Indirect dark matter detection: recent results and perspectives - - PowerPoint PPT Presentation

Indirect dark matter detection: recent results and perspectives

Piero Ullio SISSA & INFN (T rieste)

“Frontiers of Fundamental Physics 14”, Marseille, July 15, 2014

Outline:

Disclaimer: a review making no attempt to produce an exhaustive list of references on all recent results

• Recent experimental/theoretical highlights on WIMP indirect

detection; any clean signal and/or clean signature?

• In case clean signatures are not available, the complementarities among

different messengers and targets may be the key to solve the dark matter puzzle; combining different information however is non-trivial: an exercise to illustrate this point.

• Short introduction on the dark matter problem to put the WIMP

prejudice in perspective. A review focussed on the weakly interacting massive particles (WIMPs) as dark matter candidates:

• The WIMP paradigm facing the null detection so far of physics beyond

the Standard Model at the LHC: indirect (or direct) detection as guideline?

it accounts for the gravitational potential wells in which CMB acoustic oscillations take place:

Credit: W . Hu website

2 10 50 1000 2000 3000 4000 5000 6000

D[µK2]

90 18 500 1000 1500 2000 2500

Multipole moment,

1 0.2 0.1 0.07

Angular scale

(Planck, 2013)

Relying on the assumption that GR is the theory of gravity; still, it is very problematic to explain, e.g., the prominence of the third peak in an alternative theory of gravity and matter consisting of baryons only Plenty of evidence for non-baryonic cold (or coldish - as opposed to hot) DM being the building block of all structures in the Universe. E.g.:

Dark matter indirectly (gravitationally) detected!

Connection to a particle dark matter framework?

the DM term is treated as a classical, cold, pressure-less fluid subject to gravitational interactions only (no coupling to ordinary matter or photons, no self-coupling); tests of such gravitational coupling determine with accuracy its mean density: (Planck, 2013 + WMAP 7 yr pol.)

ΩDM = 0.1199 ± 0.0027

and the spectrum of its perturbations (nearly scale invariant, as expected from inflation). Reformulating the DM problem in terms of elementary particles in the dilute limit (two-body interactions dominating over multi-body interactions) is an assumption, and not the only possible extrapolation, e.g.: the recent attention on primordial black holes as DM; the recent interest on the possibility that DM is the form of (or, in certain regimes behaves like) a

• condensate. Will it be possible to single out this possibilities?

The standard model for cosmology, the ΛCDM model, does not aim to address questions regarding the nature of the DM component:

Observations and particle properties of DM

Assuming a particle formulation, astro/cosmo observables provide mainly informations on the properties that DM does not have, e.g.: it needs to be non-baryonic, non-relativistic at the phase of matter-radiation equality, … This is enough to say that DM is NOT within the SM of particle physics. The mass scale is essentially unconstraint, admitting ultralight bosons ( with macroscopic de Broglie wavelength), fermions at the level of about (Gunn-T remaine bound from phase space density limits), and no relevant upper limits (up to the MACHO range tested via lensing searches and even beyond)

10−22 eV

50 eV

The interaction scale has very tight limits with photons (DM millicharge, electric and magnetic dipole moments severely suppressed), significant with baryons, rather weak for self-interactions (from galaxy clusters morphologies and mergers, such as from the Bullet cluster - early claims of evidence for self interaction from the Musket Ball cluster not confirmed) At the same time, loose bounds on the properties which are crucial for devising a detection strategy for DM particles - the mass and coupling to ordinary matter.

Observations and particle properties of DM

Assuming a particle formulation, astro/cosmo observables provide mainly informations on the properties that DM does not have, e.g.: it needs to be non-baryonic, non-relativistic at the phase of matter-radiation equality, … This is enough to say that DM is NOT within the SM of particle physics. At the same time, loose bounds on the properties which are crucial for devising a detection strategy for DM particles - the mass and coupling to ordinary matter. Particle models cover a large part of the available range of masses and interactions: sub-eV axions, keV sterile neutrinos, GeV-TeV WIMPs, supermassive DM close to the Planck scale; gravitinos with gravitational interactions, numerous weakly interacting DM candidates, mirror DM with strong self-interactions, …

i-

• f

and

s, photons

Kim& Carosi, 2010

Guidelines to narrow the DM problem?

Focussing corresponds almost always to ratify a prejudice. Possible criteria to support such option include:

• A clean production mechanism, e.g.: thermal production (symmetric,

asymmetric), non-thermal states (e.g. from heavier state decays), production as a condensate, gravitational production, …

• A motivation from an open problem in the SM of particle physics, e.g.:

the naturalness problem, the violation of CP in strong interactions, a mechanism to explain neutrino masses, …

• An impact on observables in cosmology or astrophysics, in connection,

e.g., to the possibly discrepancies of the SM with observations on small scales (non-linear regime; galactic and sub-galactic scales; central over densities and abundance of substructures), e.g.: W arm DM, self- interacting DM, DM carrying macroscopic quantum effects, DM with with non-standard couplings with photons or baryons. Numerical N- body simulations are starting to touch these cases; still to be cleared is the role of baryons in DM numerical simulations.

Guidelines to narrow the DM problem?

• An “aesthetic” motivation in analogy to other counterparts, e.g.:

Asymmetric DM relying on a mechanism explaining the reason why the density of baryons and DM are comparable.

• A “pragmatic” motivation: lacking incontrovertible evidence for new

physics at accelerators, DM may be the only window for new physics.

• A “contingent” motivation: given some “anomaly” (e.g. an excess in the

radiation detected towards the GC) you study the class of compatible candidates (mass, interaction, annihilation or decay mode) without (necessarily) a reference particle framework.

• A systematic evaluation of what is experimentally accessible

“Historical” DM candidates - (SUSY) WIMPs, axions, sterile neutrinos, … - have mainly been motivated as relying on a natural production mechanism and, at the same time, carrying a particle physics motivation; should one give up on such approach?

Ωχh2 ' 3 · 10−27cm−3s−1 hσAviT =Tf

WIMP “miracle”

Ωχh2 Mχ s0 Y eq

χ (Tf)

ρc/h2 Mχ s0 ρc/h2 H(Tf) s(Tf)σAvTf Mχ Tf g

χ

geff 1 · 10−27cm−3s−1 σAvT =Tf

with: Mχ/Tf ∼ 20 (freeze-out + entropy conservation) (standard rad. dominated cosmology)

The WIMP recipe to embed a dark matter candidate in a SM extension: foresee an extra particle that is stable (or with lifetime exceeding the age

• f the Universe), massive (non-relativistic at freeze-out) and weakly
• interacting. Plenty of frameworks in which it is viable to apply this recipe.

χ

1 10 100 1000 0.0001 0.001 0.01

Yχ ≡ nχ s

Γ(Tf) = neq

χ (Tf)σAvT =Tf H(Tf)

CDM particles as thermal relics

Thermal equilibrium of enforced via:

χ

q q _

! !

_

annihilation

CP

q q _

! !

_

production

tests at LHC

scattering

q

! !

q crossing symmetry

direct detection

WIMP coupling to ordinary matter:

crossing symmetry

Early Universe

≈ halo annihilations

A model independent

approach to WIMP detection?

! !

_

annihilation

Early Universe

≈ halo annihilations

! !

_

production

tests at LHC

scattering

q

! !

q crossing symmetry ???

direct detection

Back to WIMP coupling to ordinary matter:

crossing symmetry ??? SM

CP ???

??? p p ___ SM “light” ??? ???

Details in the model

may become critical for detection strategy

LHC BSM null detection and the WIMP framework

The viability of a WIMP DM is significantly reshaped but not ruled out. E.g.: among the viable option for MSSM neutralino DM prior the LHC:

• 1. light Bino (SU(2) singlet) annihilating into fermions via t- & u-channel

exchange of moderately light sfermion (bulk region of the CMSSM);

• 2. annihilation on a s-channel Higgs resonance;
• 3. coannihilation with a sfermion quasi degenerate in mass;
• 4. well-tempering of Bino-Higgsino fraction;
• 5. pure Higgsino (SU(2) doublet) of 1.1 TeV mass or pure Wino (SU(2)

triplet) of 2.5 TeV mass.

• nly 1. has been wiped out by the LHC (but was already in trouble because
• f flavour observables and Higgs mass limits); 2. has been reshaped by the

(SM) Higgs discovery; 3., 4. & 5. have not (or marginally) addressed.

Direct detection target Indirect detection target

The viewpoint that new states close to the EW scale are needed to address the hierarchy problem, a pillar of beyond-SM searches, is severely shaking.

The current (pragmatic) tendency is to replace the model building attitude which was applied up to around early 2000 or so: Construct a natural theory; discover that WIMPs are predicted as part of the spectrum of such natural theory; argue that there is a mechanism enforcing the stability of the lightest of these; compute its thermal relic density and discover that you can solve the DM puzzle! (SUSY , large extra dimensions, Randall-Sundrum model, …) with: Give up on naturalness; construct ab-initio the model assuming the existence of a stable WIMP (sometimes protecting the stability by ah-hoc symmetry) with thermal relic density matching the DM density; match lab constraints, most often making the theory looking SM-like under flavour and EW observables. (split SUSY , minimal DM models, …) sometimes preserving extra virtue (e.g.: gauge coupling unification), sometimes as DM as the only target! DM detection data as THE guideline!

LHC BSM null detection and the WIMP framework

Pair annihilations

• f WIMPs in

the early Universe (i.e. at T= T )

Indirect detection of WIMP dark matter

f.o.

A chance of detection stems from the WIMP paradigm itself:

Pair annihilations

• f WIMPs in

DM halos (i.e. at T≅0)

! ! species particles SM lighter stable annihilation 2-body final state into, e.g., a fragmentation and/or decay process

Indirect detection of WIMP dark matter

A chance of detection stems from the WIMP paradigm itself:

Focus on: antiprotons, positrons, antideuterons, gamma-rays, (neutrinos) WIMP DM source function (sum over all processes : )

Qi(~ r, E) = (Av)T =0 X

f

dN f

i

dE (E) BfNpairs(~ r) χχ → f ¯ f

Is is fair to assume:

(σAv)T =0 ⇠ hσAviT =Tf

? Counterexamples: coannihilations, non-thermal WIMPs, ...

Pair annihilations

• f WIMPs in

DM halos (i.e. at T≅0)

! ! species particles SM lighter stable annihilation 2-body final state into, e.g., a fragmentation and/or decay process

Indirect detection of WIMP dark matter

A chance of detection stems from the WIMP paradigm itself:

Focus on: antiprotons, positrons, antideuterons, gamma-rays, (neutrinos) WIMP DM source function (sum over all processes : )

Qi(~ r, E) = (Av)T =0 X

f

dN f

i

dE (E) BfNpairs(~ r) χχ → f ¯ f

: learn it from dynamical

• bservations (?) or numerical simulations (?); huge scale mismatch

with respect to the DM clumping scale. A signature out of it?

Npairs(~ r ) ∝ [⇢χ(~ r )]2 ≡ [⇢DM(~ r )]2

Pair annihilations

• f WIMPs in

DM halos (i.e. at T≅0)

! ! species particles SM lighter stable annihilation 2-body final state into, e.g., a fragmentation and/or decay process

Indirect detection of WIMP dark matter

A chance of detection stems from the WIMP paradigm itself:

Focus on: antiprotons, positrons, antideuterons, gamma-rays, (neutrinos) WIMP DM source function (sum over all processes : )

Qi(~ r, E) = (Av)T =0 X

f

dN f

i

dE (E) BfNpairs(~ r) χχ → f ¯ f

What final state? Hard or soft yields? The WIMP paradigm relies on a generic coupling between WIMPs and thermal bath particles: it does not address in any way these issues!!! A signature out of it?

Hints of particle DM indirect detection?

GeV γ-ray excess at the GC?

several authors, 2009-14

Flux (cnts s

• 1 keV
• 1)
• 0.02
• 0.01

Residuals 3 3.2 3.4 3.6 3.8 4

Energy (keV)

• Eff. Area (cm

3.57 ± 0.02 (0.03) XMM-MOS Full Sample 6 Ms

3.5 keV X-ray line?

Bulbul et al., 2014

• r away from a

GeV/TeV DM state to a keV , e.g., sterile neutrino

• r a axion-like

particle:

1 10

2

10

AMS-02

• 1

10

PAMELA Fermi

a DM induced positron excess?

Pamela, Fermi, AMS02, 2009-14

Fermi 130 GeV γ-ray line? Weniger, 2012 + others, e.g.: the 511 keV GC line (Knödlseder et al.,

2003), connections to the

light-mass direct detection puzzle, …

A γ-ray excess at ~ GeV energies towards the GC ?

When comparing the flux measured by Fermi in the Galactic center region against models accounting for diffuse emission and point sources, a residual at the level of 10% of the total intensity emerges, Morselli & Vitale

(Fermi Coll.), 0912.3828 (+ Goodenough & Hooper, 0910.2998):

e 7◦×7◦ ROI.

diffuse emission according to a physical propagation model (with Galprop) tuned on data away from GC; isotropic background from all sky data; point sources from the Fermi catalogue. Still: it is tricky to include properly systematic effects and uncertainties in the modelling of each component. An exercise which has been repeated over the years with different assumptions, with the goal of addressing whether this is a DM excess:

A γ-ray excess at ~ GeV energies towards the GC ?

The attractive feature is the morphology signature, as expected from the enhancement in DM density towards the Galactic center: Hooper & Linden,

1110.0006 (building up from 0910.2998 and 1010.2752) use a template fitting

procedure to claim a DM signal in the inner few degrees: consistent with , e.g. from N-body result + mild adiabatic contraction from baryon infall

ρDM ∝ r−1.3 ρNFW ∝ r−1

Spectrum with ~ 10 GeV cutoff

A γ-ray excess at ~ GeV energies towards the GC ?

Result confirmed in other independent works, considering slightly different approaches to model the background, see Abazajian & Kaplinghat,

1207.6047 and Gordon & Macias, 1306.5725:

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

distance [deg]

50 100 150 200 250

surface brightness [counts/pixel] NFW (γ = 1.2) NFW (γ = 1.3) Point source model

again cuspy profile, not point source

100 101 102

Eγ [GeV]

10−7

E2dN/dE [GeV cm−2 s−1]

2 × 10−8

Inner slope: γ = 1.2

MDM = 10 GeV, 100% b¯ b MDM = 30 GeV, 100% b¯ b MDM = 10 GeV, 100% τ +τ −

again a sharp cut-off

5 10 15 20 25 30 35 40 45

MDM [GeV]

10−27 10−26 10−25

< σv > [cm3/s]

100% b¯ b 50% b¯ b, 50% leptons 10% b¯ b, 90% leptons

Inner slope: γ = 1.3 1σ CI, this work 2σ CI, this work Hooper & Linden (2011)

best fit with WIMP annihilations into b-quarks

• r leptons or a

combination

A γ-ray excess at ~ GeV energies towards the GC ?

Residuals searched for and found also in different parts of the sky; in particular Hooper & Slatyer, 1302.6589 find consistent energy spectrum and morphology at slightly higher latitudes, in the Fermi bubbles region, where assumptions on the background needs to be different, but still very uncertain (see also results from Huang et al., 1307.6862): final state

ρDM ∝ r−1.2

Mχ = 10 GeV

τ +τ −

Hooper et al., 1305.0830 show that this is inconsistent with an unresolved

population of millisecond pulsars (MSPs) with the same spectral features as those measured by Fermi for MSPs in the sun neighbourhood. This point has been questioned in other analyses, see Yuan & Zhang,1404.2318.

A γ-ray excess at ~ GeV energies towards the GC ?

Latest update from Daylan et al., 1402.6703: better angular tagging (Fermi CTBCORE parameter discrimination), template fitting on large sky patch all the way from |b|>1° to the Fermi bubble region, higher statistics:

0.5-1 GeV residual

• 20
• 10

10 20

• 20
• 10

10 20 5 10 15 20 5 10 15 20 10-6 counts/cm2/s/sr

1-2 GeV residual

• 20
• 10

10 20

• 20
• 10

10 20 2 4 6 8 10 2 4 6 8 10 10 counts/cm /s/sr

2-5 GeV residual

• 20
• 10

10 20

• 20
• 10

10 20 1 2 3 4 5 1 2 3 4 5 10-6 counts/cm2/s/sr

5-20 GeV residual

• 20
• 10

10 20

• 20
• 10

10 20 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 10 counts/cm /s/sr

A γ-ray excess at ~ GeV energies towards the GC ?

Minor changes for slightly different angular and/or energy cuts; signal robustly associated with an approximately spherical component (not a contaminant from the disc) and extending up at least 10° (hardly compatible with a MSP component). Latest update from Daylan et al., 1402.6703: better angular tagging (Fermi CTBCORE parameter discrimination), template fitting on large sky patch all the way from |b|>1° to the Fermi bubble region, higher statistics:

A γ-ray excess at ~ GeV energies towards the GC ?

… however, on top of the MSP hypothesis, there are other possible (and plausible) explanations:

Daylan et al., 1402.6703 DM bump Carsor & Profumo, 1405.7685 hadronic

model from recent SNR bursts

Petrovic, Serpico & Zaharijas, 1405.7928:

leptonic model recent-past from injection from GC black hole

0.5 1.0 2.0 5.0 10.0 20.0 1 ¥ 10-7 2 ¥ 10-7 5 ¥ 10-7 1 ¥ 10-6 2 ¥ 10-6 5 ¥ 10-6

E @GeVD

E F @GeV cm-2s-1sr-1D

b0, t0, E0 0.3 b0, 3 t0, 2 E0 3 b0, 0.3 t0, 0.5 E0

morphology can be adjusted in all such cases

φγ(r) ∝ 1 r2.4 or 1 r2.6

A γ-ray excess at ~ GeV energies towards the GC ?

To some extent, a scary analogy with the EGRET GeV excess:

” well”

Dark matter solution

de Boer et al., astro-ph/0508617

data fit improves adding a 70 GeV WIMP annihilating into bb _ models for diffuse emission (Galprop) under-predicting the “measured” flux

” well”

E [MeV]

The 1997-2005 puzzle turned out to be related to a miscalculation of the EGRET sensitivity above a few GeV and Fermi did not confirm the excess (Atwood et al., 0902.1089); Fermi has systematics errors which are difficult to address below 1 GeV: a new mission in soft γ-rays (maybe on the same satellite as Gamma-400?) to understand the 2009-14 GeV excess? pointed out first in:

Hunter et al., 1997

A γ-ray excess at ~ GeV energies towards the GC ?

To some extent, a scary analogy with the EGRET GeV excess:

” well”

Dark matter solution

de Boer et al., astro-ph/0508617

data fit improves adding a 70 GeV WIMP annihilating into bb _ models for diffuse emission (Galprop) under-predicting the “measured” flux

” well”

E [MeV]

For the 1997-2005 puzzle alternative explanations were also available, and the dark matter interpretation in tension with (or at least not too naturally accounted for by) other indirect detection signals; is this repeating with the 2009-14 GeV puzzle? see later … pointed out first in:

Hunter et al., 1997

A γ-ray line at 130 GeV in FERMI data ???

Bringmann et al., 1203.1312; Weniger,1204.2797 look at the Galactic center,

• ptimizing the search region with respect to the assumption on the DM

density profile (assuming a simple power-law background) and find a 3.2σ statistical significance (if “look elsewhere” effect included) for a monochromatic signal at about 130 GeV:

3 σ 1 σ

Compatible with line limits from the whole sky: Fermi-LAT coll., arXiv:1205.2739, as well as from dwarfs: Geringer-Sameth & Koushiappas,

arXiv:1206.0796

A γ-ray line at 130 GeV in FERMI data ???

Su & Finkbeiner, arXiv:1206.1616 use a template fitting method and claim

“strong evidence”, with local significance of 5 o 6 σ for 2 lines at 111 & 129 GeV! Template for the DM cusp

• ff-centered by 1.5º (200 pc)

Off-center due to a density wave excitation by the stellar components? Matching a hydrodynamical N-body result Kuhlen et al., arXiv:1208.4844

Hektor et al., 2012 find evidence for 2 lines at 3.6σ from stacked analysis of 18

galaxy clusters; Su & Finkbeiner, 2012 at 3.3σ from unassociated LAT sources

A γ-ray line at 130 GeV in FERMI data ???

Fermi Coll., arXiv:1305.5597 addressed the issue with reprocessed data,

updated 2D energy reconstruction and more data, as well as optimized regions:

(GeV)

χ

m

10

2

10

)

• 1

s

3

95% CL Limit (cm

γ γ

v> σ <

• 30

10

• 29

10

• 28

10

• 27

10

• 26

10

• 25

10 3.7 year R16 Einasto Profile Observed Upper Limit Expected Limit Expected 68% Containment Expected 95% Containment Weniger [20] Limit

Limits falling within the expected bands

Energy (GeV) Events / 5.0 GeV

10 20 30 40 50 60 70

= 133.0 GeV

γ

P7_REP_CLEAN R3 2D E = 17.8 evts

sig

n σ = 3.3

local

s = 276.2 evts

bkg

n = 2.76

bkg

Γ (c) Energy (GeV)

60 80 100 120 140 160 180 200 220

) σ

• Resid. (
• 4
• 2

2 4

Local 3.3σ corresponding to only a global 1.5σ

Excesses in FERMI γ-ray data ???

A monochromatic signal + continuum counterpart in a model with physical background: Cholis, Tavakoli & P.U., arXiv:1207.1468 A fit with several degeneracies, given the many components in the fit: Diffuse emission + point sources + DM component Significantly away from the typical 1-loop over tree-level ratio: definite guideline for the DM model?

upper limit on: lines best fit Zγ + 2γ Zγ + Hγ μ μ + −

W W

+ −

b b _

τ τ + − + − e e

Sample model fitting the data:

A γ-ray line at 130 GeV in FERMI data ???

Other puzzles: about a 3σ evidence for a 130 GeV line in low-incidence- angle Earth limb data (Finkbeiner et al. & Hektor et al., 2012; Fermi Coll., arXiv:

1305.5597) and within 5° from the position of the Sun (Whiteson 2013)!

UPDATE!

Evolution of the effect in time (Weniger, ICTP DM workshop 2013): Dashed/dotted lines: 68% and 95% CL containment regions for real signal and statistical fluke. Fermi modified survey strategy to address the issue on a shorter timescale. The effect as initially claimed possibly confirmed by HESS II at 5σ for systematics under control (Bergström et al. 2012) by october this year!

Present and future of γ-ray lines

A new analysis by Fermi

• n lines below 10 GeV

,

Albert et al., 1406.3430:

‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡

This Work Hstat. + syst.L This Work Hstat. onlyL Fermi-LAT 3.7 yr Hstat. onlyL Vertongen, Weniger Hstat. onlyL EGRET Galactic Centre Hstat. onlyL

all limits at 95% CL

0.1 0.3 1 3 10 30 10-32 10-31 10-30 10-29 10-28 DM Mass mDM HGeVL DM Annihilation Cross Section <sv>gg Hcm3 s-1L

A revival of theoretical models having very large monochromatic signals, such as, e.g., leptophylic (e.g.:

Bergström, 2012) or scalar DM

(e.g.: Giacchino et al., 2013)

0.1 1 10−4 10−3 10−2 10−1 100 101

x = Eγ/mχ x2 dNγ/dx µ = 9

f ¯ fγ γγ γZ Total

scalar DM + exotic vector-like fermion mediating coupling to SM (Ibarra et al., 2014) ΔE/E = 10% “GeVish" lines fading away,

Present and future of γ-ray lines

A new analysis by Fermi

• n lines below 10 GeV

,

Albert et al., 1406.3430:

‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡

This Work Hstat. + syst.L This Work Hstat. onlyL Fermi-LAT 3.7 yr Hstat. onlyL Vertongen, Weniger Hstat. onlyL EGRET Galactic Centre Hstat. onlyL

all limits at 95% CL

0.1 0.3 1 3 10 30 10-32 10-31 10-30 10-29 10-28 DM Mass mDM HGeVL DM Annihilation Cross Section <sv>gg Hcm3 s-1L

“GeVish" lines fading away, A new generation of instruments with the potential of searching for this smoking-gun signal;

• CTA, air Cherenkov telescope, 2018?, effective area ~ 10 m &

Δθ ~ 0.05° at 1 TeV , 100 GeV - few TeV;

• DAMPE, Chinese satellite, 2016?, training mission for HERD;
• Gamma-400, Russian/Italian satellite, 2018?, ΔE/E ~ 1% & Δθ ~ 0.01°

at E > 100 GeV , effective area ~ 0.4 m , 100 MeV - 3 TeV;

• HERD, Chinese space station, 2020?, ΔE/E ~ 1% & Δθ ~ 0.01° at

E ~ 100 GeV , effective acceptance ~ 4 m sr

2 2 2 6

Searches for a γ-ray flux from MW dwarf satellites

Signature: identify a γ-ray signal from objects which are DM dominated, have gas and plasma components below detectable levels and hence a very low internal contamination from standard astrophysical backgrounds. Unfortunately there no dwarf is “bright” in γ-rays: upper limits from null detections, from, e.g., Hess (1012.5602), and most recently Fermi (1310.0828):

UMi Leo IV Her Sex Seg 1 UMa I Dra Com Leo I CMa Wil 1 For Sgr Scl CVn II Seg 2 Car Boo II Leo II Psc II Boo III CVn I UMa II Leo V Boo I

the 25 dwarfs discovered so far

Searches for a γ-ray flux from MW dwarf satellites

Signature: identify a γ-ray signal from objects which are DM dominated, have gas and plasma components below detectable levels and hence a very low internal contamination from standard astrophysical backgrounds. Unfortunately there no dwarf is “bright” in γ-rays: upper limits from null detections, from, e.g., Hess (1012.5602), and most recently Fermi (1310.0828):

UMi Leo IV Her Sex Seg 1 UMa I Dra Com Leo I CMa Wil 1 For Sgr Scl CVn II Seg 2 Car Boo II Leo II Psc II Boo III CVn I UMa II Leo V Boo I

103 104 105

10−7 10−6 10−5

Energy (MeV)

Energy Flux (MeV cm−2 s−1)

bin-by-bin energy flux limits and expected sensitivities from 500 MeV to 500 GeV & 4 yrs of LAT data

Draco

the 25 dwarfs discovered so far

Searches for a γ-ray flux from MW dwarf satellites

Signature: identify a γ-ray signal from objects which are DM dominated, have gas and plasma components below detectable levels and hence a very low internal contamination from standard astrophysical backgrounds. Unfortunately there no dwarf is “bright” in γ-rays: upper limits from null detections, from, e.g., Hess (1012.5602), and most recently Fermi (1310.0828):

UMi Leo IV Her Sex Seg 1 UMa I Dra Com Leo I CMa Wil 1 For Sgr Scl CVn II Seg 2 Car Boo II Leo II Psc II Boo III CVn I UMa II Leo V Boo I

103 104 105

10−7 10−6 10−5

Energy (MeV)

Energy Flux (MeV cm−2 s−1)

Ursa Minor

bin-by-bin energy flux limits and expected sensitivities from 500 MeV to 500 GeV & 4 yrs of LAT data the 25 dwarfs discovered so far

Searches for a γ-ray flux from MW dwarf satellites

Signature: identify a γ-ray signal from objects which are DM dominated, have gas and plasma components below detectable levels and hence a very low internal contamination from standard astrophysical backgrounds. Unfortunately there no dwarf is “bright” in γ-rays: upper limits from null detections, from, e.g., Hess (1012.5602) and most recently Fermi (1310.0828): For this limit you need to assume a dynamical model for the tracer stellar population and a density profile for the DM; a number of simplifying assumptions implemented. Slightly weaker limits than in the previous analysis (1108.3546) on 10 dwarfs; still touching thermal DM candidates.

101 102 103 mDM (GeV) 10−22 10−23 10−24 10−25 10−26 hσvi (cm3 s−1) Combined τ +τ −

Maximum Likelihood Bayesian Median Expected 68% Containment 95% Containment UMi Leo IV Her Sex Seg 1 UMa I Dra Com Leo I CMa Wil 1 For Sgr Scl CVn II Seg 2 Car Boo II Leo II Psc II Boo III CVn I UMa II Leo V Boo I

the 25 dwarfs discovered so far

stacking of 15 dwarfs

τ +τ −

What about the lepton puzzle?

1 10

2

10

AMS-02

• 1

10

PAMELA Fermi Energy (GeV) 0.5 1 2 3 4 10 20 100 200 10002000 )

2

GeV

• 2

m

• 1

sr

• 1

(s

3

E × Flux 1 10

2

10

PAMELA HEAT94+95 AMS CAPRICE94 MASS91 ATIC Kobayashi BETS Fermi HESS

positron fraction electron, electron+positron fluxes

What about the lepton puzzle?

From the spectacularly precise recent measurements of the leptonic components in cosmic ray, by AMS02, Pamela, Fermi, ..., we learned that the picture with electrons as primaries from SNRs and positrons as secondaries from the interaction of CRs on the ISM is wrong!

positron fraction

Y

• u need extra (hard) positron sources and they need to be close to us

(because 100 GeV - few TeV leptons lose energy on short timescale). In most cases WIMP DM does not have the correct spectrum; you saturate the

• bserved spectra, e.g., with pulsar sources and get a competitive WIMP

limit.

electron+positron fluxes

AMS02 2014 update AMS02 2014 update

Although harder with AMS, fair agreement with data, can still be obtained from toy models of annihilating WIMPs, e.g.: Cholis & Hooper, arXiv:1304.1840

MΧ500 GeV 1.6 TeV ΧΧΦΦ2e, 2Μ, 2Π at 1:1:2

1 5 10 50 100 0.02 0.05 0.10 0.20 0.50 1.00 E GeV eee

MΧ500 GeV 1.6 TeV ΧΧΦΦ2e, 2Μ, 2Π at 1:1:2

1 10 100 1000 1 5 10 50 100 500 1000 E GeV ee E3x diff. flux GeV2m2 s sr1

What about saturating the extra component with DM only? large boost factors, leptophilic channels Or you can saturate the flux with some other component and set limits

• n DM bumps, see, e.g.,

Bergström et al., 2013; Ibarra et al. 2014:

for ¡” ”, ¡ – …

Testing the DM hypothesis against other possibilities?

• V

ery hard from CR lepton data alone; possible falsification of the DM hypothesis from the detection of angular anisotropies in the flux (still one should be confident about modeling propagation in the local environment).

• In principle possible by looking at the radiative emissions associated to

the extra lepton components. Sources confined to the disc (as in case of pulsars) or spread out in the whole diffusive halo (as for DM annihilations) produce very different vertical lepton density profiles:

• 4
• 3
• 2
• 1

1 2 3 4 z [kpc] 10

• 4

10

• 3

10

• 2

10

• 1

10 Φe [MeV cm

• 2 s
• 1sr
• 1]

~ ρ

2 (arbitrary normalization)

DMe CR primary secondary at source secondary in ISM Ee = 200 GeV R = 8 kpc

Regis & P.U., arXiv: 0904.4645

Primary/secondary astrophysical components mostly localized at z≅0 versus a DM term extending to much larger z High-latitude inverse Compton and synchrotron profiles are sensibly different in the two case and should be distinguishable in the future.

Testing the DM hypothesis against other possibilities?

• Consistency checks are also possible looking at radiative emissions from

the central region of the Galaxy, however these are much more model

• dependent. In particular they heavily rely on what extrapolation one takes

for the dark matter distribution from the local neighborhood to the the Galactic center and on magnetic fields + energy losses models; in case of Einasto profile, the tension with currently available radio data is very severe (e.g. Bertone et al. 2009).

• Limits from “polluting” the early Universe with DM yields:

Slatyer et al., arXiv: 0906.1197

CMB limits: mainly from ionization of the thermal bath, Ly-α excitation of Hydrogen and heating

• f the plasma

What about the broader multi-messenger picture?

• measurements
• f the local

antiproton flux;

• measurements
• f the local

lepton fluxes;

• γ-rays at low,

intermediate and high latitudes. In most cases a given DM yield impacts several WIMP indirect detection channels at the same time: a multi-messenger and multi-wavelength

• pportunity to find a signal and cross check limits. A problem requiring a

fully self-consistent approach when comparing different observables. E.g.: within a given WIMP yield, a given DM distribution, a given model for propagation of CRs in the galaxy, a given model for (radiative) γ-ray emissivities (connected to the gas and ISRF models), you can extract limits from (Tavakoli, PU et al., 1308.4135):

What about the broader multi-messenger picture?

• measurements
• f the local

antiproton flux;

• measurements
• f the local

lepton fluxes;

• γ-rays at low,

intermediate and high latitudes.

Daylan et al., 1402.6703

3σ GC signal region; shown to guide the eye

• nly!!

Inconsistent frameworks! In most cases a given DM yield impacts several WIMP indirect detection channels at the same time: a multi-messenger and multi-wavelength

• pportunity to find a signal and cross check limits. A problem requiring a

fully self-consistent approach when comparing different observables. E.g.: within a given WIMP yield, a given DM distribution, a given model for propagation of CRs in the galaxy, a given model for (radiative) γ-ray emissivities (connected to the gas and ISRF models), you can extract limits from (Tavakoli, PU et al., 1308.4135):

Model dependence of DM antiprotons predictions

Predictions for secondary (background) antiprotons, stemming from CR propagation models calibrated on other secondary/primary CR ratios, are fairly robust; the same is not true primaries from DM annihilations: secondary/primary B/C secondary p (given primary p) _ primary p ( ) _

Mχ = 200 GeV, W +W −

Kraichnan Kolmogorov convective “thick” “thin” factor ≈ 50 Einasto Burkert (cored) factor ≾ few

sample set of CR propagation model: sample set of DM profiles:

Evoli, PU et al., 1108.0664

More on multi-messengers and the GeV γ-ray excess

A few recent analyses have tried to cross-correlate multi messenger limits with the putative GC γ-ray signal. E.g.: Cirelli et al., 1407.2173 point to ps: _

20 40 60 80 100 10-27 10-26 10-25 10-24

MDM @GeVD Xsv\ @cm3 s-1D

THN CON KOL THK KRA Benchmark propagation models

φ¯

p F = φp F fixed

upper limits for b-b final state: _

Daylan et al., 1402.6703

3σ GC signal region

2

20 40 60 80 100 10-27 10-26 10-25 10-24

MDM @GeVD Xsv\ @cm3 s-1D

THN CON KOL THK KRA Benchmark propagation models

φ¯

p F 2 [0.1, 1.1] GV

different solar modulation assumptions Bringmann et al., 1406.6027

stress the role of radio data (result depending

• n magnetic field

assumptions):

bb cc qq 80% t+t- G=1.04 rc =2pc 10 20 30 15 10-26 2 3 5

mc@GeVD sv@cm3êsD

Improving limits and/or detecting DM with p & D ?

With the upcoming AMS data on nuclei, our understanding of the CR propagation model will be refined. A window for singling out DM?

_

Evoli, Cholis, Grasso, Maccione & PU, arXiv:1108.0664

_

Tough, but not inconceivable, in the p channel with AMS:

_

Possibly more promising in the D channel:

_

today AMS GAPS Donato, Fornengo & Maurin, 2008 WIMP secondary

_

NOTE: much cleaner DM signature compared to p

More on DM γ-ray predictions and limits

Rather than low, intermediate & high latitude, follow the strength of the limit on the whole sky: plotting: (σv)3σ/(σv)3σ

min

with: The darker the region, the tighter the constraint gas density scaled up to a factor of 2

1.19 18.1 18.6 52.1 12.6 1.00 3.93 36.3 18.2 34.3 20.4 1.94 2.83 20.9 21.4 15.4 18.8 2.22 4.57 11.5 35.7 23.5 17.7 2.51 43.7 49.9 51.9 54.5 44.0 36.0 1.69 34.8 8.84 103 12.1 1.10 3.43 29.9 17.9 36.2 15.1 2.82 9.29 47.4 89.1 14.3 13.1 2.25 2.39 12.7 22.3 21.8 8.30 2.14 40.0 42.0 36.2 29.7 41.0 44.8

150 100 50 50 100 150 50 50 l b

1.32 113 840 115 14.9 1.06 30.3 81.6 671 50.0 27.7 2.32 36.8 47.4 272 16.0 41.5 2.50 59.1 43.3 88.1 84.4 52.6 46.3 137 111 100 112 184 149 1.82 46.7 714 267 15.5 1.00 9.44 62.2 165 62.7 18.0 2.82 9.29 47.4 89.1 14.3 13.1 2.94 31.4 24.6 104 32.3 10.2 17.6 58.2 56.8 57.1 37.2 48.6 84.8

150 100 50 50 100 150 50 50 l b

Mχ = 10 GeV, b¯ b

(σv)3σ

min = 2.5 · 10−27cm3s−1

reference model with “fixed” gas

Tavakoli, PU et al., 1308.4135

Rather than low, intermediate & high latitude, follow the strength of the limit on the whole sky: plotting: (σv)3σ/(σv)3σ

min

with:

1.32 113 840 115 14.9 1.06 30.3 81.6 671 50.0 27.7 2.32 36.8 47.4 272 16.0 41.5 2.50 59.1 43.3 88.1 84.4 52.6 46.3 137 111 100 112 184 149 1.82 46.7 714 267 15.5 1.00 9.44 62.2 165 62.7 18.0 2.82 9.29 47.4 89.1 14.3 13.1 2.94 31.4 24.6 104 32.3 10.2 17.6 58.2 56.8 57.1 37.2 48.6 84.8

150 100 50 50 100 150 50 50 l b

(σv)3σ

min = 2.5 · 10−27cm3s−1

in the “reference” case.

• 2.20

0.39 0.055 1.94 2.07 2.31 0.31 1.16 0.066 1.68 1.80 2.03 0.19 0.82 0.19 2.34 0.62 0.46 0.19 0.50 1.12 0.68 0.73 0.13 0.78 0.77 1.27 0.80 0.59 0.59 2.26 1.82 0.030 1.64 1.91 2.670.51 0.83 0.16 1.68 0.28 0.47 0.33 0.77 1.34 1.97 1.85 0.51 0.19 0.60 0.52 1.64 1.99

0.16

1.68 1.80 1.54 2.01 2.06 1.29

150 100 50 50 100 150 50 50 l b

plotting the largest departure from 1 of: (σv)3σ

i /(σv)3σ ref

where “i” labels a set of models with different assumptions on the gas (reddish regions) or on the ISRF (greenish regions). The brightest the color, the less robust the limit:

Mχ = 10 GeV, b¯ b

More on DM γ-ray predictions and limits

Tavakoli, PU et al., 1308.4135

Project the limit into latitude bins and translate them from the sample Einasto halo profile into other possibilities:

ρEin ∝ exp  − 2 α (xα − 1)

• ρBur ∝

1 (1 + x)(1 + x2)

clumpy: Npairs ∝ ρ All normalized to a local halo density: ρ(R) = 0.4 GeV cm3

10

• 27

10

• 26

10

• 25

10 20 30 40 50 60 70 80 90

b [degrees] v [cm3 s-1]

Einasto Burkert clumpy M = 10 GeV, bb

– final state, free gas

10

• 27

10

• 26

10

• 25

10 20 30 40 50 60 70 80 90

b [degrees] v [cm3 s-1]

Einasto Burkert clumpy M = 10 GeV, bb

– final state, fixed gas

More on DM γ-ray predictions and limits

Play it even harder and define the density profile as log-log interpolation

• f a set of discrete values at the galactocentric distances

corresponding to the radii at the tangential points in the latitude bins. Assume also that the profile is monotonic and that:

ρ(r) = ρEin(r) for r > R

ri ρi

Fix the annihilation rate, and generate a random sample of , testing whether each configuration is excluded by the flux limits in all latitude bins. For all surviving models, consider the bin encompassing the GC and compute the line of sight integration factors obtained by imposing that the density profile is constant below . Plot the maximum

• f in the sample and compare it to

the analogous quantity for the preferred parametric profile:

ρi Ji ri Ji

10

• 2

10

• 1

1 10 1 2 3 4 5 6 7 8 9

R [ kpc ] Ji max [ GeV2 cm-6 kpc sr ]

Einasto Burkert 3 10-27 cm3 s-1 10-26 cm3 s-1

M = 10 GeV, bb

– final state, fixed gas

More on DM γ-ray predictions and limits

Play it even harder and define the density profile as log-log interpolation

• f a set of discrete values at the galactocentric distances

corresponding to the radii at the tangential points in the latitude bins. Assume also that the profile is monotonic and that:

ρ(r) = ρEin(r) for r > R

ri ρi

Fix the annihilation rate, and generate a random sample of , testing whether each configuration is excluded by the flux limits in all latitude bins. For all surviving models, consider the bin encompassing the GC and compute the line of sight integration factors obtained by imposing that the density profile is constant below . Plot the maximum

• f in the sample and compare it to

the analogous quantity for the preferred parametric profile:

ρi Ji ri Ji

10

• 2

10

• 1

1 10 1 2 3 4 5 6 7 8 9

R [ kpc ] Ji max [ GeV2 cm-6 kpc sr ]

Einasto Burkert 3 10-27 cm3 s-1 10-26 cm3 s-1 3 10-26 cm3 s-1

M = 10 GeV, bb

– final state, free gas

More on DM γ-ray predictions and limits

Summary and conclusions

• Plenty of indirect (gravitational) evidence for DM from cosmological and

astrophysical observations, but loose connection to specific DM particle physics scenarios. May the indirect evidence for dark matter turn into evidence for dark matter particles? Can we trust any guideline (prejudice) and/or (re)focus the DM problem in directions previously overlooked?

• W

ere clean WIMP signature unavailable, the key would be to efficiently combine multi-messenger and multi-targets signals; such synthesis however may be particularly delicate.

• The LHC is challenging the beyond SM extensions based on naturalness,

and is reshaping the WIMP scenario: DM shifting from byproduct to key element for discussing SM extensions.

• A few “hints” of WIMP indirect detection have been claimed, but the

picture is far from being satisfactory. Clean WIMP DM signatures are being tested or will be tested in the near future