Baryons in the Outskirts of the Nearest, Brightest Galaxy Cluster

Baryons in the outskirts of the nearest, brightest galaxy cluster

Aurora Simionescu (KIPAC)

Steve Allen, Adam Mantz, Norbert Werner, Yoh Takei, Glenn Morris, Andy Fabian, Jeremy Sanders, Paul Nulsen, Matt George Why study clusters to large radii?

Accurate measurements of the properties of galaxy clusters out to large radii provide critical insight into

• physics of the ICM and pre-virialized IGM (the formation of largest scale structure `as it happens’) • use of clusters as cosmological probes (calibration of X-ray mass proxies; benchmark for hydro. simulations) Until recently, detailed thermodynamic studies of clusters out to r ~rvir have proved extremely challenging

• inherently low surface brightness of cluster outskirts. • relatively high particle backgrounds of Chandra/XMM-Newton.

2/3 of cluster volumes practically unexplored!

Suzaku enables these studies by providing a lower and more stable background. From Akamatsu et al. 2011 (additional data from Hoshino et al. 2010, George et al. 2009, Kawaharada et al. 2010, Bautz et al. 2009, Reiprich et al. 2009)

To maximize the signal-to-noise and minimize the systematics related to the modest PSF of Suzaku, we must observe the outskirts of the nearest, brightest clusters, making the Perseus Cluster an ideal target. Results from the Perseus Cluster observations:

The ﬁrst two arms: analysis of E & NW mosaics (total 260 ks) reported by Simionescu et al. 2011, Science, 331, 1576 Surface brightness images of the NW and E arms: Spectral analysis Spectral analysis I: background

Background model based on ﬁts to ROSAT and Suzaku outer pointings

• unabsorbed LHB 0.1 keV thermal

absorbed GH 0.2 keV thermal

•

absorbed 0.6 keV thermal AWM7 • Perseus

Algol • absorbed Γ=1.41 power law • expected particle background (subtracted)

50 51 54 60 71 95 140 231 414 777 1499 Spectral analysis II: stray light

Stray light spectrum softens with radius Exclude parts of each spectrum where stray>0.2*(data-modelCXB) - usually >1.5 keV Results Projected temperature and metallicity proﬁles:

excellent agreement with Chandra data

detailed proﬁles spanning 3 decades in radius r500

proﬁles between r500 and r200 resolved for the ﬁrst time

metallicity profile measured for the first time until the virial radius CXB systematics are small: Deprojected thermodynamic profiles:

r 500 shallow decline of electron density at large radii E cold front entropy appears to ﬂatten at large radii compared to the expected power-law

pressure at large radii greater than predicted by numerical simulations (ﬁtted to XMM data inside r500 by Arnaud et al. 2010) Comparison with ROSAT

ROSAT data from Ettori et al. 1998 Gas mass fraction proﬁle towards the NW:

NW arm highly relaxed ➛ use hydrostatic equilibrium to infer gas and total mass proﬁles (E arm excluded due to cold front at 30’) Underlying mass distribution assumed to follow NFW proﬁle; no other parametrizations (e.g. for ne, kT) were used! good agreement with previous observations and numerical simulations at r<0.4r200

fgas value matches cosmic mean at r~r500

no missing baryons in clusters Gas mass fraction proﬁle towards the NW:

fgas exceeds cosmic mean at large radii (r>0.6-0.7r200)

most likely cause: the gas is clumpy, thus ne predicted from the X-ray surface brightness is biased high

bottom panel shows the ﬁrst measurements of the gas clumping factor

important implications for future studies at very large radii in clusters, e.g. X-ray+SZ Corrected thermodynamic proﬁles:

correcting for clumping (red lines) brings measurements into agreement with expected trends other mechanisms, e.g. Te ≠Ti would explain entropy ﬂattening but not explain pressure and fgas proﬁles AComparisonofCosmologicalCodes 11

AComparisonofCosmologicalCodesIs the clumping factor realistic? 11

numerical simulations by Vazza et al. 2011 Figure 9. Mass weighted proﬁles of gas density (left column), gas temperature (center column)andgasclumpingfactor(right column)forClusterAatvarious resolutions. GADGET runs are in the upper row, TVD runs are in the middle and ENZO runs are in the bottom row. Vertical dashedlinesshowtheminimum radius enclosing the minimum mass suitable for convergence studies, as introduced in Sec.4.

Figure 10. Volume weighted proﬁles of gas entropy (in arbitrary code un its) for Cluster A at various resolutions. The vertical dashed lines show the minimum radius enclosing the minimum mass suitable for convergence studies, as introduced in Sec.4.

15 10 M!/h and Rvir =2.32Mpc/h,inafairlyrelaxed We preliminary checked that the total masses at all dynamical stage; resolutions and in all codes are in agreement within a Figure 9. Mass weighted proﬁles of gas density (left• columncluster), gas B: temperature a system (center of column total)andgasclumpingfactor( mass M =1.64right· column∼)forClusterAatvarious6 per cent level within Rvir,sothatthegeneralpa- resolutions. GADGET runs are in the upper row, TVD15 runs are in the middle and ENZO runs are in the bottom row. Vertical dasherametersdlinesshowtheminimum deﬁning the systems are nearly identical in all radius enclosing the minimum mass suitable for10 convergenceM!/hstudies,and as introducedRvir =2 in. Sec.4.47Mpc/h,inanongoing investigated resolutions. merger phase.

"c 0000 RAS, MNRAS 000,000–000

15 10 M!/h and Rvir =2.32Mpc/h,inafairlyrelaxed We preliminary checked that the total masses at all dynamical stage; resolutions and in all codes are in agreement within a • cluster B: a system of total mass M =1.64 · ∼ 6 per cent level within Rvir,sothatthegeneralpa- 15 rameters deﬁning the systems are nearly identical in all 10 M!/h and Rvir =2.47Mpc/h,inanongoing investigated resolutions. merger phase.

"c 0000 RAS, MNRAS 000,000–000 To conﬁrm gas clumping, we need to directly detect and study the clumps with Chandra

Moreover, simulations predict azimuthal variations in clumping ➙ crucial to have measurements along other directions / in other systems! Look forward to:

S and W arms have been observed - data reduction is under way

NNE, NE, SE, SW arms will be observed in AO-6. Surface brightness image of the combined mosaic to date (preliminary!)

0 0.0002 0.00059 0.0014 0.003 0.0061 0.012 0.025 0.05 0.1 0.2 More deprojected thermodynamic proﬁles (preliminary!): Also look forward to:

Abell 2199 Coma

Coma A2199

0 0.069 0.21 0.49 1 2.2 4.4 8.7 18 35 70 0 0.04 0.12• extension0.28 0.59 1.2 to2.5 other5 nearby,10 20 bright40 clusters (Coma, A2199) to study system-to-system variations Conclusions:

•We have obtained the ﬁrst observational proofs for gas clumping in cluster outskirts.

•Clumping provides a new window onto the virialization and equilibration processes and the physics of cluster outskirts -> numerical simulations will be a key to understand this further.

•Knowledge of the radial dependence and azimuthal variance of clumping is critical for robust measurements of thermodynamic quantities, e.g. density, entropy, pressure.

•Along one relaxed arm of Perseus, we have measured a very accurate gas mass fraction proﬁle. Our results indicate that there are no “missing” baryons in clusters.

Perseus NW spectrum 0.95-1.05r200 data and folded model

0.05 1 ï 0.02 keV 1 ï 0.01

5×10ï3

ï3 normalized counts s 2×10

10ï3 3

2 ratio

12 5 Energy (keV)

aurora 14ïMarï2011 23:55 Stray light systematics are small: gNFW mass model