no.6, 063009

Valentina De Romeri - IFIC UV/CSIC Valencia

Estimation of the flux sensitivity (2FHL)

17

• Interplay between the shape of the DM spectrum and the energy dependence of the

background.

• The sensitivity flux threshold grows (up to mDM = 1 TeV) because the energy threshold for

2FHL is 50 GeV. In this energy range the spectrum has a very soft shape and shows a peak at E<50 GeV.

• For mDM>1TeV the peak falls inside the 2FHL range, and the sensitivity flux threshold reaches

a plateau. The shape of the energy spectrum is similar for different mDM.

MDM from 100 GeV to 20 TeV

Sensitivity flux threshold as a function of DM mass

bb

2FHL

Calore, VDR et al. Phys.Rev. D96 (2017) no.6, 063009

Valentina De Romeri - IFIC UV/CSIC Valencia

18

Detectable subhalos

100 MC realisations: ~2000 SHs for single Hydro realisation

• For all the SHs we compute

the gamma-ray flux above a given energy assuming an Einasto DM density profile.

• Compare the predicted flux

with the sensitivity flux threshold.

• A SH is detectable if the

predicted flux is larger than the sensitivity flux threshold.

• Small numbers of detectable

SHs: compatible with the fact that no emission from the direction of known dwarf galaxies has been observed yet.

Calore, VDR et al. Phys.Rev. D96 (2017) no.6, 063009

Valentina De Romeri - IFIC UV/CSIC Valencia

19

Detectable subhalos

• SH distance from the observer as a function of the SH mass.
• Smallest rs values correspond to undetectable SHs, independently of MSH. Detectable SHs

can have rs ranging from 0.4 kpc to 3 kpc

• Detectability depends on the interplay between rs (small rs —> large concentration) and SH

distance.

• The farthest SHs have to be extremely massive in order to be detectable. Such far and

massive SHs are also very likely to have a stellar counterpart and therefore to be detected in the optical wavelength as dwarf galaxies.

3FGL 3FGL

Calore, VDR et al. Phys.Rev. D96 (2017) no.6, 063009

MDM = 100 GeV, thermal ⟨σv⟩ Dark SHs

Valentina De Romeri - IFIC UV/CSIC Valencia

20

• The small (or even null) number of detectable DM SH candidates among the Fermi-LAT

unassociated sources allows to set upper limits on the DM annihilation cross section ⟨σv⟩.

Limits on dark matter annihilation cross section

• Upper limit: value of

⟨σv⟩ for which the number of detectable SHs is smaller than a given number of SH candidates.

• NCandidate

=NUnassociated gives the most conservative limits.

• The limits derived

assuming NCandidate = 0 are very tight and competitive with limits from dwarfs galaxies.

Calore, VDR et al. Phys.Rev. D96 (2017) no.6, 063009

Valentina De Romeri - IFIC UV/CSIC Valencia

21

• The source count distribution N as a function of the integrated flux F is an important

characterisation of astrophysical source populations and it can provide information also about the faintest end of the flux distribution.

Source count distribution (Log N - Log F) of SHs

Calore, VDR et al. Phys.Rev. D96 (2017) no.6, 063009

• The Log N - Log F of DM subhalos

shows a sharp cutoff at high fluxes, that corresponds to few very bright subhalos.

• The numerous faint and undetectable

subhalos populate the Log N - Log F at low fluxes.

• Regardless of the choice of the

integration energy threshold, the SHs source count is strongly subdominant with respect to the observed flux distribution of AGN in both the 3FGL and 2FHL catalogs.

MDM = 100 GeV, ⟨σv⟩ = 10−25 cm3/s, bb blazar population 1FGL

3FGL

simulated DM SHs (integrated F>0.1 GeV)

Valentina De Romeri - IFIC UV/CSIC Valencia

Detecting extended SHs

22

• We calculate the flux for different

angular distances from the center of the SH and infer the distance d68 within which the 68% of the flux is contained.

• We use the extension of 3FGL sources

to estimate the angular extension sensitivity of the telescope.

• Conservative approach: reference angle

is the size of W44 (Θext = 0.16º)

• Optimistic approach: Θext = 0.1º.
• If d68 >Θext the SH is considered

extended.

• We find 0.5 (0.8) extended sources in

the conservative (optimistic) approach.

Calore, VDR et al. Phys.Rev. D96 (2017) no.6, 063009

• Should an unassociated source be detected by LAT with a non-zero spatial extension

at high latitude, it would be a tantalising hint of a signal from DM SH.

Θext = 0.1º MDM = 40 GeV, ⟨σv⟩ = thermal, bb

• The extended SHs would have mSH > 2x10^7 Msun and dSH<80 kpc.

Θext = 0.16º mSH = 1.9x109 Msun, dSH =46 kpc mSH = 4.7x109 Msun, dSH =80 kpc

Valentina De Romeri - IFIC UV/CSIC Valencia

Conclusions

• First realistic estimation of the detectability of Galactic dark matter subhalos in the Fermi-LAT

3FGL and 2FHL catalogs.

• Based on one of the most recent hydrodynamic simulations for structure formation (Hydro-

Aquarius simulation)

• Although baryons affect the abundance and internal structure of sub-halos (especially the more

massive ones), these discrepancies do not substantially alter the predictions on the scale radius.

• Sensitivity: we fully account for dependence of the sensitivity flux threshold on the dark matter

annihilation channel, the dark matter mass and the subhalo position in the main halo.

• Our results show that the largest number of detectable subhalos, that might already be among

the unassociated sources of the 3FGL catalog, is at most 0.9 ± 0.8 for MDM = 8 GeV – with ⟨σv⟩ fixed to the upper limit derived from the latest analysis of dwarf spheroidal galaxies. The prediction for the 2FHL catalog is lower: NDetectable = 0.0 ± 0.2 for MDM = 10 TeV.

• Competitive bounds on ⟨σv⟩ if Candidate = 0.
• Future gamma-ray exps such as CTA, e-ASTROGAM will improve on the sensitivity to detect DM
• SHs. Relentless effort in numerical simulations will clarify how to reliably include baryons in the

formation of galactic structures.

23

Valentina De Romeri - IFIC UV/CSIC Valencia

• First realistic estimation of the detectability of Galactic dark matter subhalos in the Fermi-LAT

3FGL and 2FHL catalogs.

• Based on one of the most recent hydrodynamic simulations for structure formation (Hydro-

Aquarius simulation)

• Although baryons affect the abundance and internal structure of sub-halos (especially the more

massive ones), these discrepancies do not substantially alter the predictions on the scale radius.

• Sensitivity: we fully account for dependence of the sensitivity flux threshold on the dark matter

annihilation channel, the dark matter mass and the subhalo position in the main halo.

• Our results show that the largest number of detectable subhalos, that might already be among

the unassociated sources of the 3FGL catalog, is at most 0.9 ± 0.8 for MDM = 8 GeV – with ⟨σv⟩ fixed to the upper limit derived from the latest analysis of dwarf spheroidal galaxies. The prediction for the 2FHL catalog is lower: NDetectable = 0.0 ± 0.2 for MDM = 10 TeV.

• Competitive bounds on ⟨σv⟩ if Candidate = 0.
• Future gamma-ray exps such as CTA, e-ASTROGAM will improve on the sensitivity to detect DM
• SHs. Relentless effort in numerical simulations will clarify how to reliably include baryons in the

formation of galactic structures.

Conclusions

24

Thank you!

equations of general rel- ativity and derive equations of motion of self-gravitating nonrelativistic particles in the expanding Universe can be treated using the collisionless Boltzmann equation paired with the Poisson equation for the gravitational po- tentia

http://iopscience.iop.org/article/10.1088/0004-637X/749/1/90/ pdf

Valentina De Romeri - IFIC UV/CSIC Valencia

26

The excellent performances of Fermi-LAT have allowed the exploration for a DM component in the Milky Way, in extragalactic nearby objects, as well as in cosmological structures.

At high galactic latitudes, a faint γ-ray irreducible emission has been measured, and shown to be isotropic on large angular scales.

Fermi-LAT sky (+ CRs bkg)

=

• Undetected sources (blazars, AGNs, GRBs High-latitude

pulsars, milli-second pulsars..)

• Diffuse processes (interactions of UHE cosmic rays with

the EBL, Intergalactic shocks..)

• DARK MATTER???

{

The extragalactic γ-ray emission (IGRB)

Valentina De Romeri - IFIC UV/CSIC Valencia

• Almost not absorbed/attenuated when propagating through halo
• Point directly to the sources: clear spatial signatures
• Clear spectral signatures to look for

but… need careful study of:

• astrophysical background
• properties of unresolved sources (number, distribution…)
• uncertainties related to the DM model
• spatial distribution of DM

27

Gamma-rays from WIMPs?

DM DM γ γ γ

Valentina De Romeri - IFIC UV/CSIC Valencia

γ-rays from DM: search targets

28

Dwarf spheroidal galaxies: Known location and DM content Low statistics Galaxy clusters: Low bkg but low statistics Astrophysical contamination Galactic center: Large statistics Large background Milky Way halo: Large statistics Diffuse background (low background at high galactic latitudes) + Isotropic background: Large statistics, but astrophysics, galactic diffuse background + Spectral lines Little or no astrophysical uncertainties, good source id, but low sensitivity because of expected small branching ratio

Credit: NASA/DOE/Fermi LAT Collaboration, using nine years of data collected from 2008 to 2017.