Dark Matter, LSP wino dark matter, satellite data, moduli and nonthermal cosmological history, and LHC Gordy Kane May 2009 PAMELA Physics Workshop Over two decades ago recognized that one way to “see” dark matter was via products of annihilation of pairs of DM particles in the galaxy Traditionally the lightest superpartner, LSP, a very good DM candidate Annihilate into everything, but positrons, antiprotons, gammas should be easier to see over backgrounds so look for those Recently reported possible satellite signals and relevant data: PAMELA, Fermi/GLAST, etc Dark matter! Not only learn what DM is – could also be the discovery of supersymmetry (if indeed LSP) – may also point toward underlying theory – probes cosmological history of universe! -Worth lot of effort to untangle the situation, test interpretations – does an LSP give a satisfactory description of the data? maybe OUTLINE OF TALK Survey of DM approaches PAMELA data light wino LSP? -- issues -- description of data Dark matter relic density from non-thermal-equilbrium cosmological history of the universe – but still “wimp miracle”! High scale underlying theory of this phenomenology, moduli Comments on LHC implications Concluding remarks Describing the data with dark matter annihilation (a) thermal cosmological history, and describe all data with one component MLSP > 2 TeV -- stable, < v> ≈3x10-26 cm3 s-1 – need enhancement Arkani-Hamed, Weiner et al – heavy hidden sector DM Bergstrom, Edjso, Zaharijas 0905.0333 Meade, Papucci, Strumia, Volansky 0905.0480 “non-conventional DM” -- decaying DM, lifetime 1026 second Arvanitaki, Dimopoulos et al, 0904.2789 – DM is singlet of SO(10) 16, like RH Shirai, Takahashi, Yanagida – wino LSP with mass 3 TeV Typically positron ratio rises significantly above 100 GeV (b) non-thermal cosmological history -- < v> ≈ 3x10-24 cm3 s-1 , wino LSP, M 200 GeV WINO LSP VERY WELL MOTIVATED FOR DARK MATTER -- two decades -- anomaly mediated supersymmetry breaking (Randall, Sundrum…) -- “split” supersymmetry (Arkani-Hamed, Dimopoulos, …) -- Z’ mediation (Wang, Langacker, Yavin, Paz, Verlinde … ) -- M theory compactified on G2 manifold (Acharya, Kane, …) -- MSSM scan – (Hewett, Rizzo et al) Wino LSP DM annihilation provides most positrons, most energetic positrons compared to other forms of LSP My perspective today: -- study whether wino LSP provides a satisfactory description of PAMELA, Fermi etc data -- no comments on other approaches CAN A WINO LSP DESCRIBE THE PAMELA DATA? Grajek, Kane, Phalen, Pierce, Watson 0812.4555, plus Ran Lu, Cheng Peng recently • Rate? – density too small with thermal equilibrium cosmology, standard propagation parameters, though wino annihilates well into e+ [better than bino, higgsino etc] wino w+ e+ , + , quarks ~+ w wino w – But in comprehensive theories, e.g. Planck scalestring theories that have TeV physics, stabilized moduli, consistency with other data, etc, non-thermal cosmological history is probably the default – We normalize to local relic density (use 0.3 GeV/cm3) – This is the right procedure if LSPs of non-thermal origin, e.g. moduli decay Also, for DM annihilation, energy dependent “boost factors” are better motivated than none – actually inevitable – improves the agreement with data for wino LSP Lavalle, Salati, Brun, Donato, Fornengo, Taillet et al 0809.5268 etc [“boost factor” is not good terminology here since average not increased] Antiprotons – Naively expect signal but not apparent in PAMELA data – however, • antiprotons from quark fragmentation, so soft – lose energy poorly so soft antiprotons get to detector • present in old data, so old data was background + “signal” • old data fitted as if just background, result used as background in recent analyses Gammas? Fermi/Glast data ? Synchrotron radiation – OK for some magnetic field parameters OTHER ISSUES • Profile of galaxy DM – use NFW everywhere – results a little better if profile a little softer, and that is probably preferred by astrophysics -- relevant for antiprotons and gammas, not much for positrons • Run Galprop, vary 6-7 parameters – not yet scan since computing time long – did > 200 simulations so far, will have >103 in a month • Mwino = 200 GeV so far, trying other values – only parameter of underlying physics • Region below 10 GeV poorly described – little wino DM signal there, so only relevant to be sure no systematic problem – assume solar modulation • Direct detection very small for wino, very sensitive to higgsino mix Density enhancement of positron 1 10-1 L = 4 kpc 28 K0 = 4 10 cm2 s-1 = 0.54 vconv = 5 km s-1 kpc-1 vAlfven = 32 km s-1 =1 f=1 10-2 10 102 Antiproton Flux Ratio With Density Enhancement signal(enhanced)+background signal+background background p/(p+ p) 10-4 L = 4 kpc 28 K 0 = 410 cm2 s-1 = 0.54 v conv = 5 km s-1 kpc-1 v Alfven = 32 km s-1 =1 f=1 Modulation Field = 800 MV 10-5 1 10 102 Kinetic Energy (GeV) 3 10 Note signal persists to low energies as expected for light quark fragmentation B/C Ratio GALPROP 0.35 HEAO Data 0.3 B/C Ratio 0.25 0.2 0.15 0.1 0.05 10-1 L = 4 kpc 28 K0 = 410 cm2 s-1 = 0.54 vconv = 5 km s-1 kpc-1 vAlfven = 32 km s-1 =1 f=1 1 10 Kinetic Energy (GeV/nuc) 102 3 10 0807.1508 [UB /(Urad + UB )=0.1] Synchrotron radiation constraint Allowed for masses with wino line below limit, e.g. green line FERMI DATA? Note wino LSP works well up to about 100 GeV note hint of two components THUS a wino LSP with mass 200 GeV is a promising candidate to describe PAMELA data including constraints, and Fermi data with additional component or modified Galprop primary spectra beyond 100 GeV IMPROVED EXPERIMENTAL TESTS COMING VERY SOON • PAMELA higher energy positrons 100-200 GeV – must see “turnover” (or flattening if one bin) if wino LSP (M 200 GeV) • PAMELA higher energy electrons – important to separate e+ from e− • Fermi/GLAST – gammas SOON • LHC • AMS-02 THEORY ISSUES: MUST HAVE NON-THERMAL COSMOLOGY TO GET RELIC DENSITY IF WINO LSP – PROBLEM? GOOD? [wino annihilation cross section 3x10-24 cm3 s-1 but thermal history implies cross section about 100x larger] -- Non-thermal cosmology – generic in any string theory with stabilized moduli, TeV scale, any high scale supersymmetric theory? – (Acharya, GK, Piyush Kumar, Scott Watson in preparation) -- similarly for supersymmetric flat directions, Q-balls (Fujii Hamaguchi), axinos, saxions, kination (Salati …, Chung, Everett…) -- speculative version of moduli decay Moroi-Randall ph/9906527 -- existence proof: M-theory compactified on G2 manifold wino LSP, relic density about right from first principles -- 0804.0863 (Acharya, Kane, Kumar, Shao, Watson) has detailed calculations, can calculate moduli masses and widths, thermal DM diluted by entropy from moduli decay DM from moduli decay CONSISTENT FRAMEWORK COSMOLOGICALLY o Consider theories with stabilized moduli and weak scale, wino LSP o Moduli coupling Planck suppressed long lifetime 10-3 sec o They dominate universe before decay, after freezeout, before BBN o Decays produce many LSPs, entropy, dilute thermal LSPs o LSPs annihilate “Non-thermal 2 nLSP 3HnLSP v nLSP wimp miracle” so LSPs will annihilate down to 3H / v but now at decay temperature rather than at freezeout temperature o Assuming large initial abundance and large annihilation cross section, results independent of initial abundance Relic density 1 165MeV mLSP 3x107 GeV 2 LSP h rh 10 Tmod 200GeV v 2 rh Tmod 30MeV , Trh 100MeV increases by ratio of reheat temp at decay to that at freezeout Dark matter at LHC Is what is seen at LHC same as in indirect data? Test wino hypothesis Get more info to calculate the relic density For G2-MSSM squarks heavy so predict no signal here, but see winos in gluino decay LHC phenomenology of light wino LSP well known Early 1999, 2000 • Moroi-Randall • Feng Moroi Randall Strassler • Ghergetta, Giudice, Wells Recent • Moroi, Yanagida et al ph/0610277 • Acharya et al 0801.0478 LSP nearly degenerate, so hard to see – missing energy from “escaping” chargino, LSP • Chargino, • mass difference 200 MeV, so decay has 1-2 soft pions • Find via gluino, its decays as trigger – Mgluino ranges from about 2 to about 9 Mwino in models • in our (Acharya, Bobkov, Kane, Kumar, Shao) string construction with M theory compactified on a manifold of G2 holonomy, the LSP was wino, and gluino decay was mainly to C1, N2 (not N1) so can look for tracks in vertex detector – also gluino pair gives four tops so easy to trigger on LHC implications of TeV LSPs or decaying DM? -- no LHC signal in some cases -- indirect (and very model dependent) in others, need certain superpartner spectrum or other matter to get annihilation or decay products to describe PAMELA, not contradict something -- sometimes new collider phenomenology – new light vector bosons with mass 1 GeV, decay into electrons so get “lepton jets” – model of Arkani-Hamed, Weiner et al “Lepton jets” Delayed decays possible, also missing energy along lepton jet IF THE PAMELA EXCESS IS INDEED DUE TO A LIGHT WINO LSP THE IMPLICATIONS ARE REMARKABLE • Would have learned that the dark matter, about a fifth of the universe, is (mainly) the W superpartner, and its approximate mass • Discovery of supersymmetry! -- guarantees can study superpartners at LHC • Would have learned that the universe had a non-thermal cosmological history, one we can probe • Suggests moduli dominated “UV completion” – M-Theory “G2 – MSSM” construction a concrete example Tests, information from more data soon and then from LHC, AMS02 Gamma Ray Emission With Dark Matter signal+background 10-2 background E 2 [MeV 2 cm-2 s-1 MeV -1] FERMI 10-3 10-4 L = 4 kpc 28 K0 = 410 cm2 s-1 = 0.54 v conv = 5 km s-1 kpc-1 v Alfven = 32 km s-1 =1 f=1 o o o o 0 l 360 , 10 |b| 20 10-5 1 10 Energy (GeV) 102 3 10 Positron Flux Ratio With Density Enhancement signal+background background PAMELA - e +/(e +e +) 0.1 L = 4 kpc 28 K 0 = 4 10 cm2 s-1 = 0.54 v conv = 5 km s-1 kpc-1 v Alfven = 32 km s -1 =1 f=1 Modulation Field = 800 MV -3 Ne - =3.6 10 at E = 34.5 GeV 0.01 1 10 Energy (GeV) 100 1000 Positron and Electron With Density Enhancement signal+background background E - E 3(e +e+) (GeV 2m-2s-1sr -1) e E -1.40 - 270 GeV -9 Additional flux: = 1.8 10 (( e ) e GeV 30% Positron 100 FERMI(statistical error) FERMI(systematical error) L = 4 kpc 28 K 0 = 4 10 cm2 s-1 = 0.54 v conv = 5 km s-1 kpc-1 v Alfven = 32 km s -1 =1 f=1 Modulation Field = 800 MV -3 Ne - =3.6 10 at E = 34.5 GeV 10 100 Energy (GeV) 1000 10 Be/ 9Be Ratio GALPROP 0.4 ISOMAX Data 10 9 Be/ Be Ratio 0.35 0.3 0.25 L = 4 kpc 28 K 0 = 4 10 cm2 s-1 = 0.54 v conv = 5 km s-1 kpc-1 vAlfven = 32 km s-1 =1 f=1 0.2 0.15 0.1 10-2 10-1 1 10 Kinetic Energy (GeV/nuc) 102 3 10 Energy-dependent and particle-dependent annihilation enhancements from density fluctuations • Galaxies are built from little galaxies – density fluctuations inevitable • Keep average local density, but <n2> <n>2 and annihilations n2 • Positrons lose energy rapidly so mainly come from nearer us • Antiprotons lose energy poorly, come from farther away • Different distances feel profile differently, different amount of clumps On the antimatter signatures of the cosmological dark matter subhalos. Julien Lavalle . Dec 2008. arXiv:0812.3576 [astro-ph] Galactic secondary positron flux at the Earth T. Delahaye, F. Donato, N. Fornengo, J. Lavalle, R. Lineros, P. Salati, R. Taillet arXiv:0809.5268 [astro-ph] Antimatter cosmic rays from dark matter annihilation: First results from an Nbody experiment. J. Lavalle (Turin U. & INFN, Turin) , E. Nezri, E. Athanassoula (Provence U.) , F.-S. Ling (Brussels U.) , R. Teyssier (Saclay) Phys.Rev.D78:103526,2008. arXiv:0808.0332 [astro-ph] Clumpiness of dark matter and positron annihilation signal: computing the odds of the galactic lottery. Julien Lavalle (Marseille, CPPM) , Jonathan Pochon (Annecy, LAPP) , Pierre Salati, Richard Taillet (Savoie U. & Annecy, LAPTH) . astro-ph/0603796 B/C Ratio GALPROP 0.35 HEAO Data 0.3 B/C Ratio 0.25 0.2 0.15 0.1 0.05 10-1 L = 4 kpc K0 = 4.31028 cm2 s-1 = 0.45 vconv = 5 km s-1 kpc-1 vAlfven = 31 km s-1 =1 f=1 1 10 Kinetic Energy (GeV/nuc) 102 3 10 Consistent motivated picture PAMELA positrons wino LSP Large wimp v 3x10-24 cm3sec-1 DM non-thermal in origin Moduli dominate universe after inflation, matter dominated Gravity mediated susy M3/2 10-100 TeV moduli masses Moduli decay before BBN Spectrum: -- scalars M3/2 -- higgsinos from Giudice-Masiero term M3/2 -- gauginos including LSP light [field whose F-term dominates susy is not the field whose vev gives the gauge coupling – maybe consequence of approximate R symmetry] G2 – MSSM is a concrete example of a UV completion of these generic arguments!