Searches for Dark Matter signatures in the Segue 1 dwarf spheroidal galaxy with the MAGIC-I telescope

2011 Fermi Symposium, Roma., May. 9-12 1 Searches for Dark Matter signatures in the Segue 1 dwarf spheroidal galaxy with the MAGIC-I telescope arXiv:1110.6775v1 [astro-ph.HE] 31 Oct 2011 S. Paiano Università di Padova and INFN, I-35131 Padova, Italy, and Astronomy Department, Padova, Italy S. Lombardi Università di Padova and INFN, I-35131 Padova, Italy M. Doro Universitat Autònoma de Barcelona, E-08193 Bellaterra, Spain D. Nieto Universidad Complutense de Madrid, E-28040 Madrid, Spain on behalf of the MAGIC Collaboration http://magic.mppmu.mpg.de/ M. Fornasa Instituto de Astrofı́sica de Andalucı́a (IAA-CSIC), E-18008 Granada, Spain Despite the interest in Dark Matter (DM) searches is currently more focused on underground experiments, a signature of DM annihilation/decay in gamma-rays from the space would constitute a smoking gun for its identification. Here, we present the results of the survey of Segue 1 by the MAGIC-I telescope performed in 2008 and 2009. This source is considered by many as the most DM dominated Milky Way satellite galaxy known so far. The nearly 43 hours of data taken constitute the deepest observation ever made on a single dwarf galaxy by Cherenkov telescopes. No significant gamma-ray emission was found above an energy threshold of 100 GeV. Integral upper limits on the gamma-ray flux were calculated assuming various power-law spectra for the possible emission spectrum and for different energy thresholds. We also discuss a novel analysis that fully takes into account the spectral features of the gamma-ray spectrum of specific DM models in a SuperSymmetric scenario. I. INTRODUCTION In the ΛCDM cosmological scenario about 80% of the matter of the Universe is believed to be composed of non-baryonic matter, called Dark Matter (DM). The most popular DM candidates are the WIMPs (weakly interacting massive particles) supposed to be cold, electrically neutral, stable, and massive [1]. Among the huge plethora of WIMP candidates, the best motivated ones are related to the SuperSymmetrical (SUSY) and Extra Dimensional extensions of the Standard Model of particle physics [2]. In the Minimal SuperSymmetric extension of the Standard Model (MSSM), the neutralino χ represents an excellent cold DM candidate with a relic density compatible with the WMAP bounds. Since the neutralino is a Majorana particle, pairs of χ can annihilate into Standard Model particles, e.g., quarks, leptons, and W bosons. The subsequent hadronization of those particles results in a continuum emission of gammarays characterized by a cut-off at the neutralino mass and by possible spectral features like bumps or a hardening of the spectral slope. The expected gamma-ray flux from DM-annihilating astrophysical objects, as function of the energy threshold E0 and the integration region ∆Ω, within which the signal is integrated, can be factorized in two terms: Φ(> E0 , ∆Ω) = ΦP P (> E0 )J(∆Ω). (1) The so-called particle physics factor ΦP P depends on the features of the DM particle, and can be written as: Z n 1 < σann v > mχ X i dNγi PP Φ (> E0 ) = B dE, 4π 2m2χ dE E0 i=1 (2) where < σann v > is the velocity averaged annihilation cross-section, and B i is the particular branching ratio for the i-th annihilation channel. The term J(∆Ω) (the so-called astrophysical factor) is given by the line–of–sight integral over the DM density squared within a solid angle ∆Ω, and depends on the density profile of the DM halo of the source: Z Z ρ2 (r(s, Ω))dsdΩ. (3) J(∆Ω) = ∆Ω los Since the gamma-ray flux of DM annihilation is proportional to the square of the DM density, only sources MAGIC Coll. (2011) 9000 8000 7000 6000 Ton = 29.42 [h] Non = 53928 ± 232 Noff = 54207 ± 233 Nexc = -279 ± 329 σLiMa = -0.85 5000 4000 3000 2000 1000 0 0 10 30 40 50 60 70 80 90 | α | [°] MAGIC Coll. (2011) 10-10 spectral index Γ = -1 spectral index Γ = -1.5 spectral index Γ = -1.8 spectral index Γ = -2 Φ spectral index Γ = -2.2 spectral index Γ = -2.4 -11 10 II. 20 FIG. 1: α-plot from 29.4 hours of Segue 1 observation above 100 GeV. Red points represent the signal (ON distribution), black points the background (OFF distribution), and green points their difference. The vertical dashed line at α = 14◦ is the fiducial region below which the excess event number is estimated. Image taken from [10]. UL with high expected DM densities are good targets for DM indirect searches. Among these, the dwarf spheroidal satellite galaxies (dSphs) of the Milky Way (MW) are interesting objects thanks to their relative proximity to the Earth, to their high mass–to–light ratio (with values within tens and thousands of M⊙ /L⊙ ) and to the expected absence of conventional gammaray sources within the system [3, 4]. So far, around two dozen dSphs have been identified. Segue 1, discovered in 2006 by the SDSS [5], is located at 28 kpc from the Galactic Center, at (RA,DEC)=(10.12h , 16.08◦ ). Kinematics studies applied to 66 member stars allowed to estimate its mass–to–light ratio to be in the range 1320-3400 M⊙ /L⊙ [6], highlighting Segue 1 as the most DM dominated dSph known so far. The MAGIC-I telescope is a 17 m dish Imaging Atmospheric Cherenkov Telescope (IACT), located at the Roque de los Muchachos Observatory, in the Canary Island of La Palma (2200 m a.s.l.). Thanks to its low energy threshold (∼60 GeV at Zenith), high flux sensitivity (1.8% of the Crab Nebula flux in 50 hour of observations above ∼250 GeV), and good angular and energy resolution (0.1◦ and 30% respectively, at 100 GeV) [7], MAGIC-I is a suited instrument for the indirect search for DM candidates with energy of the neutralino mass or the Kaluza-Klein state. counts 2011 Fermi Symposium, Roma., May. 9-12 [cm-2 s-1] 2 MAGIC-I OBSERVATION AND ANALYSIS RESULTS 10-12 A search for a possible DM gamma-ray signal coming from Segue 1 was performed by the MAGIC-I telescope between November 2008 and March 2009, for a total of 29.4 hours of observation time (after data selection). The data analysis was performed using the standard MAGIC-I analysis and reconstruction software [8]. The number of gamma-ray candidates events from the direction of the source was estimated using the distribution of |α| angles, which are related to the orientation of the showers. The overall analysis cuts were optimized and cross-checked for point-like sources with the aid of contemporaneous Crab Nebula data. In Figure 1, the |α|-plot above 100 GeV is shown. The number of excess events was computed in a fixed fiducial signal region with |α| < 14◦ and resulted to be Nexc (> 100 GeV) = −279 ± 329, corresponding to a significance of −0.85σ, computed using eq.(17) of Li&Ma [9]. Since results were consistent with no signal over the background, we derived Upper Limits (ULs) on the flux, calculated using the Rolke method [11] at 95% confidence level, and assuming a 30% systematic uncertainty. Figure 2 shows the integral ULs achieved by the MAGIC-I observation of Segue 1 considering different energy thresholds E0 and different power-law spectra with spectral index Γ = −1, −1.5, −1.8, −2, −2.2, −2.4. It is worth noting that, using the Rolke method, the ULs on the num- 10-13 102 3 10 E_0 [GeV] FIG. 2: Integral flux ULs from Segue 1. The arrows indicate the integral flux upper limits for different power-law spectra and energy thresholds. The dashed lines indicate the corresponding integral ULs if zero significance σLi,M a is assumed. Image taken from [10]. ber of the excess events, and consequently the integral flux ULs, are affected by statistical fluctuations quantified by the significance of the observation σLi,Ma . This is an intrinsic feature of the statistical method exploited in the analysis and it should be taken into account when comparing ULs from different analyses. To show this effect, in Figure 2 we plot also the ULs (dashed lines) computed assuming a value for σLi,Ma equal to zero (with number of ON events equal to the number of OFF events in the signal region of |α|plot) for different values of spectral index and energy threshold. III. CONSTRAINTS ON DARK MATTER MODELS Assuming a particular form for Segue 1 DM halo, and a given particle model for the DM candidates, we can translate the integral ULs derived from the Segue 1 observation into constraints on the DM annihilation rate. Motivated by results from cosmological simulation, the DM halo around Segue 1 was modeled by using the Einasto radial profile [12] with σs =1.1×108 M⊙ kpc−3 , rs =0.15 kpc, and n=3.3. With those parameters, the total astrophysical factor of Segue 1 results to be J(∆Ω) = 1.78×1019 GeV2 cm−5 sr. Since the analysis was performed assuming point-like source cuts (corresponding to an angular integration of 0.14◦ above 100 GeV), we estimated the effective astrophysical factor within the analysis cuts to be used in the following analye sis, being its value J(∆Ω) = 1.14×1019 GeV2 cm−5 sr (corresponding to the 64% of the total astrophysical factor). Concerning the particle physics, we restricted ourselves to the case of a SUSY model in which the presence of a discrete symmetry (R–parity) guarantees that the Lightest SuperSymmetric Particle (LSP) is stable over cosmological timescales and, therefore, a good DM candidate. We considered a 5-dimensional subspace of the MSSM called mSUGRA [13], for which the basic parameters are the universal masses of the gauginos (m1/2 ) and scalars (m0 ), the trilinear coupling (A0 ), the ratio of the vacuum expectation values of the two Higgs fields (tanβ) and the sign of the Higgsino mass term (sign(µ)). In order to study the phenomenology of mSUGRA we performed a grid scan over the parameter space, for a total of 5×106 points (for the details see [10]). The full circles of Figure 3 represent all the models of the scan i) where the lightest SUSY particle is a neutralino, ii) that survive the Standard Model constraints and iii) with a relic density compatible with the value derived by WMAP data within three times its experimental error σW MAP [14]. For each DM model of the scan, we computed the integral flux UL (above an energy threshold E0 ), using the Segue 1 data and the specific gamma-ray spectrum derived from the individual DM model. Since the spectra for each DM model have different shapes and cut-offs, the value of the optimal energy threshold E0 was computed individually for each DM mass. We then converted the flux ULs into ULs on the velocity averaged cross-section to have a direct comparison of experimental data with the theoretical predictions. The results are plotted in Figure 3 as function of the neutralino mass: each DM model of the scan (full circle) is compared to its own UL (square). For each point we defined an enhancement factor (ENF) as the ratio between the UL on the velocity averaged cross-section <σv> [ cm3 s-1 ] 2011 Fermi Symposium, Roma., May. 9-12 10 10 10 3 MAGIC Coll. (2011) -18 Upper limits mSUGRA (<WMAP) mSUGRA (WMAP) 4 ENF<10 5 4 105<ENF<106 10 <ENF<10 6 ENF>10 -19 -20 10 -21 10-22 10 -23 10 -24 10 -25 10 -26 10 -27 10 -28 10 -29 0 200 400 600 800 1000 1200 mχ [GeV] FIG. 3: Velocity averaged annihilation cross-section ULs from Segue 1 MAGIC-I data computed for individual points in the scan. Grey crosses indicate the velocity averaged annihilation cross-section value for those points in the scan that pass the SM constraints and with a relic density lower than WMAP bound. The full circles only consider models within 3σW M AP from WMAP bounds. For each of these full circles the UL on the cross-section can be computed from the Segue 1 data (after energy threshold optimization) and it is indicated here by a square. Circles and squared are colored in term of the enhancement factor. Image taken from [10]. and the value predicted by mSUGRA. This quantity quantifies how far away we are from excluding some portions of the mSUGRA parameters space. From Figure 3 it can be seen that ENFs for model compatible with the WMAP bounds are typically above 103 , while typical values are of the order of 104−5 . IV. IMPACT ON PAMELA PREFERRED REGION We tested our ULs on some of the models proposed in the literature that can explain the PAMELA data [15] for the energy spectrum of the positron fraction e+ /(e+ + e− ) as due to DM annihilation into leptons. The regions in the (mχ , < σann v >) plane that provide a good fit to the PAMELA measurements [16, 17] for three different channels of DM annihilation are shown in Figure 4. The annihilation channels χχ → µ+ µ− , χχ → τ + τ − have been taken from SuperSymmetry, while for χχ → φ+ φ− → 2e+ e− the existence of a new dark force, mediated by the carrier φ that decays into leptons [18], has been assumed. In Figure 4 we plot the ULs obtained from the Segue 1 data, using again the specific DM annihilation spectra. We can see that, in this case, the ENFs needed to meet the PAMELA-favoured region 2011 Fermi Symposium, Roma., May. 9-12 <σ v> [ cm3 s-1 ] 4 10 MAGIC Coll. (2011) -19 µ µ + - 10 τ+τ- -20 + - φφ→2e e 10-21 10 10 -22 -23 10-24 10 -25 10 -26 10 -27 PAMELA µ +µ PAMELA τ+τ+ PAMELA φφ→2e e- 0 200 400 600 800 1000 1200 mχ [GeV] FIG. 4: Exclusion lines for a neutralino DM annihilating exclusively into µ+ µ− (green lines) or τ + τ − (blue line), and for a DM candidate interacting with a light intermediate state φ decaying into a pair of electrons (pink line). The same annihilation channels (with the same color coding) are considered to draw the regions in the plane that provide a good fit to the PAMELA measurement of the energy spectrum of the positron fraction. Image taken from [10]. are much smaller than those found for mSUGRA, and in the case of annihilation into τ + τ − our ULs are probing the relevant regions. However, it is worth mentioning that, since the uncertainty in the Segue 1 astrophysical factor is quite large [12], an improvement in accuracy of astrophysical factor value could be able to put more stringent constrains and to confirm the exclusion of the PAMELA region for DM particle annihilating in τ + τ − . on the gamma-ray emission were computed assuming different power-law energy spectra. Within the mSUGRA scenario, a large scan of neutralino models was performed over the parameter space. Subsequently for each simulated DM model, the ULs on the velocity averaged annihilation cross-section (mχ , < σann v >) were derived separately for each point in the scan in order to account for the dependence on the specific spectra. Results indicate that a general exclusion plot cannot be drawn to constrain the parameter space, so we provide the results in terms of enhancement factors. A minimum boost on the flux is found of the order of 103 (for models compatible with WMAP) while the typical values are at 104−5 . MAGIC-I data of Segue 1 can be useful to put constraints on those DM models that are provided in the literature to explain the PAMELA data. Our ULs are probing the PAMELA region for the DM models annihilating into τ + τ − but the robustness of this result could be improved, decreasing the uncertainty in the astrophysical factor. Although the MAGIC-I observation did not result in a detection, and the ULs require still high flux enhancement factors to actually match the experiment sensitivity, an analysis like the one presented here is able to point out details and features that can be important for future deep exposures of this or similar objects, with next-generation Cherenkov experiments. Acknowledgments A search for a possible DM gamma-ray signal coming from Segue 1 was performed by the MAGIC-I telescope. No hints of signal were found above the background for energies larger than 100 GeV. Integral ULs We would like to thank the Instituto de Astrofı́sica de Canarias for the excellent working conditions at the Observatorio del Roque de los Muchachos in La Palma. The support of the German BMBF and MPG, the Italian INFN, the Swiss National Fund SNF, and the Spanish MICINN is gratefully acknowledged. This work was also supported by the Marie Curie program, by the CPAN CSD2007-00042 and MultiDark CSD200900064 projects of the Spanish Consolider-Ingenio 2010 programme, by grant DO02-353 of the Bulgarian NSF, by grant 127740 of the Academy of Finland, by the YIP of the Helmholtz Gemeinschaft, by the DFG Cluster of Excellence “Origin and Structure of the Universe”, by the DFG Collaborative Research Centers SFB823/C4 and SFB876/C3, and by the Polish MNiSzW grant 745/N-HESS-MAGIC/2010/0. [1] D. N. Spergel et al., ApJS 170 (2007) 377. [2] G. Bertone, D. Hooper, J. Silk, Phys. Rept. 405 (2005) 279-390. [3] M. Sanchez-Conde et al., AIP Conf. Proc.1166 (2009) 191-196. [4] G. Gilmore et al., Astrophys.J.663 (2007) 948-959. [5] V. Belokurov et al. [SDSS Colaboration], Astrophys.J. 654 (2007) 897-906 [astro-ph/0608448]. [6] J.D. Simon et al., (submitted) (2010) astroph.GA/1007.4198. [7] J. Albert et al. [MAGIC Collaboration], Astrophys.J. 674 (2008) 1037-1055 [arXiv:0705:3244]. [8] T. Bretz et al., Proc. 29th ICRC 2005, Tsukuba. [9] T.P. Li, Y.Q. Ma, Astrophys.J. 272 (1983) 317-324. [10] J. Aleksić et al. [MAGIC Collaboration], JCAP 06 (2011) 035. [11] W. A. Rolke, A. M. Lopez, J. Conrad, Nucl.Instrum.Meth. A551 (2005) 493-503 [physics/0403059]. [12] R. Essig, N. Sehgal, L. E. Stringari, M. Geha, V. CONCLUSIONS 2011 Fermi Symposium, Roma., May. 9-12 J.D. Simon, Phys.Rev D82 (2010) 123503 [astroph/1007.4199]. [13] A. H. Chamseddine, R. L. Arnowitt, P. Nath, Phys.Rev.Lett 49 (1982) 970. [14] E. Komatsu et al. [WMAP Collaboration], Astrophys.J.Suppl. 192 (2011) 18 [astro-ph.CO:1001.4538]. [15] O. Adriani et al. [PAMELA Collaboration], Nature 458 (2009) 607-609 [astro-ph/0810-4995]. [16] I. Cholis et al., JCAP 0912 (2009) 007 [astroph/0810.5344]. [17] I. Cholis et al., Phys.Rev D80 (2009) 123518 [astroph/0811.3641]. [18] N. Arkani-Hamed et al. Phys.Rev D79 (2009) 015014 [hep-ph/0810.0713]. 5