Carsten Rott. mps. ohio-state. edu. (for the IceCube Collaboration)

Transcription

1 Searches for Dark Matter from the Galactic Halo with IceCube Carsten Rott mps. ohio-state. edu (for the IceCube Collaboration) Center for Cosmology and AstroParticle Physics (CCAPP) The Ohio State University CCAPP Symposium Columbus OH, October 12 14, Carsten Rott 1

2 IceCube Halo WIMP analysis Outline Introduction Dark Matter Searches with IceCube Neutrinos from the Dark Matter Halo IceCube Sensitivity Conclusions and Outlook Carsten Rott, CCAPP Symposium 10/13/ 2

3 Season Deployed 07/08 06/07 IC string strings strings IC strings strings 09/10 05/06 10/11 08/09 04/05 CCAPP Symposium Carsten Rott Carsten - ICRC09 Rott 3

4 Dark Matter Understanding Observational Evidence Particle Nature Non-baryonic Cold massive Not strongly interacting Stable (long lived) WIMP Mass? Cross-sections? Self-annihilation < Av> Interaction with matter Theoretical Model Measure SUSY, LED,... Carsten Rott, CCAPP Symposium 09 10/13/ 4

5 Solar Earth Halo Neutrino Flux, Scattering Neutrino Flux,? Neutrino Flux, Selfannihilation cross-section cross-section Muon neutrinos Muon neutrinos Muon neutrinos, Cascades Background off-source onsourcsource Background simulations Background off-source on- Excess Excess Anisotropy, Spectrum IceCube ( + DeepCore) IceCube ( + DeepCore) DeepCore ( + IceCube) Solar/Earth WIMPs probe WIMP-nucleon scattering cross-section (compare to direct detection experiments) Halo WIMPs compare to -ray indirect detection experiments Carsten Rott 5

6 Expected Flux Annihilation ~ 2 Decay ~ Carsten Rott, CCAPP Symposium '09 10/13/ 6

7 Moore NFW Kravtsov Einasto Rsc Moore, NFW, Kravtsov - cusp rs - scale radius Rsc - 8.5kpc Einasto 7

8 Line of sight integral: Moore NFW Kravtsov Expected differential neutrino Flux:? DarkSUSY Measure integrated flux Isotropic emission 8

9 Event selection / dataset days of livetime collected with IceCube operating in the 22-string configuration ( ) 5114 Events after selection from -5 to +85 declination Track selection criteria have been well established for the IceCube point source search, for simplicity and minimization of systematic effect we apply the same selection criteria (Astrophys.J.701:L47- L51,.) No energy weighting factors have been applied The following slides are for simulation, matching the IC22 dataset. We compute the sensitivity based on this dataset Carsten Rott, CCAPP Symposium 10/13/ 9

10 Halo Profile Dependence Background MC scaled 5114 events Signal: Line spectrum of 5TeV WIMP Kravtsov Einasto NFW Moore Kravtsov Einasto NFW Moore Atmospheric background Atmospheric background Signal has varying intensity, while background is essentially flat ra. Example neutrino induced muon flux 1GeV 50GeV 100GeV 10

11 Signal Region Optimization Calculate flux signal flux as function of the distance to the GC on Northern Hemisphere (-5 o 85 o ) Optimize 75 o 50 o 25 o -5 o Atm. background Signal GC Kravtsov NFW Moore Optimal signal extend of the region shows some dependence on the halo model, annihilation channel and WIMP mass. Their overall behavior is however very similar: Larger regions are better and S/sqrt(B) flattens out or declines beginning with GC 11

12 Analysis Strategy Rel. flux intensity Three regions: Galactic Center: RA = 17:45.6, Dec = -28:56 (2000.0) transition <Nbg>=1300 <Nbg>=1300 on source off source on-source signal region transition control region off-source background estimate Symmetric on/off source region simplistic approach minimize systematic uncertainties ra. 12

13 How large can the self-annihilation cross-section < A v> be? Yuksel, Horiuchi, Beacom, Ando (2007) KKT unitarity bound BBM Theoretical/cosmological constraints: KKT - Kaplinghat, Knox, Turner (2000) DM annihilation flattens cusp Unitarity bound Unitarity of the scattering matrix Natural scale DM is thermal relic of early universe Derived limits/sensitivity: BBM - Beacom, Bell, Mack (2008) Cosmic time-integrated annihilation MW Halo Isotropic J( )(immediate neighborhood) MW Halo Average Average flux from halo MW Halo Angular 30 o cone around GC Carsten Rott, CCAPP Symposium 10/13/ 13

14 IC22 Sensitivity Average upper limit at 90%C.L. N90 = 65.2 events (This is a limit on the difference between on and off-source) soft(bb) hard(ww) hard( ) line ( ) DarkSUSY Line width: Uncertainty due to halo profile Central line: - Einasto Carsten Rott, CCAPP Symposium 14

15 IC22 Sensitivity for neutrino line spectrum All final states except neutrinos will p r o d u c e g a m m a - r a y s Following the argumentation of BBM, one can place a model independent l i mi t on the t ot al WIMP sel f - a n n i hi l a t i o n c r o s s -s e c t i on by assuming only the least visible c h a n n e l is p r o d u c e d. IceCube data can be used to constrain the WIMP self-annihilation crosssec ti on above abo ut 10 0GeV, DeepCore will extend the reach significantly to lower energies down to and below accelerator bounds. Carsten Rott, CCAPP Symposium 15

16 Systematic uncertainties Signal acceptance: Uneven exposure (negligible) Ice properties / DOM efficiency (~20%) MC/data disagreement (horizontal events) Background: Majority of systematics cancel out (as we use the data itself) existing large scale anisotropy - Uneven exposure - Neutrino anisotropy caused by cosmic ray anisotropy Carsten Rott, CCAPP Symposium 16

17 Uneven exposure Track reconstruction efficiency varies in detector coordinates In equatorial coordinates this reconstruction efficiency is smeared out (as the detector rotates) Uneven detector up-time can however reduce this smearing effect Detector down-time correlates with satellite visibility (maintenance mode) Detector uptime in sidereal days defines this impact 100% 90% IceCube up-time sketch 17

18 Track reconstruction efficiency shows mirror symmetry Uneven exposure transition Normal + 180deg rotated on source off source Uneven exposure systematic uncertainty is on the order of the difference between the two graphs ~0.1% Total systematic effect ~ 3% N/N ~ 0.1% 18

19 Celestial Cosmic Ray intensity map at different energies (TIBET) 4 TeV 6.2 TeV 12 TeV 50 TeV Science 314: ,2006. Neutrino energy is roughly 1/10-1/30 of the shower energy (cosmic ray primary) most relevant energy region 300 TeV See also Abbasi et al ICRC 09 Phys.Rev.D53: ,

20 Cosmic ray anisotropy Cosmic ray anisotropy could also cause anisotropy in atmospheric neutrinos At relevant energies the anisotropy of cosmic rays is a fraction of a percent On/off-source region has a background expectation of 1300 neutrino candidate events For an anisotropy of 0.2%, a maximum effect of 2.6events -> 4% syst. uncertainty can be expected 20

21 Summary / Conclusions Neutrino Telescopes can probe Dark Matter selfannihilation cross-section We have computed IceCube s sensitivity to Galactic halo WIMPs, using the IC22 string up-going neutrino sample We find that IceCube will be able to set very completive constraints on the Dark Matter selfannihilation cross-section < Av> Preliminary study of systematics shows they are well under control DeepCore and the larger IceCube detector will have significantly improved sensitivities Carsten Rott 21

22 Outlook IceCube 22 string sensitivity for galactic halo WIMPs has been evaluated Data is available and analysis is getting finalized Beginning with the IceCube 40 string configuration, a starting track filter has been active This allows to probe the Galactic Center directly Halo WIMP sensitivity is expected to be significantly larger Larger uncertainty due to halo models Principle of veto reconstruction (Schulz, Euler, Grant, ICRC ) 22

23 Backup 23

24 IC22 Point Source Analysis arxiv: Carsten Rott, CCAPP Symposium 24

25 IceTop 80 Stations (2 tanks each) Surface air shower array 300TeV threshold IceCube (InIce) 80 Strings with 60 DOMs each Hexagonal pattern with an interstring distance of 125 m Vertical DOM spacing of 17 m Optimized for TeV range AMANDA 19 strings with 677 modules total m vertical spacing m horizontal spacing Deep Core 6 Strings with 60 High Quantum Efficiency DOMs (vert. spacing 7 m) Low Energy extension (20-100GeV) IceCube will instrument a volume of one cubic kilometer of Antarctic ice by 2011 Carsten Rott - ICRC09 25

26 Selection Criteria Optimization IC22 point source track selection criteria are used... can we optimize the cuts further for our purposes? 5.5TeV WIMP Line spectrum 811 GeV WIMP Line spectrum Keep cuts as they are to avoid becoming model specific and keep broad acceptance to large signal parameter space (generality) 26

27 How large can the self-annihilation cross-section be? Most often assumed natural scale 3 x cm 3 /s KKT Kaplinghat, Knox, Turner (2000) cusp profiles Unitarity Limit from Q.M. The probabilities for elastic and inelastic scattering must sum to 1 Unitarity of the scattering matrix [Beacom,Bell, Mack 2008]] Carsten Rott 27

28 Model Independent Limit on DM Annihilation Cross Section Assume DM annihilations only produce SM final states Stringent upper limit on total annihilation cross section can be obtained by assuming only neutrinos are produced in final states (worst case, least visible) Any other final state would be more visible [Mack, Jacques, Beacom, Bell, Yuksel 2008] 28

29 Background Estimation 29

30 IC22 Point Source Analysis Gulliver 32-iter SPE Reduced LogL (g32rlogl) < 9.5 Paraboloid Fit Status == 0 Paraboloid Sigma < 3 deg 32iter "Umbrella Fit" loglikelihood - 32iter Normal Fit loglikelihood > 15 32iter Bayes Fit loglikelihood - 32iter Normal Fit loglikelihood > 30 1-iter split fits 1 and 2 both have zenith > 70 degrees days of livetime Astro-ph