![Arxiv:1303.1406V1 [Astro-Ph.HE] 6 Mar 2013 the flux of Gamma-Rays from Dark Matter Annihilation Takes a Surprisingly Simple Form 2](http://data.docslib.org/img/0024f12da74efc22c151c46685f0943d-1.webp)
4th Fermi Symposium : Monterey, CA : 28 Oct-2 Nov 2012 1
The VERITAS Dark Matter Program Alex Geringer-Sameth∗ for the VERITAS Collaboration Department of Physics, Brown University, 182 Hope St., Providence, RI 02912
The VERITAS array of Cherenkov telescopes, designed for the detection of gamma-rays in the 100 GeV-10 TeV energy range, performs dark matter searches over a wide variety of targets. VERITAS continues to carry out focused observations of dwarf spheroidal galaxies in the Local Group, of the Milky Way galactic center, and of Fermi-LAT unidentified sources. This report presents our extensive observations of these targets, new statistical techniques, and current constraints on dark matter particle physics derived from these observations.
1. Introduction Earth’s atmosphere as a target for high-energy cosmic particles. An incoming gamma-ray may interact in the The characterization of dark matter beyond its Earth’s atmosphere, initiating a shower of secondary gravitational interactions is currently a central task particles that travel at speeds greater than the local of modern particle physics. A generic and well- (in air) speed of light. This entails the emission of motivated dark matter candidate is a weakly in- ultraviolet Cherenkov radiation. The four telescopes teracting massive particle (WIMP). Such particles of the VERITAS array capture images of the shower have masses in the GeV-TeV range and may inter- using this Cherenkov light. The images are analyzed act with the Standard Model through the weak force. to reconstruct the direction of the original particle as Searches for WIMPs are performed at particle accel- well as its energy. erators [Fairbairn et al. 2007], where dark matter may VERITAS is sensitive to showers initiated by be produced in high-energy collisions, and by low- gamma-rays with energies from around 100 GeV to background direct detection experiments which look more than 30 TeV. Because of the steep energy spec- for the scattering of dark matter particles off nuclei tra of typical sources, fluxes in the VERITAS en- [Gaitskell 2004]. In astrophysics, the problem is be- ergy band are much smaller than those seen with ing addressed through the field of indirect detection Fermi. However, the difficulty of measuring such small [Jungman et al. 1996, Bergstr¨om2000]. Here the goal fluxes is mitigated by the enormous effective areas of is the detection of the end-products of dark matter ACTs. VERITAS is sensitive to gamma-ray show- annihilation which can take place throughout the Uni- ers incident over an (energy-dependent) area of ap- verse. proximately 109 cm2 (compare with ∼ 104 cm2 for the Dark matter annihilation into Standard Model par- LAT). At 1 TeV the energy resolution of VERITAS is ticles generically gives rise to high-energy gamma- approximately 15%. The array has an angular resolu- rays. There are several classes of experiments cur- tion of about 0.14◦ at 200 GeV and 0.1◦ at 1 TeV. rently searching for hints of such gamma-ray emis- sion. The Large Area Telescope (LAT) onboard the Fermi Gamma-ray Space Telescope constantly surveys the entire sky at energies of about 100 MeV up to a 3. Dark matter targets few hundred GeV. At higher energies, ground-based atmospheric Cherenkov telescopes perform searches for emission from specific targets. In this contribu- The particle physics governing the annihilation of tion I review the dark matter searches currently being cold dark matter is the same everywhere. Therefore, carried out by the Very Energetic Radiation Imaging any location in the Universe that hosts a suitably high Telescope Array System (VERITAS). dark matter density is a target for an indirect search. arXiv:1303.1406v1 [astro-ph.HE] 6 Mar 2013 The flux of gamma-rays from dark matter annihilation takes a surprisingly simple form 2. VERITAS dF hσvi dNγ (E) dJ(θ) (E, θ) = 2 . (1) The VERITAS array [Weekes et al. 2002, Holder dEdΩ 8πMχ dE dΩ et al. 2006] consists of four 12-meter imaging atmo- spheric Cherenkov telescopes (ACTs) located at the Here, dF/dEdΩ is the flux of gamma-rays per energy base of Mount Hopkins at the Fred Lawrence Whip- per solid angle, hσvi is the velocity-averaged cross ple Observatory in Arizona. Such telescopes use the section for dark matter self-annihilation and Mχ is the mass of the dark matter particle. The quantity dNγ /dE is the energy spectrum of gamma-rays arising from a single annihilation. The last factor quantifies ∗Electronic address: [email protected] the distribution of dark matter along the line of sight.
eConf C121028 2 4th Fermi Symposium : Monterey, CA : 28 Oct-2 Nov 2012 1178 ACCIARI ET AL. Vol. 720
10−10The so-called J value is defined as 10 −20 Draco Willman I Z Bootes I 10 −21 dJ 2 Draco = dlρ , Willman I (2) χ −22 Ursa Minor
) dΩ 10 1 −11 Ursa Minor − 10 Bootes I −23 s 10 2 where ρ is the dark matter density and the integral 1 χ −
s −24 3
is taken along a particular line of sight. − 10
Note that the dependence on the environment is −25 10−12 10
completely captured in the J value; the other quanti- cm v> −26 σ 10 ties are universal properties of the dark matter parti- < 10 −27 dN/dE (ergs cm cle. In particular, the energy spectrum of annihilation 2 −13 E 10 −28 is governed by the annihilation channel (e.g annihila- 10
tion into quarks, leptons, photons, etc). Only at cos- −29 10 mological distances is this spectrum expected to be −30 −14 10 2 3 10 attenuated−1 by absorption or pair-production. 10 10 10 1 10 Mχ (GeV) Unlike charged particles,E (TeV) gamma-rays travel unde- flected from their point of emission. The excellent2 1 Figure 2. Exclusion regions in the (Mχ , σv )parameterspacebasedonthe Figure 1.angular95% CL upper resolution limits on theof spectral VERITAS energy densitytherefore (erg cm makes− s− ) it resultsFigure of 1: the observations. Upper limits It is on computed the dark# according$ matter to annihilation Equation (4)usinga as a function of gamma-ray energy for the four dSphs. compositecross section neutralino derived spectrum from (see VERITAS Wood et al. 2008 observations)andthevaluesof of fourJ from (A colora version powerful of this figure instrument is available because in the online it journal.) can target specific TableMilky1.BlackasterisksrepresentpointsfromMSSMmodelsthatfallwithin Way dwarf galaxies. Asterisks correspond to a locations with a high dark matter density (large J model3 standard scan deviations of MSSM of the parameter relic density space measured subject in the three-yearto a relic WMAP value). The Collaboration is currently focused on four data± set (Spergel et al. 2007). the dwarf galaxy center, ρ is the DM mass density, and λ is abundance constraint. Figure taken from Acciari et al. classes of dark matter targets: nearby dwarf galaxies, the line-of-sight distance to the differential integration volume. [2010]. the Milky Way galactic center, nearby galaxy clusters, The astrophysical contribution to the flux can be expressed by hour, significantly constraining upper limits on σ v could be E. ALIU et al. 4 # $ PHYSICAL REVIEW D 85, 062001 (2012) the dimensionlessand unidentified factor J sources in the Fermi catalog. established if J is on the order of 10 . Because the factor, J, is proportional to the DM6 density by Rolke et al. [85] is used in our analysis. As discussed 1 squared, it is subject to considerable uncertainty in its exper- J ( ) d ρ2(λ, ) dλ, (3) in the following sections, these ULs will serve for the 3.1.∆Ω Dwarf galaxies2 Ω Ω = ρ RH imental determination. For example, the mass of a DM4 halo is computation of integrated flux ULs and for constraining ! c " #∆Ω # 17 Milky Way dwarf spheroidal galaxies are dark- determined by the interaction of a galaxy with its neighbors and some dark matter models. To make the computation of which has been normalized to the product of the square of the is concentrated in the outer regions of the galaxy. Unlike the matter dominated systems30 that orbit3 inside the grav- 2 integrated flux ULs robust, we define a minimum energy, critical density, ρc 9.74 10− gcm− and the Hubble DM halo mass, the neutralino annihilation flux is determined]
itational potential= of× the Galaxy [Walker 2012]. They [deg] above which the energy reconstruction bias is less than 5%. radius, RH 4.16 Gpc following Wood et al. (2008). by the inner regions of the galaxy where the density is highest. are= excellent targets for indirect detection because The energy reconstruction bias as a function of the recon- Based on Equation (2), the upper limits on the gamma- ForJ2000 these regions, the stellar kinematic data do not fully con- they are nearby, have large dark matter densities, and 0 ray rate, Rγ (95% CL), constrain the WIMP parameter space strain16 the DM density profile due to limited statistics. Various structed -ray energy has been studied with Monte Carlo perhaps most importantly, contain no known gamma- (mχ , σ v )accordingto parameterizations of the DM mass density profile have been put simulations. The -ray selection cuts used in this analysis # ray$ sources. Therefore, any detection of gamma-rays -2 forward (Navarro et al. 1997;Kazantzidisetal.2004;deBlok[ significance set the minimum reconstructed energy to Emin
Rfromγ (95% this CL class) of objectsJ becomesσ intriguingv evidence declination ¼ > # $ et al. 2001;Burkert1995)basedonempiricalfitsandstudies 300 GeV. The ULs on the number of -rays computed forhr the1 annihilation1.09 of10 dark4 3 matter.10 26 cm3 s 1 of simulated cold dark matter (CDM) halos. We adopt the as- with the Rolke prescription are displayed in Table I, along − ! − − " -4 In the absence of× a signal, constraints× 2 can be placed sumption15 of the NFW profile (Navarro et al. 1997) given in A(E) 300 GeV c 2 with the analysis results. on the dark∞ matter mass and annihilation− cross sec- Equation (1) which describes a smooth distribution of DM with 8 2 tion× so long0 5 as10 onecm has an estimatemχ of the J value. asinglespatialscalefactorrs. The astrophysical factor,-6 J, is # × ! " h m h m III. FLUX UPPER LIMITS Currently,EdN the/dE dark(E,m matter) dE distribution in the dwarfs then given by 10 10 10 05 χ , (4) is modeled using2 the motions of their member stars right ascension [hours] The analysis of the data did not show any significant × 10 E J2000 − 2 1 λmax 2 (e.g. Strigari et al. [2007, 2008]). 2πρs r(λ) − excess over the background at the nominal position of where A(E)istheenergy-dependentgamma-raycollectingarea. J (∆Ω) VERITAS has published the results of observations FIG.Figure 1 2:(color Significance= online).ρ2R map Significance of the region map surrounding obtainedr from the Segue 1. The ULs on the number of -rays in the signal The expression has been cast as a product of dimensionless ! c H " #cos(0.115◦) #λmin ! s " of Bo¨otesI (14 hr), Draco (18 hr), Ursa Minor (19 VERITASdwarf galaxy observations Segue 1. The of Segue cross 1is4 atafter the location-ray selection of the and search region can then be converted to ULs on the integral factors with the variables normalized to representative quan- r(λ) − hr), and Wilman 1 (14 hr) [Acciari et al. 2010] and backgrounddwarf galaxy subtraction. and the1+ circles The black denote cross regionsd indicatesλ d(cos excludedθ the), position of(5) -ray flux. The number of -rays detected by an array of tities, e.g., the cross section times velocity is normalized to when determining× the backgroundr (one centered on deep26 observations3 1 of Segue 1 (48 hr) [Aliu et al. 2012]. Segue 1. The black circles corresponds to the two exclusion regions IACTs above a minimum energy Emin is related to the 3 10− cm s− which is a rough generic prediction for σ v Segue 1, the other on$ a bright! "% star). Figure taken from × Figure 1 shows the upper limits on the dark matter# $ used for the background determination. See text for further details. source integral flux È E Emin by: for a WIMP thermal relic in the absence of coannihilations for whereAliu et the al. lower [2012]. integration bound of 0◦.115 corresponds to the ð Þ annihilation2 cross section as a function of dark mat- mχ > 100 GeV c− (cf. Figure 2). The main contribution to size of the signal integration region. The galactocentric distance, E dN dE ter mass derived from observations of the first four of E1min Aeff dE the integral comes from the energy range in the vicinity of the rTwo(λ), iscircular determined regions, by of radius 0.2 centered on the target N E Emin Tobs ð Þ È E Emin ; these dwarf galaxies. The energy spectrum of gamma- dN energy threshold (E 300 GeV for observations in this paper) ð Þ¼ ÂR 1 dE  ð Þ rays from an annihilation, dN /dE, was derived from positionmap of theand region of radius surrounding 0.3 centered the dwarf on the galaxy bright is star Emin dE where A(E)changesrapidly.ForVERITAStheeffectivearea& γ 2 2 -Leonis (withr(λ) apparentλ + R magnitude2λRdSph in thecos θ visible, band(6) at 300 GeVa representative is 6 108cm set2.ForarepresentativeMSSMmodel, of MSSM parameters. The aster- shown in Fig. 2.= Here, thedSph cross− marks the location of R (1) isks in∼ the× figure represent a scan over MSSM parame- MSegueV 3 1:5 and, and the located circles& 0.68 denote from regions the positionexcluded of when Segue EdN/dE at 300 GeV is a function of neutralino mass, mχ ,and where¼λ is the line-of-sight distance and RdSph is the distance of where T is the observation time, dN =dE the assumed ter space with the2 constraint1 that each model predicts2 1),determining were excluded the background. for the background One is determination. centered on the obs it changes in the range 10− –10− for mχ from 300 GeV c− to a the dwarf galaxy from the Earth. a dark2 matter candidate with a relic density within 3 dwarfThe galaxy, analysis the of other the data on anresulted unrelated in the bright selection star. of source differential energy spectrum, and Aeff E is the few TeV c− .AlthoughEdN/dE is a rapid function of mχ ,this Although the integration limits, λmin and λmax,aredetermined ð Þ standard deviations of that determined by three-year2 2 NAsON with1082 the other -ray dwarf candidates galaxies, in the upper signal limits search can region be instrument effective area. The effective area Aeff E is the dependence is nearly compensated by the (300 GeV c− /mχ ) by the¼ tidal radius of the dSph (rt 7kpcwasusedforthese instrument response function to the collection of ð -raysÞ of prefactor.WMAP The product observations of these [Spergel two contributions et al. 2007]. and, conse- calculations;andplacedNOFF on the S12479anchez-Conde´ darkbackground matter et annihilation al. events2007= ), the in cross themain background section.contribution No significant gamma-ray excess was detected from ringThese region, limits¼ with are showna normalization in Fig. 3, factor where each0 curve:084, rep- result- energy E, and it depends on the zenith angle of the obser- quently, the overall integral value is weakly dependent on the to J (∆Ω)comesfromtheregionsr TABLE I. Analysis results of the VERITAS observations of Segue 1. Nexc E E is the ð minÞ number of excess events in the signal search region with energies E Emin, after background subtraction. N95% CL E E is the 95% confidence level (CL) upper limit on the number of ð minÞ -rays with energies E Emin in the signal search region, computed according to the Rolke [85] prescription. Live time ( min) E (GeV) Nexc E E Significance N95% CL E E min ð minÞ ð minÞ 2866 - 30.4 0.9 135.9 2866 300 31.2 1.4 102.5 062001-4 4th Fermi Symposium : Monterey, CA : 28 Oct-2 Nov 2012 3 VERITAS DEEP OBSERVATIONS OF THE DWARF ... PHYSICAL REVIEW D 85, 062001 (2012) FIG.Figure 3. 95% 3: CL Annihilation ULs from the cross VERITAS section observationsupper limits of derived Segue 1from on the observations WIMP velocity-weighted of the dwarf galaxy annihilation Segue cross-section 1. Each curvev as a functionrepresents of the a WIMPdifferent mass, annihilation considering channel. different The final cross state section particles. required The grey to band reproduce area represents the observed a range relic of abundance generic values is h fori the shown by the gray bands. Figures taken from Aliu et al. [2012]. annihilation cross-section in the case of thermally produced dark matter. Left: hadronic channels WþWÀ, bb and þ À. Right: leptonic channels eþeÀ and þÀ. The Astrophysical Journal,757:123(14pp),2012October1 Arlen et al. ] The gray region indicates the cross section required to -1 Table 8 s 3 23 3 1 χχ → bb Upper Limits on the DM Annihilation Cross Section Times Velocity σv muonreproduce decay the darke matter abundanceand observede today. has of 10 cm s at 1 TeV. The current ULs on +v- are ! " À À e þ þ e À À χχ → W W from VERITAS Observations of the Coma Cluster This represents! a lower limit to the allowed! cross sec- +h - i been included in the þÀ -ray spectra [91,92]. 2 orders of magnitude above the predictionsχχ for→ τ thermallyτ 3 1 tion in generic WIMP models: WIMPs with a smaller v> [cm -19 Channel R (deg) mχ (GeV) σv (cm s− ) Figure 3 shows the 95% CL exclusion curves on producedσ 10 WIMP dark matter. ! " < + 20 cross section would be overabundant today. There- W W− 0 2000 1.1 10− v as a function of the dark matter particle mass for × 21 h fore,i the parameter space for general WIMP models is 0.2 1900 4.3 10− the five channels considered above, using Eq. (3) and (7). C. Lower limits on the decay lifetime × 21 bounded at both ends. Further observations by VER- 0.4 1900 8.4 10− -20 × 20 For the WþWÀ channel, the 95% CL UL on the velocity- 10If we assume that dark matter is a decaying particle, bb 0 3500 1.2 10− ITAS and many other experiments will95% continue CL to ¯ × 21 weighted annihilation cross-section is v 8 LLs on the lifetime of dark matter can be derived. In 0.2 3400 4.4 10− tighten the upper limits on the cross section. × 21 24 3 1 h i  0.4 3500 8.7 10− 10À cm sÀ at 1 TeV. This limit is the most constraining decaying dark matter scenarios, the dark matter particle + × 21 τ τ − 06702.4 10− reported so far for any dSph observations in the VHE -ray can either-21 be bosonic or fermionic. The LLs are computed × 22 10 0.2 650 9.1 10− × 21 band. The bb and þ À exclusion curves illustrate the using Eq. (7) and making the appropriate substitutions to 0.4 660 1.8 10− range3.2. of Clusters uncertainties on the v ULs from the dark Eq. (3), as explained in Sec. IVA. For bosonic dark matter × matter particle physics model.h Concerningi the lepton particles, the same channels as in the annihilating dark Notes. Upper limits are shown for different integration regions and DM particle While further away, galaxy clusters are much larger mass m and are derived from the VERITAS gamma-ray flux upper limits channels eþeÀ and þÀ, the limits are at the level matter case are considered:3 WþWÀ, bb, þ À, 4 eþeÀ, χ than dwarf galaxies. In fact, these systems are the 10 10 presented in this work. mχ [GeV] largest dark matter structures in the Universe. Be- Figure 6. Limits on the DM annihilation cross section σv from VERITAS cause of their high dark matter densities and dense observationsFigure 4: Upperof the Coma limits Cluster on the as dark a function matter of theannihilation! DM" particle mass 6. DISCUSSION AND CONCLUSIONS substructures they represent an attractive class of in- mcrossχ derived section from derived the VERITAS from gamma-ray observations flux ofupper the limits Coma (0◦.2 aperture) direct detection targets. However, clusters contain presentedgalaxy cluster.in this work. Different lines are upper limits assuming We have reported on the observations of the Coma Cluster of large reservoirs of hot gas which will interact with different annihilation channels. Figure taken from Arlen galaxies in VHE gamma rays with VERITAS and complemen- et al. [2012]. Table 7 cosmic rays. This results in the production of pions Astrophysical Factors tary observations with the Fermi-LAT. VERITAS observed the which then decay into gamma-rays. This gamma-ray Coma Cluster of galaxies for a total of 18.6 hr of high-quality 2 5 2 5 R (deg) J signal (GeV cm− sr) α J bkg (GeV cm− sr) live time between March and May in 2008. No significant ex- emission, highly interesting in its own right, consti- ! " ! " tutes a background for dark matter searches. In such 053.3. Galactic.7 center1016 1.3 1014 (negligible) cess of gamma rays was detected above an energy threshold × 16 × 14 0.2 8.1 10 4.4 10 (<0.01 J signal,negligible) of 220 GeV. The Fermi-LAT has observed the Coma Cluster searches there is often a tradeoff between high dark × × ! " 0.4 9.4 1016 1.3 1015 ( 0.01 J ,negligible) in all-sky∼ survey mode since its launch in 2008 June. We have matter densities and astrophysical backgrounds. The Milky Way galactic center is expectedsignal to be, × × & ! " used all data available from launch to 2012 April for an updated VERITAS has published comprehensive results by far, the brightest source of dark matter annihila- Notes. J bkg is the astrophysical factor calculated for the background region analysis compared to published results (Ackermann et al. 2010). ! " from a 19 hour observation of the Coma cluster [Arlen (ringtion. method) This and is due is used to to its estimate proximity the level (8 of kpc) gamma-ray and its contamination very Again, no significant excess of gamma rays was detected. We et al. 2012]. No significant gamma-ray excess was fromlarge DM expected annihilation. darkα is the matter size ratio density. of the ON- Because and OFF-source baryons regions. have used the VERITAS and Fermi-LAT data to calculate flux seen. This allows interesting constraints to be placed dominate the mass distribution in this region there is upper limits at the 99% confidence level for the cluster core on models of Coma’s cosmic ray population and mag- contaminationa great deal from of uncertainty DM annihilation on the in the dark background matter dis- region. (considered as both a point-like source and a spatially extended netic fields as well as on dark matter annihilation. Fig- Astribution long as near the DM the contributionGalactic center. to the Nonetheless, event number the in the emission region) and for three member galaxies. The flux upper ure 4 shows the upper limits on the dark matter anni- J value for the galactic center is likely several orders FIG. 4. 95% CL LLs from the VERITAS observations of Segue 1 onbackground the decay lifetime region as is a negligible, function of the the upperdark matter limits particle derived mass. here limits obtained were then used to constrain properties of the hilation cross section derived from the null-detection of magnitude larger than those of the nearby dwarfs. 1 cluster. Left: Bosonic dark matter decaying to two identical particles: WþWÀdirectly, bb, þ scaleÀ, eþ witheÀ and the astrophysicalþÀ. The black factor, star and UL( theσ v black) crossJ − . of Coma. In this analysis no gamma-ray background The search at the Galactic center is complicated! " ∝ by We have employed various approaches to constrain the CR denote the best fits to the Fermi and PAMELA data considering the TheþÀ analysisand þ usesÀ channels, a ring respectively, region to estimate and are taken the background from [58]. from cosmic rays was assumed, making the limits con- a very large astrophysical0 background of gamma-rays population and magnetic field distribution that are complemen- Right: Fermionic dark matter decaying to two different particles: WÆin‘Ç an(where ON‘ region.e, We, ) have and Z to. compute The black the triangle expected indicates level the of servative. arising from¼ known and unknown point sources and tary in their assumptions and hence well suited to assessment best fit to the Fermi and PAMELA data considering the channel WÆgamma-rayÇ, taken fromemission [58]. from DM annihilation in the ring region in order to check that it is negligible with respect to the level of of the underlying Bayesian priors in the models. (1) We used gamma-ray emission from DM annihilation in the ON region. asimplified“isobaricCRmodel”thatischaracterizedbya eConf C121028 This is equivalent to compute the astrophysical factor of the constant CR-to-thermal pressure fraction and has a power-law 062001-7ON- and OFF-source regions since this quantity is related to the momentum spectrum. While this model is not physically justi- rate of DM annihilations. fied a priori, it is simple and widely used in the literature and The resulting exclusion curves in the ( σ v ,mχ )parameter captures some aspects of more elaborate models such as (2) the space are shown in Figure 6 for three! different" DM self- simulation-based analytical approach of Pinzke & Pfrommer + + annihilation channels, W W−,bb, and τ τ −.Dependingonthe (2010). The latter is a “first-principle approach” that predicts ¯ 20 DM annihilation channel, the limits are on the order of 10− the CR distribution spectrally and spatially for a given set of as- 21 3 1 to 10− cm s− .Theminimumforeachexclusioncurveand sumptions. It is powerful since it only requires the density profile corresponding DM particle mass is listed in Table 8.Westress as input due to the approximate universality of the CR distri- that these limits are derived with conservative estimates of the bution (when neglecting CR diffusion and streaming). Note, astrophysical factor J.Theydonotincludeanyboosttothe however, that inclusion of these CR transport processes may annihilation rate possibly due to DM substructures populating be necessary to explain the radio-halo bimodality. (3) Finally, the Coma halo, which could scale down the present limits by a we used a pragmatic approach that models the CR and mag- factor O(1000) in the most optimistic cases (Pinzke et al. 2011; netic distributions in order to reproduce the observed emission Gao et al. 2012). profile of the Coma radio halo. While this approach is also not We also note that when the size of the integration region is physically justified, it is powerful because it shows what the increased, the limits on σ v result from a competition between “correct” model has to achieve and can point in the direction of the gain in the astrophysical! " factor J and the integrated the relevant physics. ! " background. For integration regions larger than 0◦.2inradius, Within this pragmatic approach, we employ two different the astrophysical factors no longer compensate for the increased methods. First, adopting a high magnetic field everywhere in number of background events, and the signal-to-noise ratio the cluster (B BCMB)yieldstheminimumgamma-rayflux deteriorates. in the hadronic) model of radio halos, which we find to be 12 4 4th Fermi Symposium : Monterey, CA : 28 Oct-2 Nov 2012 VERITAS. Variable sources as well as those close to the Galactic plane are ruled out as they are not likely to be dark matter halos. Sources with hard spectra and those detected at the highest energies by the LAT are preferred. Figure 6 shows significance maps for the two sources which have been observed. They are 2FGL J0312.8+2013 (10 hr) and 2FGL J0746.0-0222 (9 hr). No significant detection can be reported from either at this time. 4. New analysis techniques Along with deeper observations of multiple dark matter targets, the VERITAS collaboration is explor- ing new analysis techniques designed to optimally ex- tract the particle physics from the observations. So far, all dark matter limits have been constructed sepa- Figure 5: Preliminary significance map of the Galactic rately for each individual target. However, the physics center based on 46 hours of observation with VERITAS. of dark matter annihilation is the same across all tar- Sources from the Fermi catalog are shown as light blue gets. Therefore, it makes sense to combine the data circles. Sources and diffuse emission detected by H.E.S.S. taken on multiple observations into a single dark mat- are shown with solid and dotted black contours. ter result. This is being performed using the frame- work developed in Geringer-Sameth and Koushiappas [2011, 2012]. In this analysis, detected events from from cosmic ray interactions with gas. different fields are weighted according to how likely Figure 5 shows a preliminary significance map de- they are to be due to dark matter as opposed to back- rived from 46 hours of VERITAS observations of the ground processes (i.e. a Segue 1 event is given more Galactic center. The locations of sources in the Fermi weight than a Bo¨otesI event since Segue 1 has a larger catalog Nolan et al. [2012] are shown with light blue J value). This will allow the complete collection of circles. Sources detected by H.E.S.S. as well as dif- observations to be reduced into a single dark matter fuse emission detected by H.E.S.S. [Aharonian et al. search or cross section upper limit. 2006] are shown with black solid and dotted con- tours. Emission from the galactic center is detected at 19σ. However, it cannot be unambiguously claimed 5. Conclusions that this signal to due to dark matter annihilation. The H.E.S.S. atmospheric Cherenkov telescope has VERITAS continues to perform dark matter reported the detection of diffuse emission along the searches across multiple classes of targets. Signifi- Galactic ridge. Dark matter is not expected to clus- cant additional exposure has already been devoted to ter in the Galactic plane and so this diffuse emission is Draco (35 hr), Ursa Minor (36 hours), and Segue 1 almost certainly the result of other astrophysical pro- (48 hours) with more observations planned. This will cesses. The expected spherical morphology of the dark result in a data set of nearly 200 hours of observation matter emission, however, can be used in an indirect across several dwarf galaxies. The full data set will be search. The idea is to look toward the Galactic center analyzed simultaneously, resulting in the most sen- but slightly away from the Galactic plane, avoiding sitive possible search given the available data. The much of the astrophysical background. This analysis VERITAS collaboration is proceeding with observa- is ongoing within the VERITAS collaboration. tions and analysis of the Galactic center and of Fermi unidentified sources and will publish its results in the 3.4. Fermi unidentified objects near future. The Fermi LAT has detected many gamma-ray sources that have no known counterpart at other wavelengths. It has been suggested that such uniden- Acknowledgments tified objects may be nearby dark matter halos. To explore this possibility two of these sources were se- This research is supported by grants from the U.S. lected for observation by VERITAS. The selection cri- Department of Energy Office of Science, the U.S. Na- teria are based on how likely the source is to be as- tional Science Foundation and the Smithsonian Insti- trophysical as well as on the detection prospects for tution, by NSERC in Canada, by Science Foundation eConf C121028 4th Fermi Symposium : Monterey, CA : 28 Oct-2 Nov 2012 5 Figure 6: Preliminary significance map of two Fermi unidentified sources observed with VERITAS. Left: 2FGL J0312.8+2013. Right: 2FGL J0746.0-0222. Ireland (SFI 10/RFP/AST2748) and by STFC in the Astroparticle Physics 25, 391 (2006), arXiv:astro- U.K. We acknowledge the excellent work of the tech- ph/0604119. nical support staff at the Fred Lawrence Whipple Ob- M. G. Walker, ArXiv e-prints (2012), 1205.0311. servatory and at the collaborating institutions in the L. E. Strigari, S. M. Koushiappas, J. S. Bullock, and construction and operation of the instrument. The M. Kaplinghat, Phys. Rev. D 75, 083526 (2007), National Radio Astronomy Observatory is a facility of arXiv:astro-ph/0611925. the National Science Foundation operated under co- L. E. Strigari, S. M. Koushiappas, J. S. Bullock, operative agreement by Associated Universities, Inc. M. Kaplinghat, J. D. Simon, M. Geha, and B. Will- I would like to thank Savvas Koushiappas, Ben man, ApJ 678, 614 (2008), 0709.1510. Zitzer, Karen Byrum, Bob Wagner, Andy Smith, Jim V. A. Acciari, T. Arlen, T. Aune, M. Beilicke, W. Ben- Buckley, Jamie Holder, Martin Pohl, and Ken Ragan bow, D. Boltuch, S. M. Bradbury, J. H. Buckley, for their support in preparing this presentation. AGS V. Bugaev, K. Byrum, et al., ApJ 720, 1174 (2010), is supported by a Galkin Fellowship at Brown Univer- 1006.5955. sity. E. Aliu, S. Archambault, T. Arlen, T. Aune, M. Beil- icke, W. Benbow, A. Bouvier, S. M. Bradbury, J. H. Buckley, V. Bugaev, et al., Phys. Rev. D 85, 062001 (2012), 1202.2144. References D. N. Spergel, R. Bean, O. Dor´e, M. R. Nolta, C. L. Bennett, J. Dunkley, G. Hinshaw, N. Jarosik, E. Komatsu, L. Page, et al., ApJS 170, 377 (2007), M. Fairbairn, A. C. Kraan, D. A. Milstead, arXiv:astro-ph/0603449. T. Sj¨ostrand,P. Skands, and T. Sloan, Phys. Rep. T. Arlen, T. Aune, M. Beilicke, W. Benbow, A. Bou- 438, 1 (2007), arXiv:hep-ph/0611040. vier, J. H. Buckley, V. Bugaev, K. Byrum, A. Can- R. J. Gaitskell, Annual Review of Nuclear and Particle non, A. Cesarini, et al., ApJ 757, 123 (2012), Science 54, 315 (2004). 1208.0676. G. Jungman, M. Kamionkowski, and K. Griest, P. L. Nolan, A. A. Abdo, M. Ackermann, M. Ajello, Phys. Rep. 267, 195 (1996), arXiv:hep-ph/9506380. A. Allafort, E. Antolini, W. B. Atwood, M. Ax- L. Bergstr¨om,Reports on Progress in Physics 63, 793 elsson, L. Baldini, J. Ballet, et al., ApJS 199, 31 (2000), arXiv:hep-ph/0002126. (2012), 1108.1435. T. C. Weekes, H. Badran, S. D. Biller, I. Bond, F. Aharonian, A. G. Akhperjanian, A. R. Bazer- S. Bradbury, J. Buckley, D. Carter-Lewis, Bachi, M. Beilicke, W. Benbow, D. Berge, M. Catanese, S. Criswell, W. Cui, et al., As- K. Bernl¨ohr,C. Boisson, O. Bolz, V. Borrel, et al., troparticle Physics 17, 221 (2002), arXiv:astro- Nature 439, 695 (2006), arXiv:astro-ph/0603021. ph/0108478. A. Geringer-Sameth and S. M. Koushiappas, Physical J. Holder, R. W. Atkins, H. M. Badran, G. Blay- Review Letters 107, 241303 (2011), 1108.2914. lock, S. M. Bradbury, J. H. Buckley, K. L. Byrum, A. Geringer-Sameth and S. M. Koushiappas, D. A. Carter-Lewis, O. Celik, Y. C. K. Chow, et al., eConf C121028 6 4th Fermi Symposium : Monterey, CA : 28 Oct-2 Nov 2012 Phys. Rev. D 86, 021302 (2012), 1206.0796. eConf C121028