64 CHAPTER 4． OB8ERVATτON AND DL4工4 REDσ（フ質ON
胃
．9
言
§
e
ぺ 這 甘 拐 口 あ 守 自 已
Φ 自
昆
奉
ぢ
査 国
㌔
．9
ぢ
①
．iコ
勺
．日bD 甘
6
①
肩 遷
○ ヱヒ 日 づ
d
① oりぢ 巴 書
q
c＼1＜◎
1＞一・マ
oOg oo
r「寸
（ F→（
寸め
Φ「寸 o．（Nめo．
〔qeカ＾ご乃 σ） r、
ζ＼1ぴoo oo
o．o OOo．
卜一→
一（N
））
oO oO
←→ †一｛
＼＼
叶（あ＼＼
▼一→o頃）㊤oo oo
c＼1（N I I
吟卜
H 〒■→
4．1．SAMPLE CLσSTER．5 65
temperature out to r500， Errors are at 90％con6dence for olle illteresting parametel．、恥 applied Bahlcinska−Church et al．（1998）for the photoelectric absorption cross−sections，Gelleralizing to a spherical collapse for ollr adopted cosmology at the red−shift of A 14ユ3 gives an over density oξ112 for the virialized region． However for Al413 r1ユ2 is only 18％
larger than rl80（Hellry 2000）． So for co皿parisoll with previous work we adopt the latter as our definition of the virial radius and will Ineasure it in this paper．
4．12 A2204
（a）A2204（XMM） （b）A2204（Suzaku）
＋54σ
＋530
＋52σ
＋5 40
＋5、3σ
16ト34［@ 16h33m205 16h32m40〜 16h32m 16h33、05 ．†6h32叩305 16 31m50
Fig．4．2：（a）Xル∫M−Neω亡oηMOS1十MOS2 image of A2204 in the O．35−1．25 keV band．
The image is corrected for exposure． non−viglletting but backgrouIld is subtracted with blank sky lpixel is 50 ．The white box is FOV for Suzaku XIS． The green circle shows the
、irial radius of 1．8 Mpc．（b）Suzaku FI十BI image for outskirts of A2204 in tlle O．5 5．O keV band smoothed by a Gaussiall withσ＝16 、 The image is not corrected fbr exposure time alld nol1−vignetting but background is llot subtracted． COR＞8 GV scree111ng is apPlied．55Fe calibration source regions are also included in the丘gure， but the sources has Ilo photons at this energy band． White circles denote 3 ．5膓7 ，11 ．5 alld 15 ．5 from the surface brightness peak．
A2204 is the second massive cooling flow cluster of the systematic red−shif口＝0．1523
（Struble＆Rood 1987）， wllich yields an anglllar diameter distance of 548，0力司Mpc、
lumhlosity distallce of 727．6ん冠Mpc and a scale of 159．4方品kpc per arcmillute for
our assulned cosnlolog｝㌧The neutral hydrogell colulnn density in this direction is 6．07×1020cm−2（Dickey＆Locknlal11990）．
Though A2204 is high z． but Inorphologically symllletric and relaxed cluster． Snowden et al，（2008）included their calalog of 78 clusters with A2204 which is observed with Xハf〃−
Ne磁oη． Reiprich et aL（2009）reported the first result of A2204 with 5鵬α丸obse官ation
66 CHAPTER 4． OB8ERVA質ON AND DATA REDσCTION
which is observed within r200． Basu et al．（2009）has also observed A22040f submillimeter band with APEX−SZ by utilizing SZ−efFect．
We observed the northern region of A2204 at radii丘om O／to 19，．5丘om the XMM surface brightness peak with the Suzaku XIS de七ec七〇rs． In table 4．3， we give七he details of our observation， and in figure 4．2（a）， we show the XIS field of view（FOV）superimposed
on the XMM−Newton image of A1413． The XIS instrument consists of 4 CCD chips；one
back−illuminated（BI：XIS 1）and three front−illuminated（FI：XISO， XIS2， XIS3）， with each is combined with an X−ray telescope（XRT）． The IR／UV blocking filters had accumulated asigni丘cant contamination by the time of七he observation since its launch（July 2005）；we include its effects on the effective area in our analysis． The XIS was operated wi七h
normal clocking mode， in 5×50r 3×3editing modes． The spaced−row charge injection
（SCI）was not applied， and all the four CCDs were working at the time of七he observation．
We show the FI十BI image ill the O．5−5 keV energy band in figure 4．2（b）． Data reduction method is same as the case of A1413． We show data log in table 4，3．
4．2 Data Reduction
4．2．1 5μzαicμ
Procedure of Analysis
We used HEAsoft ver 6．4．1 and dALDB version 2008−06−21 for all the Suzaku analysis presented here． We started七he event screening from the cleaned event丘1e， We extracted pulse−height spectra in annular regions from七he XIS event files．
Here， we explain the procedure used for the spectral analysis of each cluster． However，
we explain background analysis in the latter chap七er．
1．The observation data are screened with ELV＞5deg， COR2＞8GV and 100 counts／s ＜PINUD＜300 counts／s，
2．From observed images， we excluded bright point sources avoiding contamination by 七hem．
3．The spectra are extracted from the annular regions centered on clusters，
4．For each annular region，七he NXB model spectrum is created using PINUD． These
NXB spectra are subtracted from the observed cluster spectra ex七racted in step 2．5．We generated the RMF for the epoch of the clus七er observation using the斑r唖geπ （version 2007−05−14）．
6．We analyzed point source spectra and estimated且ux con七amination by all of them
in FOV（A1413 and AWM7）．
7．By assuming surface brightness pro丘1es， we created simulated cluster images．
4．2．DAZ14 REDσC質ON 67 8．We generated uniform ARFs for NXB， CXB， and galactic components， and cluster ARFs for each region by⑰55Xmα噛eηwith 2M photons．
9．We carried out simoltanious丘tting with all regions， all detectors（FI十BI）， and all
background model components（CXB and galactic components）．
10．We estima七ed stray light 』ctions for each region by流5乞m．
11．We looked into all of uncertainties for CXB Huctuations， NXB且uctuations， Con−
taminations on IR／UV blocking filters．
12．We simula七ed NXB， CXB， galactic components， and also ICM with痂3仇↓七〇create surface brightness profiles wi七h photon counts with the best−fit results．
Spatial and Spectral Responses
We need to prepare the spatial and spectral responses which are necessary for reducing and analyzing our observa七ions of A1413． These responses have complica七ed properties for extended sources。 Indeed七hey depend on the surface brightness dis七ribution of the source and so are unique for each annular region． Monte Carlo simulators are used to generate some of the responses． The X−ray telescope十XIS simulator is called￠τ552m，
and the ARF generator using the simulator is called斑52mα物εη（Ishisaki et al．2007）．
We used version 2008−04−050f the simula七〇r．
Asurface brightness distribution is necessary for励55仇z and励55τmα加εη， because
the point spread function（PSF）of七he XRT produces an e伍ciency that is correlated
among adjacen七spatial cells． Since the XIS FOV did not include the brightness peak ofA1413， we used the KBB model of Pratt＆Arnaud（2002）to genera七e the ARF We numerically projected七he KBB 3−dimensional model of the gas density to generate the
input surface brightness distribution． Since the ARF describes the detection ef五ciency as afunction of energy， no particular spectral shape is required for input． The effect of the XIS IR／UV blocking filter contamination is included in the ARF based on the calibra七ionof November 2006． The normalization of the ARF is such that the measured且ux in a
spectral fit for a given spatial region is the且ux from the entire input surface brightness．The且ux just from the spatial region is the fit flux times the痂52mα吻eηoutput parameter SO［硯OE−R、4冗0」認G（七able 5．7）． The surface brightness ffom a given spatial region is then the usual flux from the region divided by the solid angle that subtends from七he observer．
4．2．2 XMM−1Ve庇oη
For the morphological analysis， we utilized XMM−1Veω舌oηobservation data set． To cre−
ate images and surface brightness profiles， we due七〇the standard analysis me七hod in Hayakawa 2006 with the initial processing with mo5−∫言1‡erwhich is a part of XMM−Newtion
68 CHAPTER 4． OB8ERVATZON AND DA工4 REDσC質ON
Extended Spectral Analysis Software（XMM−ESAS）1．We utilized blank sky data to es−
timate background illtensities． Hayakawa 2006 carried out background subtraction with
blank sky data． However， they did not utilize exposure−corrected images for blank sky．In the outer region， we must care for background subtraction in slite photon counts． We carried out correction of background with exposure map．
In§3．2．3， we described the EPIC background properties， and found七ha七the back−