Indirect Evidence for Dark Matter Annihilation from Diffuse Galactic Gamma Rays Wim de Boer, Christian Sander, Valery Zhukov Univ. Karlsruhe Outline (see astro-ph/0408272) Annihilation cross section determined by Hubble constant and WMAP relic density EGRET Data on diffuse Gamma Rays show excess in all sky directions with the SAME energy spectrum WIMP mass between 50 and 100 GeV from spectrum of EGRET excess Halo distribution from sky map Data consistent with Supersymmetry December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 1
Dark Matter annihilation cross section determined by Hubble constant! December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 2
Hubble const. determines WIMP annihilation x-section Thermal equilibrium abundance Jungmann,Kamionkowski, Griest, PR 1995 Comoving number density Actual abundance T=M/25 x=m/t Boltzmann equation: H-Term takes care of decrease in density by expansion. Right-hand side: Annihilation and Production. T>>M: f+f->m+m; M+M->f+f T<M: M+M->f+f T=M/25: M decoupled, stable density (wenn annihilation rate @ expansion rate, i.e. G=<sv>nc @ H!) WIMP annihilation is a strong source of antiprotons, positrons and gammas by annihilation into quarks. Present number density (W h 2 =0.113 0.009) requires <sv>=2.10-26 cm 3 /s December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 3
DM annihilation in Supersymmetry c c ~ f f f c c A f f c c Z f f c c c W c c 0 W c Z Z Dominant diagram for WMAP cross section: c + c A b bbar quark pair B-fragmentation well studied at LEP! Yield and spectra of positrons, gammas and antiprotons well known! December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 4
10-27 Annihilation cross sections in m 0 -m 1/2 plane (µ > 0, A 0 =0) tan=5 tan=50 bb t t 10-24 bb t t EGRET WMAP tt WW tt WW For WMAP x-section of <sv>@2.10-26 cm 3 /s one needs large tanβ in bulk region (no coannihilation, no resonances) December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 5
SUSY Mass spectra in msugra LSP largely Bino DM may be supersymmetric partner of CMB Charginos, neutralinos and gluinos light December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 6
Unification of gauge couplings Update from Amaldi, db, Fürstenau, PLB 260 1991 SM SUSY With SUSY spectrum from EGRET data perfect gauge coupling unification possible for a s =0.123 as determined from R l =Γ had /Γ lepton This value is independent of luminosity. Note a s =0.115 from Γ had alone, but this value is probably low because of systematic uncertainty in LEP lumi. If lumi increased by 3s, a s =0.122 and N v =3.0! (db, Sander, PL B585 (2004)) December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 7
EGRET excess interpreted as DM consistent with WMAP, Supergravity and electroweak constraints Stau LSP 0 Stau coannihilation m A resonance MSUGRA can fulfill all constraints from WMAP, LEP, b->sg, g-2 and EGRET simultaneously, if DM is neutralino with mass in range 50-100 GeV and squarks and sleptons are O(1 TeV) Incomp. with EGRET data WMAP No EWSB Bulk EGRET December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 8
Indirect Dark Matter detection December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 9
Indirect detection of DM with gamma rays First question: how to tell source of a gamma ray? Answer: one cannot, since gammas observed along line of sight, and many sources in a given direction. How to distinguish then between sources? Point sources from intensity ( hot spots ), if strong enough Diffuse gamma ray sources (pp->p 0 +x->gg+x, eg->eg, en->egn, Extragalactic Background, DMA) all have quite different ENERGY spectra AND different SKYMAPS Strategy: fit SHAPES of various contributions to energy spectrum (FREE NORMALIZATIONS) in 360 SKY DIRECTIONS -> CHECK if normalizations and intensities in various sky directions are consistent with models, BUT WE DO NOT DEPEND ON Galactic models or supersymmetric parameters!!!!! Beware: errors in EGRET data dominated by systematic uncertainties-> strong correlations -> to be taken into account in fits. (as in D Agostini, db, Grindhammer, Phys.Lett.B229:160,1989) Published data not sufficient. Need direct analysis of public EGRET data. December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 10
Executive Summary Expected Profile Observed Profile z xy xz xy v 2 xz Halo profile M/r=cons. and r M/r 3 r 1/r 2 for const. rotation curve Divergent for r=0? NFW 1/r Isotherm const. Rotation Curve Rotation Curve x y DM halo halo disk 2003, Ibata et al, Yanny et al. bulge disk Inner Inner Ring Ring F R F G Outer Outer Ring Ring December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 11
Do other galaxies have bumps in rotation curves? Sofue & Honma December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 12
EGRET DATA on diffuse Gamma Rays December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 13
EGRET on CGRO (Compton Gamma Ray Observ.) 9 yrs of data taken in space! (1991-2000) December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 14
Origins of diffuse gamma rays Gamma Ray Flux from WIMP annihilation: Similar expressions for: pp->p 0 +x->gg+x, (r given by gas density, highest in disc) eg->eg, en->egn, (r given by electron/gamma density, highest in disc) Extragalactic Background (isotropic) DM annihilation (r 1/r 2 for flat rotation curve) All have very different, but known energy spectra. Cross sections known. Densities not well known, so keep absolute normalization free for each process. Fit shape of various contributions with free normalization, but normalization limited by experimental overall normalization error, which is 15% for EGRET data. Point-to-point errors @ 7% (yields good c 2 ). December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 15
Strong, Moskalenko, Reimer, APJ.613, astro-ph/0406254 Excess of Diffuse Gamma Rays above 1 GeV (first publications by Hunter et al, Sreekumar et al. (1997) A B C p 0 Inv. Compton Bremsstrahlung 2004 D E F A: inner Galaxy (l=±30 0, b <5 0 ) B: Galactic plane avoiding A C: Outer Galaxy D: low latitude (10-20 0) E: intermediate lat. (20-60 0 ) F: Galactic poles (60-90 0 ) December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 16
Excess of Diffuse Gamma Rays has same spectrum in all directions compatible with WIMP mass of 50-100 GeV Egret Excess above extrapolated background from data below 0.5 GeV Statistical errors only Excess same shape in all regions implying same source everywhere Important: if experiment measures gamma rays down to 0.1 GeV, then normalizations of DM annihilation and background can both be left free, so one is not sensitive to absolute background estimates, BUT ONLY TO THE SHAPE, which is much better known. December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 17
Diffuse Gamma Rays for different sky regions Good Fits for WIMP masses between 50 and 100 GeV A: inner Galaxy B: outer disc C: outer Galaxy D: low latitude E: intermediate lat. F: galactic poles 3 components: galactic background + extragalactic bg + DM annihilation fitted simultaneously with same WIMP mass and DM normalization in all directions. Boost factor around 70 in all directions and statistical significance > 10s! December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 18
Uncertainties in background and signal shapes RED=flux from DM Annihilation Yellow = backgr. Blue = BG uncert. WIMP MASS 50-100 GeV p 0 WIMPS IC p 0 WIMPS IC 65 100 Extragal. Bremsstr. Extragal. Bremsstr. Blue: uncertainty from background shape Blue: uncertainty from WIMP mass December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 19
Local electron and proton spectra determine shape of gamma background No SM No SM Protons Quarks from WIMPS Electrons Quarks in protons December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 20
Background uncertainties T. Kamae et al, 2004, astro-ph/0410617 III=double break III II BG I Main results on halo profile, substructure, and WIMP mass not affected after renormalization to data between 0.1 and 0.5 GeV. Note: point-to-point errors only half of plotted errors of 15%. Statistical errors negligible. December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 21
E -n n=2.5 2.7 T. Kamae et al, 2004, astro-ph/0410617 n=3.0 December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 22
Energy loss times of electrons and nuclei t -1 = 1/E de/dt t univ Protons diffuse for long times without loosing energy! If centre would have harder spectrum, then hard to explain why excess in outer galaxy has SAME shape (can be fitted with same WIMP mass!) December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 23
Optimized Model from Strong et al. astro-ph/0406254 Change spectral shape of electrons AND protons December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 24
Optimized Model from Strong et al. astro-ph/0406254 Change spectral shape of electrons AND protons 100 MeV 150 MeV 300 MeV 500 MeV 1000 MeV 2000 MeV December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 25
Optimized Model from Strong et al. astro-ph/0406254 Change spectral shape of electrons AND protons Probability of optimized model if c2 measured in 360 sky directions and integrate data with E>0.5 GeV: c2= 962.3/324 for Optimized Model without DM +correl. errors: Prob = < 10-14 In this sense it WILL go a long way towards realizing a model reproducing astronomical and directly-measured data on cosmic rays in the context of a single model of the high-energy Galaxy WITHOUT DM. I believe it is impossible! c2= 306.5/324 for Optimized Model +DM +correl. errors: Prob = 0.736 Boostfactor=25 for OM vs 70 for CM December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 26
Fit results of halo parameters DM Gamma Ray Flux:(<sv> from WMAP) =2 Enhancement of rings over 1/r 2 profile 2 and 7, respectively. Mass in rings 1.3 and 0.3% of total DM 1/r 2 2 Gaussian rings H 2 H R [kpc] 14 kpc coincides with ring of stars at 14-18 kpc due to infall of dwarf galaxy (Yanny, Ibata,..) 4 Boostfactor @ 50-70 Dokuchaev et al: 10<B<100, IDM2004 4 kpc coincides with ring of neutral hydrogen molecules! December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 27
Longitude and Latitude for Gammas BELOW 0.5 GeV In the plane (± 5 0 in lat. ) Out of the plane (± 30 0 in long.. ) December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 28
Longitude and Latitude for Gammas ABOVE 0.5 GeV Longitude: azimuthal angle Latitude: angle out of the plane December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 29
Longitude fits for 1/r 2 profile with/w.o. rings WITH 2 rings WITHOUT rings DISC 10 0 <b<20 0 DISC 10 0 <b<20 0 5 0 <b<10 0 20 0 <b<90 0 5 0 <b<10 0 20 0 <b<90 0 Halo parameters from fit to 180 sky directions: 4 long. profiles for latitudes <5 0, 5 0 <b<10 0, 10 0 <b<20 0, 20 0 <b<90 0 (=4x45=180 directions) December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 30
Normalization of background: Compare with GALPROP, which gives absolute prediction of background from gas distributions etc. December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 31
Background scaling factors Elements in our Galaxy Background scaling factor = Data between 0.1 and 0.5 GeV/GALPROP GALPROP = computer code simulating our galaxy (Moskalenko, Strong) December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 32
Normalization of DMA -> Boost factor (= enhancement of DMA by clustering of DM) December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 33
Clustering of DM -> boosts annih. rate Annihilation DM density squared! Clumps in N-body simulations with kpc resolution Clumps in analyt. calculations with pc resolution Clustersize: 10 14 cm =10x solar system סּM M min @ 10-8 Cluster density @ 25 pc -3 Halo mass fraction in clumps: 0.002 Clumps with M min give the dominant contribution to DM annihilation -> many in a given direction -> similar boost factor in all directions Boost factor ~ <r 2 >/<r> 2 ~ 10-100, i.e. effective annihilation cross section is 20-200 10-26 cm 3 s -1 Fitted boostfct: 20-70 (depending on background) December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 34
Rotation curves with different backgrounds Rotation Curve (CM) Rotation Curve (different BG) disk DM halo bulge Inner Ring Outer Ring סּM 8.10 10 Outer Ring סּM 4.10 10 סּM Rotation curve shows there is a ring of CDM with a mass of (4-8).10 10 Not enough baryons in outer ring to explain mass (only 10 9 סּM in stars)! December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 35
More on Halo profiles Do not expect perfectly spherical halo but triaxial one with short axis aligned with angular momentum of disc! z b c y a x Angular momentum along short axis c as expected from N-body simulations (Bailin, Steinmetz, astro-ph/0408163)) Disc in flattened rugby ball with L disc along short axis! Furthermore expect from N-body simulations visible mass/dm mass to be few % December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 36
EGRET data compatible with prolate isothermal halo profile with b/a @ 0.9, c/a @ 0.8 W. de Boer et al., astro-ph/0408272 December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 37
Galactic Parameters Jimenez, Verde,,Peng Oh; astro-ph/0201352 M disc /M halo Isothermal profile N-body simulations EGRET Concentration M halo M halo Further parameters: R 200 סּM Baryonic: 6.10 10 ;סּM =295 kpc; DM: 3.10 12 Mass of outer (inner) ring»1.6 (0.3)% of total DM. Only so important, while nearby!!!!!!!!! December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 38
What do we have? DMA describes all sky directions with single WIMP mass and WIMP annihilation into mono energetic quarks DM distributed as 1/r 2 profile with substructure at positions with enhanced gravitational potential Rotation curve can be reconstructed from GAMMA Rays! (First time real explanation for change of slope in outer RC) What is the problem? Many aspects from different fields-> hard to judge all aspects simultaneously. December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 39
Objections from astrophysicists Astrophysicists: cannot believe collisionless DM is self-annihilating and that particle physicists even believe to know spectral shapes of final states! They prefer their local bubbles. Answer: too bad, but their local bubbles or optimized models do not work if they learn what covariance matrices are. December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 40
Objections from particle physicists Particle physicists: you assume spectrum of protons to be the same everywhere in galaxy. How can you be sure? Answers: 1) If centre of galaxy would be different because of SN there,no reason to get same shape of excess in outer galaxy (very few SN) 2) energy loss times > age of universe, very hard to image strong changes in spectral shape if protons diffuse from centre to us in» 10 8 yr December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 41
Objections from astronomers Astronomers: outer rotation curve determined with different method than inner rotation curve. Hard to compare. Furthermore slope depends on distance to centre (R 0 ) v Sofue &Honma Inner rotation curve Outer rotation curve R 0 =8.3 kpc R 0 =7.0 Answer: First points of outer rot. curve in perfect agreement with inner rot. curve. CHANGE in slope for any R 0 R/R 0 Black hole at centre: R 0 =8.0 0.4 kpc December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 42
Positron fraction and antiprotons from present balloon exp. Positrons Antiprotons SAME Halo and WIMP parameters as for GAMMA RAYS but fluxes dependent on propagation! DMA can be used to tune models: at present no convection, nor anisotropic diffusion in spiral arms. December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 43
Future Experiments SPACE Experiments: Pamela (2005) antiprotons, positrons GLAST (2007) gammas (higher energies) AMS (2008) all nuclei, gammas positrons, antiprotons Accelerators: LHC all SUSY particles? ILC(500) Charginos, top, Higgs Direct DM searches: CDMS (taking data), Edelweiss, DAMA, ZEPLIN, Cresst, Genius,.. December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 44
Dark Matter Searches with AMS-02 AMS: Alpha Magnetic Spectrometer 2008 Large acceptance detector (0.8 m 2 sr) with excellent particle identification by Silicon Tracker in SC magnet, RICH, TRD, TOF, EM Calorimeter December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 45
AMS-02 capabilities Beryllium Boron Helium 1 year 1 year 6 months 1 day 6 months 1 day 10 Be (t 1/2 =1.5Myr) / 9 Be will allow to estimate the propagation time and size of the ISM B is secondary produced in nuclear interaction, C is primary produced in stars. B/C is sensitive to the diffusion constant 3 He/ 4 He ratio is sensitive to the density of the ISM December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 46
LHC Chargino/Neutralino associated production x-sections Calculated with Pythia (Isasugra) (Martin Niegel) Pythia: Total cc X-section 11 pb. c+c-: 2l, 3pb c+c0: 3l, 6pb December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 47
Comparison with direct DM Searches Spin-independent Spin-dependent DAMA ZEPLIN CDMS Edelweiss Projections Predictions from EGRET data assuming Supersymmetry December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 48
Physics Questions Astrophysicists: What is the origin of GeV excess of diffuse Galactic Gamma Rays? Astronomers: Why a change of slope in the galactic rotation curve above R 0 =8.3 kpc? Why ring of stars at 14 kpc so stable? Why ring of molecular hydrogen at 4 kpc so stable? Cosmologists: How is Cold Dark Matter distributed? Particle physicists: Is DM annihilating as expected in Supersymmetry? December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 49
Answers 1. DM made of WIMPS annihilating into quarks, which yield hard gammas from p 0 decays 2. Annihilation cross section is consistent with HUBBLE constant! Gamma excess shows substructure: 3. a ring of DM at 14-18 kpc coinciding with ring of stars thought to originate from the infall of a dwarf galaxy 4. and ring of DM at 4 kpc correlated with molecular H 2 5. From the ENERGY SPECTRUM of the excess of gamma rays DM: WIMP mass 50-100 GeV 6. From INTENSITY: halo distribution and rotation curve 7. WIMP has properties of the lightest supersymmetric particle Conventional models based on local bubbles cannot explain the stability of the ring of stars at 14 kpc nor the shape of the rotation curve nor the good agreement for all the answers SIMULTANEOUSLY! December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 50
Summary EGRET excess shows all key features from DM annihilation: Excess has same shape in all sky directions: everywhere it is perfectly (only?) explainable with superposition of background AND mono-energetic quarks of 50-100 GeV Excess follows expectations from galaxy formation: 1/r 2 profile, visible matter/dm»0.02, anisotropic infall, high surface density, Ldisc along minor axis of triaxial halo Excess connected to MASS, since it can explain peculiar shape of rotation curve These combined features provide FIRST EVIDENCE that DM is not so dark and follows all expectations one has been imagining so far. December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 51
Is Dark Matter Supersymmetric? Gammas cannot prove the supersymmetric nature of DM, but cross sections for fitted WIMP mass range of 50-100 GeV ARE PERFECTLY CONSISTENT WITH SUPERSYMMETRY. LHC experiments will tell in 2008 if DM is supersymmetric! December 2 2004 Seminar Bonn, W. de Boer, Univ. Karlsruhe 52