Indirect Dark Matter Searches with VERITAS
aa r X i v : . [ a s t r o - ph . H E ] O c t PROCEEDINGS OF THE 31 st ICRC, Ł ´OD ´Z 2009 1
Indirect Dark Matter Searches with VERITAS
R.G. Wagner ∗ for the VERITAS Collaboration †∗ High Energy Physics Division, Argonne National Laboratory, Argonne, IL 60439-4815, USA † see R.A. Ong et al. (these proceedings) or http://veritas.sao.arizona.edu/conferences/authors?icrc2009 Abstract . A leading candidate for astrophysicaldark matter (DM) is a massive particle with a mass inthe range from 50 GeV to greater than 10 TeV and aninteraction cross section on the weak scale. The self-annihilation of such particles in astrophysical regionsof high DM density can generate stable secondaryparticles including very high energy gamma rayswith energies up to the DM particle mass. Dwarfspheroidal galaxies of the Local Group are attractivetargets to search for the annihilation signature of DMdue to their proximity and large DM content. Wereport on gamma-ray observations taken with theVery Energetic Radiation Imaging Telescope ArraySystem (VERITAS) of several dwarf galaxy targetsas well as the globular cluster M5 and the local groupgalaxies M32 and M33. We discuss the implicationsof these measurements for the parameter space ofDM particle models
Keywords : VERITAS Gamma-ray Dark Matter
I. I
NTRODUCTION
The existence of astrophysical non-baryonic dark mat-ter (DM) has been established by its gravitational effecton galaxy rotation [1] and the velocity dispersions ofobjects from dwarf galaxies [2] through large galaxyclusters [3]. Additional evidence of DM existence comesfrom cosmic microwave background measurements [4]and gravitational lensing of galaxies by foregroundgalaxy clusters, for example, the evident separation ofdark and baryonic matter in 1E0657-558, the “bulletcluster” [5].At present though, the existence of dark matter issolely inferred from its gravitational influence. The par-ticle nature of dark matter is yet to be revealed throughdirect detection in terrestrial dark matter searches, itsproduction in particle accelerators, or indirect dark mat-ter searches looking for evidence of a flux of knownparticles produced by annihilation of dark matter particlepairs. Here we report on an indirect dark matter searchcarried out using VERITAS located at the Fred LawrenceWhipple Observatory in southern Arizona, USA, at anelevation of 1268m [6]. Gamma rays can be producedeither directly from the pair annhilation of weakly inter-acting massive particles (WIMPs) or as secondary decayproducts from the primary particles produced in WIMPannihilation. The former would produce a monoenergeticsource of gamma rays equal to the WIMP mass andwould constitute definitive evidence for particle dark
TABLE IS
UMMARY OF O BSERVATION P ERIOD AND O BSERVED T IME FOR I NDIRECT
DM S
EARCH
Source Period Hours ObservedDraco 2007 Apr-May 22.3Ursa Minor 2007 Feb-May 26.0Willman 1 2007 Dec-2008 Feb 13.7M5 2009 Feb-Mar 15.6M33 2007 Nov-2008 Feb 15.8M32 2008 Oct-2009 Jan 13.2 matter. The latter mechanism would produce a spectrumof gamma-ray energies with a cutoff at the WIMP mass.For astrophysical targets, the WIMP annihilation rateand associated gamma-ray flux are highly uncertaindue to theoretical uncertainties and limited observationalconstraints on the DM halo profile. Therefore, VERITAShas surveyed a variety of possible sources in its darkmatter search. We present here results from observationsof three dwarf spheroidal galaxies (dSph): Draco, UrsaMinor, and Willman 1; the globular cluster M5; and thelocal group galaxies M32 and M33. The emphasis of theprogram has been to target dSphs due to their relativeproximity and low expected background from knownastrophysical gamma-ray sources.VERITAS can detect and measure gamma rays inthe ∼
100 GeV - 30 TeV energy range with an energyresolution of 15-20%, and angular resolution of ∼ . ◦ per event. Further technical description of VERITAS canbe found in [7].II. D WARF G ALAXY R ESULTS
Dwarf spheroidal galaxies typically have stellar ve-locity dispersions implying a high mass-to-light (
M/L )ratio indicative of their dynamics being dominated bydark matter. Observations of the Draco, Ursa Minor, andWillman 1 dSph galaxies were performed in 2007 and2008 in three and four telescope array configurations.Details of the observations are summarized in Table I.Gamma-ray candidates are selected from stereo recon-structed events on the basis of the summed digital pulseheight from the camera phototubes, the image distancefrom the center of field of view, and mean scaled widthand length cuts optimized for a signal that is 3% that ofthe Crab Nebula. As the data were acquired in “wobble”mode in which the telescope array was pointed at anoffset of . ◦ from the targeted source, the net signalis calculated using the “reflected region” backgroundmodel [8]. Using the Li and Ma eqn. 17 method [9] tocalculate the significance, no excess above background WAGNER et al.
VERITAS DARK MATTER SEARCH
Fig. 1. Upper limits on h σv i as a function of neutralino mass, m χ using a composite neutralino spectrum (see [11] and the values of J from Table III. Black triangles represent points from MSSM modelsthat fall within ± standard deviations of the relic density measuredin the 3 year WMAP data set [4]. was detected from any of the three dSphs. Gamma-rayflux upper limits were calculated at the 95% confidencelevel using the bounded profile likelihood ratio statisticdeveloped by Rolke [10]. Table II summarizes the resultsfor each of the three dwarf galaxies. Following theformalism of [11] we set limits in the WIMP parameterspace ( m χ , h σv i ) : h σv i × − cm s − 95% C.L. ) (cid:16) m χ 100 GeV (cid:17) × (cid:18) . × GeV J (cid:19) × (cid:26) φ Z ∞ 200 GeV A ( E ) (cid:20) dN ( E, m χ ) /dE − GeV − (cid:21) dE (cid:27) − , where φ = 6 . × − cm − s − is 1% of theintegral Crab Nebula flux above 100 GeV [12], A ( E ) is the energy-dependent effective collecting area, and J is a dimensionless astrophysical factor normalizedto the product of the square of the critical density, ρ c = 9 . × − g cm − and the Hubble radius, R H = 4 . Gpc. We provide limits on the WIMPparameter space based on the assumption of a smoothNFW profile [13] with J values given in Table III.This provides conservative estimates of the expectedflux. Significant “boosts” of the flux with respect tothis smooth halo assumption are possible from halosubstructure that can produce enhancements as large asa factor of 100 [14].Figure 1 shows h σv i limits as a function of neutralinomass using the expression given above. Also plotted arevalues from Minimal Supersymmetric Standard Mod-els (MSSM) generated with DarkSUSY [15] that areconsistent with the WMAP bounds on the relic DMdensity [4]. The limits indicate that a boost factor of ∼ would be necessary to produce a signal withinour present sensitivity. III. G LOBULAR C LUSTER AND L OCAL G ROUP G ALAXY R ESULTS As part of the VERITAS indirect dark matter searchprogram, we also targeted the globular cluster, M5 andthe Local Group large galaxies, M32 and M33. Details ofthe exposure are given in table I. Analysis of these datawere carried out in a similar manner to that describedin section II. The mean scaled length and width, andthe summed pulse height selections applied to M5 arestricter than those for the dSphs, M32, and M33. Thedetails of the M5 analysis are given in [16]. Globularclusters and large galaxies may have significant sourcesof background from standard very high energy gamma-ray producing phenomena such as supernova remnantsand compact binary objects. Thus, while they are worth-while targets for an indirect DM search, interpretationof any signal would be more complicated. However,we found no gamma-ray excess above background ofany significance for these three targets. The results aresummarized in table IVIV. C ONCLUSIONS We have carried out a search for very high en-ergy gamma rays from three dwarf spheroidal galaxies:Draco, Ursa Minor, and Willman I; the globular cluster,M5; and the local galaxies M32 and M33 as part of anindirect dark matter search program on the VERITASIACT array. No significant excess above backgroundwas observed from any target. We set upper limits onthe flux and, for the dwarf spheroidal galaxies limitson the cross section times velocity, h σv i , for neutralinopair annihilation as a function of neutralino mass. The h σv i limits indicate that a substantial boost factor abovesmooth dark halo expectations would be required ofMSSM-type models.We will continue our program in the future with anemphasis on further dSph observations. Next generationIACT arrays now being planned such as the AdvancedGamma-ray Imaging System (AGIS) and the CherenkovTelescope Array (CTA) will provide an order of magni-tude increase in sensitivity over current arrays such asVERITAS, MAGIC II, and HESS and can be expectedto constrain some models of both supersymmetric darkmatter or Kaluza-Klein dark matter associated with mod-els of universal extra dimensions even in the conservativesmooth DM halo paradigm. Along with observations bythe Fermi Gamma-ray Space Telescope and new particlesearches at the Large Hadron Collider, prospects forunderstanding the nature of dark matter over the nextdecade look to be promising.This research is supported by grants from the USDepartment of Energy, the US National Science Foun-dation, and the Smithsonian Institution; by NSERC inCanada; by Science Foundation Ireland; and by STFCin the UK. We acknowledge the excellent work of thetechnical support staff at the FLWO and the collaborat-ing institutions in the construction and operation of theinstrument. ROCEEDINGS OF THE 31 st ICRC, Ł ´OD ´Z 2009 3 TABLE IIS UMMARY R ESULTS OF D WARF G ALAXY O BSERVATIONS Quantity Draco Ursa Minor Willman IExposure (hr) 18.38 18.91 13.68Signal Region (events) 305 250 326Total Background (events) 3667 3084 3602Number Backgrd. Regions 11 11 11Significance a -1.51 -1.77 -0.0895% c.l. (counts) b ) 12518 16917 33413Energy Threshold (GeV) 200 200 200Flux Limit 95% c.l. (cm − s − ) . × − . × − . × − a Li and Ma eqn. 17 method [9] b Rolke method [10] TABLE IIIP ARAMETERS USED FOR THE ASTROPHYSICAL FACTOR , J CALCULATION .Quantity Draco Ursa Minor Willman I R dSph (kpc) a 80 66 38 r t (kpc) b ρ s ( M ⊙ / kpc ) c . × . × × r s (kpc) d J e a Earth-dwarf galaxy distance b The value of J is negligibly changed for tidal radius, r t ,as low as 0.9 kpc. c scale density d scale radius e J is expressed as a dimensionless value normalized to thecritical density squared times the Hubble radius, . × GeV cm − . TABLE IVS UMMARY R ESULTS OF G LOBULAR C LUSTER AND L ARGE G ALAXY O BSERVATIONS Quantity M5 M32 M33Exposure (hr) 15.0 11.29 11.83Signal Region (events) 25 262 147Total Background (events) 251 2156 992Number Backgrd. Regions 11 7 7Significance a -0.3 0.59 0.4195% c.l. (counts) b − ) . × − . × − . × − a Li and Ma eqn. 17 method [9] b Rolke method [10] R EFERENCES[1] Rubin, V.C., Ford, W.K., & Thonnard, N., 1980, AJ, , 471[2] Walker, M.G. et al. , 2007, ApJ, , L53[3] Zwicky, F., 1937, ApJ , 217[4] Spergel, D.N. et al. , 2007, ApJS, , 377[5] Clowe, D. et al. , 2006, ApJ, , L109[6] Weekes, T.C. et al. (VERITAS collaboration). 2002, Astropart.Phys., , 221[7] Acciari, V.A., et al. (VERITAS collaboration) 2008, ApJ, ,1427[8] Berge, D., Funk, S., & Hinton, J. A&A, , 1219[9] Li, T-P. and Ma, Y-Q, 1983, ApJ, , 317[10] Rolke, W.A., Lopez, A.M. & Conrad, J. 2005, arXiv:0403059v4[11] Wood, M. et al. , 2008, ApJ , 594[12] Hillas, A. M., et al. , 744[13] Navarro, J.F., Frenk, C.S. & White, S.D.M. 1997, ApJ, , 493[14] Strigari, L.E. et al. , 2007, Phys. Rev., D75 , 083526[15] Gondolo, P. et al. , 2004, JCAP, , 008[16] see M. McCutcheon et al.et al.