Dark matter, black holes, and the Fermi-LAT GeV excess

Transcription

1 Dark matter, black holes, and the Fermi-LAT GeV excess Jessie Shelton University of Illinois, Urbana-Champaign Mainz Institute for Theoretical Physics July 18, 2014

2 interesting-ness Abbreviated history of the excess Goodenough & Hooper (2011): excess over power-law background models in the Galactic Centre Independent analyses by multiple groups with increasingly sophisticated methodology find similar results (Abazajian & Kaplinghat (2012, 2014), Gordon & Macias (2013), Fermi Collaboration (?) ) time???

3 interesting-ness Abbreviated history of the excess Goodenough & Hooper (2011): excess over power-law background models in the Galactic Centre Independent analyses by multiple groups with increasingly sophisticated methodology find similar results (Abazajian & Kaplinghat (2012, 2014), Gordon & Macias (2013), Fermi Collaboration (?),... ) time???

4 interesting-ness Abbreviated history of the excess Goodenough & Hooper (2011): excess over power-law background models in the Galactic Centre Independent analyses by multiple groups with increasingly sophisticated methodology find similar results (Abazajian & Kaplinghat (2012, 2014), Gordon & Macias (2013), Fermi Collaboration (?),... ) time??? Hooper & Slatyer (2013) establish spatial extent out to ~10 o from the GC Daylan, Finkbeiner, Hooper, Linden, Portillo, Rodd, & Slatyer (2014): increased angular resolution tightens case

5 interesting-ness Abbreviated history of the excess Goodenough & Hooper (2011): excess over power-law background models in the Galactic Centre Independent analyses by multiple groups with increasingly sophisticated methodology find similar results (Abazajian & Kaplinghat (2012, 2014), Gordon & Macias (2013), Fermi Collaboration (?),... ) time??? Hooper & Slatyer (2013) establish spatial extent out to ~10 o from the GC Daylan, Finkbeiner, Hooper, Linden, Portillo, Rodd, & Slatyer (2014): increased angular resolution tightens DM case

6 The Galactic Centre Excess Spatially uniform continuum spectrum Annihilation in an adiabatically contracted NFW halo Overall rate Daylan, Finkbeiner, Hooper, Linden, Portillo, Rodd, Slatyer

7 Mass, cross section, final states: Daylan, Finkbeiner, Hooper, Linden, Portillo, Rodd, Slatyer

8 A cautionary note stacked MSPs Known astrophysical objects with similar energy cutoff: millisecond pulsars Energy spectrum? Spatial distribution? Need novel luminosity distribution to avoid over-predicting resolved MSPs at high latitude By this argument: < 5-10 % of excess Hooper, Cholis, Linden, Siegal-Gaskins, Slatyer

9 Mediating annihilations in the GC SM SM

10 Mediating annihilations in the GC SM SM h Z SM SM SM SM

11 Mediating annihilations in the GC SM SM h Z SM SM SM SM PDG

12 New mediators X SM SM m X >m annihilation, direct detection amenable to contact operator description collider signals may require explicit introduction of new d.o.f. pseudoscalar X requires 2HDM or similar

13 New mediators X X X SM SM SM SM m X >m annihilation, direct detection amenable to contact operator description collider signals may require explicit introduction of new d.o.f. m X <m dominant DM annihilation mode is to mediators lifetime of mediators bounded by BBN pseudoscalar X requires 2HDM or similar

14 New mediators DM L int SM Thermal history fixes this coupling SM The coupling to the SM can be parametrically small SM Pospelov, Ritz

15 Final states in cascade annihilations SM singlets couple through portals: Higgs: H 2 s 2 hypercharge: B µ V µ 1 gluons: M sgµ G µ,... 1 M agµ e Gµ m DM 2 m DM 2 1 ± r 1 m med m DM

16 Cascade annihilation spectra 80 8 E 2 dn de max HGeVL, bb mdm HGeVL m med HGeVL Martin, JS, Unwin

17 Portals into the SM: Hypercharge E 2 dn de max HGeVL, vector L int = V µ B µ 40 3 kinetically mixed U(1) gets mass from dark SSB mdm HGeVL for m. 20 GeV, coupling to SM fermions approximately proportional to charge m med HGeVL Martin, JS, Unwin

18 Portals into the SM: Higgs 80 8 E 2 dn de max HGeVL, higgs portal L int = s 2 H s inherits Yukawa couplings through Higgs mixing after SSB mdm HGeVL 40 4 radiative corrections: HDECAY for branching fractions 20 2 fit results also apply to pseudoscalar a m med HGeVL Martin, JS, Unwin

19 Portals into the SM: gluons E 2 dn de max HGeVL, scalar HggL L int = 1 G µ G µ, 1 ag µ G µ 60 8 mdm HGeVL new colored matter at high scales similar operators with electroweak field strengths can generate gamma ray boxes: potentially interesting signal for future m med HGeVL Martin, JS, Unwin

20 Many ways to fit a spectrum m DM = 22 GeV m med = 12 GeV h vi = cm 3 /s m DM = 40 GeV m med = 20 GeV h vi = cm 3 /s /s/sr) 2 dφ/de (GeV/cm 2 E vector scalar(higgs) scalar(gg) 0 m DM = 60 GeV m med = 40 GeV h vi = cm 3 /s log (E /GeV) γ 10 Martin, JS, Unwin

21 A sample model Consider fermionic DM and a dark vector mediator: X V f X X V f f ' mvêmc f ' 0.2 Requiring h 2 =0.11 determines D m c HGeVL D (10 3 ) Martin, JS, Unwin

22 A sample model Consider fermionic DM and a dark vector mediator: mchgevl LUX then constrains admissible kinetic mixing m V HGeVL log 10 Martin, JS, Unwin

23 A sample model 10-1 EWPM CMS 10-3 Collider constraints e E774 a m, 5 s a m,±2 s favored a e APEX HADES E141 KLOE A1 BaBar BrHhÆZZ D L=10-6 PROMPT Strongest bounds: quarkonia LHC signals: exotic Higgs decays? competiveness with LUX depends on structure of dark Higgs sector, model dependent 10-5 NON-PROMPT Orsay U m Curtin, Essig, Gori, Jaiswal, Katz, T. Liu, Z. Liu, McKeen, JS, Strassler, Surujon, Tweedie, Zhong

24 Another sample model Consider scalar DM annihilating to a dark scalar: Let s get a VEV; induces Higgs mixing msêmf After SSB: Freezeout sets to y /m s ) 4, m s,, 4 (insensitive m f HGeVL 4 (10 2 ) Martin, JS, Unwin

25 Another sample model Consider scalar DM annihilating to a dark scalar: mfhgevl Direct detection less constraining than in vector case (and / y 4 ) for LUX (y = 1) m s HGeVL Martin, JS, Unwin

26 Another sample model 1.2 In fact, the most stringent limits come from exotic Higgs decays Constraint on total exotic branching fraction assuming SM production e (direct observation will be hard at LHC) m s HGeVL Martin, JS, Unwin

27 And another sample model Scalar DM annihilating to a dark scalar, no condensation: Direct detection proceeds at one loop: no signals Higgs portal coupling could mediate h! 4g msêmf m f HGeVL 5 4 (10 2 ) Martin, JS, Unwin

28 No silver bullets soon Many proposed models predict signals right around the corner at LHC and/or LUX (and/or flavour), but others (e.g. ours) do not Gamma-ray signals from other objects (e.g. dwarfs) could add weight to DM interpretation, but statistically difficult in Fermi Antiprotons potentially powerful discriminant, but systematics are challenging DM gamma ray signal from Sagittarius A*

29 No silver bullets soon Many proposed models predict signals right around the corner at LHC and/or LUX (and/or flavour), but others (e.g. ours) do not Gamma-ray signals from other objects (e.g. dwarfs) could add weight to DM interpretation, but statistically difficult in Fermi Antiprotons potentially powerful discriminant, but systematics are challenging DM gamma ray signal from Sagittarius A*

30 No silver bullets soon Many proposed models predict signals right around the corner at LHC and/or LUX (and/or flavour), but others (e.g. ours) do not Gamma-ray signals from other objects (e.g. dwarfs) could add weight to DM interpretation, but statistically difficult in Fermi Antiprotons potentially powerful discriminant, but systematics are challenging DM gamma ray signal from Sagittarius A*

31 No silver bullets soon Many proposed models predict signals right around the corner at LHC and/or LUX (and/or flavour), but others (e.g. ours) do not Gamma-ray signals from other objects (e.g. dwarfs) could add weight to DM interpretation, but statistically difficult in Fermi Antiprotons potentially powerful discriminant, but systematics are challenging DM gamma ray signal from Sagittarius A*

32 Black hole-induced density spikes A black hole growing adiabatically in an NFW halo gives rise to a steep spike: inside zone of influence density rises like rb where spike index depends on NFW profile when annihilation becomes important, spike levels out to a shallow r sp r b =0.2 M BH v 2 0 rin 1/2 r sp = 9 r in : (r) = 2 c 4 c m h vi Gondolo and Silk Vasiliev

33 A fiducial model for the spike CDM inner halo plus spike: best fit parameters sp =2.36 r HGeVêcm 3 L r HpcL assume GeV reference DM candidate, annihilating to bb Fields, Shapiro, JS

34 Parametric dependence steep power law inputs ) strong dependence on v 0 M BH inner NFW index DM properties: potential DM signal allows (relatively) precise statements:, c ann c ann = m h vi

35 Parametric dependence F spike êf halo Hinner 1 o L r ann Æ 2 r ann g c = 1.4 r ann Æ 1 r 2 ann g c = DM velocity dispersion v 0 HkmêsecL Fields, Shapiro, JS

36 The point source at the centre of the galaxy 2FGL J E 2 dfêde Hphotons GeVêcm 2 secl E HGeVL Second Fermi Point Source Catalog

37 The point source at the centre of the galaxy Inner 1 o of halo E 2 dfêde Hphotons GeVêcm 2 secl E HGeVL Fields, Shapiro, JS

38 The point source at the centre of the galaxy canonical adiabatic spike: sp =2.36. E 2 dfêde Hphotons GeVêcm 2 secl E HGeVL Fields, Shapiro, JS

39 The point source at the centre of the galaxy vector portal gluon portal 1.50 E 2 dfêde H10-6 photons GeVêcm 2 secl (i) (iv) Ê Ê 2FGL J Ê Ê (ii) (iii) E HGeVL 80% taus, 20 % b s Fields, Shapiro, JS

40 Stellar Heating dense stellar population in Galactic nucleus could alter spike: heating enhanced DM capture by BH limiting spike: sp =1.5 existence of such a stellar cusp still unsettled Merritt Primack and Gnedin

41 The point source at the centre of the galaxy non-equilibrium spike: sp =1.8. E 2 dfêde Hphotons GeVêcm 2 secl E HGeVL limiting spike: sp =1.5. Fields, Shapiro, JS

42 The spiky takeaway You can t have both a DM interpretation of the excess a canonical adiabatic spike GeV-range spectrum of Sgr A* is interestingly close to a heated spike + halo DM interpretations of excess don t harmonize well with other spiky scenarios (cores, sudden) Discovery of e.g. a pulsar associated with 2FGL J would significantly weaken case for DM spike

43 Conclusions GC excess, if due to DM, suggests new interactions Minimal cascade possibilities: fermionic DM + vector; scalar DM + (pseudo-)scalar mediators; make more natural but less minimal models by adding additional new species Cascade annihilations: parametrically explain lack of deviations from SM...at the cost of making terrestrial confirmation harder weakest couplings may be best constrained astrophysically Milky Way s SMBH spike adds point-like contribution to DM gamma-ray signal Pick at least one: not DM, or: no canonical adiabatic spike

44 Backup

45 A sample model Cosmological history: V still decays long before BBN e = 10-6 t H10-11 secl m V HGeVL For m GeV, values of mixing down to V & are safe

46 A sample model Cosmological history: Thermal decoupling? estimate thermal decoupling at: Tdec 2 (10 MeV) 2 mv 4 D 10 GeV 20 GeV P m gi Q 2 i assume reheating yields T DM = T SM initially provided T fo >T P T, relatively insensitive