THE WIMP IS DEAD! LONG LIVE THE WIMP!
Dan Hooper – Fermilab and the University of Chicago Light Dark World 2018, KAIST, Daejeon December 17, 2018 Dan Hooper – The WIMP is Dead! Long Live the WIMP! The Case for WIMPs § To simplify the discussion, lets make two well-motivated assumptions: 1) Standard expansion history in the early universe (radiation dominated) 2) The dark matter was in thermal equilibrium (at some point in time)
§ The abundance of dark matter that emerges from the Big Bang depends on the dark matter’s mass and couplings § A thermal relic with very small couplings will exceed the measured dark matter density, while one with very large couplings will constitute only a small fraction of the dark matter – to accommodate the measured abundance, something comparable to the weak force is required § Under these well-motivated assumptions, viable dark matter candidates must fall within a mass range of ~10 MeV to ~100 TeV § From this perspective, dark matter candidates with weak-scale masses and interactions – “WIMPs” – are particularly well motivated § This argument has also played a significant role in guiding our experimental program Dan Hooper – The WIMP is Dead! Long Live the WIMP!
Putting the WIMP Paradigm to the Test Dan Hooper – The WIMP is Dead! Long Live the WIMP! Direct Detection
• Over the past 15 years, constraints from direct detection experiments have improved with a Moore’s-law like behavior (a factor of 2 every 15 months) Dan Hooper – The WIMP is Dead! Long Live the WIMP! Direct Detection
• Over the past 15 years, constraints from direct detection experiments have improved with a Moore’s-law like behavior (a factor of 2 every 15 months) • Some important benchmarks: -1990s: Excluded the cross sections predicted for a WIMP that scatters and annihilates through Z-exchange -Last few years: Testing WIMPs that interact through Higgs exchange (including many SUSY models) Dan Hooper – The WIMP is Dead! Long Live the WIMP! Dark Matter at the LHC § With the exception the Higgs Boson, no evidence of new physics has appeared at the Large Hadron Collider § For the first time, much of the electroweak scale has now been explored; although there still exists room for SUSY (or your favorite EW scale theory)
Selected CMS SUSY Results* - SMS Interpretation ICHEP '16 - Moriond '17
~ ~ ~ ∼ 0 pp →g g , g → qq χ SUS-16-014 SUS-16-033 0l(MHT) ~ ~ ~ ∼ 01 pp →g g , g → qq χ SUS-16-015 SUS-16-036 0l(MT2) ~ ~ ~ ∼ 01 pp →g g , g → bb χ SUS-16-014 SUS-16-033 0l(MHT) ~ ~ ~ ∼ 01 pp →g g , g → bb χ SUS-16-015 SUS-16-036 0l(MT2) ~ ~ ~ ∼ 01 pp →g g , g → bb χ SUS-16-016 0l( α ) T ~ ~ ~ ∼ 01 pp →g g , g → tt χ SUS-16-014 SUS-16-033 0l(MHT) ~ ~ ~ ∼ 01 pp →g g , g → tt χ SUS-16-015 SUS-16-036 0l(MT2) ~ ~ ~ ∼ 01 pp →g g , g → tt χ SUS-16-016 0l( α ) T ~ ~ ~ ∼ 01 Preliminary pp →g g , g → tt χ SUS-16-019 SUS-16-042 1l( ∆ φ) ATLAS SUSY Searches* - 95% CL Lower Limits ATLAS ~ ~ ~ ∼ 01 pp →g g , g → tt χ SUS-16-020 SUS-16-035 2l same-sign ~ ~ ~ ∼ 01 December 2017 √s =7,8,13TeV
Gluino pp →g g , g → tt χ SUS-16-022 SUS-16-041 Multilepton ~ ~ ~ ∼ 01 miss 1 pp →g g , g → tt χ SUS-16-030 0l Model e,µ,τ, γ Jets E ! dt[fb− ] Mass limit √s =7,8TeV √s =13TeV Reference ~ ~ ~ ∼ 01 T pp →g g , g → tt χ SUS-16-037 1l(MJ) L ~ ~ ~ ~ ∼ 01 pp →g g , g → t t → t c χ SUS-16-030 0l (M - M = 20 GeV) 0 0 Mother LSP ˜ χ˜ st nd ~ ~ ~ ∼1± q˜q˜, q˜ qχ˜ 0 2-6 jets Ye s 36.1 q 1.57 TeV m( 1)<200 GeV, m(1 gen. ˜q )=m(2 gen. ˜q ) 1712.02332 pp →g g , g → bt χ SUS-16-033 0l(MHT) (M - M = 5 GeV) 1 1 ∼ ± LSP → 0 0 ~ ~ ~ ∼ ± ∼ 0 χ χ˜ mono-jet 1-3 jets Ye s 36.1 q˜ 710 GeV m(q˜)-m(χ˜ )<5GeV 1711.03301 pp →g g , g → qqχ → qq W χ SUS-16-019 SUS-16-042 1l( ∆ φ) x=0.5 1 q˜q˜, q˜ q 1 (compressed) 1 1 1 → 0 0 ~ ~ ~ ∼ ± ∼ 0 g˜ χ˜ pp →g g , g → qqχ → qq W χ SUS-16-020 SUS-16-035 2l same-sign x=0.5 g˜g˜, g˜ qq¯χ˜ 1 0 2-6 jets Ye s 36.1 2.02 TeV m( 1)<200 GeV 1712.02332 ~ ~ ~ ∼ 1 ∼ 01 → 0 0 0 pp →g g , g → qqχ ± → qq W χ SUS-16-020 SUS-16-035 2l same-sign (M - M = 20 GeV) χ˜ χ˜ 2-6 jets g˜ χ˜ χ˜± χ˜ 1712.02332 Interm. LSP g˜g˜, g˜ qq 1± qqW± 1 0 Ye s 36.1 2.01 TeV m( 1)<200 GeV, m( )=0.5(m( 1)+m(g˜)) ~ ~ ~ ∼ ∼ 0 1 ∼ 01 pp →g g , g → qq( χ ±/ χ ) → qq (W/Z) χ SUS-16-014 SUS-16-033 0l(MHT) x=0.5 → →0 0 g˜g˜, g˜ qq¯(ℓℓ)χ˜ ee,µµ 2 jets Ye s 1 4 . 7 g˜ 1.7 TeV m(χ˜1)<300 GeV, 1611.05791 ~ ~ ~ ∼ ±1 ∼ 02 ∼ 01 1 pp →g g , g → qq( χ / χ ) → qq (W/Z) χ SUS-16-022 SUS-16-041 Multilepton x=0.5 → 0 0 1 2 1 - g˜ χ˜ = g˜g˜, g˜ qq(ℓℓ/νν)χ˜1 3 e,µ 4 jets 36.1 1.87 TeV m( 1) 0GeV 1706.03731 → 0 0 ~ ~ ~ ∼ 0 χ˜ 0 7-11 jets Ye s 36.1 g˜ 1.8 TeV m(χ˜ ) <400 GeV 1708.02794 pp →t t , t → t χ SUS-16-014 SUS-16-033 0l(MHT) g˜g˜, g˜ qqWZ 1 1 ~ ~ ~ ∼ 01 → ˜ ˜ pp →t t , t → t χ SUS-16-015 SUS-16-036 0l(MT2) GMSB (ℓ NLSP) 1-2 τ +0-1ℓ 0-2 jets Ye s 3.2 g 2.0 TeV 1607.05979 ~ ~ ~ ∼ 01 pp →t t , t → t χ SUS-16-016 0l( α ) GGM (bino NLSP) 2 γ - Ye s 36.1 g˜ 2.15 TeV cτ(NLSP)<0.1 mm ATLAS-CONF-2017-080 1 T ~ ~ ~ ∼ 0 Inclusive Searches pp →t t , t → t χ SUS-16-027 SUS-17-001 2l opposite-sign 0 GGM (higgsino-bino NLSP) γ 2 jets Ye s 36.1 g˜ 2.05 TeV m(χ˜1)=1700 GeV, cτ(NLSP)<0.1 mm, µ>0 ATLAS-CONF-2017-080 ~ ~ ~ ∼ 01 pp →t t , t → t χ SUS-16-028 SUS-16-051 1l 1/2 ˜ 4 = = ~ ~ ~ ∼ 01 Gravitino LSP 0 mono-jet Ye s 20.3 F scale 865 GeV m(G)>1.8 10− eV, m(g˜) m(q˜) 1.5 TeV 1502.01518 pp →t t , t → t χ SUS-16-029 SUS-16-049 0l × ~ ~ ~ ∼ 01 pp →t t , t → t χ SUS-16-030 0l ¯χ˜ 0 g˜ χ˜0 1711.01901 ~ ~ ~ 01 g˜g˜, g˜ bb 1 03b Ye s 36.1 1.92 TeV m( 1)<600 GeV ∼ gen. pp →t t , t → c χ SUS-16-032 0l (Max exclusion for M - M LSP < 80 GeV) med. → 0 0 1 Mother χ˜ 0-1 e,µ g˜ χ˜ 1711.01901 ~ ~ ~ 0 rd ¯ 3 b Ye s 36.1 1.97 TeV m( )<200 GeV ˜ g˜g˜, g˜ tt 1 ∼ g 1 pp →t t , t → c χ SUS-16-036 0l(MT2) (Max exclusion for M - M LSP < 80 GeV) 3 → ~ ~ ~ ∼ 01 Mother pp →t t , t → c χ 0 0 SUS-16-049 0l (Max exclusion for M - M LSP < 80 GeV) ˜ ˜ ˜ χ˜ 0 36.1 b˜ 950 GeV χ˜ 1708.09266 ~ ~ ~ ∼ 0 1 Mother b1b1, b1 b 1 2 b Ye s 1 m( 1)<420 GeV pp →t t , t → b f f χ (4-body) SUS-16-025 SUS-16-048 2l soft (Max exclusion for M - M < 80 GeV) Mother LSP CMS Preliminary→ ˜ 0 0 1 χ χ χ ~ ~ ~ ∼ 0 b˜1b˜1, b˜1 tχ˜1± 2 e,µ(SS) 1 b Ye s 36.1 b1 275-700 GeV m( ˜1)<200 GeV, m( ˜1±)= m( ˜1)+100 GeV 1706.03731 pp →t t , t → b f f χ (4-body) SUS-16-029 SUS-16-049 0l (Max exclusion for M - M < 80 GeV) ~ ~ ~ Mother LSP → 0 0 ∼ 01 ˜ ˜ ˜ χ˜± 0-2 e,µ 1-2 b Ye s 4 . 7 / 1 3 . 3 t˜1 117-170 GeV 200-720 GeV m(χ˜±)=2m(χ˜ ), m(χ˜ )=55 GeV 1209.2102, ATLAS-CONF-2016-077 pp →t t , t → b f f χ (4-body) SUS-16-031 1l soft (Max exclusion for M - M LSP < 80 GeV) t1t1, t1 b 1 1 1 1 ~ ~ ~ 1 Mother Squark ∼ ± ± ∼ 0 → 0 0 ˜ 0 pp →t t , t → χ b → b W χ SUS-16-028 SUS-16-051 1l x=0.5 t˜ t˜ , t˜ Wbχ˜ or tχ˜ 0-2 e,µ 0-2 jets/1-2 b Ye s 2 0 . 3 / 3 6 . 1 t1 90-198 GeV 0.195-1.0 TeV m(χ˜1)=1 GeV 1506.08616, 1709.04183, 1711.11520 1 1 1 1 1 1 1 ~ ~ ~ ∼ ± ± ∼ 0 → 0 pp →t t , t → χ b → b W χ SUS-16-029 SUS-16-049 0l x=0.5 s = 13TeV 0 ˜ χ˜ = t˜1t˜1, t˜1 cχ˜1 0 mono-jet Ye s 36.1 t1 90-430 GeV m(t˜1)-m( 1) 5GeV 1711.03301 ~ ~ ~ ∼ 1 ∼ 01 pp →t t , t → χ ± b → b W ± χ SUS-16-036 0l(MT2) x=0.5 → 0 gen. squarks ˜ ˜ ˜ χ˜ ~ ~ ~ ∼ 1 ∼ 01 t1t1(natural GMSB) 2 e,µ(Z) 1 b Ye s 20.3 t1 150-600 GeV m( 1)>150 GeV 1403.5222 pp →t t , t → χ ± b → b W ± χ SUS-17-001 2l opposite-sign x=0.5 -1 -1
rd 0 ~1 ~ ~ 01 t˜ t˜ , t˜ t˜ + Z 3 e,µ(Z) t˜ χ˜ = 1706.03986 3 1 b Ye s 36.1 2 290-790 GeV m( ) 0GeV ∼ direct production 2 2 2 1 1 pp →b b , b → b χ SUS-16-014 SUS-16-033 0l(MHT) L = 12.9 fb L = 35.9 fb 1 → 0 ~ ~ ~ ∼ 0 t˜2t˜2, t˜2 t˜1 + h 1-2 e,µ 4 b Ye s 36.1 t˜2 320-880 GeV m(χ˜ )=0GeV 1706.03986 pp →b b , b → b χ SUS-16-015 SUS-16-036 0l(MT2) → 1 ~ ~ ~ ∼ 01 pp →b b , b → b χ SUS-16-016 0l( α ) 0 0 ~ ~ ~ T ˜ χ˜ = ∼ 01 ℓ˜L,Rℓ˜L,R, ℓ˜ ℓχ˜1 2 e,µ 0 Ye s 36.1 ℓ 90-500 GeV m( 1) 0 ATLAS-CONF-2017-039 pp →b b , b → b χ SUS-16-032 0l + + 1 ~ ~ ~ ~ ~ ~ → 0 0 ~ ~ ~ ∼ 0 χ˜ χ˜−, χ˜ ℓν˜ (ℓν˜) 2 e,µ 0 Ye s 36.1 χ˜ ± 750 GeV m(χ˜ )=0, m(ℓ˜, ν˜)=0.5(m(χ˜± )+m(χ˜ )) ATLAS-CONF-2017-039 pp →q q , q → q χ SUS-16-014 SUS-16-033 0l(MHT) q +q (u,d,c,s ) ~ 1 1 1 1 1 1 1 01 R L ~ ~ ~ ~ ~ 0 →+ 0 0 0 ~ ~ ~ ∼ - χ˜ ± χ˜ = χ˜ χ˜ pp →q q , q → q χ SUS-16-015 SUS-16-036 0l(MT2) q +q (u,d,c,s ) χ˜ 1±χ˜1∓/χ˜2, χ˜1 τν˜ (τν˜), χ˜ 2 ττ˜ (νν˜) 2 τ Ye s 36.1 1 760 GeV m( 1) 0, m(τ˜, ν˜)=0.5(m( 1± )+m( 1)) 1708.07875 1 R L 0 → → ˜ ˜ 0 0 0 0 χ˜ ±χ˜ ℓ˜ νℓ˜ ℓ(˜νν), ℓν˜ℓ˜ ℓ(˜νν) 3 e,µ 0 Ye s 36.1 χ±, χ 1.13 TeV m(χ˜1±)=m(χ˜2), m(χ˜1)=0, m(ℓ˜, ν˜)=0.5(m(χ˜1± )+m(χ˜1)) ATLAS-CONF-2017-039 ∼ 0 ∼ ∼ 0 ∼ 0 1 2 L L L 1 2 pp →χ χ ± → lll ν χ χ SUS-16-024 SUS-16-039 Multilepton (flavour democratic) x=0.5 0→ 0 0 0 0 0 2 1 1 1 χ˜ χ˜ χ˜ χ˜ 2-3 e,µ 0-2 jets Ye s 36.1 χ˜ ±, χ˜ 580 GeV m(χ˜±)=m(χ˜ ), m(χ˜ )=0, ℓ˜ decoupled ATLAS-CONF-2017-039 ∼ 0 ∼ ± ∼ 0 ∼ 0 1± 2 W 1Z 1 1 2 1 2 1 pp →χ χ → lll ν χ χ SUS-16-039 Multilepton + 2l same-sign (flavour democratic) x=0.95 EW 0→ 0 0 0 0 0 direct ˜ ˜ ∼ 02 ∼ 1 ∼ 01 ∼ 01 χ˜ ±χ˜ Wχ˜ h χ˜ , h bb¯/WW/ττ/γγ e,µ,γ 0-2 b Ye s 20.3 χ±, χ 270 GeV m(χ˜1±)=m(χ˜2), m(χ˜1)=0, ℓ˜ decoupled 1501.07110 pp →χ χ ± → ll τ ν χ χ 1 2 1 1 1 2 SUS-16-039 Multilepton (tau enriched) x=0.5 0 0 →0 → 0 0 0 0 0 0 ∼ 02 ∼ ±1 ∼ 01 ∼ 01 χ˜ χ˜ χ˜ ˜ 4 e,µ χ˜ χ˜ χ˜ χ˜ ˜ χ˜ χ˜ 1405.5086 pp →χ χ → τττ ν χ χ SUS-16-039 Multilepton (tau dominated) x=0.5 2 3, 2,3 ℓRℓ 0 Ye s 20.3 2,3 635 GeV m( 2)=m( 3), m( 1)=0, m(ℓ, ν˜)=0.5(m( 2)+m( 1)) ∼2 0 ∼1 ∼10 ∼10 → 0 pp → χ χ ± → W Z χ χ SUS-16-024 SUS-16-039 Multilepton GGM (wino NLSP) weak prod., χ˜ γG˜ 1 e,µ+ γ - Ye s 20.3 W˜ 115-370 GeV cτ<1mm 1507.05493 2 1 1 1 For decays with intermediate mass, 1 ∼ 0 ∼ ± ∼ 0 ∼ 0 0 → pp → χ χ → W H χ χ SUS-16-039 Multilepton GGM (bino NLSP) weak prod., χ˜ γG˜ 2 γ - Ye s 36.1 W˜ 1.06 TeV cτ<1mm ATLAS-CONF-2017-080 ∼ 02 ∼ 1 ∼ 01 ∼ 01 1 pp → χ χ ± → W Z χ χ SUS-16-025 SUS-16-048 2l soft (Max exclusion for M - M < 40 GeV) m = x⋅ m +(1-x)⋅ m → Mother LSP Intermediate + 2 1 1 1 Mother LSP χ˜ 0 Direct χ˜ χ˜ − prod., long-lived χ˜ ± Disapp. trk 1 jet Ye s 36.1 ± 460 GeV m(χ˜1±)-m(χ˜1) 160 MeV, τ(χ˜1±)=0.2 ns 1712.02118 EWK Gauginos 1 1 1 1 + 0 ∼ - χ˜ ± Direct χ˜1 χ˜ 1− prod., long-lived χ˜ 1± dE/dx trk Ye s 18.4 1 495 GeV m(χ˜1±)-m(χ˜1) 160 MeV, τ(χ˜1±)<15 ns 1506.05332 0 200 400 600 800 1000 1200 1400 1600 1800 2000 0 ∼ Stable, stopped g˜ R-hadron 0 1-5 jets Ye s 27.9 g˜ 850 GeV m(χ˜1)=100 GeV, 10 µs<τ(˜g)<1000 s 1310.6584 *Observed limits at 95% C.L. - theory uncertainties not included Mass ScaleStable g˜[GeV]R-hadron trk --3.2 g˜ 1.58 TeV 1606.05129 Metastable g˜ R-hadron -- g˜ χ˜0 Only a selection of available mass limits. Probe *up to* the quoted mass limit for m ≈0 GeV unless stated otherwise dE/dx trk 3.2 1.57 TeV m( 1)=100 GeV, τ>10 ns 1604.04520 0 - g˜ = χ˜0 = LSP Metastable g˜ R-hadron, g˜ qqχ˜1 displ. vtx Ye s 32.8 2.37 TeV τ(˜g) 0.17 ns, m( 1) 100 GeV 1710.04901 particles → 0 Long-lived 0 -- χ˜ 10
The Motivation for Indirect Searches
§ To account for the observed dark matter Fermi abundance, a thermal relic must have an annihilation cross section (at freeze-out) of σv~2x10-26 cm3/s § Although many model-dependent factors can cause the dark matter to possess a somewhat lower or higher annihilation cross section today, most models predict current annihilation rates that are within an order of magnitude or so of this estimate AMS-02 § Indirect detection experiments that are sensitive to dark matter annihilating at approximately this rate will be able to test a significant fraction of WIMP models Dan Hooper – The WIMP is Dead! Long Live the WIMP! Recent Constraints from Indirect Detection
§ A variety of gamma-ray strategies (GC, dwarfs, IGRB, etc.) as well as cosmic-ray antiproton and positron measurements from AMS, are sensitive to dark matter with the annihilation cross section predicted for a simple thermal relic, for masses up to ~100 GeV § This program is not a fishing expedition, but is testing a wide range of well-motivated dark matter models – we are testing the WIMP paradigm! 4
Bergstr¨om et al. (2013) + − 10−23 ing a band around the e e constraint, corresponding dashed: Fermi LAT −3 to the range Urad + UB =(1.2 2.6) eV cm ,and solid: AMS-02 (this work) ⊙ −3 − 10−24 WMAP7 ρχ =(0.25 0.7) GeV cm [61, 74] (note that the form of the DM− profile has a much smaller impact). Uncer- 10−25 tainty bands of the same width apply to each of the other ] 1 −
s final states shown in the figure, but are not explicitly 3 10−26 shown for clarity. Other diffusion parameter choices im- [cm ⟩ v pact our limits only by up to 10%, except for the case σ ⟨ − + ∼ 10 27 τ τ − of low DM masses, for which the effect of solar modula- + µ µ− tion may be increasingly important [53, 75]. We reflect −28 + 10 e e−γ this in Fig. 3 by depicting the limits derived in this less + + e e− certain mass range, where the peak of the signal e flux 10−29 101 102 (as shown in Fig. 1) falls below a fiducial value of 5 GeV, mχ [GeV] with dotted rather than solid lines. For comparison, we have also considered a collection of physical background models in which we calculated FIG.Bergstrom, 3. Upper limits (95% CL) et on al, the DM annihilation cross Fermithe expected Collaboration, primary and secondary lepton fluxesCuoco using , et al., arXiv:1610.03071 section, as derived from the AMS positron fraction, for various arXiv:1611.03184GALPROP, and then added the contribution from all finalarXiv states (this:1306.3983 work), WMAP7 (for ℓ+ℓ−)[44]andFermi Cui, et al. arXiv:1610.03840 + − + − galactic pulsars. While this leads to an almost identical LAT dwarf spheroidals (for µ µ and τ τ )[43].Thedotted description of the background at high energies as in the portions of the curves are potentially affected by solar modu- −26 3 −1 phenomenological model, small differences are manifest lation. We also indicate ⟨σv⟩therm ≡ 3 × 10 cm s .The AMS limits are shown for reasonable reference values of the at lower energies due to solar modulation and a spec- local DM density and energy loss rate (see text), and can vary tral break [55, 76, 77] in the CR injection spectrum at a by a factor of a few, as indicated by the hatched band (for few GeV (both neglected in the AMS parameterization). clarity, this band is only shown around the e+e− constraint). We cross-check our fit to the AMS positron fraction with lepton measurements by Fermi [64]. Using these physical background models in our fits, instead of the phenomeno- our upper bound on the annihilation cross section to logical AMS parameterization, the limits do not change e+e− is approximately two orders of magnitude below significantly. The arguably most extreme case would be σv therm. If only a fraction f of DM annihilates like the appearance of dips in the background due to the su- assumed,⟨ ⟩ limits would scale like f −2 (and, very roughly, perposition of several pulsar contributions, which might −1 σv therm f ). We also show in Fig. 3 the upper conspire with a hidden DM signal at almost exactly the ⟨bounds⟩ obtained∝ for other leptonic final states. As ex- same energy. We find that in such situations, the real lim- pected, these limits are weaker than those found in the its on the annihilation rate could be weaker (or stronger) case of direct annihilation to electrons – both because by up to roughly a factor of 3 for any individual value of part of the energy is taken away by other particles (neu- mχ. See the Appendix [45] for more details and further trinos, in particular) and because they feature broader discussion of possible systematics that might affect our and less distinctive spectral shapes. These new limits analysis. + − + − on DM annihilating to µ µ and τ τ final states are Lastly, we note that the upper limits on σv (mχ)re- ⟨ ⟩ still, however, highly competitive with or much stronger ported in Fig. 3 can easily be translated into upper limits than those derived from other observations, such as from on the decay width of a DM particle of mass 2mχ via ⊙ the cosmic microwave background [44] and from gamma- Γ σv ρ /mχ. We checked explicitly that this sim- ≃⟨ ⟩ χ ray observations of dwarf galaxies [43]. Note that for ple transformation is correct to better than 10% for the the case of e+e−γ final states even stronger limits can L =4 kpc propagation scenario and e+e− and µ+µ− final be derived for mχ ! 50 GeV by a spectral analysis of states over the full considered energy range. gamma rays [73]. We do not show results for the ¯bb Conclusions. In this Letter, we have considered a channel, for which we nominally find even weaker lim- possible dark matter contribution to the recent AMS cos- its due to the broader spectrum (for mχ 100 GeV, mic ray positron fraction data. The high quality of this −24 3 −1 ≃ about σv " 1.1 10 cm s ). In fact, due to de- data has allowed us for the first time to successfully per- ⟨ ⟩ · generacies with the background modeling, limits for an- form a spectral analysis, similar to that used previously nihilation channels which produce such a broad spectrum in the context of gamma ray searches for DM. While we of positrons can suffer from significant systematic uncer- have found no indication of a DM signal, we have derived tainties. For this reason, we consider our limits on the upper bounds on annihilation and decay rates into lep- e+e− channel to be the most robust. tonic final states that improve upon the most stringent Uncertainties in the e± energy loss rate and local DM current limits by up to two orders of magnitude. For density weaken, to some extent, our ability to robustly light DM in particular, our limits for e+e− and µ+µ− fi- constrain the annihilation cross sections under consid- nal states are significantly below the cross section naively eration in Fig. 3. We reflect this uncertainty by show- predicted for a simple thermal relic. When taken together Dan Hooper – The WIMP is Dead! Long Live the WIMP! So, is the WIMP Paradigm Dead? Dan Hooper – The WIMP is Dead! Long Live the WIMP! So, is the WIMP Paradigm Dead?
No.
Despite the very stringent constraints that have been placed on the nature of dark matter, there remain many viable options for WIMP model building Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints: Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints:
Common Theme: Mechanisms that deplete the dark matter abundance in the early universe without leading to large elastic scattering rates with nuclei or large annihilation rates in the universe today
Unsuppressed Suppressed
X SM X X
X SM N N Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints: 1) Co-annihilations between the dark matter and another state
Griest, Seckel (1991) Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints: 1) Co-annihilations between the dark matter and another state 2) Annihilations to W, Z and/or Higgs bosons Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints: 1) Co-annihilations between the dark matter and another state 2) Annihilations to W, Z and/or Higgs bosons; scattering with nuclei only through highly suppressed loop diagrams
Hisano, et al., arXiv:1007.2601, 1104.0228, 1504.00915; Hill, Solon, arXiv:1309.4092, 1409.8290; Berlin, DH, McDermott, arXiv:1508.05390 Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints: 1) Co-annihilations between the dark matter and another state 2) Annihilations to W, Z and/or Higgs bosons; scattering with nuclei only through highly suppressed loop diagrams 3) Interactions which suppress elastic scattering with nuclei by powers of velocity or momentum Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints: 1) Co-annihilations between the dark matter and another state 2) Annihilations to W, Z and/or Higgs bosons; scattering with nuclei only through highly suppressed loop diagrams 3) Interactions which suppress elastic scattering with nuclei by powers of velocity or momentum 4) Dark matter that is lighter than a few GeV (evading direct constraints) Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints: 1) Co-annihilations between the dark matter and another state 2) Annihilations to W, Z and/or Higgs bosons; scattering with nuclei only through highly suppressed loop diagrams 3) Interactions which suppress elastic scattering with nuclei by powers of velocity or momentum 4) Dark matter that is lighter than a few GeV (evading direct constraints) 5) Departures from radiation domination in the early universe (early matter domination; late-time reheating, etc.) which result in the depletion of the dark matter’s relic abundance Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints: 1) Co-annihilations between the dark matter and another state 2) Annihilations to W, Z and/or Higgs bosons; scattering with nuclei only through highly suppressed loop diagrams 3) Interactions which suppress elastic scattering with nuclei by powers of velocity or momentum 4) Dark matter that is lighter than a few GeV (evading direct constraints) 5) Departures from radiation domination in the early universe (early matter domination; late-time reheating, etc.) which result in the depletion of the dark matter’s relic abundance 6) The dark matter annihilates into unstable non-Standard Model states (ie. hidden sector models) Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints: 1) Co-annihilations between the dark matter and another state 2) Annihilations to W, Z and/or Higgs bosons; scattering with nuclei only through highly suppressed loop diagrams 3) Interactions which suppress elastic scattering with nuclei by powers of velocity or momentum 4) Dark matter that is lighter than a few GeV (evading direct constraints) 5) Departures from radiation domination in the early universe (early matter domination; late-time reheating, etc.) which result in the depletion of the dark matter’s relic abundance 6) The dark matter annihilates into unstable non-Standard Model states (ie. hidden sector models) Dan Hooper – The WIMP is Dead! Long Live the WIMP!
Dark Matter Within a Hidden Sector
§ Lets hypothesize that the dark matter is one of several particle species within a hidden sector, which is entirely uncharged under the Standard Model § Even without any direct couplings between these two sectors, small “portal” interactions could very plausibly have kept them in kinetic equilibrium with each other in the early universe
Standard Portal Hidden Model Sector Dan Hooper – The WIMP is Dead! Long Live the WIMP!
Dark Matter Within a Hidden Sector
§ The dark matter, X, freezes-out of thermal equilibrium entirely within its own hidden sector (annihilating to produce lighter particles within the hidden sector, Y) § These lighter hidden sector particles then decay through portal interactions into Standard Model particles
Dark matter annihilates …and then those hidden sector within the hidden sector… annihilation products decay through portal interactions Dan Hooper – The WIMP is Dead! Long Live the WIMP!
Dark Matter Within a Hidden Sector
Hidden sector dark matter models offer a number of attractive features: 1) Elastic scattering with nuclei is highly suppressed 2) Production at colliders is highly suppressed 3) Relic abundance is easily accommodated; similar to ordinary WIMPs 4) Simple model building possibilities, including the three renormalizable “portal” interactions: the vector portal, Higgs portal, and lepton portal Dan Hooper – The WIMP is Dead! Long Live the WIMP!
Dark Matter Within a Hidden Sector
Hidden sector dark matter models offer a number of attractive features: 1) Elastic scattering with nuclei is highly suppressed 2) Production at colliders is highly suppressed 3) Relic abundance is easily accommodated; similar to ordinary WIMPs 4) Simple model building possibilities, including the three renormalizable “portal” interactions: the vector portal, Higgs portal, and lepton portal
For example, consider the vector portal scenario, which is described by the following Lagrangian:
§This includes a new vector gauge boson, Z’, which undergoes kinetic mixing with the Standard Model photon and/or Z §This portal interaction allows the Z’ to decay into the Standard Model, but leads to negligible direct detection and collider constraints (for small �) Dan Hooper – The WIMP is Dead! Long Live the WIMP! A Well-Motivated Variant of Hidden Sector Dark Matter § In most discussions of hidden sector dark matter scenarios, the portal interactions are assumed to be strong enough for kinetic equilibrium to be reached between the two sectors; this need not be the case § Lets hypothesize that the dark matter resides within a heavy (> TeV) hidden sector, which is not only uncharged under the Standard Model, but is highly decoupled from the Standard Model bath § Both the Standard Model sector and the hidden sector are populated during reheating (after inflation); we treat the ratio of their initial temperatures as a free parameter and an initial condition
(very feeble) Standard Portal Hidden Model Sector Berlin, DH, Krnjaic, arXiv:1609.02555, 1602.08490 Dan Hooper – The WIMP is Dead! Long Live the WIMP! A Well-Motivated Variant of Hidden Sector Dark Matter
§ As the universe expands, the temperatures of the two sectors remain independent of each other, and evolve according to entropy conservation § The dark matter freezes-out of thermal equilibrium entirely within its own hidden sector, annihilating to produce lighter particles within the hidden sector, which become non-relativistic and evolve to increasingly dominate the energy density of the early universe § When this hidden sector state ultimately decays, it can reheat the universe, diluting the abundances of any relics, including dark matter
Berlin, DH, Krnjaic, arXiv:1609.02555, 1602.08490 Dan Hooper – The WIMP is Dead! Long Live the WIMP! Dark Matter Within a Heavy Decoupled Hidden Sector Here’s an example of a vector portal dark matter scenario:
Berlin, DH, Krnjaic, arXiv:1609.02555, 1602.08490 Dan Hooper – The WIMP is Dead! Long Live the WIMP! Dark Matter Within a Heavy Decoupled Hidden Sector Here’s an example of a vector portal dark matter scenario: § For a sizeable portal interaction strength (� ~10-7 or higher), kinetic equilibrium between the sectors is reached, and the early universe undergoes a standard thermal history
Berlin, DH, Krnjaic, arXiv:1609.02555, 1602.08490 Dan Hooper – The WIMP is Dead! Long Live the WIMP! Dark Matter Within a Heavy Decoupled Hidden Sector Here’s an example of a vector portal dark matter scenario: § For a sizeable portal interaction strength (� ~10-7 or higher), kinetic equilibrium between the sectors is reached, and the early universe undergoes a standard thermal history § For very weak portal interactions, however, there is an early matter- dominated era, followed by late-time reheating § This dilution can easily facilitate an acceptable dark matter abundance for masses as high as ~10-100 PeV (well above of the range for a standard Berlin, DH, Krnjaic, WIMP) arXiv:1609.02555, 1602.08490 Dan Hooper – The WIMP is Dead! Long Live the WIMP!
Prospects Moving Forward?
I’ve just described several ways in which the prospects for dark matter detection could be almost arbitrarily suppressed
Does this mean that efforts to detect dark matter particles are likely to be doomed from the start? Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints: 1) Co-annihilations between the dark matter and another state 2) Annihilations to W, Z and/or Higgs bosons; scattering with nuclei only through highly suppressed loop diagrams 3) Interaction which suppress elastic scattering with nuclei by powers of velocity or momentum 4) Dark matter that is lighter than a few GeV (evading direct constraints) 5) Departures from radiation domination in the early universe (early matter domination; late-time reheating, etc.) which result in the depletion of the dark matter’s relic abundance 6) The dark matter annihilates to unstable non-Standard Model states (ie. hidden sector models) Dan Hooper – The WIMP is Dead! Long Live the WIMP! An (Incomplete) List of Ways to Reconcile WIMP Dark Matter With All Current Constraints: 1) Co-annihilations between the dark matter and another state 2) Annihilations to W, Z and/or Higgs bosons; scattering with nuclei only through highly suppressed loop diagrams 3) Interaction which suppress elastic scattering with nuclei by powers of velocity or momentum 4) Dark matter that is lighter than a few GeV (evading direct constraints) 5) Departures from radiation domination in the early universe (early matter domination; late-time reheating, etc.) which result in the depletion of the dark matter’s relic abundance 6) The dark matter annihilates to unstable non-Standard Model states (ie. hidden sector models)
Although potentially invisible to both underground detectors and colliders, many of these scenarios are testable with indirect searches In this sense, the lack of such signals has strengthened the motivation for gamma-ray and cosmic-ray searches for dark matter Dan Hooper – The WIMP is Dead! Long Live the WIMP! The Galactic Center GeV Excess
10 10 § A bright and highly statistically significant excess of gamma-rays has been observed from the region surrounding the Galactic Center § This signal is difficult to explain with astrophysical sources or mechanisms, but is very much like the signal predicted from annihilating dark matter
Among other references, see: DH, Goodenough (2009, 2010) DH, Linden (2011) Abazajian, Kaplinghat (2012) Gordon, Macias (2013) Daylan, et al. (2014) Calore, Cholis, Weniger (2014) Murgia, et al. (2015) FIG. 10: The raw gamma-ray maps (left)FIG. and 10: the The residual raw gamma-ray maps after maps subtracting (left) and the the best-fit residual Galactic maps di after↵use subtracting model, 20 cmthe best-fit Galactic di↵use model, 20 cm Ackermann et al. (2017) template, point sources, and isotropic templatetemplate, (right), point sources, in units and of photons/cm isotropic template2/s/sr. (right), The right in units frames of clearly photons/cm contain2/s/sr. a The right frames clearly contain a significant central and spatially extendedsignificant excess, peaking central at and1-3 spatially GeV. Results extended are excess, shown peaking in galactic at coordinates,1-3 GeV. Results and all are maps shown in galactic coordinates, and all maps ⇠ ⇠ have been smoothed by a 0.25 Gaussian.have been smoothed by a 0.25 Gaussian.
ing to a statical preference for such aing component to a statical at the preferenceas shown for such in the a component right frame at of the Fig. 2,as respectively). shown in the The right frame of Fig. 2, respectively). The level of 17 . In Fig. 8, we show thelevel spectrum of 17 of. theIn Fig.normalizations 8, we show the of spectrum the Galactic of the Centernormalizations and Inner Galaxy of the Galactic Center and Inner Galaxy dark-matter-like⇠ component, for valuesdark-matter-like of =1⇠ .2 (left component,signals for are values compatible of =1 (see.2 Figs. (left 6 andsignals 8), although are compatible the (see Figs. 6 and 8), although the frame) and =1.3 (right frame). Shownframe) for and compari- =1.3 (rightdetails frame). of this Shown comparison for compari- depend ondetails the precise of this mor-comparison depend on the precise mor- son is the spectrum predicted from ason 35.25 is theGeV spectrum WIMP predictedphology from that a is 35.25 adopted. GeV WIMP phology that is adopted. annihilating to b¯b. The solid line representsannihilating the contribu- to b¯b. The solid line represents the contribu- We note that the Fermi tool gtlikeWedetermines note that the the Fermi tool gtlike determines the tion from prompt emission, whereas thetion dot-dashed from prompt and emission, whereas the dot-dashed and quality of the fit assuming a givenquality spectral of shape the fit for assuming a given spectral shape for dotted lines also include an estimate fordotted the contribution lines also include an estimate for the contribution the dark matter template, but does notthe generally dark matter provide template, but does not generally provide from bremsstrahlung (for the z =0.15from and bremsstrahlung0.3 kpc cases, (for the z =0.15 and 0.3 kpc cases, a model-independent spectrum for thisa model-independent or other compo- spectrum for this or other compo- Dan Hooper – The WIMP is Dead! Long Live the WIMP! Dark Matter and the GeV Excess
10 5 10 5 ⇥ ⇥ 1.2 1.0 I II The observed characteristics of the excess 0.8 0.6 0.4 are in good agreement with the 0.2 0.0 0.2 10 6 10 6 expectations of annihilating dark matter 4 ⇥ ⇥ III IV ssr)] 3 2
(cm 2 § Spectrum: Well fit by a ~40-70 GeV / 1 [GeV dE dN 0 particle annihilating to quarks, and is 2 E 1 10 6 10 6 uniform across the Inner Galaxy ⇥ ⇥ 1.5 V VI
1.0 § Morphology: Approximately spherically 0.5 0.0
0.5 symmetric, with a flux that falls as 10 6 10 6 -2.4 3.0 ⇥ ⇥ 2.5 VII VIII ~r out to at least ~10°, ssr)] 2 2.0
(cm 1.5 consistent with a DM halo only / 1.0
[GeV 0.5 0.0 dE dN
2 0.5
E slightly steeper than NFW 1.0 10 6 10 6 1.5 ⇥ ⇥ IX X § Intensity: Requires an annihilation 1.0 0.5 -26 3 cross section of σv ~ 10 cm /s, 0.0 0.5 1.0 near the value of a thermal relic 100 101 102 100 101 102 E [GeV] E [GeV]
Figure 16. Same as figure 14, but from a fit with the segmented GCE template as illustrated in figure 15. We show results for GDE model F (black dots), as well as the envelope for all 60 GDE models (blue dotted lines) and the systematic errors that we derived from fits in 22 test regions along Daylan, Finkbeiner, Hooper,the Galactic Linden, disk (yellow boxes Rodd,, in analogy to Slatyer figure 12). See figure(2014)28 below for the spectra of all Calore, Cholis, Wenigercomponents.; Calore, Cholis, McCabe, Weinger (2014);
– 32 – Dan Hooper – The WIMP is Dead! Long Live the WIMP! Millisecond Pulsars and The Galactic Center Gamma-Ray Excess
The Main Arguments in Favor of Pulsars: § The gamma-ray spectrum of observed pulsars § Small-scale power in the gamma-ray emission from the Inner Galaxy § Pulsars are known to exist Small-scale power in the gamma-ray emission from the Inner Galaxy Dan Hooper – The WIMP is Dead! Long Live the WIMP! Millisecond Pulsars and The Galactic Center Gamma-Ray Excess
The Main Arguments in Favor of Pulsars: § The gamma-ray spectrum of observed pulsars § Small-scale power in the gamma-ray emission from the Inner Galaxy § Pulsars are known to exist Small-scale power Arguments Against Pulsars: § The lack of any pulsars detected in the Inner Galaxy § The measured luminosity function of gamma-ray pulsars § The lack of low-mass X-ray binaries in the Inner Galaxy in the gamma-ray emission from the Inner Galaxy Dan Hooper – The WIMP is Dead! Long Live the WIMP! Small Scale Power Among Inner Galaxy �-Rays
§ In 2015, two groups found that the ~GeV photons from the direction of the Inner Galaxy are more clustered than predicted from smooth backgrounds, suggesting that the GeV excess might be generated by a population of unresolved point sources § Lee et al. used a non-Poissonian template technique to show that the photon distribution within ~10° of the Galactic Center (masking within 2° of the Galactic Plane) is clumpy, potentially indicative of an unresolved point source population § Bartels et al. reach a qualitatively similar conclusion employing a wavelet technique Small-scale power in the gamma-ray emission from the Inner Galaxy
Lee, Lisanti, Safdi, Slatyer, Xue, arXiv:1506.05124 Bartels, Krishnamurthy, Weniger, arXiv:1506.05104 Dan Hooper – The WIMP is Dead! Long Live the WIMP! Small Scale Power Among Inner Galaxy �-Rays
§ A typical Fermi Inner Galaxy analysis might include the following spatial templates: 1) Galactic diffuse emission 2) Fermi Bubbles 3) Isotropic background 4) Known point sources 5) Dark matter annihilation products (generalized NFW2)
§ Lee et al. then add a number of non-Possionian templates to model the distribution of unresolved point sources: 5) Isotropically distributed point sources 6) Disk-correlated point sources 7) NFW2 correlated point sources
Lee, Lisanti, Safdi, Slatyer, Xue, arXiv:1506.05124 Dan Hooper – The WIMP is Dead! Long Live the WIMP! 4 Small Scale Power Among Inner Galaxy �-Rays
Excess-Like Population
18 0 0 3 5 9 Disk-Like Population
2 2 0
(1.9-11.9 GeV)
FIG. 2: (Left) Best-fit source-count functions within 10 of the GC and b 2 , with the 3FGL sources unmasked. The Lee,median Lisanti, and Safdi 68%, Slatyer confidence, Xue, arXiv:1506.05124 intervals are shown for each of the following PS| | components: NFW (dashed, orange), thin-disk (solid, blue), and isotropic (dotted, green). The number of observed 3FGL sources in each bin is indicated. The normalization for the di↵use emission in the fit is consistent with that at high latitudes, as desired. (Right) Posteriors for the flux fraction within 10 of the GC with b 2 arising from the separate PS components, with 3FGL sources unmasked. The inset shows the result of removing the NFW| | PS template from the fit. Dashed vertical lines indicate the 16th,50th,and84th percentiles.
FIG. 3: Same as Fig. 2, except with 3FGL sources masked.
sources. When the NFW PS template is omitted (inset), The flux fraction attributed to the NFW PS component +1.0 the fraction of flux absorbed by the disk PS population is is 5.3 1.1%, while the NFW DM template absorbs no +0.7 essentially unchanged at 6.8 0.9%, and the DM template significant flux. +0.7 absorbs 7.7 0.8% of the flux. The DM flux obtained in In the masked analysis, the Bayes factor for a model absence of an NFW PS template is consistent with other that contains an NFW PS component, relative to one estimates in the literature [12, 14]. The model including that does not, is 102, substantially reduced relative to the NFW PS contribution is preferred over that without the result for the⇠ unmasked case. Masking the 3FGL by a Bayes factor 106.4 sources removes most of the ROI within 5 of the GC, ⇠ ⇠ When the 3FGL sources are masked, the NPTF proce- reducing photon statistics markedly, especially for any dure yields a best-fit source-count function given by the signal peaked at the GC. Furthermore, in the masked orange band in the left panel of Fig. 3. Below the break, ROI, non-NFW PS templates can absorb a substantial the source-count function agrees well with that found by fraction of the excess. For example, if only disk and the unmasked fit. In this case, the contributions from the isotropic PS templates are added, the flux fraction at- +0.70 isotropic and disk-correlated PS templates are negligible. tributed to the disk template is 2.5 0.62%, while that +1.6 attributed to NFW DM is 2.2 2.2% (the flux attributed to isotropic PSs is negligible). When no PS templates are included in the fit, the NFW DM template absorbs 4 +1.1 For reference, this corresponds to test statistic 2 ln 36. 4.1 1.2% of the total flux. As we will discuss later, this L ⇡ Dan Hooper – The WIMP is Dead! Long Live the WIMP! 4 Small Scale Power Among Inner Galaxy �-Rays
Excess-Like Population
18 0 0 3 5 9 Disk-Like Population
Approximate 2 Fermi 2 Detection 0 Threshold
(1.9-11.9 GeV) Bottom Line: A population of ~103 points sources with luminosities near FIG. 2: (Left) Best-fit source-count functions within 10 of the GC and b 2 , with the 3FGL sources unmasked. The Fermi’smedian detection and 68% threshold confidence could intervals potentially are shown account for for each the of GeV the Excess following PS| | components: NFW (dashed, orange), thin-disk (solid, blue), and isotropic (dotted, green). The number of observed 3FGL sources in each bin is indicated. The normalization for the di↵use emission in the fit is consistent with that at high latitudes, as desired. (Right) Posteriors for the flux fraction within 10 of the GC with b 2 arising from the separate PS components, with 3FGL sources unmasked. The inset shows the result of removing the NFW| | PS template from the fit. Dashed vertical lines indicate the 16th,50th,and84th percentiles.
FIG. 3: Same as Fig. 2, except with 3FGL sources masked.
sources. When the NFW PS template is omitted (inset), The flux fraction attributed to the NFW PS component +1.0 the fraction of flux absorbed by the disk PS population is is 5.3 1.1%, while the NFW DM template absorbs no +0.7 essentially unchanged at 6.8 0.9%, and the DM template significant flux. +0.7 absorbs 7.7 0.8% of the flux. The DM flux obtained in In the masked analysis, the Bayes factor for a model absence of an NFW PS template is consistent with other that contains an NFW PS component, relative to one estimates in the literature [12, 14]. The model including that does not, is 102, substantially reduced relative to the NFW PS contribution is preferred over that without the result for the⇠ unmasked case. Masking the 3FGL by a Bayes factor 106.4 sources removes most of the ROI within 5 of the GC, ⇠ ⇠ When the 3FGL sources are masked, the NPTF proce- reducing photon statistics markedly, especially for any dure yields a best-fit source-count function given by the signal peaked at the GC. Furthermore, in the masked orange band in the left panel of Fig. 3. Below the break, ROI, non-NFW PS templates can absorb a substantial the source-count function agrees well with that found by fraction of the excess. For example, if only disk and the unmasked fit. In this case, the contributions from the isotropic PS templates are added, the flux fraction at- +0.70 isotropic and disk-correlated PS templates are negligible. tributed to the disk template is 2.5 0.62%, while that +1.6 attributed to NFW DM is 2.2 2.2% (the flux attributed to isotropic PSs is negligible). When no PS templates are included in the fit, the NFW DM template absorbs 4 +1.1 For reference, this corresponds to test statistic 2 ln 36. 4.1 1.2% of the total flux. As we will discuss later, this L ⇡ Dan Hooper – The WIMP is Dead! Long Live the WIMP! Evidence For Unresolved Point Sources?
§ It is difficult to tell whether these clustered gamma-rays result from unresolved sources, or from backgrounds that are less smooth than are being modeled Dan Hooper – The WIMP is Dead! Long Live the WIMP! Evidence For Unresolved Point Sources?
§ It is difficult to tell whether these clustered gamma-rays result from unresolved sources, or from backgrounds that are less smooth than are being modeled
Observed
Smooth and well-modeled background
Flux Point source dominated excess signal
Direction Dan Hooper – The WIMP is Dead! Long Live the WIMP! Evidence For Unresolved Point Sources?
§ It is difficult to tell whether these clustered gamma-rays result from unresolved sources, or from backgrounds that are less smooth than are being modeled
Observed Observed
Smooth and well-modeled Poorly-modeled background, background including points sources or other small scale structure (gas) Flux Point source dominated Flux excess signal Smooth (dark matter) dominated excess signal
Direction Direction Dan Hooper – The WIMP is Dead! Long Live the WIMP! Evidence For Unresolved Point Sources?
§ It is difficult to tell whether these clustered gamma-rays result from unresolved sources, or from backgrounds that are less smooth than are being modeled § These clusters consist of only a few photons each, on top of large and imperfectly known backgrounds
Lee, Lisanti, Safdi, Slatyer, Xue, arXiv:1506.05124 (see also Bartels, Krishnamurthy, Weniger, arXiv:1506.05104) Dan Hooper – The WIMP is Dead! Long Live the WIMP! Evidence For Unresolved Point Sources?
§ It is difficult to tell whether these clustered gamma-rays result from unresolved sources, or from backgrounds that are less smooth than are being modeled § These clusters consist of only a few photons each, on top of large and imperfectly known backgrounds § Gamma-ray point source identification is difficult in the Galactic Center region – even for bright sources – and the contents of source catalogs depend strongly on how one treats diffuse backgrounds (try comparing the membership of Fermi’s 3FGL, 1FIG, and 2FIG catalogs)
Lee, Lisanti, Safdi, Slatyer, Xue, arXiv:1506.05124 (see also Bartels, Krishnamurthy, Weniger, arXiv:1506.05104) Dan Hooper – The WIMP is Dead! Long Live the WIMP! Comparison With The Measured MSP Luminosity Function § The MSP populations observed in globular clusters and in the Galactic Disk exhibit a gamma-ray luminosity function that is very different from that indicated by the analysis of Lee et al. § The measured MSP luminosity function is very broad and extends over several orders of magnitude and up to at least ~1035 erg/s § If the small scale power identified by these analyses does in fact originate from MSPs, this is a very different population than those observed elsewhere Dan Hooper – The WIMP is Dead! Long Live the WIMP! Gamma-Ray Bright MSPs in The Inner Galaxy?
§ The most direct way to prove that MSPs generate the GeV excess would be to detect a significant number of individual pulsars in the Inner Galaxy Dan Hooper – The WIMP is Dead! Long Live the WIMP! Gamma-Ray Bright MSPs in The Inner Galaxy?
§ The most direct way to prove that MSPs generate the GeV excess would be to detect a significant number of individual pulsars in the Inner Galaxy § Last year, the Fermi Collaboration posted a paper which purported to present strong evidence (~7�) for a large centrally located pulsar population
Fermi Collaboration, arXiv:1705.00009v1 +4.89 5 NB =5.03 2.52 10 ⇥
+0.42 6 106 ND =1.06 0.34 10 ⇥ ⇥ 4 2.
D 6 1. N 8 0. +0.05 .0 z0 =0.08 0.03 0 .0 Dan Hooper – The WIMP1 is Dead! Long Live the WIMP! 8 0.
0 6 0. Evidence of a Central Pulsarz Population? 4 0. § In examining this paper, my collaborators and I 2 0. +0.08 found that we were unable to reproduce these +4.89 5 =2.11 0.07 NB =5.03 2.52 10 ⇥ results; our fit favored only a ~2� preference for a 4 2. central source component .1 § As a result of the ensuing discussions with the 2 8 Fermi Collaboration, an error was identified in their 1. 5 code, and a replacement (v2) of their paper was 1. posted in conjunction with our paper 2 +0.42 6 6 1. ND =1.06 0.34 10 10 ⇥ ⇥ 0 8 6 4 0 8 6 4 2 4 6 8 0 2 5 8 1 4 0. 0. 1. 2. 0. 0. 1. 2. 0. 0. 0. 0. 1. 1. 1. 1. 2. 2. 4 6 6 2. 10 10 NB ⇥ ND ⇥ z0
D 6 1. N 8 0. +0.05 0 z0 =0.08 0. 0.03 0 1. 8 0. Bartels, DH, Linden, Mishra-Sharma, Rodd, Safdi, Slatyer, arXiv:1710.10266
0 6 0. z 4 0. 2 0. +0.08 =2.11 0.07 4 2. 1 2.
8 1. 5 1. 2 1. 0 8 6 4 0 8 6 4 2 4 6 8 0 2 5 8 1 4 0. 0. 1. 2. 0. 0. 1. 2. 0. 0. 0. 0. 1. 1. 1. 1. 2. 2. 106 106 NB ⇥ ND ⇥ z0 Dan Hooper – The WIMP is Dead! Long Live the WIMP! Evidence of a Central Pulsar Population? § In our paper, we also note that masking the pulsar candidate sources contained in the 2FIG Fermi catalog does not impact the characteristics of the excess; a negligible fraction of the excess emission originates from these sources
2.0 p6v11 Analysis Default Analysis Bulge PSRs Masked /s/sr] 2 1.5 [GeV/cm 6 1.0 10 ⇥
0.5 dN/dE 2 E
0.0 100 101 102 E [GeV]
Bartels, DH, Linden, Mishra-Sharma, Rodd, Safdi, Slatyer, arXiv:1710.10266 Dan Hooper – The WIMP is Dead! Long Live the WIMP!
Fermi Observations of Dwarf Galaxies § Current Fermi dwarf constraints are based on stacks of dwarf galaxy candidates, aided by recent discoveries by DES and other surveys § At this time, these constraints are compatible with dark matter interpretations of the Galactic Center excess § That being said, if the Galactic Center signal is coming from Region favored by annihilating dark matter, we the GeV Excess should expect gamma rays to be detected from dwarfs soon
Fermi Collaboration, arXiv:1611.03184 (see also 1503.02641) Dan Hooper – The WIMP is Dead! Long Live the WIMP! Fermi’s Observations of Dwarf Galaxies Intriguing (but inclusive) excesses have been observed from the recently discovered and nearby dwarf galaxies Reticulum II and Tucana III (Geringer-8Sameth et al., Drlica-Wagner, et al., DH & Linden)
b¯b Indus II 8 Tucana III Range Favored Reticulum II by Galactic Center Tucana IV 6 84% Containment 97.5% Containment
TS 4
2
0 101 102 103 104 DM Mass (GeV)
Figure 4. Local detection significance,Fermi Collaboration expressed as, arXiv:1611.03184 a log-likelihood test statistic (TS), from the broad-band analysis of each target in + Table 1 assuming DM annihilation through the b¯b (left) or ⌧ ⌧ (right) channels. The bands represent the local one-sided 84% (green) and 97.5% (yellow) containment regions derived from 300 random sets of 45 blank-sky locations. Curves corresponding to targets with peak significance larger than the local 95% expectation from blank-sky regions are explicitly colored and labeled, while other targets are shown in gray.
Table 2 Targets with the Largest Excesses above Background
(1) (2) (3) (4) (5) (6) (7) Name Channel Mass (GeV) TS plocal ptarget psample + Indus II ⌧ ⌧ 15.8 7.4 0.01 (2.3 )0.04(1.7 )0.84(-1.0 ) + Reticulum II ⌧ ⌧ 15.8 7.0 0.01 (2.3 )0.05(1.7 )0.88(-1.2 ) + Tucana III ⌧ ⌧ 10.0 6.1 0.02 (2.1 )0.06(1.5 )0.94(-1.6 ) + Tucana IV ⌧ ⌧ 25.0 5.1 0.02 (2.1 )0.09(1.3 )0.98(-2.1 ) Note. — (1) Target name (2) best-fit DM annihilation channel (3) best-fit DM particle mass (4) highest TS value (5) local p-value calibrated from random blank regions (6) target p-value applying a trials factor from testing multiple DM annihilation spectra (7) sample p-value applying an additional trials factor from analyzing 45 targets. The Gaussian significance associated with each p-value is given in parentheses. More details can be found in Section 3. using the spectroscopic J-factors from Geringer-Sameth We reiterate that this analysis assumes that the newly et al. (2015b) as opposed to those from Martinez (2015). discovered systems are DM-dominated, similar to the The two data sets give compatible results (see DW15); known population of ultra-faint dSphs. Some of the more however, the J-factors derived by Geringer-Sameth et al. compact systems might actually be faint outer-halo star (2015b) rely on fewer assumptions about the popula- clusters. Some of the larger systems also may be subject tion of dSphs and provide slightly more conservative esti- to tidal stripping, in which case the distance-based esti- mates for the predicted J-factors. The predicted J-factor mation described above may not apply. On-going spec- for each stellar system is shown in Table 1. troscopic analyses seek to robustly determine the DM In addition to predicting the value of the J-factor we content of new systems and identify those that have com- approximate the uncertainty achievable with future ra- plicated kinematics. dial velocity measurements. The uncertainty on the J-factor derived from spectroscopic observations depends 5. DARK MATTER CONSTRAINTS on several factors, most importantly the number of stars We use the spectroscopically determined J-factors for which radial velocities have been measured. For ultra- (when possible) and predicted J-factors (otherwise) for faint dSphs that are similar to the dSph candidates, spec- each confirmed and candidate dSph to interpret the - tra have been measured for 20–100 stars. Additional ray flux upper limits within a DM framework. Figure 6 sources of uncertainty include the DM density profile summarizes the observed flux and v upper limits de- and dynamical factors such as the velocity anisotropy rived for individual confirmed andh candidatei dSphs, as- of member stars. We consider characteristic J-factor un- suming a DM particle with a mass of 100 GeV annihilat- certainties, log = 0.4, 0.6, 0.8 dex, for the newly ¯ 6 10 J { } ing through the bb-channel. We find that the observed discovered ultra-faint satellites lacking spectroscopically upper limits are consistent with expectations from blank- determined J-factors. Note that these uncertainties re- sky regions. We also show the median expected upper fer to characteristic measurement uncertainties on the J-factor for a typical dSph, and do not reflect any in- 6 Results for both channels as well as bin-by-bin likelihood func- trinsic scatter that may exist in a larger population of tions for each target are available in machine-readable format at: satellites. http://www-glast.stanford.edu/pub_data/1203/. Dan Hooper – The WIMP is Dead! Long Live the WIMP! The plot I see in my dreams… II Ray Flux Flux Ray - Recticulum III 1 Tucana Willman Berenices Coma Major II Measured Gamma Measured Ursa Segue Segue I J-factor (proportional to the predicted annihilation signal) Dan Hooper – The WIMP is Dead! Long Live the WIMP! Annihilating Dark Matter and Cosmic-Ray Antiprotons § The AMS-02 Collaboration has observed a small excess in the cosmic-ray antiproton spectrum at ~10-20 GeV § The excess is well fit by a ~50-90 GeV dark matter particle with an annihilation cross section of ~10-26 cm3/s (for bb), in good agreement with the Galactic Center excess § Although statistically significant at face value (~4.5σ), the systematics associated with the antiproton production cross section and the effects of solar modulation are difficult to quantify Cuoco, et al., arXiv:1610.03071 § Searches for cosmic-ray anti-deuterons Cui, et al., arXiv:1610.03840 and anti-helium may help to clarify Reinert, Winkler, arXiv:1712.00002 Cui, et al., arXiv:1803.02163 Cholis, Hooper and Linden (in prep.) Dan Hooper – The WIMP is Dead! Long Live the WIMP! Summary § Direct detection experiments have improved in sensitivity at an exponential rate over the past 15 years, and have ruled out many well-motivated dark matter models; many others will be explored over the next decade § The LHC is our most comprehensive tool to explore the TeV-scale; many different analyses and search strategies are of relevance to dark matter § Indirect searches using gamma rays and cosmic rays are currently testing the range of annihilation cross sections that are predicted for a thermal relic, for masses up to ~100 GeV § These experiments are collectively testing the WIMP paradigm § Although the WIMP is far from dead, the lack of a clear discovery has redirected the field and has lead to an explosion in beyond-WIMP model building § The Galactic Center’s GeV excess remains compelling: highly statistically significant, robust, extended, spherical, and difficult to explain with known or proposed astrophysics; future gamma-ray observations of dwarf galaxies and radio searches for millisecond pulsars will provide critical tests to establish the origin of this signal