DM Workshop@UT Austin May 7, 2012 Constraints on dark matter annihilation cross section with the Fornax cluster Shin ichiro Ando University of Amsterdam Ando & Nagai, arxiv:1201.0753 [astro-ph.he]
Galaxy clusters: Why interesting? Bullet cluster (1E0657-56) The largest virialized darkmatter structure The largest number of dark matter particles Presence of collisionless dark matter clearly seen in bullet cluster Good probe of cosmological parameters
23 Cluster constraints on DM properties Ackermann et al., JCAP 1005, 025 (2010) <!v> [cm 3 s -1 ] 10-20 10-21 10-22 10-23 10-24 10-25 10-26 WMAP models All SUSY models Fornax NGC4636 Centaurus Coma AWM7 M49 Annihilation Wino dark matter 11-month data Higgsino dark matter 100 1000 WIMP Mass [GeV] 100 1000 DM mass [GeV] Factor >30 gap from Fig. 1. Most Constraints stringent on the constraint decay lifetime as a fu : Left: Upper canonical limits cross on section annihilation cross asection b b final decay state σv (left lifetime for panel) a b b andfinal a µ state as a + µ final state ass, obtained from six nearby clusters. The constraints are compared to the DM lifetime [s] Dugger et al., JCAP 1012, 015 (2010) 10 27 10 26 10 25 Ugrav =2x10-8 bb final state - constraints from clusters AWM7 Centaurus Coma Fornax M49 NGC4636 Ugrav =0.5x10-8 Decay New Constraints MNP =1016 B GeV Galactic + EG Previous Constraints level upper limits on the gamma-ray flux from 11 MNP =4x1016 B GeV DM lifetime [s]
Annihilation boost in substructure Gao et al., arxiv:1107.1916 10-5 Mmin = 10 12 Msun 10-6 S(R) [photons cm -2 s -1 sr -1 ] 10-7 10-8 10-9 Mmin = 5x10 7 Msun Mmin = 10 6 Msun Extrapolation >13 dex gap Resolution limit Millennium Simulation 10-10 Flux boosted by a factor of ~1000 10-11 0.1 1.0 10.0 100.0 1000.0 R [ kpc ]
Annihilation boost in substructure Huang et al., arxiv:1110.1529 Dwarf limits (no subhalos) Cluster limits with subhalos Figure 7: Left panel: individual upper limits on annihilation rate into b b, includingasignalboost from dark matter substructures as discussed in Section 2.3 (solid lines). For comparison, the dotted Ackermann et al., arxiv:1108.3546 lines show the limits on the unboosted signal, cp. Fig. 4. Right panel: the same, but for annihilation Geringer-Sameth & Koushiappas, arxiv:1108.2914 into τ + τ.
Motivation Does stacking help? If so, how much? There are many more clusters than dwarfs!! No! It doesn t help What is the effect of baryons (stars+gas)? Baryons dominate gravitational potential at central regions This should modify dark matter profile (adiabatic contraction) It improves limits by a factor ~4
Dark matter annihilation in galaxy clusters Gamma-ray intensity from annihilation I γ (θ,e)= 1 4π 1 σv (1 + z) 2 2m 2 χ dn γ ((1 + z)e) de dl ρ 2 (r(l, θ)) Cosmological redshift Particle-physics factor Astrophysical factor Depends on three factors Particle physics: annihilation cross section and dark-matter mass; depends on SUSY models, etc. Astrophysics: density profile and subhalos Cosmological redshift: straightforward if redshift is measured
Astrophysical factor: density profile Umetsu et al., Astrophys. J. 738, 41 (2011) Numerical simulations imply universal form of density profile: NFW ρ s ρ = (r/r s )(r/r s + 1) 2 ρ ~ r 1 for small radii, and ρ ~ r 3 for large radii NFW profile is confirmed with lensing observations
Gamma-ray intensity NFW Mmin = 10 6 Msun Subhalos Mmin = 5x10 7 Msun Fornax Coma z Mvir (10 14 h 1 Msun) 0.005 1.2 0.023 9.6 Intensity due to subhalos is much more extended than the smooth component Subhalo boost factor is ~1000 for cluster-size halos, if minimum subhalos are of Earth size
Analysis of Fermi-LAT data We analyze data of Fermi-LAT for 2.8 years around 49 relatively large galaxy clusters DIFFUSE and DATACLEAN class of photon data between MET = 239557417 s and 329159098 s 23 clusters from X-ray (Reiprich & Boehringer 2002) and 34 from cosmology catalogs (Vikhlinin et al. 2009); 3 are found in both and 5 are at low Galactic latitudes We first perform likelihood analysis of the data using the known sources (from 2FGL catalog) as well as both Galactic and extragalactic backgrounds Use photons between 1 GeV and 100 GeV, and divide them into 20 energy bins equally spaced logarithmically Models are convolved with P6_V11 instrumental response functions
Fermi-LAT data and best-fit model for Fornax 48-25.0 Counts Map (Fornax) -25.0 Model Map (Fornax) 24 12-30.0-30.0 6-34.5-35.0 Counts Map (Fornax) Declination -35.0 Declination -35.0 3 1.5 0.71 Declination -35.5-40.0-40.0 0.33 0.14-36.0 65.0 60.0 55.0 50.0 45.0 Right ascension 65.0 60.0 55.0 50.0 45.0 Right ascension 0.048 0 55.5 55.0 54.5 54.0 53.5 Right ascension Integrated maps for 1 100 GeV There is no gamma-ray source at cluster location We then add cluster component at the center of the best-fit model map, to put upper limit on that component
Upper limits on cluster component Analyze With -34.5 Data Counts Map (Fornax) -34.5 Fornax (host halo only) Cluster model 8.67e-07 7.80e-07 6.94e-07-34.5 Fornax (with subhalos) 1.20e-05 1.11e-05 1.02e-05-35.0-35.0 6.07e-07-35.0 9.31e-06 Declination -35.5 Declination -35.5 5.20e-07 4.34e-07 3.47e-07 Declination -35.5 8.40e-06 7.49e-06 6.58e-06-36.0-36.0 2.60e-07-36.0 5.67e-06 1.73e-07 4.75e-06 55.5 55.0 54.5 54.0 53.5 Right ascension 55.5 55.0 54.5 54.0 53.5 Right ascension 8.68e-08 3.96e-10 55.5 55.0 54.5 54.0 53.5 Right ascension 3.85e-06 2.94e-06 1 Outcome Declination -34.5-35.0-35.5 Model Map with Cluster (Fornax) 0.5 0.26 0.13 0.069 0.038 0.022 Best-fit model (almost isotropic background) and cluster component (no subhalos) allowed by 5% -36.0 55.5 55.0 54.5 54.0 53.5 Right ascension 0.014 0.0099 0.008 0.007 95% CL upper limits on annihilation cross section
Limits on annihilation cross section from Fornax NFW halo with no subhalos Ando & Nagai, arxiv:1201.0753 [astro-ph.he]
Cross section limits for all clusters NFW _ halo only bb channel
Cross section limits from stacking analysis Procedure Remove clusters with >3σ excess compared with (fixed) background This reduces to 38 clusters to be analyzed Result Stacking does not help Better to model Fornax more precisely Ando & Nagai, arxiv:1201.0753 [astro-ph.he]
The Fornax cluster M ~ 1014 Msun D ~ 20 Mpc (l, b) = (236.72 deg, 53.64 deg) The second largest cluster locally Central massive elliptical (cd) galaxy: NGC 1399 http://heritage.stsci.edu/2005/09/supplemental.html
Baryons in Fornax HST photometry (4 x4 region) ROSAT observations Stars Gas Precision measurement down to 1 pc! Lauer et al. Astron. J. 129, 2138 (2005) Paolillo et al. Astrophys. J. 565, 883 (2002)
Density profiles of Fornax Surface brightness luminosity profile density profile Profiles from DMonly simulations Stars dominate the gravitational potential at the central region What is the feedback effect of this deepened potential?
Adiabatic contraction Angular momentum conservation: ri Blumenthal et al. Astrophys. J. 301, 27 (1986) rf M i (r i )r i = M f (r f )r f M f (r f )=[M dm,f (r f )+M b,f (r f )]r f =[M dm,i (r i )+M b,f (r f )]r f
Modified halo contraction Gnedin et al., Astrophys. J. 616, 16 (2004); arxiv:1108.5736 M i ( r i )r i =[M dm,i ( r i )+M b,f ( r f )]r f r 0.03r vir = A 0 r 0.03r vir w A0 = 1.6, w = 0.8 well explain simulation results Uncertainty range: w = 0.6 1 There is no firm observational evidence for/ against this effect yet
Effect of halo contraction Ando & Nagai, arxiv:1201.0753 [astro-ph.he] Canonical contraction model (A0=1.6, w=0.8) Density is enhanced at the center for both NFW and Einasto profiles
Gamma-ray intensity enhanced Ando & Nagai, arxiv:1201.0753 [astro-ph.he] Contraction produces sub- PSF structure at 10 4 10 3 deg (30 300 pc) Gamma-ray flux is boosted by a factor of ~4 (NFW) ~2 (Einasto)
Cross section upper limits Limits improve by a factor of 4.1 (NFW) 2.4 (Einasto) This is almost independent of mass and annihilation channel <σv> < (2 3)x10 25 cm 3 /s for low-mass WIMPs Ando & Nagai, arxiv:1201.0753 [astro-ph.he]
Other model parameters Ando & Nagai, arxiv:1201.0753 [astro-ph.he] Adiabatic : A0=1, w=1 Break : no contraction within 1 kpc (~ current resolution limit) Uncertainty range of the boost: 2 6
How important is this?: Compare with subhalos Ando & Nagai, arxiv:1201.0753 [astro-ph.he] x4 DM only To boost the limit by a factor of 4, the minimum subhalo mass has to be smaller than 1 Msun Otherwise, one cannot ignore the effect of halo contraction
Conclusions Galaxy clusters are potentially strong source of gamma rays from dark matter annihilation We showed that stacking ~50 clusters does not improve the limits obtained with Fornax The detailed mass modeling of Fornax is therefore important We computed the halo contraction of Fornax and showed that the cross section limits improved by a factor of ~4 The limits for low-mass WIMPs are within a factor of 10 from the canonical annihilation cross section after ~3 years