1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{aastex}
3: \usepackage{emulateapj5}
4:
5: %\usepackage{graphicx}
6: %\usepackage{times}
7:
8: \makeatletter
9: \newenvironment{inlinetable}{%
10: \def\@captype{table}%
11: \noindent\begin{minipage}{0.999\linewidth}\begin{center}\footnotesize}
12: {\end{center}\end{minipage}\smallskip}
13: \newenvironment{inlinefigure}{%
14: \def\@captype{figure}%
15: \noindent\begin{minipage}{0.999\linewidth}\begin{center}}
16: {\end{center}\end{minipage}\smallskip}
17: \makeatother
18:
19: \newcommand{\lsim}{\mbox{\hspace{.2em}\raisebox{.5ex}{$<$}\hspace{-.8em}\raisebox{-.5ex}{$\sim$}
20: \hspace{.2em}}}
21: \newcommand{\gsim}{\mbox{\hspace{.2em}\raisebox{.5ex}{$>$}\hspace{-.8em}\raisebox{-.5ex}{$\sim$}\hspace{.2em}}}
22:
23: \def\asca {{\em ASCA}\/}
24: \def\suzaku {{\em Suzaku}\/}
25: \def\chandra {{\em Chandra}\/}
26: \def\xmm {XMM-{\em Newton}\/}
27: \def\rosat {{\em ROSAT}\/}
28: \def\sax {{\em BeppoSAX}\/}
29: \def\hst {{\em HST}\/}
30: \def\iras {{\em IRAS}\/}
31: \def\wmap {{\em WMAP}\/}
32: \def\mydegree{$^\circ\mskip-5mu$}
33: \def\myarcmin{$^\prime\mskip-5mu$ }
34: \def\myarcsec{$^{\prime\prime}\mskip-5mu$}
35: \def\rv {$r_{\rm vir}$}
36:
37: \begin{document}
38:
39: \title{\chandra\ studies of the X-ray gas properties of galaxy groups}
40:
41: \author{
42: M.\ Sun\altaffilmark{1},
43: G.\ M.\ Voit\altaffilmark{1},% $^{\!}$\altaffilmark{1}
44: M.\ Donahue\altaffilmark{1},%$^{\!}$\altaffilmark{1}
45: C.\ Jones\altaffilmark{2},%$^{\!}$\altaffilmark{2}
46: W.\ Forman\altaffilmark{2},%$^{\!}$\altaffilmark{2}
47: A.\ Vikhlinin\altaffilmark{2}%$^{\!}$\altaffilmark{2}
48: }
49:
50: \altaffiltext{1}{Department of Physics and Astronomy, MSU, East Lansing, MI 48824; msun@virginia.edu}
51: \altaffiltext{2}{Harvard-Smithsonian Center for Astrophysics,
52: 60 Garden St., Cambridge, MA 02138}
53:
54: \shorttitle{Properties of galaxy groups}
55: \shortauthors{SUN ET AL.}
56:
57: \begin{abstract}
58:
59: We present a systematic analysis of 43 nearby galaxy groups ($kT_{500} = 0.7 - 2.7$
60: keV or $M_{500} = 10^{13} - 10^{14} h^{-1}$ M$_{\odot}$, 0.012 $<z<$ 0.12), based
61: on \chandra\ archival data. With robust background subtraction and modeling,
62: we trace gas properties to at least $r_{2500}$ for all 43 groups. For 11 groups,
63: gas properties can be robustly derived to $r_{500}$. For an additional 12 groups,
64: we derive gas properties to at least $r_{1000}$ and estimate properties at $r_{500}$
65: from extrapolation. We show that in spite of the large variation in temperature
66: profiles inside 0.15 $r_{500}$, the temperature profiles of these groups are similar
67: at $>$ 0.15 $r_{500}$ and are consistent with a ``universal temperature profile.''
68: We present the $K - T$ relations at six characteristic radii (30 kpc, 0.15 $r_{500}$,
69: $r_{2500}$, $r_{1500}$, $r_{1000}$ and $r_{500}$), for 43 groups from this work and
70: 14 clusters from the Vikhlinin et al. (2008) sample. Despite large scatter in the
71: entropy values at 30 kpc and 0.15 $r_{500}$, the intrinsic scatter at $r_{2500}$
72: is much smaller and remains the same ($\sim$ 10\%) to $r_{500}$. The entropy excess
73: at $r_{500}$ is confirmed, in both groups and clusters, but the magnitude is smaller
74: than previous \rosat\ and \asca\ results. We also present scaling relations for the
75: gas fraction. It appears that the average gas fraction between $r_{2500}$ and $r_{500}$
76: has no temperature dependence, $\sim$ 0.12 for 1 - 10 keV systems. The group gas
77: fractions within $r_{2500}$ are generally low and have large scatter.
78: This work shows that the difference of groups from hotter clusters stems from the
79: difficulty of compressing group gas inside of $r_{2500}$. The large
80: scatter of the group gas fraction within $r_{2500}$ causes large scatter in the group
81: entropy around the center and may be responsible for the large scatter of the
82: group luminosities. Nevertheless, the groups appear more regular and more like clusters
83: beyond $r_{2500}$, from the results on gas fraction and entropy.
84: Therefore, mass proxies can be extended into low mass systems. The $M_{500} - T_{500}$
85: and $M_{500} - Y_{\rm X, 500}$ relations derived in this work are indeed well behaved
86: down to at least 2 $\times 10^{13}$ h$^{-1}$ M$_{\odot}$.
87:
88: \end{abstract}
89:
90: \keywords{cosmology: observations --- dark matter --- galaxies: clusters: general --- X-rays: galaxies: clusters}
91:
92: \section{Introduction}
93:
94: Most baryons in clusters reside in the hot ICM,
95: and most clusters are low-mass groups and poor clusters because the mass function
96: of virialized systems is steep.
97: Studies of galaxy groups are thus especially important for understanding the
98: gravitational and thermal evolution of the bulk of matter in the Universe.
99: In contrast to hot clusters, galaxy groups are systems where baryon physics
100: (e.g., cooling, galactic winds and AGN feedback) begins to dominate
101: over gravity. Groups are not just scaled-down version of massive clusters.
102: Cluster scaling relations (e.g., $L - T$ and entropy - $T$) show
103: deviations from the self-similar relations at the low-mass end (reviewed
104: in Voit 2005), reflecting the importance of baryon physics, which
105: is essential to shape the properties of the cluster galaxies
106: and ICM. Cooling sets a characteristic entropy threshold
107: in the ICM (Voit \& Ponman 2003) and is required to understand the cluster
108: scaling relations (Voit 2005). On the other hand, cooling has to be balanced
109: by feedback to preserve the observed cluster stellar fraction and the galaxy luminosity
110: function. The most important feedback may be the AGN outflows driven by the
111: central SMBH. AGN outflows can simultaneously explain the antihierarchical
112: quenching of star formation in massive galaxies, the exponential cut-off
113: at the bright end of the galaxy luminosity function, and the quenching of
114: cooling-flows in cluster cores (e.g., Scannapieco et al. 2005; Croton et al. 2006).
115: Along with SN winds, they act to suppress cooling and structure formation.
116: They pump thermal energy into the surrounding ICM as the host galaxy and the
117: central SMBH formed
118: and evolve. These imprints are recorded in the ICM and are reflected in
119: the cluster scaling relations.
120:
121: There are two important ICM properties that directly reflect the role of baryon physics.
122: The first is ICM entropy.
123: With \rosat\ and \asca\ data, Ponman et al. (2003) showed that at 0.1 $r_{200}$,
124: the ICM entropy ($K$) obeys a simple relation with the cluster temperature ($T$),
125: $K \propto T^{\sim 0.65}$, which is different from the self-similar
126: relation ($K \propto T$). Ponman et al. (2003) and Voit \& Ponman (2003) demonstrated
127: that this $K - T$ relation at 0.1 $r_{200}$ can be understood with a simple model
128: involving a fixed cooling threshold, which may be related to galaxy formation.
129: However, the details of cooling and feedback that determine this threshold and its scatter
130: are still unclear. Finoguenov et al. (2002) and Ponman et al. (2003) showed that groups and
131: clusters have significant excess entropy at $r_{500}$. Voit \& Ponman (2003) suggested
132: that smooth accretion (via e.g., galactic winds in subhalos) can boost entropy
133: higher than lumpy hierarchical accretion to produce the observed excess. However,
134: Borgani et al. (2005) show that galactic winds alone are not able to boost
135: entropy very much at radii beyond $r_{500}$. The AGN-like feedback algorithm in
136: Kay (2004) has a more substantial effect on entropy at large radii, which may
137: imply that radio outflows have an important impact on group properties.
138: Thus, it is important to robustly constrain the entropy around $r_{500}$ with
139: observations and to connect the dispersion with other group properties. The
140: Ponman et al. (2003) results rely on extrapolations based on assumed density
141: ($\beta$-model) and temperature (polytropic-model) profiles, which may not be
142: adequate and may bias the derived ICM properties (e.g., Borgani et al. 2004; Vikhlinin
143: et al. 2006, V06 hereafter). These early results can be examined with \chandra\
144: and \xmm. There has been some work on the ICM entropy of groups with the \chandra\
145: or \xmm\ data (e.g., Mushotzky et al. 2003; Sun et al. 2003; Pratt et al. 2005,
146: 2006; Mahdavi et al. 2005; Finoguenov et al. 2007). Our much larger sample allows
147: us to study the scatter in group properties. As the ICM properties are traced to at least
148: $r_{2500}$ for each group, the $K - T$ relation at $\gsim r_{2500}$ can be better
149: constrained from our studies.
150:
151: Another important ICM property is the gas fraction. Groups have low gas
152: fractions within $r_{2500}$ (0.03-0.07, V06; Gastaldello et al. 2007, G07 hereafter).
153: However, few groups have gas fraction measured to $r_{500}$. V06 and G07
154: derived gas fraction within $r_{500}$ for four low-temperature systems
155: ($kT <$ 2.7 keV). The results, 0.06 - 0.15, span a large range. Gas fraction
156: in simulated groups and clusters is directly related to the strength of cooling and
157: star formation (e.g., Kravtsov et al. 2005), and a small gas fraction in groups
158: may imply efficient cooling and star formation. The enclosed gas fraction can
159: also be modified by the AGN feedback (e.g., Puchwein et al. 2008), thus
160: bearing the imprint of the feedback history. It is therefore of great interest to know how
161: gas fractions and total baryon fractions (gas + stars) of groups at $r_{500}$
162: compare with those of clusters.
163: The $f_{gas} - T$ relation is also important for cosmology, e.g., determining
164: cosmological parameters, and using the Sunyaev-Zel'dovich flux as a proxy for
165: the cluster mass (e.g., Vikhlinin et al. 2008, V08 hereafter).
166: Besides the science related to entropy and gas fraction, better determination of
167: the group gas properties are also essential to constraining the low-mass ends of
168: other important scaling relations (e.g., $M - T, M - Y_{\rm X}$ and $c - M$).
169: $M - T$ and $M - Y_{\rm X}$ relations are essential for using clusters to study
170: cosmology (e.g., Kravtsov et al. 2006; V08),
171: while the $c - M$ relation is important for understanding the formation and evolution
172: of the dark matter halos (e.g., Buote et al. 2007).
173:
174: We started a systematic analysis of galaxy groups in the \chandra\ archive to better
175: constrain the ICM properties in groups and to better understand the difference
176: between groups and clusters. In this paper, we present the results on 43 galaxy
177: groups that appear relaxed beyond the central core. Our sample is larger
178: than in the previous work on 3 - 16 groups with \chandra\ and \xmm\ data
179: (Mahdavi et al. 2005; Pratt et al. 2005; Finoguenov et al. 2007; G07; Rasmussen \&
180: Ponman 2007). The data are homogeneously analyzed with results on temperature, entropy,
181: gas mass and total mass. The sample size also allows us to measure the scatter in
182: various interesting ICM properties.
183: We focus on the data reduction and the derived scaling relations in this
184: paper, while more extensive discussions with modeling and work on an extended
185: sample including irregular groups will be presented in future papers.
186: The group sample is defined in $\S$2. The data analysis is presented
187: in $\S$3, including spatial and spectral analysis. We especially discuss the
188: \chandra\ background and our method of background subtraction in this section and
189: Appendix. In $\S$4, we define four tiers of groups with different data coverage.
190: Different characteristic radii ($r_{2500}$, $r_{1500}$, $r_{1000}$ and $r_{500}$)
191: are reached in different tiers. We also define the system temperatures
192: in $\S$4 ($T_{500}$ etc.) and derive their empirical relations.
193: The group temperature profiles are discussed
194: in $\S$5, while $\S$6 is about the ICM entropy. In $\S$7, we discuss $M - T$ and
195: $M - Y_{\rm X}$ relations, gas fraction, concentration parameter, baryon fraction and fossil
196: groups in this sample. There are groups with signs of AGN heating and
197: groups with strong central radio sources in this sample. We discuss them in $\S$8.
198: Systematic errors are discussed in $\S$9. $\S$10 is the summary and conclusions.
199: We assumed H$_{0}$ = 73 km s$^{-1}$ Mpc$^{-1}$, $\Omega$$_{\rm M}$=0.24,
200: and $\Omega_{\rm \Lambda}$=0.76.
201:
202: \section{The group sample}
203:
204: The groups in the sample are mainly selected from the \chandra\ archive as of September, 2008.
205: We also include proprietary data on three groups from the \chandra\ program 09800349
206: (PI: Vikhlinin). We started to collect groups from several previous
207: group samples constructed from \rosat\ observations: Mahdavi et al. (2000),
208: Mulchaey et al. (2003) and GEMS (Osmond, \& Ponman 2004). We have also searched
209: for low temperature systems ($T < 3$ keV) in the on-line database
210: BAX\footnote{http://bax.ast.obs-mip.fr/}. However, most \rosat\ groups with
211: well-determined temperatures are quite nearby. The most distant group in GEMS is at
212: $z$=0.0282 (NGC~6338). As we want to constrain gas properties out to at least $r_{2500}$
213: (ideally $r_{500}$) for each system, many nearby systems in these \rosat\
214: catalogs are excluded. Thus, we have had to extensively examine the data in the
215: \chandra\ archive seeking groups with good observations out to these large radii.
216: Our selection criteria are: 1) full-frame Advanced CCD Imaging Spectrometer (ACIS)
217: data at the focal plane temperature of -120C (after Jan. 29, 2000);
218: 2) $T_{500} \lsim 2.7$ keV and a global temperature constrained to better
219: than 15\%; 3) 0.015 $<z<$ 0.13 and group emission traced to at least $r_{2500}$
220: with the \chandra\ data;
221: 4) group emission well centered around the cD and not significantly
222: elongated or disturbed beyond the group core.
223: $T_{500}$ is the temperature measured between 0.15 $r_{500}$ and $r_{500}$
224: (defined and discussed in $\S$4). The upper limit on $T_{500}$ is determined
225: from the $M_{500} - T_{500}$ relation in V06 to make sure that we are studying low-mass
226: systems with $M_{500}$ of $< 10^{14} h^{-1}$ M$_{\odot}$. We understand that
227: there is not a well-defined temperature boundary separating galaxy groups and clusters.
228: Many people may consider groups to be systems with temperatures of $<$ 2 keV.
229: Nevertheless, we refer to all the systems in our sample as groups for
230: convenient comparison with the clusters in V08. The requirement to well constrain
231: $T_{500}$ allows us to derive a temperature profile for each group.
232: The constraint on redshift enables a first cut to make sure $r_{2500}$ can
233: be reached. We used the $r-T$ relation in V06 as the first guess. It turns out
234: the derived $r-T$ relations from our work are close to V06's. With the $r_{2500} - T$
235: relation in V06, $r_{2500}$ = 11.6$'$ for an 1 keV group at $z=0.015$. An ACIS
236: pointing can reach this radius although the coverage is partial.
237: The only exception to the redshift requirement is NGC~1550 at $z=0.0124$. There
238: are two offset observations for this luminous group that allow us to reach $r_{1000}$.
239: Clearly some hotter systems at $z\gsim0.015$ may not meet our criterion to reach
240: $r_{2500}$ and are excluded after their temperatures are constrained.
241: We indeed have examined many more systems than the ones in our final sample.
242: With \chandra's superior angular resolution, almost all groups have substructures
243: around the center, and in some cases these features can be very significant and
244: striking, like the central 60 kpc of IC~1262 ($\sim 0.1 r_{500}$).
245: Nevertheless, they are included as long as the group emission is well centered
246: and regular at $\gsim 0.15 r_{500}$, where our main interests are.
247: Our science goals concerning gas entropy are also not affected by the dynamic state of the
248: group. Moreover, at least the $Y_{\rm X}$ mass proxy is still robust in unrelaxed systems
249: (Kravtsov et al. 2006).
250:
251: The \chandra\ archival search not only includes targets in the cluster and galaxy
252: categories, but also targets in the AGN category, as many radio galaxies are in
253: group environments. There are also targets selected optically. Many faint
254: systems are included, as shown by the wide range of gas entropy at 30 kpc and
255: 0.15 $r_{500}$ ($\S$6). Thus, our final sample is not much biased to the X-ray
256: luminous systems. The final sample includes 43 groups, listed in Table 1.
257: It is hardly a homogeneous or a complete sample. But the wide spread of the gas
258: densities and entropy values
259: at $r < 0.15 r_{500}$ indicates a wide luminosity range as cool cores in groups contribute
260: a large portion of their luminosities (See $\S$7.2).
261: If there were groups that are much fainter or poorer in gas than the faint systems in this sample,
262: it is difficult to constrain their gas properties with the current
263: X-ray instruments anyway. Because of the layout of ACIS, full
264: coverage at large radii (defined as $>$ 80\% coverage for the outermost spectral
265: extraction bin, as there is always area missing from extended and point sources and chip gaps)
266: is only achieved in 7 groups (Table 2).
267: However, the coverage at $r \lsim r_{2500}$ is much better. Previous analysis
268: for nearby clusters or groups with the \chandra\ data all had partial
269: coverage at large radii (e.g., V06; Rasmussen et al. 2007; V08) and our sample
270: is about 3 times bigger than these previous ones.
271:
272: \section{The data analysis}
273:
274: \subsection{\chandra\ observations \& calibration steps}
275:
276: All observations were performed with ACIS.
277: Standard data analysis was performed which includes
278: the corrections for the slow gain change
279: \footnote{http://cxc.harvard.edu/contrib/alexey/tgain/tgain.html}
280: and Charge Transfer Inefficiency (for both the FI and BI chips).
281: We investigated the light curve of source-free regions (or regions with a small
282: fraction of the source emission) to identify and exclude time intervals with particle
283: background flares, including weak flares.
284: The relevant information on the \chandra\ pointings is listed in Table 1.
285: We corrected for the ACIS low energy quantum efficiency (QE) degradation
286: due to the contamination on ACIS's Optical Blocking Filter
287: \footnote{http://cxc.harvard.edu/cal/Acis/Cal\_prods/qeDeg/index.html},
288: which increases with time and is positionally dependent.
289: The dead area effect on the FI chips, caused by cosmic rays, has also been
290: corrected. As the background subtraction is important for this project,
291: we present it in detail in the next section and Appendix.
292: We do not use any data on the S4 chip, because of the residual streaks
293: often seen at low levels (after running the CIAO tool ``DESTREAK'') and the
294: lack of the stowed background data.
295: The ``readout artifact'', seen in groups with very bright cores, is
296: also corrected (see e.g., V06). We used CIAO3.4 for the data analysis.
297: The calibration files used correspond to \chandra\ Calibration Database
298: (CALDB) 3.4.3 from the \chandra\ X-ray Center, released in March, 2008.
299: We are aware of a possible over-correction of the \chandra\ effective area
300: beyond the Iridium M-edge ($\sim$ 2 keV) in CALDB 3.4.3 and before
301: (see David's presentation in the 2007 \chandra\ Calibration Workshop)
302: \footnote{http://cxc.harvard.edu/ccw/proceedings/07\_proc/presentations/david},
303: which can bias \chandra\ temperatures to higher values, especially for hot
304: clusters ($T > 4$ keV). The difference between \chandra\ and \xmm\ temperatures
305: is also shown in Snowden et al. (2008). The calibration work to incorporate
306: this into the \chandra\ CALDB is ongoing. However, as shown in both
307: David's presentation and Snowden et al. (2008), the agreement between \chandra\ and
308: \xmm\ temperatures for $\lsim$ 4-5 keV systems is very good. The highest gas
309: temperature at any radii in our sample is $\sim$ 3.3 keV for several groups at the center,
310: while temperatures at large radii are much lower and mainly determined by
311: the iron L hump (instead of the continuum slope as for hot clusters). Thus, any
312: changes in our results from this correction should be smaller than the current
313: measurement errors and scatter.
314: In the spectral analysis, a lower energy cut of 0.4 keV (for the BI data)
315: and 0.5 keV (for the FI data) is used to
316: minimize the effects of calibration uncertainties at low energy.
317: The solar photospheric abundance table by Anders \& Grevesse
318: (1989) is used in the spectral fits. Uncertainties quoted in this paper are 1 $\sigma$.
319:
320: \subsection{Determining the local background}
321:
322: Proper background subtraction is important for the analysis in the low surface
323: brightness regions of the groups. A detailed discussion of the \chandra\ background
324: and the relevant dataset is presented in the Appendix.
325: We determine the local background based on the stowed background data
326: and modeling. The corresponding stowed background of each observation is
327: scaled according to the flux ratio in the 9.5 - 12 keV band (e.g., V05).
328: We adopt a 3\% uncertainty on the particle background (PB) normalization (5\% for BI
329: data in period E, see Appendix) in the error budgets.
330:
331: The local cosmic X-ray background (CXB) can be modeled.
332: The best fit of the hard CXB component is determined with an absorbed power
333: law with a photon index of 1.5. The absorption is determined from spectral fits
334: to the group spectra, which involves iterations and is present in $\S$3.3.
335: The soft CXB component is adequately described by two thermal components
336: at zero redshift, one unabsorbed component with a fixed temperature of 0.1 keV and
337: another absorbed component with a temperature either derived from spectral fits
338: or fixed at 0.25 keV (see Appendix). Abundances of both components are fixed at Solar.
339: We can compare our CXB model with other work. V05 used the blank sky
340: background and corrected for differences between a local control field and
341: the blank sky background. As the scaling factors of the PB are
342: close to unity in their sample, V05 ignored the correction for the unresolved hard CXB.
343: However, the correction for hard CXB becomes more important after the middle of 2004 when
344: the PB flux is 30\% - 50\% higher than that in the blank sky
345: background (Appendix). V05 found that generally a single thermal component with a
346: temperature of $\sim$ 0.2 keV and solar abundance can fit the soft CXB excess or
347: decrement well. In regions with high RASS R45 values, a second thermal
348: component with a temperature of $\sim$ 0.4 keV and solar abundance is required.
349: Humphrey \& Buote (2006) and G07
350: also used two thermal components with solar abundances to describe the soft CXB. The
351: CXB temperatures are fixed at 0.07 keV and 0.2 keV. As discussed in the Appendix,
352: Snowden et al. (2008) used three thermal components to describe the soft CXB. Thus,
353: our model of the soft CXB only has less freedom than the model by Snowden et al. (2008).
354: However, our model balances the requirement of having enough components to fit the local
355: CXB and avoiding parameter degeneracy in generally low-statistics data. First, the two
356: $\sim$ 0.1 keV components will naturally be mixed in our energy band when the
357: absorption is low. Second,
358: statistics for most of our groups at large radii are not very good, so uncertainties
359: are mainly statistical errors. As long as the background uncertainties are folded
360: into the final error budgets, very detailed modeling of the soft CXB component
361: is not crucial. Third, we test different models of the soft CXB for groups in our sample
362: for which a large group-emission-free region is available. All spectra can be well
363: fit by our two-component model, while the three-component model by Snowden et al. (2008)
364: makes little or no improvement. In fact, the soft CXB can be well fitted by a single
365: thermal component in most cases. However, the abundance usually has to be free
366: and the best-fit value is usually very close to zero. Therefore, we conclude that
367: the two-component soft CXB model is adequate and also necessary for our analysis.
368:
369: The \chandra\ observations in our sample can be classified into two categories, one with
370: regions free of group emission in the FOV, another with group
371: emission detected to the edge of the field.
372: Cool systems at higher redshifts (e.g., $z > 0.04$) usually have group-emission-free
373: regions in the off-center chips (e.g., S2 for ACIS-I observations,
374: I2/3 and S1 for ACIS-S observations). There are also two groups with adjacent
375: \chandra\ pointings for unrelated targets in coincidence (ESO~306-017 and A1692).
376: We took the following approach to look for group-emission-free regions.
377: The radial surface brightness profile is derived with the exposure
378: correction and background subtraction using the scaled stowed background.
379: Different pointings for the same group are combined.
380: The obtained surface brightness profile has group emission plus the CXB.
381: The region where CXB is more dominant than the group emission, if present, can be
382: determined from the flattened portion at the outer region of the surface brightness profile.
383: We used a power law plus a constant at large radii to determine whether a significant
384: group-emission-free region exists and the radial range of that region.
385: If such a region is found, we have a control field with only CXB.
386: In this work, the inner radius of the group-emission-free region is
387: $> 3'$ larger than the outermost radius for the group temperature profile.
388: This control field may be a single region on the S2 chip (for the ACIS-I data)
389: or two separate regions on the S1 and I2/I3 chips respectively (for the ACIS-S
390: data). Although we could simply fit the spectrum (or spectra) of this control
391: field to determine the local CXB, the statistics are generally not sufficient. Moreover,
392: the hard CXB in this control field may be larger than that in the outermost
393: bin for the group temperature profile, as the flux of the hard CXB depends on
394: point source excision. Thus, we fit the spectra of the control field and the
395: outermost bin for the group temperature profile together to better constrain the
396: local CXB. The normalizations of the soft CXB in different regions
397: are linked by the ratios of their covered solid angles of the sky, while the normalizations
398: of the hard CXB are not linked. Therefore, we can determine the local CXB in the
399: outermost bin for the group temperature profile. In this work, we assume that the
400: soft CXB is constant across the examined group area.
401: Accurate determination of the local CXB is only important for large radii.
402: The covered area at large radii tends to be in the same direction
403: from the group center so the assumption should be reasonable. There is a complication
404: that the hard CXB may be smaller in inner radial bins but it is a small effect
405: and can be corrected (also discussed later in this section).
406:
407: However, many groups in our sample have group emission detected to the edge of the
408: field, so we have to fit the CXB components with the group emission together.
409: Since these groups are limited by the \chandra\ FOV, the group emission is generally
410: still significant near the edge of the field. We fit the spectra from the two outermost
411: radial bins together, with the normalizations of the soft CXB components linked.
412: Generally in this work, we are conservative and do not
413: include the outermost bin in the temperature profile as the uncertainties are generally
414: large. However, there are a few groups that remain bright to the edge of the field (e.g.,
415: NGC~1550, A262 and MKW4). They all have S1 data, so the soft CXB component can be easily
416: separated from the group emission (e.g., Fig. 1), owing to the good response
417: of the BI chips at $<$ 1 keV. As the temperatures of the cool gas can be well
418: constrained from the iron hump centroid, group emission can be robustly separated from the
419: soft CXB even to the edge of the field.
420: Thus, we derive temperature profiles to the edge of the field for these groups.
421:
422: As we separate the PB from the CXB, the absolute CXB flux can be
423: derived in each group field. In Table 2, we list the flux of the local background
424: components, for both the soft CXB and the hard unresolved CXB (in the outermost bin).
425: We expect the derived soft CXB is on average
426: correlated with the RASS R45 flux in the surrounding area (excluding the source region,
427: see Table 2), which is true as shown in Fig. 2.
428: One should be aware of the uncertainty in the R45 flux, as the RASS soft X-ray background
429: maps have 0.2 deg pixels and the various uncertainties combined are not small (e.g.,
430: variable SWCX emission).
431: There are also uncertainties related to cross-calibration and
432: cosmic variance of both the soft and hard CXB, so a detailed one-to-one comparison
433: is hardly meaningful.
434: The derived 2- 8 keV flux of the hard CXB component depends on the limiting flux
435: of the observation. As discussed in Appendix, there is an empirical
436: relation between the limiting flux for point sources and the average
437: 2 - 8 keV flux of unresolved hard CXB (K07). We estimated the average limiting flux
438: of the group observations in the outermost spectral bin.
439: The regions we used to control the local background generally have an area of about one ACIS
440: chip. As shown in Hickox \& Markevitch 2006 (HM06 hereafter) and K07, the cosmic
441: variance in this angular scale (depending
442: on the two-point angular correlation function of point sources) is 20\% - 30\% and
443: the total hard CXB flux has $\sim$ 10\% uncertainty. Thus, the expected hard CXB flux
444: from the empirical relation is only meaningful in an average sense. We use the CIAO
445: tool MKPSF to generate several PSFs (at an energy of 1.4 keV) in the outermost
446: spectral bin. We then derive the 90\% enclosed power aperture and measure the 3$\sigma$
447: limits at these regions. Their average is taken as the estimate for the limiting flux
448: for point sources. Compared with the average growth curve determined by K07 (their Fig. 19), we have an estimate of the unresolved hard CXB, which is also listed in Table 2.
449: We can see from Table 2 that the general agreement is quite good, while the
450: uncertainties for the hard X-ray CXB are much larger than the difference.
451: It is also true that the limiting flux for point sources depends on the off-axis angle.
452: The \chandra\ limiting flux changes little within the central 6-7$'$ from the
453: aimpoint, but increases rapidly beyond 10$'$. This can be corrected from the slope
454: of limiting flux vs. unresolved hard CXB in K07 and the absolute flux in the
455: outermost bin. The correction is small, also because the errors of the hard CXB are not small.
456:
457: Besides the 3\% uncertainty on the PB mentioned above (5\% for BI data in period E),
458: we also included the following error budgets and added them (and the PB uncertainty)
459: to the statistical uncertainties of temperatures, abundances and surface brightness
460: in quadrature. For the hard CXB, we set the photon index at 1.1 and 1.9 (e.g., HM06)
461: and repeat the analysis to estimate the local background in those situations. For the
462: soft CXB component, if the temperature of the hotter component cannot be derived from
463: the data and is fixed at 0.25 keV, we change it to 0.2 keV and 0.3 keV and repeat the
464: analysis to estimate the local background in those situations. Thus, we are
465: conservative in the uncertainties for the local CXB.
466:
467: \subsection{Spectral analysis and deprojection}
468:
469: Once the local CXB is determined, we proceed to derive the projected temperature
470: profile. The radial bins for temperature measurement are decided from the outermost
471: bin, by requiring the temperature to be constrained to better than 30\% (1 $\sigma$
472: with all error budgets) and $r_{out} / r_{in}$ = 1.3 - 1.75. After determining the
473: outermost bin, the inner bins are determined progressively with $r_{out} / r_{in}$ =
474: 1.25 - 1.6. Point sources and sub-clumps are excluded. The absorption column is
475: determined from the MEKAL fit to the integrated spectrum between $0.1 r_{500}$
476: and $0.4 r_{500}$. In this process, we also tried the VMEKAL model with extra free
477: parameters of O, Ne, Mg, Si, S, Fe and Ni abundances, as a lower O/Fe ratio may
478: cause excess absorption (Buote et al. 2003). This process and the determination of
479: the local CXB are done in iterations to obtain the final value of the absorption.
480: If the derived absorption is consistent with the Galactic absorption from the HI
481: survey (Table 2) within 1 $\sigma$, the Galactic value is used. Only when both MEKAL
482: and VMEKAL fits show excess absorption, the X-ray absorption from the MEKAL fit
483: (always consistent with the VMEKAL fit within 1 $\sigma$) is adopted (Table 2).
484: About 44\% of groups show excess absorption relative to the Galactic value
485: and indeed we find \iras\ 100 $\mu$m enhancement in many of these groups (see Appendix
486: for discussions of some groups). We also examined the absorption variation with
487: radius in each group. Beyond the central 20 kpc, no significant absorption variation
488: is found for any group so we used a fixed absorption column for each group. In several
489: cases, we observed an absorption increase within the central 20 kpc, but this analysis
490: is complicated by the possible multi-phase gas around the center. Nevertheless,
491: it has little effect on any of our results. We also discuss the systematic error related to
492: absorption in $\S$9.
493:
494: We used the MEKAL model to fit the spectra of the group emission. The free parameters
495: are temperature, abundance and normalization, once the absorption is determined.
496: For spectra around the center, we also include a component to account for the LMXB
497: emission from the cD galaxy. The LMXB component is represented by a power law with
498: an index of 1.7, within $D_{25}$ aperture obtained from HyperLeda.
499: The total LMXB luminosity is fixed from the $L_{\rm X} - L_{\rm Ks}$ relation
500: derived in Kim \& Fabbiano (2004), where $L_{\rm Ks}$ is the total $K_{\rm s}$ band
501: luminosity of the cD from 2MASS. We also assume that the LMXB emission follows
502: the $K_{\rm s}$ band light (see also Gilfanov 2004).
503:
504: The projected group temperature profiles are shown in Fig. 3 - 5.
505: We used the algorithm determined by Vikhlinin (2006) to derive the deprojected
506: temperature profiles in a parametric way, which was first applied in V06. The required
507: inputs are the three-dimensional (3D) or deprojected profiles
508: of gas density and abundance, and the projected temperature profile.
509: V06 simply used the projected abundance profile as their sample is dominated
510: by hot clusters. However, the emissivity of $\lsim$ 2 keV plasma is sensitive
511: to the chemical abundances so the 3D abundance profile is required for our work.
512: We applied the non-parametric geometrical deprojection (summarized in e.g.,
513: Pizzolato et al. 2003; G07) to derive
514: deprojected abundances in wider radial bins, generally merging 2 - 3 adjacent
515: bins for the temperature profile. The deprojected abundance is an emission-weighted
516: average in each bin so an effective radius in each bin is required.
517: We define the effective radius as the emissivity-weighted radius in each bin,
518: where the plasma emissivity, $\varepsilon (T, Z, r)$, depends on the 3D profiles of
519: temperature, abundance and density.
520:
521: Since the determination of the 3D profiles of temperature, abundance and density
522: depends on each other, iterations have to be done to derive the best-fit 3D profiles.
523: Usually at most three iterations are required to stabilize the best-fits of these
524: profiles.
525: We assumed the following 3D abundance profile:
526:
527: \begin{eqnarray}
528: Z(r) = Z_{0} + Z_{1} exp (-(r/r_{Z})^{\alpha_{\rm Z}})
529: \end{eqnarray}
530:
531: This simple form can fit the abundance profiles of all groups in the sample.
532: An example is shown in Fig. 6 that represents the best-constrained 3D abundance
533: profile in this sample (thus the most difficult to fit, as the errors are the smallest).
534: Most groups in this sample only have 3D abundances constrained
535: in 3 - 5 bins with larger errors so good fits can always be achieved with this simple
536: function. Once the best-fits of all profiles are achieved, we applied 1000 Monte Carlo
537: simulations to address the uncertainties of the 3D abundance profile. The simulation is
538: realized by scattering the 3D abundance profile according to the measurement
539: errors. In this process,
540: to be conservative, we also include an uncertainty of 10\% of the bin size on the
541: effective radius of each bin for the abundance profile.
542: Thus, besides the best-fit 3D abundance profile, we have 1000 simulated profiles to
543: cover the error ranges, which will be used to estimate
544: the uncertainties of the 3D temperature and density profiles.
545:
546: We used the same form of the 3D temperature profile as used in V06:
547:
548: \begin{eqnarray}
549: T(r) = \frac{T_{0}(r/r_{cool})^{a_{cool}}+T_{min}}{(r/r_{cool})^{a_{cool}}+1}
550: * \frac{(r/r_{t})^{-a}}{[1+(r/r_{t})^{b}]^{c/b}}
551: \end{eqnarray}
552:
553: The exceptions are A1139, A1238 and A2092, which are faintest 2 - 3 keV systems
554: in this sample so the errors on temperatures are large. Their 3D temperature
555: profiles are modeled as: $T(r) = T_{0} [1+(r/r_{t})^{a}]^{b/a}$, which fits their
556: temperature profiles very well (Fig. 4 and 5).
557: The inner boundary of the fit is 5 - 20 kpc, depending on the quality of the fits
558: at the center. As we focus on properties at large radii, the detailed choice of the
559: inner boundary (5 - 20 kpc) is not a concern.
560: Because the sizes of bins at large radii are not small, each bin is further divided into 6
561: sub bins. The temperature modeling is done at the center of each sub bin, only added up
562: (with the algorithm by Vikhlinin 2006) at the end to obtain the expected
563: temperature for each original bin of spectral analysis. In this way, the binsize effect
564: is minimized and we don't need to worry about the accurate determination of effective
565: radius for each radial bin, as most previous work had to address.
566: The uncertainty of the 3D temperature profile
567: is derived from 1000 Monte Carlo simulations. In each simulation, the measured projected
568: temperature profile is scattered according to the measurement errors, and a new simulated
569: abundance profile is input. The density profile is fixed at the best-fit value, as the
570: temperature error is the dominant error source.
571: The reconstructed 3D temperature profiles, with 1 $\sigma$ uncertainties from simulations,
572: are also shown in Fig. 3-5. There are groups where a central corona exists
573: (e.g., Sun et al. 2007), so naturally the temperature gradient at 5 - 20 kpc
574: is large in these cases (e.g., A2462, A160, ESO~306-107, HCG~51 and NGC~6269). Sometimes
575: the 3D temperature appears too low within 5 - 10 kpc radius (e.g., A160 and NGC~6269),
576: which however affects none of our results in this work as core properties are
577: excluded.
578:
579: \subsection{Gas density}
580:
581: We extract the surface brightness profile in the 0.7 - 2 keV band, as suggested by
582: V06 to avoid the 0.6 keV hump in the soft CXB. Point sources and chip gaps
583: are excluded. The scaled stowed background is subtracted so only the group emission
584: and the local CXB is left. Some previous analysis on a surface brightness profile involved
585: the correction of a single exposure map.
586: However, the \chandra\ exposure map is energy and position dependent.
587: It is not accurate to use a single exposure map (even one convolved with the group
588: spectrum) as there is spectral variation in a group. The most accurate approach is
589: to generate response files for each bin of the surface brightness and convert the
590: raw count rate (without corrections on vignetting and other response) to density, from
591: the derived 3D temperature and abundance profiles and response files.
592: To achieve that, we use XSPEC to generate an MEKAL emissivity matrix that depends on temperature,
593: abundance and position of the radial bin for surface brightness (or response files there),
594: once the absorption is determined. The ranges for
595: temperatures and abundances cover the observed ranges for any particular group.
596: This emissivity matrix provides the conversion factor needed to transfer the
597: observed count rate to the emission measure.
598: We assume the following density model (with 11 free parameters):
599:
600: \begin{eqnarray}
601: n_{e}^{2} = n_{0}^{2} \frac{(r/r_{c})^{-\alpha}}{[1+(r/r_{c})^{2}]^{3\beta-\alpha/2}}
602: \frac{1}{[1+(r/r_{s})^{\gamma}]^{\delta/\gamma}} \nonumber \\
603: + n_{02}^{2} [1-(r/r_{c2})^{\alpha_{2}}]^{3\beta_{2}/\alpha_{2}} (r < r_{c2})
604: \end{eqnarray}
605:
606: This model combines the profile used in V06 and the profile proposed by Ettori (2000) for
607: cool cores. We find that this model provides very good fits for all groups in our sample.
608: The density profile is then converted to the emissivity profile from the emissivity matrix.
609: The emissivity profile is then projected
610: along line of sight with the formula of geometrical deprojection (e.g.,
611: McLaughlin 1999). The local CXB can be added later with the known CXB spectra and the radial
612: set of the response files. The resulting surface brightness profile can be compared with
613: the observed one.
614: To avoid the binsize effect that may especially affect outskirts as wider bins are required
615: there, we further divide bins within 100 kpc to 3 sub bins and bins beyond the central
616: 100 kpc to 6 sub bins. The conversion and deprojection is done in these sub bins.
617: They are later merged to compare with the observed profile.
618: The density errors are estimated from 1000 Monte Carlo simulations, with 1000 corresponding
619: simulated abundance and temperature profiles as inputs.
620:
621: This method is similar to that used in V06, but we use the 3D temperature and abundance
622: profiles to convert count rates to density. Like V06,
623: we also analyzed the \rosat\ PSPC pointed observations for the purpose of constraining
624: gas density profiles at large radii. The inclusion of the PSPC data is especially
625: useful to $z<0.04$ groups. We used the software developed by Snowden et al.
626: (1994) to produce flat-fielded PSPC images in the 0.7 - 2 keV band (more exactly, R567
627: bands, or PI channels of 70 - 201). The images are further analyzed as described in
628: Vikhlinin et al. (1999).
629: Seventeen groups have sufficiently long PSPC data (listed in Table 1 with effective exposure)
630: and we included the PSPC surface brightness profile in the modeling of the density profile.
631: The probed outermost radius of the PSPC data for each group is listed in Table 2.
632: We only used the PSPC surface brightness profiles outside the central
633: 3$'$ to avoid the PSF correction in the core. In all cases, there is good
634: agreement between the \chandra\ and PSPC surface brightness data.
635: One example is shown in Fig. 7.
636:
637: \section{Characteristic radii and definition of system temperatures}
638:
639: As stated above, we generated 1000 simulated profiles of the 3D temperature,
640: abundance and density profiles, which cover the ranges of the measurement errors.
641: Assuming hydrostatic equilibrium,
642: each set of the 3D temperature and density profiles determine a set of characteristic
643: radii of the group ($r_{500}$,
644: $r_{1000}$, $r_{1500}$ and $r_{2500}$). The profiles of interesting quantities, e.g., total
645: mass, gas mass and entropy, and their values at the characteristic radii are also
646: determined. For each quantity, the peak in the distribution from 1000 simulations
647: defines the most probable value (or ``best fit''). The 1$\sigma$ errors at two sides
648: are estimated by determining the regions that contain 68\% of realizations at each side.
649: The results are listed in Table 3 and 4.
650: $r_{500}$ is the basic characteristic radius that is used in most scaling relations in
651: this work. However, it cannot be robustly determined for all groups in our sample so
652: the $r_{500} - T$ scaling relation is required.
653: As the radial coverage of data in this sample differs from one group to
654: another, we define four tiers of groups for which group properties (entropy, mass,
655: $Y_{\rm X}$, and gas fraction) are derived to different characteristic radius.
656:
657: \begin{itemize}
658: \item Tier 1: groups with surface brightness (including the PSPC profile) derived at
659: $> 2\sigma$ levels to $> r_{500}$ (note we are conservative to estimate the errors of
660: the local CXB, see notes of Table 2) and a temperature profile derived to
661: $>$ 80\% of $r_{500}$. Eleven groups in our sample are in this tier and their
662: temperature profiles are derived to 81\% - 117\% of $r_{500}$ with a median of 97\%.
663: Groups in this tier have density profiles derived to $> r_{500}$ so the density
664: gradient at $\sim r_{500}$ is constrained. Temperature profiles are derived
665: beyond $r_{500}$ or sufficiently close to $r_{500}$ for reasonable extrapolation.
666: We note that in V06, there are three clusters and groups (A383, MKW4 and A1991)
667: with temperature profiles derived only to 73\% - 89\% of $r_{500}$, but properties
668: at $r_{500}$ are still derived from extrapolation. Thus, our criteria are similar
669: to those in V06 and V08.
670: \item Tier 2: groups with surface brightness and temperature profiles all derived
671: to at least $r_{1000}$ but not in tier 1. Twelve groups are in this tier and their temperature
672: profiles are derived to 68\% - 87\% of $r_{500}$ with a median of 77\%.
673: We consider that groups in this tier have properties well constrained close to
674: $r_{500}$ so $r_{500}$ is determined in these groups. Group properties at $r_{500}$
675: are also derived from extrapolation. However, in figures of scaling relations at
676: $r_{500}$, tier 2 groups are always marked differently from tier 1 groups. The fits
677: with or without them are both listed.
678: \item Tier 3: groups with surface brightness and temperature profiles derived
679: to at least $r_{1500}$ but not in tiers 1 and 2. Eleven groups are in this tier.
680: Nine of them have temperature profiles derived to 52\% - 72\% of $r_{500}$, which is
681: close to $r_{1000}$ ($\sim 0.73 r_{500}$ in this sample). Thus, group properties
682: at $r_{1000}$ are also derived from extrapolation. A1238 and RXJ~1206-0744 are
683: put in this tier as the temperature and density errors are large although $r_{1000}$
684: is reached in both cases.
685: \item Tier 4: nine other groups with surface brightness and temperature profiles
686: derived to at least $r_{2500}$. Group properties at $r_{1500}$ are also derived from
687: extrapolation in this tier.
688: \end{itemize}
689:
690: Previous X-ray work on clusters and groups often defined $<T>$, which is the
691: emission-weighted temperature within a certain aperture. As definitions
692: are generally different, it is necessary to use an unified definition that
693: is easily accessible from observations. In this work, we define the system temperature as:
694:
695: \begin{itemize}
696: \item $T_{500}$: the spectroscopic temperature measured from the integrated spectrum
697: in the projected 0.15 $r_{500}$ - $r_{500}$ annulus.
698: \end{itemize}
699:
700: We derive $T_{500}$ by integrating the 3D temperature profile from 0.15 $r_{500}$ to
701: 1.6 $r_{500}$ (or $\sim r_{180}$), in an annular cylinder with projected radii
702: of 0.15 $r_{500}$ - $r_{500}$ along the line of sight, with the algorithm by
703: Vikhlinin (2006). The choice of the outer radius in the 3D integration little
704: affects $T_{500}$ as it is emission-weighted. The same definition of the system
705: temperature was also used in Nagai, Kravtsov \& Vikhlinin (2007b, NKV07 hereafter),
706: Maughan (2007) and V08.
707: This definition excludes the central region, where a cool core or a locally heated
708: region may exist. Indeed the group temperature profiles are much more similar
709: at $r > 0.15 r_{500}$ ($\S$5 and Fig. 8).
710: This temperature can also be directly derived from data, provided that the full coverage of
711: $r_{500}$ is achieved (generally not the case for groups in our sample).
712: Similarly, we can define $T_{1000}$, $T_{1500}$ and $T_{2500}$, with the projected
713: inner boundary always at 0.15 $r_{500}$ and the projected outer boundary at $r_{1000}$,
714: $r_{1500}$ and $r_{2500}$ respectively (3D temperature profile still integrated to
715: 1.6 $r_{500}$). Their empirical relations can also be determined.
716:
717: As we cannot derive $r_{500}$ for groups in tiers 3 and 4, the $r_{500} - T_{500}$
718: relation needs to be determined. The $r - T$ relation is just a manifestation
719: of the $M - T$ relation so it is presented in $\S$7.1.
720: We also derived the average ratios of the characteristic radii from 23 tier 1 and 2
721: groups: $r_{1000}$ / $r_{500}$ = 0.741$\pm$0.013, $r_{1500}$ / $r_{500}$ = 0.617$\pm$0.011
722: and $r_{2500}$ / $r_{500}$ = 0.471$\pm$0.009, which are about what are expected for
723: the average $c_{500}$ of this sample ($\sim 4.2, \S$7.3), 0.727, 0.599 and 0.465
724: respectively (assuming an NFW profile).
725:
726: We also need empirical relations between $T_{500}$ and $T_{1500}$ (or $T_{2500}$)
727: to estimate $T_{500}$ for groups in tiers 3 and 4.
728: For 23 groups in the first and second tiers, we found:
729:
730: \begin{eqnarray}
731: kT_{500} / kT_{1500} = 0.93 \pm 0.02
732: \end{eqnarray}
733:
734: This ratio is not temperature dependent in our sample and the fit is very
735: good ($\chi^{2}$/dof = 6.6/22). If we only fit 11 groups in the first tier,
736: the ratio is the same. We notice that V08 derived a similar empirical
737: relation between $T_{500}$ and temperature measured at 0.15 - 0.5 $r_{500}$ (close
738: to our $T_{2500}$). V08 also included a linear term as they explored a wider
739: temperature range. We also derived $kT_{500} / kT_{2500} = 0.89 \pm 0.02$,
740: but the scatter is larger as shown by the poorer fit ($\chi^{2}$/dof = 18.6/22).
741: The empirical relation between $kT_{500}$ and $kT_{1500}$, combined with the
742: $r_{500} - T_{500}$ scaling, allows us to estimate $r_{500}$ for groups in tiers 3
743: and 4, in a few iterations. We specifically used the $r_{500} - T_{500}$ relation
744: from 23 tier 1 and 2 groups and 14 clusters in V06 and V08 (The fourth row of Table 6).
745: For reference, we also give the best fit of the $r_{2500} - T_{500}$ relation
746: for all 43 groups in this sample:
747: $(E(z) r_{2500} / {\rm 155\pm4 h^{-1} kpc}) = (T_{500} / \rm 1 keV)^{0.520\pm0.040}$,
748: which is similar to the V06 result,
749: $(E(z) r_{2500} / {\rm 146\pm3 h^{-1} kpc}) = (T_{500} / \rm 1 keV)^{0.547\pm0.020}$.
750:
751: \section{Temperature profiles}
752:
753: Scaled temperature profiles for these groups are shown in Fig. 8, in logarithmic
754: and linear scales. We scale temperatures with $T_{2500}$, which is robustly
755: determined for each group. While the scatter within the central 0.15 $r_{500}$
756: is large, the group temperature profiles are more similar beyond 0.15 $r_{500}$,
757: with a declining slope similar to that predicted from simulations (e.g., Loken et al. 2002).
758: From 0.15 $r_{500}$ to $\sim r_{500}$, we can fit the projected temperature profiles
759: with this simple form:
760:
761: \begin{eqnarray}
762: T/T_{2500} = (1.22\pm0.02) - (0.79\pm0.04) r/r_{500}
763: \end{eqnarray}
764:
765: Interestingly, V05 derived an average form of $T/<T> = 1.22 - 1.2 r/r_{180}$
766: (0.125 $< r/r_{180} <$ 0.6) for 13 systems in their sample, which is very similar
767: to ours as $r_{500} \sim 0.62 r_{180}$.
768: If we use a similar form as used in Loken et al. (2002), we can also fit the
769: projected temperature profiles at 0.15 $r_{500}$ - $r_{500}$ with this form:
770:
771: \begin{eqnarray}
772: T/T_{2500} = (1.37\pm0.03) (1 + r/r_{500})^{-(1.34\pm0.21)}
773: \end{eqnarray}
774:
775: Based on \asca\ data, Markevitch et al. (1998) first suggested that temperature
776: profiles of clusters are self-similarly declining with radius. This result
777: was later confirmed by De Grandi \& Molendi (2002) with the \sax\ data, and
778: by V05 with the \chandra\ data, and by Piffaretti et al. (2005),
779: Pratt et al. (2007) and Leccardi \& Molendi (2008, LM08 hereafter) with the \xmm\
780: data. Self-similarly declining temperature profiles
781: are also generally observed in simulations (e.g., Loken et al. 2002; Borgani et al.
782: 2004; Kay et al. 2004). The self-similar decline in groups was also suggested
783: by Sun et al. (2003), G07 and Rasmussen \& Ponman (2007).
784: Our group sample is the largest one so far with detailed studies.
785: As shown in Fig. 8, group temperature profiles are generally self-similar with
786: a slope consistent with that in simulations, although there is some scatter.
787: We also combine all data points to make a mean temperature profile (Fig. 9).
788: Fig. 9 also plots the mean temperature profile of hot clusters at $z$=0.1-0.3
789: from LM08, as well as the mean temperature profile
790: of 1-3 keV systems from the Borgani et al. (2004) simulations.
791: We do not plot the mean temperature profiles from De Grandi \& Molendi (2002),
792: V05 and Pratt et al. (2007), but they are close to LM08's (Fig. 21 in LM08).
793: Clearly, the group temperature profiles are more peaky than those of clusters,
794: starting to decline at $\sim 0.2 r_{500}$, where the mean cluster temperature
795: profile is still flat. This difference was also noticed by V06.
796: The group temperature profiles actually agree more with simulations (e.g., Loken et al. 2002;
797: Borgani et al. 2004), which generally have problems explaining the flat
798: temperature profiles of clusters at $0.15 r_{500} - 0.3 r_{500}$ (e.g., Borgani et al. 2004).
799:
800: \section{Entropy}
801:
802: The ICM entropy, defined as $K = T / n_{e}^{2/3}$, is a fundamentally important
803: quality of the ICM as summarized in Voit (2005). Entropy
804: records the thermodynamic history of the ICM, as non-gravitational processes (e.g.,
805: AGN heating, cooling and star formation) deviate the entropy relations from the
806: self-similar relations determined with only gravity and accretion shocks
807: (e.g., Ponman et al. 1999; Voit \& Bryan 2001).
808: Voit et al. (2005, VKB05 hereafter) derived the baseline ICM
809: entropy profile from simulations in the absence of non-gravitational processes,
810: $K(r) / K_{\rm 200, adi} = 1.32 (r/r_{200})^{1.1}$.
811: $K_{\rm 200, adi}$ is an entropy scale of a non-radiative cluster at $r_{200}$ (VKB05).
812: This baseline entropy profile can be compared with our observational results
813: to measure the impact of non-gravitational processes on the ICM.
814: With a weighted average of $\sim$ 4.2 for $c_{500}$ in this sample ($\S$7.3),
815: $r_{500}/r_{200}$ = 0.669 for an NFW mass profile and the baseline entropy relation is converted to:
816: $K(r) / K_{\rm 500, adi} = 1.40 (r/r_{500})^{1.1}$ (note that we use $K_{\rm 500, adi}$
817: here for the adiabatic entropy scale at $r_{500}$, see the definition of $K_{\rm 200, adi}$
818: in VKB05 and Voit 2005).
819: For the cosmology assumed in this paper and a baryon fraction ($f_{\rm b}$) of 0.165
820: (Komatsu et al. 2008),
821:
822: \begin{equation} \label{K500}
823: K_{\rm 500, adi} = 342 \hspace{0.16cm} {\rm keV cm^{2}} (\frac{M_{500}}{10^{14} M_{\odot}})^{2/3} E(z)^{-2/3} h_{73}^{-4/3}
824: \end{equation}
825:
826: Note that $K_{\rm 500, adi}$ has a different $h$ dependence from the observed entropy
827: ($K \propto h_{73}^{-1/3}$).
828: From the derived temperature and density profiles, we obtained entropy profiles
829: for each group. The best-fit scaled entropy profiles are shown in Fig. 10.
830: Substantial scatter is present within 0.3 - 0.4 $r_{500}$.
831: As Fig. 10 is basically a plot showing the large scatter of the density profiles,
832: it is clear that there is a wide range of X-ray luminosities in this sample,
833: which is also implied by the $K - T$ relations shown next.
834: We discuss the ICM entropy of these groups in the following three sections.
835:
836: \subsection{$K - T$ relations}
837:
838: We first examined the $K - T$ relations at 30 kpc, 0.15 $r_{500}$, $r_{2500}$,
839: $r_{1500}$, $r_{1000}$ and $r_{500}$ (Fig. 11). Entropy values at these radii are:
840: $K_{\rm 30 kpc}, K_{\rm 0.15 r500}, K_{2500}, K_{1500}, K_{1000}, K_{500}$
841: respectively (note the difference between $K_{500}$ and $K_{\rm 500, adi}$
842: defined earlier). We chose 30 kpc to represent the entropy level around the BCG.
843: Previous studies often discussed entropy at 0.1 $r_{200}$ (e.g., Ponman et al. 2003),
844: which is about 0.15 $r_{500}$. We also performed the BCES (Y$|$X) regression
845: (Akritas \& Bershady 1996) to these relations and the results are listed in Table 5.
846: The entropy values at 30 kpc radius hardly show any correlation with the system
847: temperature. The $K - T$ relation is stronger at 0.15 $r_{500}$, but the
848: intrinsic scatter is still substantial. The large scatter of gas entropy
849: around the center has been known for a while (e.g., Ponman et al. 2003).
850: We also examined the connection between the luminosity of the
851: central radio source (see Table 1) and the
852: entropy scatter at 0.15 $r_{500}$. No correlation is found, before or after the
853: temperature dependence of entropy is removed. However in $\S$7.2, we show that
854: the entropy scatter at 0.15 $r_{500}$ is tightly correlated with the scatter of
855: the gas fraction within $r_{2500}$.
856:
857: Despite large entropy scatter from the core to at least 0.15 $r_{500}$,
858: the $K - T$ relations are tighter at $r_{2500}$ and beyond.
859: We derived the intrinsic scatter in these relations,
860: using the method described in Pratt et al. (2006) and Maughan (2007). As shown
861: in Fig. 11, the intrinsic scatter reduces to 10\% at $r_{2500}$ and stays at that
862: level to $r_{500}$. As the derived intrinsic scatter decreases with
863: increasing measurement errors, could this trend be due to the increasing
864: measurement errors with radius? To answer this question,
865: we compared $K_{2500} - T_{2500}$ and $K_{0.15 r500} - T_{2500}$ relations.
866: If we use the relative measurement errors of $K_{0.15 r500}$ for the corresponding
867: $K_{2500}$, the intrinsic scatter only increases to 12\%. Similarly, if we use the
868: relative measurement errors of $K_{2500}$ for the corresponding $K_{0.15 r500}$, the
869: intrinsic scatter only decreases to 28\%. Therefore, the significant tightening of the
870: $K - T$ relation from 0.15 $r_{500}$ to $r_{2500}$ is not caused by different measurement
871: errors. Similarly, we find that the intrinsic scatter in entropy at radii from
872: $r_{2500}$ to $r_{1000}$ is affected little by switching measurement errors.
873: The $K_{500}$ values have large measurement errors and half of systems
874: are extrapolated from $r_{1000}$. Clearly more groups with $K_{500}$ robustly determined
875: are required, but the intrinsic scatter in the current $K_{500} - T_{500}$ relation is
876: consistent with the level from $r_{2500}$ to $r_{1000}$.
877: Thus, groups behave more regularly from $r_{2500}$ outward than inside 0.15 $r_{500}$.
878:
879: We also include 14 clusters from V06 and V08 for comparison and to constrain the
880: $K - T$ relations in a wider temperature range (Fig. 12 and Table 5). The measured
881: slopes increase from $\sim$ 0.50 at 0.15 $r_{500}$ to $\sim$ 1 at $r_{500}$, mainly
882: caused by the excess entropy of groups at their centers.
883: The slope we find at 0.15 $r_{500}$, 0.494$\pm$0.047, is consistent with what
884: Pratt et al. (2006) found at 0.1 $r_{200}$ for 10 groups and clusters with the \xmm\
885: data, 0.49$\pm$0.15, but smaller than what Ponman et al. (2003) found at 0.1 $r_{200}$
886: with \rosat\ and \asca\ data, $\sim$ 0.65. We however caution that the V08 cluster sample
887: lacks non-cool-core clusters. Pratt et al. (2006) also gave the $K - T$ relation at
888: 0.3 $r_{200}$, which is about $r_{2500}$. The slope they found, 0.64$\pm$0.11, is
889: also consistent with our result, 0.740$\pm$0.027.
890: At $r_{500}$, the slope we found is consistent with the expected value from the
891: self-similar relation (1.0).
892: Interestingly, the derived $K_{500} - T_{500}$ relation agrees very well with the
893: NKV07 simulations with cooling + star-formation (Fig. 12), although the agreement
894: is progressively worse with decreasing radius. The tightening of the $K - T$ relation
895: at $r_{2500}$ and beyond was also reported in the NKV07 simulations, but the predicted
896: $K_{2500} - T$ relation by NKV07 lies below all groups in our sample (Fig. 11 and 12).
897: The NKV07 simulations with cooling + star-formation achieve entropy amplification
898: with strong condensation to drop dense materials out of the X-ray phase. The
899: resulting stellar fraction is about twice the observed value as too much
900: material has cooled. If the star formation is suppressed with more efficient
901: feedback, entropy is lower in the NKV07 simulations, between the results from the
902: simulations with cooling + star-formation
903: and the non-radiative simulations, which further disagrees with observations.
904: It seems that the most challenging task is to explain the excess entropy of groups at $r_{2500}$.
905: As shown in Borgani \& Viel (2008), simply increasing entropy floor from pre-heating
906: produces too large voids in the Lyman-$\alpha$ forest.
907: Thus, it is still an open question on how to generate enough entropy in groups
908: from the center to $r_{500}$ and still preserve other relations like the condensed
909: baryon fraction and properties of the Lyman-$\alpha$ forest.
910:
911: \subsection{Entropy ratios}
912:
913: We derived the ratios of the observed entropy to the baseline entropy
914: from VKB05 at $r_{500}$, $r_{1000}$ and $r_{2500}$ (Fig. 13).
915: For comparison, we also include the entropy ratios
916: for 14 clusters from V06 and V08. The entropy ratios are always larger than or comparable
917: to unity at all radii. The average ratio decreases with radius, for both groups and
918: clusters, although the decrease is more rapid in groups. For groups, the weighted
919: mean decreases from 2.2 at $r_{2500}$ to 1.57 - 1.60 at $r_{500}$, while the mean
920: decreases only $\sim$ 9\% for clusters. Thus, there is still a significant entropy excess
921: over the VKB05 entropy baseline at $r_{500}$, in both groups and clusters. The weighted mean
922: entropy of groups at $r_{500}$ is $\sim$ 18\% larger than that of clusters. If we consider
923: only the $T_{500} \lsim 1.4$ keV groups, the weighted mean for 7 groups is 1.68.
924: Similar studies have been done with \asca\ and \rosat\ data (Finoguenov et al. 2002;
925: Ponman et al. 2003). When the observed entropy values at $r_{500}$ are scaled
926: with $M_{500}^{-2/3}$ (which is proportional to $K_{\rm 500, adi}^{-1}$), both works
927: found that groups have on average twice the scaled entropy of clusters. Ponman et al. (2003)
928: further concluded that the excess entropy is observed in the full mass range.
929: Thus, the results from this work, V06 and V08 confirm the previous finding of the excess
930: entropy at $r_{500}$ over the full mass range, but the magnitude of the excess is smaller
931: and the difference between groups and clusters is also smaller than previous results,
932: $\sim$ 15\% - 20\% vs. $\sim$ 100\% in Finoguenov et al. (2002) and Ponman et al. (2003).
933:
934: As pointed out by Pratt et al. (2006), the magnitude of the excess may be
935: affected by the systematics of the $M - T$ relation, or how robust the hydrostatic
936: equilibrium (HSE) mass is. The NKV07 simulations show that the HSE mass is systematically
937: lower than the real mass and the difference is biggest in low-mass systems. The real mass
938: is 45\% higher than the HSE mass for 1 keV groups, while the difference reduces to
939: 13\% for 10 keV clusters. The best-fit $M_{500} - T_{500}$ relation from this work and
940: V06 ($\S$7.1) is close to the relation for the HSE mass in NKV07. Thus, if this
941: bias is real, the actual entropy ratios at $r_{500}$ are smaller. For 1 keV groups,
942: a 45\% higher total mass means a 13\% larger $r_{500}$. For a typical entropy slope of
943: 0.7 in this sample ($\S$6.3), the entropy ratio at $r_{500}$ is reduced by 17\%. Similarly
944: for 5 keV clusters, the entropy ratio at $r_{500}$ is reduced by 8\% for an average entropy
945: slope of 0.9 at large radii. Thus, this bias can explain only part of the entropy excess
946: observed. The entropy baseline we adopted in this work is from the SPH simulations,
947: while the AMR simulations produce a baseline with ~ 7\% higher normalization (VKB05),
948: likely because of its better capability to catch shocks. However, the entropy excess
949: for groups at $r_{500}$ remains significant.
950:
951: There are several mechanisms to achieve entropy amplification at $r_{500}$.
952: The first idea relies on modification of accretion, without extra non-gravitational
953: processes (Ponman et al. 2003; Voit et al. 2003; Voit \& Ponman 2003).
954: If preheating or feedback in small subhalos that are being accreted can eject gas out
955: of the halo and thicken the filaments significantly, the accretion may be smoother
956: than the lumpy accretion in hierarchical mergers. Voit (2005) showed that the smooth
957: accretion can generate $\sim$ 50\% more entropy throughout the cluster than would
958: lumpy hierarchical accretion. However, Borgani et al. (2005) argued that this
959: entropy amplification effect is substantially reduced by cooling.
960: The other ideas resort to non-gravitational processes.
961: As discussed in the last section, although the NKV07 simulations predict the $K_{500} - T_{500}$
962: relation very well, they produce too many stars. Thus, models with only cooling may
963: not be enough. Borgani et al. (2005) showed that galactic winds from SN explosions
964: are rather localized and cannot boost entropy enough at large radii.
965: Thus, a feedback mechanism that can distribute heat in a very diffuse way is required.
966: As groups are smaller than hot clusters, AGN outflows from the central galaxy
967: can reach a larger scaled radius. One good example in this sample is 3C449 ($\S$8 and
968: Appendix), with radio lobes extending to at least 3.7 $r_{500}$.
969: Thus, it is still an open question to explain the entropy excess at $r_{500}$, especially
970: in groups. But the much reduced entropy ratios from this work largely alleviate the
971: problem.
972: It is also clear that more systems with the entropy ratio constrained at $r_{500}$
973: are required for better comparison, especially $T_{500} \lsim 1.4$ keV groups.
974:
975: \subsection{Entropy slopes}
976:
977: We also derived the entropy slopes at 30 kpc$ - 0.15 r_{500}$, 0.15 $r_{500} - r_{2500}$,
978: $r_{2500} - r_{1500}$ (or $r_{\rm det, spe}$) and $r_{1500} - r_{\rm det, spe}$ (Fig. 14).
979: The scatter is large but the average slopes are about the same ($\sim$ 0.7) beyond
980: 0.15 $r_{500}$. The slope is always shallower than that from pure gravitational processes
981: ($\sim$ 1.1).
982: Mahdavi et al. (2005) analyzed the \xmm\ data of 8 nearby groups and found their
983: entropy profiles are best fitted by a broken power law with the break radius of
984: $\sim 0.1 r_{500}$. Across the break radius, the entropy slope decreases from 0.92
985: to 0.42. As the break radius is very near the core, this kind of entropy profiles
986: may be caused by a cool core within 0.1 $r_{500}$. From 0.15 $r_{500}$, there are
987: systems in our sample with an entropy slope of around 0.42, but the weighted mean
988: is significantly larger. The measured average slopes in this work are consistent with
989: the result by Finoguenov et al. (2007) (0.6 - 0.7) for groups.
990:
991: \section{Mass and gas fraction}
992:
993: \subsection{$M - T$ and $M - Y_{\rm X}$ relations}
994:
995: One of the most important aspects of cluster science is to use clusters to study cosmology,
996: which often involves derivation of a cluster mass function. Assuming hydrostatic equilibrium,
997: X-ray observations can be used to derive the cluster mass (at least the HSE value),
998: provided the radial coverage of the data is good. Often we estimate cluster mass from
999: another X-ray observable used as a mass proxy. A frequently used
1000: mass proxy is gas temperature so it is important
1001: to understand the cluster $M - T$ relation. There had been a lot of work on the
1002: cluster $M - T$ relation before the \chandra\ and \xmm\ era (e.g., Finoguenov et al. 2001;
1003: Sanderson et al. 2003). However, as emphasized in V06, it is crucial to
1004: constrain the gas properties (both the temperature gradient and the density gradient)
1005: at large radii (e.g., around $r_{500}$). Systematics like the assumption of a polytropic
1006: equation of state and the inadequate fit of the density profile at large radii
1007: can bias the HSE mass to lower values (Borgani et al. 2004; V06).
1008: The $M - T$ relations have also been constrained with \chandra\ and \xmm\ (Arnaud et al.
1009: 2005 with the \xmm\ data; V06 and V08 with the \chandra\ data).
1010: However, the number of clusters used to
1011: constrain the $M - T$ relation is still small (10 in Arnaud et al. 2005 and 17 in V08).
1012: There are only 4 systems with temperatures of 2.1 - 3.0 keV
1013: in Arnaud et al. (2005) and only three systems with temperatures of
1014: 1.6 - 3.0 keV in V08 (two overlapping with the Arnaud et al. sample).
1015: Our paper adds more $<$ 2.7 keV systems (23 tier 1/2 groups).
1016:
1017: The total mass values and uncertainties at interesting radii are derived from
1018: the 1000 simulated density and temperature profiles. The determination of the
1019: best-fit value and the 1$\sigma$ errors (at two sides) is mentioned in $\S$4.
1020: One difference from the determination of entropy is that we include only
1021: simulations that produce physically meaningful mass density profiles (mass
1022: density always larger than zero). The derived $M_{500} - T_{500}$ relation
1023: with the tier 1+2 groups is shown in Fig. 15. The BCES fits are listed in
1024: Table 6. We also included 14 clusters from V08 to constrain the $M - T$
1025: relation in a wider mass range. Our results show that the $M_{500} - T_{500}$
1026: relation can be described by a single power law down to at least $M_{500} =
1027: 2 \times 10^{13}$ h$^{-1}$ M$_{\odot}$. At $T_{500} < 1$ keV, more systems
1028: are needed to examine whether the relation steepens or not. Our $M_{500} - T_{500}$
1029: relation is steeper than but still consistent with V08's (1.65$\pm$0.04 vs. 1.53$\pm$0.08).
1030: Our slope is consistent with the Borgani et al. (2004) simulations with non-gravitational
1031: processes included (1.59$\pm$0.05), especially if two groups at $T_{500} < 1$ keV
1032: are excluded. The derived $M_{500} - T_{500}$
1033: relation can also be compared with that by Arnaud et al. (2005), which has a slope
1034: of 1.71$\pm$0.09. Arnaud et al. (2005) defined the system temperature as the overall
1035: spectroscopic temperature of the 0.1 $r_{200} - 0.5 r_{200}$ region, which should
1036: be close to $T_{1000}$ defined in this work (note $r_{500} \sim 0.66 r_{200}$). The
1037: data of the tier 1 and 2 groups give $T_{500} / T_{1000}$ = 0.96 on average. Then
1038: we find that the \chandra\ $M_{500} - T$ relation constrained from 23 groups + 14 V08
1039: clusters is 18\% - 3\% higher than that by Arnaud et al. (2005) at 1 - 10 keV.
1040: We also notice that at 1 keV, the normalization of the \chandra\ $M_{500} - T_{500}$
1041: relation is 54\% higher than that by Finoguenov et al. (2001). This is
1042: expected from the generally higher density gradient around $r_{500}$ derived
1043: in this work (Fig. 16) than the typical values in Finoguenov et al. (2001)
1044: (also see the Appendix of V06).
1045:
1046: As shown in Fig. 15, our $M_{500} - T_{500}$ relation is offset
1047: from the true $M_{500} - T_{500}$ relation in the NKV07 simulations, from
1048: 33\% lower at 1 keV to 9\% lower at 10 keV. The agreement with the Borgani et al.
1049: (2004) simulations is better, from 18\% lower at 1 keV to 6\% lower at 10 keV.
1050: Interestingly, the $M_{\rm 500, HSE} - T_{500}$ relation from the NKV07
1051: simulations is almost the same as the \chandra\ $M_{500} - T_{500}$ relation
1052: from 23 groups + 14 V08 clusters. NKV07 attribute turbulence as the extra pressure
1053: to deviate the HSE mass from the true mass (also see Rasia et al. 2004; Kay et al.
1054: 2004). Indeed for the ICM without a magnetic field, the dynamic
1055: viscosity is roughly proportional to $T_{\rm ICM}^{2.5}$. Thus, it may not
1056: be surprising that cool groups can develop stronger turbulence. However,
1057: the ICM is magnetized and the real magnitude of the ICM turbulence is unknown.
1058: The NKV07 simulations only have numerical viscosity that is small.
1059: Simulations with viscosity at different strengths are required to better
1060: determine this bias term. High resolution X-ray spectra of the ICM may be
1061: ultimately required to constrain the turbulence pressure in the ICM.
1062:
1063: Kravtsov et al. (2006) suggested a new mass proxy, the $Y_{\rm X}$
1064: parameter (product of the gas temperature and the gas mass derived from the
1065: X-ray image, or $M_{\rm gas, 500} T_{500}$ in this work), which in simulated
1066: clusters has a remarkably low scatter of only 5\%-7\%, regardless of whether
1067: the clusters are relaxed or not. The agreement between simulations and observations
1068: is also better for the $M - Y_{\rm X}$ relation than the $M - T$ relation (NKV07).
1069: We examined the $M_{500} - Y_{\rm X, 500}$ relation for 23 groups + 14 V08 clusters
1070: (Fig. 15 and Table 7). Our results indicate that a single power law
1071: relation can fit the data very well. Our best-fit (0.57$\pm$0.01) is the same as
1072: the V08 best-fit at $Y_{\rm X} > 2\times10^{13} M_{\odot}$ keV (0.57$\pm$0.05),
1073: implying the groups aligned well with clusters.
1074: Our best-fit is also consistent with the \xmm\ result by Arnaud et al. (2007)
1075: (10 clusters at $Y_{\rm X} > 10^{13} M_{\odot}$ keV), on both the slope
1076: (0.548$\pm$0.027) and the normalization (within $\sim$ 3\%).
1077: Maughan et al. (2007) assembled 12 clusters at $z=0.14-0.6$ from their work and
1078: literature ($Y_{\rm X} > 8\times10^{13} M_{\odot}$ keV) and found that the slope of the
1079: $M_{500} - Y_{\rm X, 500}$
1080: relation is consistent with the fit to the V06 clusters (0.564$\pm$0.009).
1081: Intrinsic scatter in both the $M-T$ and $M - Y_{\rm X}$
1082: relations are consistent with zero as the measurement errors for groups are large.
1083: If we simply move the best-fit lines up or down to estimate the range of the scatter
1084: from the best-fit mass values, the scatter in the $M - Y_{\rm X}$ relation is about the
1085: half that in the $M - T$ relation. The slope is very close to the self-similar value of
1086: 0.6, especially if only tier 1 groups are included (0.588$\pm$0.012).
1087: Our best fits lie between the true mass and the HSE mass from the NKV07 simulations,
1088: but the offset is much smaller than that in the $M_{500} - T_{500}$ relation.
1089: Thus, the $Y_{\rm X}$ parameter appears to be a robust mass proxy down to at least
1090: 2 $\times 10^{13}$ h$^{-1}$ M$_{\odot}$.
1091: With the derived $M-T$ and $M - Y_{\rm X}$ relations in this work, the slope of
1092: the $M_{\rm gas, 500}-T_{500}$ relation is 1.89$\pm$0.05 (for tier 1+2 groups +
1093: V08 clusters), or 1.86$\pm$0.06 (for tier 1 groups + V08 clusters).
1094: This value of the slope is consistent with the result by Mohr et al. (1999),
1095: 1.98$\pm$0.18 (90\% confidence).
1096:
1097: \subsection{Gas fraction}
1098:
1099: We derived the enclosed gas fraction profile for each group. The enclosed gas fraction
1100: generally increases with radius and this trend continues to the outermost radius
1101: in our analysis, as generally found in V06. The enclosed gas fractions for groups at
1102: $r_{2500}$ have a large scatter (Fig. 17). For groups with similar $T_{500}$,
1103: $f_{\rm gas, 2500}$ can be different by a factor as large as 2.5.
1104: Both the weighted average and the median of $f_{\rm gas, 2500}$ in our sample is
1105: 0.043, much smaller than the typical value of $\sim$ 0.09 for V08 clusters.
1106: This mean can be compared with the average $f_{\rm gas, 2500}$ by G07 (0.050$\pm$0.011),
1107: and we note that the 16 groups in G07 are on average brighter than our
1108: systems in a similar redshift range.
1109: The intrinsic scatter of $f_{\rm gas, 2500}$ is tightly correlated with the intrinsic
1110: scatter of $K_{\rm 0.15 r500}$, after the temperature dependence of both variables are removed
1111: from their relations with temperature (Fig. 18).
1112: Groups with low $f_{\rm gas, 2500}$ have high $K_{\rm 0.15 r500}$, relative to
1113: the average relations. Thus, the large scatter of gas fraction within $r_{2500}$ for
1114: groups is tightly correlated with the large scatter of entropy at 0.15 $r_{500}$,
1115: and likely also the large scatter of X-ray luminosities for groups. Group properties (e.g.,
1116: luminosity and central entropy) have large scatter because of the large scatter of
1117: the gas fraction around the center (e.g., $r_{2500}$). Groups are on average
1118: fainter than what is expected from the self-similar $L - T$ relation because groups
1119: are generally ``gas-poor'' within $r_{2500}$, compared with clusters.
1120:
1121: The enclosed gas fraction within $r_{500}$ is also derived for tier 1 + 2 groups
1122: (Fig. 19). We added 14 clusters from V06 and V08 to constrain the $f_{\rm gas} - T_{500}$
1123: relations. The $f_{\rm gas, 500} - T_{500}$ relation has a slope of $\sim$ 0.16 - 0.22,
1124: depending on whether both tier 1 and 2 are included (Fig. 19).
1125: We also give the $f_{\rm gas, 500} - M_{500}$ relation from the BCES orthogonal fit.
1126: For tier 1 + 2 groups + clusters:
1127:
1128: \begin{equation}
1129: f_{\rm gas, 500} = (0.0616\pm0.0060) \hspace{0.1cm} h_{73}^{-1.5} (\frac{M_{500}}{10^{13} h_{73}^{-1} M_{\odot}})^{0.135\pm0.030}
1130: \end{equation}
1131:
1132: For tier 1 groups + clusters:
1133:
1134: \begin{equation}
1135: f_{\rm gas, 500} = (0.0724\pm0.0078) \hspace{0.1cm} h_{73}^{-1.5} (\frac{M_{500}}{10^{13} h_{73}^{-1} M_{\odot}})^{0.093\pm0.031}
1136: \end{equation}
1137:
1138: We notice that the
1139: NKV07 simulations predict $f_{\rm gas, 500} \propto T_{500}^{0.152}$ (or $M_{500}^{0.10}$),
1140: from their
1141: best-fit $M_{500} - T_{500}$ and $M_{500} - Y_{\rm X, 500}$ relations. The gas fraction
1142: predicted in simulations depends on the modeling of cooling (e.g., Kravtsov et al. 2005)
1143: and is often tangled with the problem of predicting the right stellar mass fraction
1144: in clusters. We also derived the enclosed gas fraction between $r_{2500}$ and
1145: $r_{500}$ for 23 groups in tiers 1 and 2 (Fig. 19). Combined with the V06 and V08 results
1146: for clusters, the average $f_{\rm gas} - T_{500}$ has little or no temperature
1147: dependence with an average value of $\sim$ 0.12, although the measurement errors are not small.
1148: $f_{\rm gas, 2500-500}$ can also be derived as:
1149:
1150: \begin{eqnarray}
1151: f_{\rm gas, 2500-500} = f_{\rm gas, 500} (\frac{1 - a \frac{f_{\rm gas, 2500}}{f_{\rm gas, 500}}}{1 - a}) \\
1152: ( a = M_{2500} / M_{500} = 5 (r_{2500}/r_{500})^3 ) \nonumber
1153: \end{eqnarray}
1154:
1155: We use: $f_{\rm gas, 2500} = 0.0347 T_{500}^{0.509}$ (from the BCES fit to 43 groups
1156: and 14 clusters, Fig. 17) and $f_{\rm gas, 500} = 0.0708 T_{500}^{0.220}$ (from the
1157: BCES fit to 23 groups and 14 clusters, Fig. 19). Combining our results on
1158: $c_{500}$ ($\S$7.3) with V06's for clusters, roughly we have
1159: $c_{500} = 5.0 (M_{500} / 10^{13} M_{\odot})^{-0.09}$.
1160: Assuming an NFW profile, the $r_{2500}/r_{500}$ ratio can be well approximated as:
1161: $r_{2500}/r_{500} \approx$ 0.322 + 0.178 $lg(1.523 c_{500})$ at $c_{500} = 1.3 - 5.0$.
1162: Thus, combined with the $M_{500} - T_{500}$ relation derived in this work, we can
1163: estimate the $f_{\rm gas, 2500-500} - T_{500}$ relation. Indeed as shown in Fig. 19,
1164: $f_{\rm gas, 2500-500}$ is nearly constant at 1 - 10 keV.
1165: The average $f_{\rm gas, 2500-500}$ is still $\sim$ 27\% lower than the universal
1166: baryon fraction (0.1669$\pm$0.0063 from Komatsu et al. 2008). However, one should
1167: be aware that the enclosed gas fraction still rises beyond $r_{500}$, as generally
1168: found in V06 and this work. We also notice that the observed $f_{\rm gas, 2500-500}$
1169: is consistent with what was found in the simulations of Kravtsov et al. (2005) with
1170: cooling and star formation. We conclude that
1171: the low gas fraction generally observed in groups is mainly due to the low
1172: gas fraction of groups within $r_{2500}$, or the generally weak ability of the
1173: group gas to stay within $r_{2500}$. Beyond $r_{2500}$, the groups
1174: are more regular and more similar to hot clusters, as also shown by the smaller
1175: scatter of their entropy values at $r \gsim r_{2500}$.
1176:
1177: A natural question motivated by these results on entropy and gas fraction is:
1178: what is the fraction of the group luminosities within 0.15 $r_{500}$ or $r_{2500}$?
1179: Detailed work on the group $L_{\rm X} - T$ relation is beyond the scope of this paper
1180: and will be presented in a subsequent paper with an extended sample (including
1181: non-relaxed groups). Here we give the curves of the enclosed count fluxes for
1182: 17 groups that $r_{500}$ is reached by \chandra\ or (and) PSPC (Fig. 20).
1183: As the previous $L_{\rm X} - T$ relations only used the global spectrum to convert
1184: the count rate to the group flux, Fig. 20 can be regarded as the growth curves of the enclosed
1185: group luminosities. At 0.15 $r_{500}$, the fraction ranges from 13\% to 69\%.
1186: At $r_{2500}$, the fraction ranges from 51\% to 94\%. Two extreme cases defining the
1187: boundaries are A1238 and AS1101, also two systems with similar $T_{500}$ but very
1188: different $f_{\rm gas, 2500}$ (Table 3). A system with a bright cool core (like
1189: AS1101) has a large fraction of the X-ray luminosity within 0.15 $r_{500}$.
1190: Its system temperature without excluding the central core (e.g., 0.15 $r_{500}$)
1191: therefore is biased to a lower value.
1192: A system without a bright cool core (like A1238) has a small fraction of the X-ray
1193: luminosity within 0.15 $r_{500}$, and its system temperature without excluding
1194: the central core (e.g., 0.15 $r_{500}$) is usually biased to a higher value
1195: (Fig. 3-5). All these factors contribute to the large scatter of the group
1196: $L_{\rm X} - T$ relation (e.g., Osmond et al. 2004). Thus,
1197: the group $L_{\rm X} - T$ relation can be significantly contaminated by the large
1198: difference in the cores (e.g., within 0.15 $r_{500}$).
1199:
1200: \subsection{$c_{500}$}
1201:
1202: We also fitted the total mass density profile with the NFW profile and derived the concentration
1203: parameter, $c_{500} = r_{500} / r_{\rm s}$, where $r_{\rm s}$ is the characteristic
1204: radius of the NFW profile. V06 used an inner radius of 0.05 $r_{500}$, since the stellar
1205: mass of the cD is dominant in the center. The V06 sample is mainly composed of clusters.
1206: The groups in our sample have $r_{500}$ of 440 - 800 kpc, so the contribution
1207: of the stellar mass at 0.05 $r_{500}$ is still significant for low temperature systems
1208: (see e.g., G07). Thus, we use a fixed inner radius of 40 kpc. The outer radius is
1209: the outermost radius for the spectral analysis ($r_{\rm det, spe}$ in Table 2).
1210: We derived the total mass with 1 $\sigma$ uncertainties at radii corresponding to
1211: the boundaries of radial bins for spectral analysis between 40 kpc and $r_{\rm det, spe}$
1212: (Fig. 3-5). The resulting mass density profile (at 4 - 10 radial points in this work) is fitted
1213: with an NFW profile and the uncertainty is estimated from 1000 Monte Carlo simulations.
1214: The results for 33 groups are present in Table 3 and are plotted with the system mass in Fig. 21.
1215: For the other 10 groups, $c_{500}$ is very poorly constrained.
1216: Our errors on $c_{500}$ are larger than those in V06 and G07 as our errors on mass
1217: are larger.
1218:
1219: We compare our results with the best-fit $c - M$ relation from G07.
1220: G07 gave the best-fit $c_{\rm vir} - M_{\rm vir}$ relation. For their range of
1221: $c_{\rm vir}$, $c_{500} \sim 0.51 c_{\rm vir}$ (for $c_{\rm vir}$ = 10.35). For
1222: $\Delta=101$, we convert G07's best-fit relation to: $c_{500} (1+z) =
1223: 3.96 (M_{500}/10^{14} M_{\odot})^{-0.226}$ (adjusted to our cosmology).
1224: As shown in Fig. 21, at $M_{500} > 4.5\times10^{13} M_{\odot}$, the G07 fit
1225: describes our results very well. But our results do not show significant mass
1226: dependence, so at $M_{500} < 4.5\times10^{13} M_{\odot}$, our results are
1227: systematically below the G07 fit, although the errors are not small.
1228: However, the difference mainly comes from three groups (NGC~1550, NGC~533 and NGC~5129)
1229: for which G07 found $r_{\rm s}$ = 41 - 46 kpc (adjusted to our cosmology), and our
1230: inner radius cut at 40 kpc prevents us from measuring such a small value of $r_{\rm s}$.
1231: In fact, the overdensity radii
1232: of these three groups agree better between G07 and this work:
1233: $r_{2500} = 206\pm2$ kpc (G07) vs. 222$\pm$6 from this work for NGC~1550,
1234: $r_{1250} = 251\pm2$ kpc (G07) vs. 275$\pm$30 for NGC~533,
1235: $r_{1250} = 217\pm7$ kpc (G07) vs. 236$\pm$13 for NGC~5129.
1236: Excluding these three groups, our results agree well with G07's. Thus, the difference
1237: mainly hinges on the determination of $r_{\rm s}$ that is related to the subtraction
1238: of stellar mass, while the results at large radii agree better.
1239: We also compare our results with the simulations of Bullock et al. (2001) and further work
1240: \footnote{http://www.physics.uci.edu/\~{{}}bullock/CVIR/}. As $c_{500}$ is sensitive
1241: to the halo formation time, smaller $\sigma_{8}$, $\Omega$$_{\rm M}$ and tilt
1242: drive $c_{500}$ smaller. We used the parameters derived in Komatsu et al. (2008)
1243: (see the caption of Fig. 21). As shown in Fig. 21, our results are generally consistent with
1244: the Bullock et al. simulations. Detailed discussions on the difference of the observed
1245: concentration parameter from the prediction can be found in Buote et al. (2007).
1246: We should point out that both our analysis and the V06 analysis do not subtract the
1247: stellar mass and the X-ray gas mass, while G07 subtracted both components.
1248: G07 also included the \xmm\ data for most of their groups.
1249: Future work on $c_{500}$ may need very good measurement of the gas properties to $> r_{500}$
1250: and careful modeling of the stellar and gas components (e.g., G07). The group dark matter
1251: mass profile may also need to be examined first to see whether a single NFW profile
1252: is the best fit.
1253:
1254: \subsection{The baryon fraction and fossil groups in this sample}
1255:
1256: We can estimate the enclosed baryon fraction from the cluster stellar fraction
1257: estimated before. Lin et al. (2003) gave: $M_{*, 500} = 7.30\times10^{11} M_{\odot}
1258: (T_{\rm X} / 1 {\rm keV})^{1.169}$ (adjusted to our cosmology) from the 2MASS
1259: data of nearby groups and clusters, which includes the stellar mass in cluster galaxies.
1260: With the $M - T$ relation derived in this work from 23 groups and 14 V08 clusters,
1261: $f_{\rm *, 500, Lin} = 0.0263 (T_{\rm X} / 1 {\rm keV})^{-0.481}$.
1262: Gonzalez et al. (2007) included the intracluster stellar mass and found:
1263: $f_{\rm *, 500, Gonzalez} = 0.0380 (M_{500} / 10^{14} M_{\odot})^{-0.64}$ (adjusted to our cosmology).
1264: With our $M_{500} - T_{500}$ relation,
1265: $f_{\rm *, 500, Gonzalez} = 0.0864 (T_{500} / 1 {\rm keV})^{-1.056}$.
1266: Adding the relation for $f_{\rm gas, 500}$, the total baryon fraction from groups
1267: to clusters can be estimated. As shown in Fig. 19, there is substantial difference
1268: for groups.
1269: We also examined the stellar mass of the cD and its relation with
1270: other group properties. As shown in Fig. 22, the stellar mass of the group cD (which is
1271: proportional to its $K_{s}$ band luminosity) is weakly correlated with the system mass.
1272: Low-mass cDs generally reside in low-mass groups. We also examined the relation between
1273: $L_{Ks}$ of the cD and $f_{\rm gas, 2500}$ but found no correlation. However, in low-mass
1274: systems with low $f_{\rm gas, 2500}$, the stellar mass of the cD can be comparable to
1275: the gas mass within $r_{2500}$.
1276:
1277: We also searched for fossil groups in this sample. From Jones et al. (2003), fossil
1278: groups are defined as a bound system of galaxies with the $R$ band magnitude difference
1279: of the two brightest galaxies within half the virial radius larger than 2 mag.
1280: In this work, because we do not have homogeneous $R$ band magnitudes for group galaxies
1281: in this sample, we used the 2MASS $K_{\rm s}$ band magnitude, which is
1282: a good indicator of the stellar mass. To ensure large and blue spirals are not left
1283: out, we also checked NED and HyperLeda to examine the $B$ band magnitude difference.
1284: Jones et al. (2003) used $r_{200}$ as the virial radius, while we use the exact
1285: definition of the virial radius for our cosmology, $r_{\rm vir}$ = $r_{\Delta}$ ($\Delta$
1286: $\sim$ 101). For the typical mass concentration of groups in this sample,
1287: 0.5 $r_{\rm vir} \sim r_{500}$. Thus, we examined the $K_{\rm s}$ and $B$ band magnitude
1288: difference for group galaxies within $r_{500}$.
1289: Six fossil groups are selected: NGC~741, ESO~306-017, RXJ~1159+5531, NGC~3402, ESO~552-020
1290: and ESO~351-021. RXJ~1159+5531 and ESO~306-017 are known fossil groups (Vikhlinin et al. 1999;
1291: Sun et al. 2004). We note that NGC~1132 was considered a fossil group (Mulchaey \& Zabludoff,
1292: 1999). However, NGC~1126 is 8.4$'$ from NGC~1132 with a velocity difference of 438 km/s
1293: ($r_{500}=16.2'$ or 440 kpc for the NGC~1132 group). The 2MASS $K_{\rm s}$ magnitude
1294: difference is 1.49 mag and the $B$ magnitude difference is 2.07 mag. Thus, we do not
1295: consider NGC~1132 a fossil group based on the Jones et al. (2003) definition.
1296:
1297: Although the sample is small and not representative, we examined whether these
1298: fossil groups populate a different position of the scaling relations than
1299: non-fossil groups. No significant difference in $K - T$, $f_{\rm gas} - T$ and $M - T$
1300: relations is found. As shown in Fig. 22, these fossil groups indeed have massive cDs.
1301: The $c_{500}$ of these fossil groups (four listed in Table 3, others poorly constrained)
1302: are $\sim$ 2.4 - 4.3, on average smaller than the average of this sample, which differs
1303: from the claim by Khosroshahi et al. (2007) that fossil groups have higher mass
1304: concentration than non-fossil systems.
1305:
1306: \section{AGN heating and Radio galaxies}
1307:
1308: \subsection{Sign of heating in entropy profiles}
1309:
1310: Besides pre-heating (from SNe and AGN) and cooling, impulsive heating from
1311: the central AGN is often required to explain the observed scaling relations
1312: like $L - T$ and $K - T$ (e.g., Lapi et al. 2005). Strong shocks driven by the
1313: central AGN may boost the ICM entropy and create an entropy bump. This transient
1314: anomaly in the entropy profile may be detected in a large sample. We have searched
1315: for such entropy features in our sample and find two promising cases: UGC~2755 and
1316: 3C~449 (Fig. 23). There are also three groups with a significant break observed in their
1317: surface brightness profiles (3C442A, IC1262 and A2462) that may be related to AGN
1318: heating. We briefly discuss them in the Appendix.
1319: Both UGC~2755 and 3C~449 host a strong FR I radio source with two-sided radio lobes.
1320: UGC~2755 has a central corona with a radius of $\sim$ 3 kpc (see Sun et al. 2007 for
1321: the connection of thermal coronae with strong radio sources).
1322: From 10 kpc to 80 kpc in radius, the surface brightness profile
1323: is very flat. Then there is a sharp break at $\sim$ 90 kpc.
1324: UGC~2755's radio lobes extend to $\sim$ 100 kpc from the nucleus in the NVSS image,
1325: which may naturally explain the entropy bump within $\sim$ 90 kpc.
1326: 3C449 also has a central corona with a radius of $\sim$ 3 kpc.
1327: Its entropy bump is at $\sim$ 40 kpc - 100 kpc and less significant as
1328: that in UGC~2755. 3C449's radio outflow is spectacular and can
1329: be traced to at least 1.6 Mpc in radius from the NVSS image (compared to its
1330: $r_{500}$ of 433 kpc estimated from $r - T$ relation).
1331: Nevertheless, the brighter inner part of the radio jets/lobes
1332: ends at $\sim$ 100 kpc from the nucleus. Thus, the entropy bump we observed
1333: in 3C~449 may represent the most recent heating event.
1334:
1335: Strong shocks are required to effectively boost entropy. Using the
1336: standard Rankine-Hugoniot conditions, the entropy increase after shocks is a
1337: sensitive function of the shock Mach number, and weak shocks have little effect
1338: on amplifying entropy. A shock with
1339: a Mach number of 1.2 only increases entropy by 0.44\%. A Mach 2 shock
1340: increases entropy by 20\%, while Mach 3.3 and 4.4 shocks increase
1341: entropy by 100\% and 200\% respectively.
1342: The entropy bump in 3C449 is only at the level of 20\%, which can be produced
1343: by a single Mach 2 shock. The entropy boost in UGC~2755 is 1.5 - 2 times,
1344: which can be produced by a single Mach 3 shock. The adiabatic sound speed
1345: in groups is not high, 540 - 630 km/s in 1.1 - 1.5 keV ICM for these two groups.
1346: In these low density groups, the ambient pressure is much lower than that in
1347: the dense cores of hot clusters. Shock deceleration may also be slower.
1348: The velocity of the outflow-driven shock
1349: is $\sim f_{P} (P_{\rm kin}/\rho r^{2})^{1/3}$, where $\rho$ is the ICM density and
1350: $f_{P}$ is a structure factor of order unity that depends on the outflow geometry
1351: and the preshock density profile (e.g., Ostriker \& McKee 1988). $n_{e} \sim
1352: 10^{-3}$ cm$^{-3}$ around the entropy bumps and $P_{\rm kin} \sim 10^{44}$
1353: ergs s$^{-1}$ from the radio luminosities of two radio AGN and the relation
1354: derived by B\^{i}rzan et al. (2004). The estimated velocity is then
1355: $\sim 1300 f_{P}$ km/s, which is comparable to the requirement of
1356: the entropy boost. Therefore, radio outflows in these two groups are capable of
1357: driving Mach 2 - 3 shocks to produce the observed entropy bumps.
1358:
1359: \subsection{Strong central radio sources in this sample}
1360:
1361: There are ten groups hosting a central radio source that is more luminous
1362: than $L_{\rm 1.4 GHz} = 10^{24}$ W Hz$^{-1}$:
1363: 3C~31, 3C~449, UGC~2755, 3C~442A, A160, A2717, AS1101, A3880, A1238 and A2462.
1364: What is the typical X-ray gas environment around these radio sources?
1365: Six of them lack large cool cores (e.g., $\gsim$ 30 kpc radius), 3C31, 3C~449, UGC~2755, A160,
1366: A1238 and A2462. However, all of them host a central corona with a radius of $\sim$
1367: 3 - 8 kpc, typical for massive cluster and group galaxies as discussed in
1368: Sun et al. (2007).
1369: This component is reflected in their temperature profile, except for
1370: A1238 as it is faint. The spectrum of A1238's central source can be
1371: described by a $\sim$ 0.8 keV thermal component. Its X-ray luminosity and
1372: $K_{\rm s}$ band luminosity fall on the typical region for coronae and
1373: its properties are similar to ESO~137-006 in A3627 (a nearby bright corona, Sun et al. 2007).
1374: Four other groups (3C442A, A2717, A3880 and AS1101) host larger cool cores
1375: with a radius of $\sim$ 30 kpc or larger. The cool cores in 3C442A and
1376: A2717 are clearly disrupted, likely by the radio sources.
1377: Thus, all these strong radio sources have low-entropy ICM ($< 30$ keV cm$^{2}$) at
1378: the center.
1379:
1380: \section{Systematic errors}
1381:
1382: We follow Humphrey et al. (2006), G07 and V08 to discuss the systematic error
1383: budgets in our results. The uncertainties of the local X-ray background are the
1384: main error budget at large radii. As shown in $\S$3.2, we have included a
1385: conservative estimate of the background uncertainties into the errors of
1386: temperatures and densities. This is the primary reason that our results
1387: at large radii have larger errors than those of V06 for the same systems.
1388:
1389: We used the LAB survey data (Kalberla et al. 2005) for the galactic hydrogen
1390: column density (Table 2) and examined the \chandra\ spectra for excess absorption.
1391: The LAB column density is on average $\sim$ 6\% lower than the column density
1392: from Dickey \& Lockman (1990). Out of 43 groups we studied, 19 show significant
1393: excess absorption. We can compare this fraction to that of V05, 6 out of 13 with
1394: excess absorption. For the five groups that were studied in V05 and this
1395: work, both works find excess absorption for the same three groups.
1396: On the other hand, G07 used the galactic hydrogen column density from Dickey \& Lockman
1397: (1990) for all groups.
1398: The incidence of excess absorption in our work increases with the galactic hydrogen
1399: column density. At $N_{\rm 21 cm} > 4\times10^{20}$ cm$^{-2}$, eight of eleven
1400: groups (most of them at $z < 0.03$) show excess absorption. This trend is
1401: qualitatively consistent with the result by Arabadjis \& Bregman (1999).
1402: We also examined the effects of a conservative $N_{\rm H}$ uncertainty of $\pm2\times10^{20}$
1403: cm$^{-2}$ (see e.g., V08) on our results.
1404: For $T <$ 1.6 keV gas, the determination of temperature is little affected by
1405: absorption. An $N_{\rm H}$ change of $\pm2\times10^{20}$ affects the temperature
1406: by $\sim \mp$ 1.4\% and the density by $\sim \pm$ 6.1\%. The subsequent changes
1407: on $K, M, Y_{\rm X}$ and $f_{\rm gas}$ at an overdensity radius are $\sim \mp$6.3\%,
1408: $\mp$2.1\%, $\pm$3.9\% and $\pm$7.6\% respectively, assuming an entropy slope
1409: of 0.7 (note that the overdensity radius also depends on mass).
1410: For $T >$ 1.6 keV gas, an $N_{\rm H}$ change of $\pm2\times10^{20}$ affects the
1411: temperature by $\sim \mp$ 5.3\% and the density by $\sim \pm$ 3.7\%.
1412: The subsequent changes
1413: on $K, M, Y_{\rm X}$ and $f_{\rm gas}$ at an overdensity radius are $\sim \mp$5.9\%,
1414: $\mp$8.0\%, $\mp$4.4\% and $\pm$9.7\% respectively.
1415: At small radii, uncertainties at these levels are not important as they are
1416: smaller than the intrinsic scatter (e.g., $K_{\rm 0.15 r500}$ and $f_{\rm gas, 2500}$).
1417: At large radii, the statistical errors (including uncertainties from the local X-ray
1418: background) overwhelm. Nevertheless, the systematic errors from $N_{\rm H}$ should
1419: be kept in mind.
1420:
1421: We use the deprojection algorithm derived by Vikhlinin (2006). The form of the
1422: 3D temperature profile is the same as that used in V06.
1423: The robustness of this deprojection algorithm has been presented in Vikhlinin (2006),
1424: with uncertainties of $\sim$ 0.05 keV. Nagai et al. (2007a) presented mock \chandra\
1425: analysis of cluster simulations, using the deprojection algorithm proposed by
1426: Vikhlinin (2006) and the form of the temperature profile used in V06. The best-fit
1427: 3D temperature profile is well consistent with the true temperature profile in simulations,
1428: with residuals in a similar level as shown in Vikhlinin (2006). This test further
1429: validates the approach used in the present work. We have also performed a test
1430: on a sub-sample of 6 groups with best-quality data. The traditional onion-peeling
1431: method (e.g., McLaughlin 1999; G07) was used. The resulting deprojected temperature
1432: profiles often show small magnitude of oscillating. If we use the form of equ (2)
1433: to fit them, the best-fits are consistent with those shown in Fig. 3 - 5 within
1434: $\sim$ 0.05 keV on average. The uncertainties on temperatures range from 2\% - 6\%,
1435: while the uncertainties on densities are 0.5\% - 3\% (both directions).
1436: Thus, this systematic error (larger for $<$ 1 keV groups) are smaller than the
1437: systematic error related to the uncertainty of $N_{\rm H}$.
1438:
1439: There are other systematic errors, like the choice of the plasma spectral codes,
1440: and departures from spherical symmetry in the group gas. Those are either small
1441: in magnitude or have little impact on the scaling relations, as discussed in
1442: Humphrey et al. (2006), G07 and V08. As mock data from simulations are starting
1443: to be analyzed in the same way as the observational data (Nagai et al. 2007a;
1444: Rasia et al. 2008), these factors are becoming better controlled.
1445:
1446: \section{Summary and conclusions}
1447:
1448: We present an analysis of 43 galaxy groups with \chandra\ observations. With inclusion of
1449: many faint systems (e.g., the ones hosting strong radio sources), our sample is not
1450: much biased to X-ray luminous groups, as shown by the wide range of ICM entropy
1451: values around the group center ($\S$6). We used the ACIS stowed background and
1452: modeled the local CXB for each group. Uncertainties of local background are folded
1453: into the derived temperature and density profiles. The projected temperature profile and
1454: the surface brightness profile are modeled with sophisticated models, which have
1455: enough freedom to describe the data from the core to the outskirts. The 3D abundance
1456: profile is also derived. The 3D temperature and density profiles are constrained
1457: through iterative fitting. The uncertainties are estimated from 1000 Monte Carlo
1458: simulations. Gas properties are derived to at least $r_{2500}$ for all 43 groups.
1459: For 11 groups, we can derive gas properties robustly to $r_{500}$. For another
1460: 12 groups, gas properties can be robustly derived to $\gsim r_{1000}$, so we
1461: extrapolate the results to $r_{500}$. The main results of this paper are:
1462:
1463: 1) We present the $M_{500} - T_{500}$ and $M_{500} - Y_{\rm X, 500}$ relations in
1464: $M_{500} = 10^{13} h^{-1} M_{\odot}$ - 10$^{15} h^{-1} M_{\odot}$, combined
1465: with the V08 results on 14 $T_{500} >$ 3.7 keV clusters. Both relations
1466: are well behaved at the low-mass end and can be well fitted with a single power
1467: law ($\S$7.1 and Fig. 15). The $M_{500} - Y_{\rm X, 500}$ relation indeed
1468: has a smaller scatter than the $M_{500} - T_{500}$ relation (about half).
1469: The $M_{500} - T_{500}$ relation from observations is still offset
1470: from simulations (e.g., NKV07). Interestingly, the \chandra\ $M_{500} - T_{500}$ relation
1471: is very close to the $M_{\rm 500, HSE} - T_{500}$ relation in the NKV07 simulations.
1472: Although it is tempting to attribute the difference to a mass
1473: bias, better understanding of the ICM viscosity is required.
1474:
1475: 2) The group gas fraction within $r_{2500}$ is on average much smaller than
1476: that of clusters (e.g., $\sim$ 0.043 for $\sim$ 1.5 keV groups vs. $\sim$ 0.09
1477: for $\sim$ 6 keV clusters, $\S$7.2 and Fig. 17), which is consistent with G07's
1478: results. The group gas fraction within
1479: $r_{2500}$ also has a large scatter, spanning a factor of $\sim$ 2 at any fixed
1480: temperature. On the other hand, the gas fraction measured between $r_{2500}$ and
1481: $r_{500}$ has no temperature dependence with an average value of
1482: $\sim$ 0.12 (Fig. 19). Thus, the generally low gas
1483: fraction in groups is due to the general low gas fraction within $r_{2500}$.
1484:
1485: 3) We derived the $K - T$ relations at 30 kpc, 0.15 $r_{500}$, $r_{2500}$,
1486: $r_{1500}$, $r_{1000}$ and $r_{500}$ ($\S$6.1, Fig. 11). The large scatter of
1487: the entropy values at 30 kpc
1488: and 0.15 $r_{500}$ reflect the wide luminosity range of groups in this sample.
1489: The $K - T$ relation is significantly tighter beyond $r_{2500}$ and the intrinsic
1490: scatter of entropy is the same at 10\% - 11\% from $r_{2500}$ to $r_{500}$.
1491: Thus, the group properties are more regular from $r_{2500}$ outward, in line with the
1492: gas fraction results.
1493: With 14 clusters from V06 and V08 included, we also present $K - T$ relations
1494: in the full temperature range (Fig. 12 and Table 5). At $r_{500}$, the slope
1495: of the $K - T$ relation is consistent with the value from self-similar
1496: relation (1.0).
1497:
1498: 4) The ratios of the observed entropy values to the baseline values (from
1499: adiabatic simulations) decrease with radius ($\S$6.2 and Fig. 13). At $r_{2500}$,
1500: the ratio ranges
1501: from $\sim$ 1.8 - 3.5, with a weighted mean of 2.2. The weighted mean decreases
1502: to $\sim$ 1.8 at $r_{1000}$ and $\sim$ 1.6 at $r_{500}$. The still significant
1503: entropy excess at $r_{500}$ in groups may require a diffuse way to distribute
1504: heat (e.g., AGN heating, see $\S$8.1), but it may also be understood with smoother accretion
1505: and the mass bias that may be especially large in groups ($\S$6.2).
1506: The entropy excess at $r_{500}$ is also detected for 14 clusters from V06 and
1507: V08 ($\sim$ 35\%, $\S$6.2). The difference in the entropy excess at $r_{500}$ between groups
1508: and clusters ($\sim$ 17\%) is not as large as previously claimed from the \rosat\
1509: and \asca\ data.
1510:
1511: 5) The entropy slopes are determined at 30 kpc - 0.15 $r_{500}$, 0.15 $r_{500}$ -
1512: $r_{2500}$, $r_{2500}$ - $r_{1500}$ and $r_{1500}$ - $r_{\rm det, spe}$ ($\S$6.3
1513: and Fig. 14) . The slopes are all shallower than 1.1 beyond 0.15 $r_{500}$.
1514: Scatter is large but the average slope is $\sim$ 0.7 beyond 0.15 $r_{500}$.
1515:
1516: 6) The group temperature profiles are similar at $> 0.15 r_{500}$, despite large
1517: scatter within 0.15 $r_{500}$ ($\S$5 and Fig. 8). The average slope is consistent
1518: with the ``Universal temperature profile'' (Markevitch et al. 1998; De Grandi \& Molendi
1519: 2002; Loken et al. 2002; V05; Pratt et al. 2007; G07; Rasmussen \& Ponman 2007; LM08)
1520: but there is still scatter. The group temperature profiles also appear more peaky
1521: than those of clusters (Fig. 9).
1522:
1523: 7) We also derived the concentration parameter ($c_{500}$) for 33 groups ($\S$7.3).
1524: Our results are generally consistent with the Bullock et al. (2001) simulations under
1525: the current WMAP5+SN+BAO cosmology.
1526:
1527: 8) We selected six fossil groups in this sample and four are new. The X-ray gas properties
1528: of these fossil groups have no significant difference from non-fossil groups
1529: ($\S$7.4) in scaling relations.
1530:
1531: 9) We found two groups with substantial entropy bumps ($\S$8.1), which may indicate
1532: a recent strong heating episode. Both host strong radio galaxies at the center and we
1533: estimate that the radio AGN is capable of driving shocks to boost entropy to the
1534: observed level.
1535:
1536: The emerging picture of groups from this work is that the main difference
1537: between groups and hotter clusters is the general weak ability of the group gas to stay
1538: within $r_{2500}$, which explains most of deviation of the group properties
1539: from the self-similar relations (e.g., entropy and luminosity).
1540: The group properties within $r_{2500}$ have large scatter, but may all be
1541: related to variations in the level of the enclosed gas fraction within $r_{2500}$.
1542: Beyond $r_{2500}$, groups are more regular and more like clusters, making
1543: them promising tools for cosmology, as shown by the well-behaved $M - T$
1544: and $M - Y_{\rm X}$ relations derived in this work.
1545:
1546: \acknowledgments
1547:
1548: This work would be impossible without the rich data sets in the \chandra\ archive.
1549: We would like to thank all PIs of the \chandra\ observations for their
1550: original efforts. We thank M. Markevitch for his help on the ACIS stowed background.
1551: We thank the referee, D. Buote, for helpful and prompt comments.
1552: We thank helpful discussions with M. Balogh, S. Borgani, D. Nagai and G. Pratt.
1553: The financial support for this work was provided by the NASA LTSA grant NNG-05GD82G.
1554: We made use of the NASA/IPAC Extragalactic Database and the HyperLeda database.
1555:
1556: \begin{appendix}
1557:
1558: \subsubsection{Components of the \chandra\ background}
1559:
1560: The \chandra\ background has been extensively discussed before (e.g., Markevitch et al.
1561: 2003; Wargelin et al. 2004; V05; HM06; Humphrey \& Buote 2006). We present here
1562: a brief summary, focusing in particular on the cosmic X-ray background (CXB). There are
1563: two basic components in the quiescent \chandra\ background, particles and photons
1564: (or CXB). The dominant background component is the charged particle background
1565: (PB), which is not vignetted. The spectrum of the \chandra\ PB has been remarkably
1566: stable since 2000 (Vikhlinin et al. 2005, V05 hereafter; Hickox \& Markevitch 2006,
1567: HM06 hereafter), although the absolute flux changes with time and is related to
1568: the solar cycle. The only exception so far is for BI data
1569: after the middle of 2005, which is discussed in the next section.
1570:
1571: The cosmic hard X-ray background is considered to be composed of unresolved
1572: X-ray point sources, mostly AGN.
1573: This component can be described by an absorbed power-law with a photon index of $\sim$ 1.5
1574: (HM06). Its flux depends on the level of point source excision or the limiting flux
1575: for point sources. HM06 analyzed the two deepest \chandra\ fields at that time,
1576: \chandra\ deep field north (a combined
1577: clean exposure of 1.01 Ms) and south (a combined clean exposure of 0.57 Ms).
1578: The unresolved hard X-ray background flux density is (3.4$\pm$1.7)$\times10^{-12}$ ergs
1579: s$^{-1}$ cm$^{-2}$ deg$^{-2}$ in the 2 - 8 keV band, which represents the lower
1580: limit of the hard CXB flux in shorter \chandra\ observations.
1581: The X-ray $logN - logS$ relation has been well studied allowing us to predict
1582: the average unresolved cosmic hard X-ray background below the point source limiting flux.
1583: Kim et al. (2007; K07 hereafter) presented the average relation between point
1584: source limiting flux and expected hard CXB flux density (Fig. 19 of K07).
1585: For the limiting flux for point sources in the outermost bins of groups in our sample,
1586: we expect a resolved fraction of 25\% - 65\% in the 2 - 8 keV band,
1587: which corresponds to a flux density of 6.1 - 13.1 $\times10^{-12}$ ergs
1588: cm$^{-2}$ s$^{-1}$ deg$^{-2}$ for the unresolved hard CXB background.
1589: However, one should be aware that the K07 relation is just an average. About 30\%
1590: variation can be expected over the angular scale we study (HM06).
1591:
1592: The soft X-ray background is composed of several components, Galactic, local bubble,
1593: geocoronal and heliospheric emission (e.g., Wargelin et al. 2004; Snowden et al. 2004;
1594: Koutroumpa et al. 2007). The latter two components are primarily from the solar wind
1595: charge exchange (SWCX) emission and are time variable, with a contribution to
1596: the O VII and O VIII lines as much as the Galactic component (Koutroumpa et al. 2007).
1597: The strength of the soft X-ray background varies
1598: with the sky position (as shown in the R45, or 3/4 keV \rosat\ all sky survey map).
1599: There has been lots of work done to quantify its spectral properties. With the
1600: \chandra\ data, Markevitch et al. (2003) and HM06 have shown that, the soft
1601: X-ray background beyond the regions with strong RASS R45 flux (e.g., the
1602: North polar spur), can be well described by a single unabsorbed thermal component with
1603: a temperature of $\sim$ 0.2 keV. Its spectrum typically shows a broad line hump
1604: around 0.6 keV, mainly from the 0.57 keV O VII and 0.65 keV O VIII lines.
1605: The soft X-ray background has also been studied with the \suzaku\ data, which
1606: have the higher spectral resolution.
1607: Fujimoto et al. (2007) analyzed the \suzaku\ data of the North Ecliptic Pole
1608: region (R45 = 140 $\times10^{-6}$ counts s$^{-1}$ arcmin$^{-2}$) and found
1609: that the soft X-ray component at the non-flare period has a temperature of 0.18 keV
1610: with over-solar abundances.
1611: Miller et al. (2008) analyzed the \suzaku\ data of the brightest
1612: region of the North Polar Spur (NPS, R45 = 748 $\times10^{-6}$ counts s$^{-1}$
1613: arcmin$^{-2}$) and found that the NPS thermal component has a temperature of
1614: $\sim$ 0.28 keV with generally sub-solar abundances, besides the assumed 0.1 keV
1615: local bubble and galactic halo components with solar abundance. Snowden et al. (2008)
1616: adopted a soft CXB model composed of an unabsorbed $\sim$ 0.1 keV component for
1617: the local hot bubble, an absorbed $\sim$ 0.1 keV component for the cooler Galactic
1618: halo emission and an absorbed $\sim$ 0.25 keV component for the hotter Galactic
1619: halo emission. All abundances are fixed at solar.
1620: Thus, we have enough knowledge to model the soft X-ray background.
1621:
1622: \subsubsection{``Blank sky background'' and stowed background}
1623:
1624: Much previous \chandra\ work on clusters used the blank-sky
1625: background data\footnote{http://cxc.harvard.edu/contrib/maxim/acisbg/}, which are
1626: good enough in high surface brightness regions.
1627: However, the averaged soft X-ray component in the standard blank-sky background
1628: data is very likely different from the actual soft X-ray foreground
1629: in any particular field, as pointed out previously by e.g., Markevitch et al. (2003),
1630: V05, Humphrey \& Buote (2006) and G07. Taking the example of the blank sky background
1631: data in period D,
1632: the ACIS-I file combines 29 pointings from Jan., 2001 to Nov. 2004, at regions where
1633: R45=90 - 150. The exposures range from 27 ks to 165 ks with a median of $\sim$ 66 ks.
1634: The ACIS-S file combines 12 pointings from Nov. 2001 to Oct. 2003, at regions where
1635: R45=90 - 150. The exposures range from 20 ks to 114 ks with a median of $\sim$ 30 ks.
1636: In 5 of 12 pointings, there are no S1 data. The ACIS response has been changed
1637: significantly from 2001 to 2004 (especially because of the contamination on the optical
1638: blocking filter). Thus, it is often inadequate to only use the blank-sky background
1639: to constrain the ICM properties at low surface brightness regions. Moreover, the PB
1640: and the CXB are not separated in the blank-sky background data. When the blank-sky
1641: background is scaled to account for the flux change of the PB, the CXB in the
1642: blank-sky background data is unphysically scaled. While the PB rate did not change
1643: much between the spring of 2000 and the spring of 2004, it has been significantly
1644: increasing ever since. In 2006, the PB rate was on average 50\% higher that the average
1645: value between the spring of 2000 and the spring of 2004 (Fig. 6.24 from \chandra's
1646: Proposers' Observatory Guide v.9\footnote{http://asc.harvard.edu/proposer/POG/}).
1647: Thus, any analysis for data taken after the middle of 2004 involves a large scaling
1648: of the PB, often resulting significant over-subtraction of the CXB.
1649: Therefore, a second correction besides subtracting the scaled blank-sky background
1650: is required (e.g., Vikhlinin et al. 2005). This ``double subtraction''
1651: is often efficient but requires the presence of source-free regions in the
1652: \chandra\ field, which is not true for many nearby groups in our sample.
1653: In this work, we utilize the newly available ACIS stowed background data
1654: to subtract the PB component. The CXB
1655: is modeled and the uncertainties are folded into the final error budgets.
1656:
1657: Since Sep., 2002, ACIS observations have been carried out twice a year in the
1658: stowed position, shielded from the sky by the science instrument module
1659: structure and away from the on-board calibration source. By the end of May, 2007,
1660: 415 ksec data had been collected.
1661: Background flares have never been observed in the stowed data.
1662: The comparison with the dark moon observations indicates that the stowed background
1663: is the same as the quiescent PB collected by the CCDs in the normal
1664: focal position (Markevitch et al. 2003; the CXC calibration website $^{8}$.).
1665: HM06 also used the stowed background (236 ks at the time of their work)
1666: to carry out absolute measurement of the unresolved CXB.
1667: They show that between Jan. 30, 2000 and Sep., 2002 when the stowed background
1668: data are not available, the spectral slope of the PB is the same.
1669: Thus, we can apply the stowed background to early data.
1670: The stowed background allows us to separate the non-vignetted PB from the vignetted CXB.
1671: The main reason for our preference for the stowed background over the blank-sky
1672: background is that we have better control of the local background for nearby
1673: groups where sources fill the whole \chandra\ field. The ``double subtraction''
1674: method with the blank-sky background cannot be applied for these nearby groups, as
1675: there is no region that is free of group emission. The derived local X-ray background
1676: based on stowed background also has a clear physical meaning and can be compared
1677: between \chandra\ observations with very different PB fluxes or observations with
1678: other telescopes like \xmm. Another subtle advantage of the stowed background
1679: is related to the telemetry limit of \chandra, especially for the VFAINT mode data.
1680: The blank-sky background data of each specific ACIS CCD may come from different
1681: combination of observations, especially for ACIS-S (e.g., S1 vs. S2+I3). Therefore,
1682: the residual or decremental background on the S1 CCD is in principle different
1683: from that on FI chips, which complicate the analysis. Similarly, even if S3 is turned
1684: on for ACIS-I observations, it cannot be used for local background study as the
1685: blank-sky background of the S3 chip (when the aimpoint is on ACIS-I) is only a subset
1686: of the ACIS-I blank-sky background data. On the contrary, with the stowed background,
1687: we are analyzing the absolute CXB in the interested field so data on different CCDs
1688: can be fitted jointly to make better constraints.
1689: Therefore, we used the stowed background to subtract the PB in our observations.
1690:
1691: The spectra of the \chandra\ PB have been very stable (e.g., HM06).
1692: However, a small change of the spectral shape on the BI CCDs (S3 and S1)
1693: has been identified from around the middle of 2005, while the
1694: spectra of the PB on the FI chips still keep the same $^{8}$.
1695: The change appears abrupt around the middle of 2005, while the spectral shape of
1696: BI data remain the same from that time to at least the middle of 2007 (private
1697: communication with Maxim Markevitch). The
1698: stowed background data have been broken into 2 periods, one with 235 ksec total
1699: exposures from five observations of Sep. 3, 2002 to June, 10, 2005 (period D), the other
1700: with 180 ksec total exposure from four observations of Nov., 13, 2005 to May, 28, 2007
1701: (period E). We emphasize that the notation
1702: adopted here is only for the purpose of this work.
1703: We examined the spectral difference between these two periods, after matching
1704: their fluxes in the 9.5 - 12 keV band. The flux of the period D background is
1705: always a little lower than that of the period E background, after re-scaling.
1706: The biggest difference is seen on the S3 chip, with a 6.0\% difference in the
1707: 0.35-7.0 keV band (Fig. 24). The residual emission is very
1708: flat (note it is non-vignetted) and can be removed by increasing the D period
1709: background by 5.7\%. The difference is smaller on the S1 chip, 2.4\% in the
1710: 0.35-5.5 keV band (note that the S1 PB increases rapidly at $>$ 5.5
1711: keV, Fig. 6.21 of \chandra's Proposers' Observatory Guide v.9). The difference
1712: is consistent with zero in FI chips, $\sim$ 1\% in the 0.5-7.0 keV band,
1713: excluding the Au line in the 2.0-2.3 keV band. Thus, we can apply the total
1714: stowed background (415 ks exposures) to the FI data. For BI data, we use the
1715: stowed background in their corresponding periods. There are only two groups in our sample
1716: with BI data taken between June, 10, 2005 and Nov., 13, 2005 (NGC~1550 and NGC~5098).
1717: Both were observed after Oct. 22, 2005 and both have earlier FI data. We used the
1718: period E stowed background for the BI data of both groups.
1719: In this work, we also take a larger uncertainty on the normalization of
1720: period E PB for BI chips (5\% compared to 3\% for FI data and the period D BI data).
1721:
1722: \section{Notes on some groups}
1723:
1724: In this section, we present notes on some groups, mainly on the comparison
1725: with previous work on the gas properties at large radii. Thus, the cited references
1726: are usually not complete for each group as the detailed dynamical and thermal
1727: structures of the group cores are beyond the scope of this work.
1728:
1729: \paragraph{NGC 1550} was examined by Sun et al. (2003), with two ACIS-I observations.
1730: Now with two additional longer ACIS-S exposures in the offset positions, the gas
1731: properties in this system can be
1732: constrained much better. NGC~1550's temperature profile is among the best determined
1733: for 1 keV groups, with the good \chandra\ coverage. We can compare our results with
1734: those from G07 who analyzed two short ACIS-I observations and an \xmm\ observation.
1735: The temperature profiles agree well although we constrain the temperature to
1736: larger radii from the ACIS-S observations. Our $r_{500}$ and $M_{500}$ are 10\% and
1737: 50\% higher than those derived
1738: by G07. Our $c_{500}$ (4.93$^{+0.50}_{-0.46}$) is smaller than G07's (9.0$\pm$0.6).
1739:
1740: \paragraph{NGC~3402} has been studied by V05 and V06. Although our temperature profile
1741: is consistent with V05, the slope of the decline is smaller.
1742: The \iras\ $100\mu$m map shows the presence of Galactic cirrus around the
1743: group. We indeed derived a higher absorption column, 1.1$\pm0.1 \times 10^{21}$
1744: cm$^{-2}$, than the Galactic value (4.0$\times10^{20}$ cm$^{-2}$ from LAB).
1745: This value is smaller than that derived in V05, 1.55$\pm0.1 \times 10^{21}$
1746: cm$^{-2}$. However, the absorption difference has little effect on the derived gas
1747: temperature. As the gas temperature is mainly determined by blended line centroid,
1748: gas temperature remains almost the same, with the higher $N_{\rm H}$ in V05.
1749: Because of higher temperatures derived at large radii, our derived $M_{2500}$ is
1750: larger than that from V06 and $f_{\rm gas, 2500}$ is smaller than in V06.
1751:
1752: \paragraph{Abell~262} is a nearby luminous system in which the X-ray emission can
1753: be traced to over 800 kpc in the 7.6 ks PSPC data. A262 was included in V06 and G07
1754: samples. We included a new deep \chandra\ exposure (110 ks) taken in 2006 in our
1755: analysis, while G07 also analyzed an \xmm\ observation. Our results of
1756: $r_{2500}, M_{2500}$ and $f_{gas, 2500}$ are consistent with those in V06 and G07.
1757: Our $c_{500}$ (3.48$^{+0.49}_{-0.45}$) is well consistent with V06's result
1758: and is close to G07's result (4.5$\pm$0.4), but smaller than the result by
1759: Piffaretti et al. (2005) (5.8$\pm$1.2).
1760:
1761: \paragraph{NGC~383} hosts a bright FRI radio source 3C31.
1762: There is a background cluster centered on 2MASX~J01065891+3209285 at
1763: $z$=0.1116. We derived the surface brightness profile centered on the
1764: background cluster and also analyzed a short exposure (ObsID 3555, 5.1 ks)
1765: centered on the background cluster. The cluster emission is detected to
1766: $\sim 4'$ radius. In the analysis for NGC~383, we
1767: excluded the region within 6.5$'$ of 2MASX~J01065891+3209285, which is a bit
1768: larger than its $r_{500}$ (5.9$'$ for $kT=2.3$ keV).
1769:
1770: \paragraph{3C~449} is located at a Galactic latitude of -16 deg and at the
1771: outskirts of a bright \iras\ 100 $\mu$m feature across several degrees,
1772: which should explain the enhanced absorption.
1773: It is one of the two groups with an entropy bump detected ($\S$8.1).
1774:
1775: \paragraph{NGC~533} was studied by Piffaretti et al. (2005) (\xmm) and G07
1776: (\chandra\ + \xmm). The explored radial range in spectral analysis is
1777: similar in all three work (up to 240 - 260 kpc).
1778: Our temperature profile agrees well with that derived by G07. Our $c_{500}$
1779: (4.58$^{+3.90}_{-2.34}$) is still consistent with the results by Piffaretti
1780: et al. (2005) (8.6$\pm$0.7) and G07 (9.0$\pm$0.7).
1781:
1782: \paragraph{MKW4} is bright enough that the group emission can be traced to
1783: the very edge of the \chandra\ field. We are able to separate the group emission
1784: and the soft background emission on the S1 spectrum because of the prominent
1785: $\sim$ 0.6 keV hump in the soft background emission and the iron L hump in
1786: the group emission. The same \chandra\ data had been analyzed by V05 and V06.
1787: Our temperature profile agrees well with V05's, as well as properties at $r_{2500}$
1788: (mass and gas fraction) with V06's. However, our results at $r_{500}$ differ from V06's.
1789: The difference should lie on the modeling of the density profile. V06
1790: derived a very steep density profile at large radii of MKW4 ($\beta_{\rm eff, 500}$
1791: = 0.92, Table 2 of V06), while we derived a value of $\sim$ 0.6, more similar
1792: to A262 and A1991 (Table 2 of V06). We notice that Vikhlinin et al. (1999)
1793: derived $\beta_{\rm outer}$ = 0.67$\pm$0.06 for MKW4, using the same PSPC
1794: data.
1795: G07 used the same \chandra\ data and also analyzed an \xmm\ observation of
1796: MKW4. Our $r_{500}$ and $M_{500}$ are consistent with G07's values, but our
1797: $c_{500}$ lies between the values of V06 and G07. G07 assumed an NFW profile for
1798: the dark matter halo and only derived gas properties to 322 $h_{73}^{-1}$ kpc.
1799: The properties at $r_{500}$ thus rely on the assumption of NFW profile and
1800: extrapolation.
1801: On the other hand, the \chandra\ data at the outermost
1802: bin only covers 10\% of the area in that annulus. Better constraints on the
1803: properties of this nearby system require more coverage at large radii.
1804:
1805: \paragraph{NGC~5129} is at the edge of the NPS so the local R45 value is very
1806: high. We indeed found a high local soft X-ray excess (Table 2). It is also near
1807: an extended feature on the \iras\ 100 $\mu$m map, which should explain the enhanced
1808: absorption. G07 presented results based on an \xmm\ observation. Our temperature
1809: profile is consistent with that from G07. However, our $c_{500}$ is smaller
1810: than G07's (3.43$^{+1.72}_{-1.22}$ vs. 7.7$\pm$1.3).
1811:
1812: \paragraph{UGC 2755} is one of the faintest and most gas poor systems
1813: in our sample ($f_{\rm gas, 2500} = 0.030\pm0.005$). It is one of the two groups with
1814: an entropy bump detected ($\S$8.1).
1815:
1816: \paragraph{NGC~4325} has a luminous cooling core.
1817: NGC~4325 was studied by G07, who also analyzed an \xmm\ observation.
1818: The explored radial range in spectral analysis is similar in both work
1819: (228 kpc vs. 232 kpc in our work).
1820: Our temperature profile agrees well with that derived by G07. Our $c_{500}$,
1821: $M_{2500}$ and $f_{\rm gas, 2500}$ are consistent with G07's within 1 $\sigma$ errors.
1822:
1823: \paragraph{3C~442A} has the most peaky temperature profile in this sample.
1824: The sharp
1825: reconstructed temperature peak at 75 - 150 kpc, and the steepening of the
1826: surface brightness at $\sim$ 120 kpc, may best be explained by the second
1827: most luminous radio source in our sample (only after Abell~2462). From the
1828: NVSS image, the two radio lobes of 3C~442A extend to $\sim 5'$ (or 152 kpc) in radius.
1829: Shock heating by the radio source may explain the high temperature peak.
1830: Because of this high temperature peak, the total mass density profile is not
1831: physically meaningful within the central $\sim$160 kpc, which casts doubt on
1832: the assumption of hydrostatic equilibrium within the central 160 kpc. Therefore,
1833: $c_{500}$ cannot be constrained reliably.
1834:
1835: \paragraph{ESO~552-020} was also studied by G07. Our temperature profile is
1836: consistent with G07's. Our $c_{500}$ is also consistent with G07's.
1837:
1838: \paragraph{IC~1262} has rich substructure within its core.
1839: However, beyond the central 60 kpc radius, it appears symmetrical and relaxed.
1840: Its surface brightness profile shows a sharp break at $\sim$ 200 kpc.
1841: Unlike 3C~442A and A2462, its radio source is faint. However,
1842: this may be caused by past AGN activity, which
1843: was also suggested to explain the rich structures in the group core by
1844: Trinchieri et al. (2007). The sharp temperature decline beyond 200 kpc
1845: also supports this scenario, as the regions between 80 and 200 kpc radius
1846: may have been recently heated (Fig. 4). The adiabatic sound speed in $\sim$ 1.9 keV gas
1847: is $\sim$ 700 km/s. Thus, a Mach 1.5 shock will travel to the current position
1848: in $\sim$ 200 Myr, which is consistent with the typical duty cycle of radio AGN.
1849:
1850: \paragraph{ESO~306-017} has been studied before (Sun et al. 2004; G07).
1851: There is an adjacent \chandra\ pointing targeted at the $z=0.64$ cluster
1852: RDCS~J0542-4100 (ObsID 914) that we used to constrain the local soft CXB.
1853: Our temperature profile covers a wider radial range than G07's. In the overlapping
1854: region, our temperature profile is consistent with G07's. Our results on $r_{500}$
1855: and $M_{500}$ are also consistent with G07's.
1856:
1857: \paragraph{NGC~5098} was also studied by G07. The \iras\ $100\mu$m map shows the
1858: presence of Galactic cirrus around the group and we indeed find extra absorption
1859: (Table 2). There is a second group in the field (Mahdavi et al. 2005) and the
1860: region around it is excluded in our analysis. Our temperature profile and
1861: results on $r_{500}$, $M_{500}$ and $c_{500}$ are consistent with G07's.
1862:
1863: \paragraph{UGC~842} was also studied by G07. Our temperature profile and
1864: results on $r_{500}$, $M_{500}$ and $c_{500}$ are consistent with G07's.
1865:
1866: \paragraph{A2717} was studied by Pratt \& Arnaud (2005), G07 and Snowden et al. (2008)
1867: with the \xmm\ data. Our temperature profile is consistent with the profiles
1868: derived by G07 and Snowden et al. (2008).
1869: Our derived $r_{500}$ and $M_{500}$ are consistent with
1870: those derived by G07, while our $c_{500}$ (2.15$^{+0.36}_{-0.32}$) is close to G07's value,
1871: 3.0$\pm$0.2 and the result by Pratt \& Arnaud (2005), 2.8$\pm$0.2 (converted from
1872: their $c_{200}$ assuming an NFW profile).
1873:
1874: \paragraph{AS1101} (or S\'{e}rsic 159-03) is the most gas-rich system in our sample.
1875: Its enclosed gas
1876: fraction at 0.1$r_{500}$ ($\sim$0.06) is almost 3 times the average of other groups
1877: in the sample. Its gas fraction at $r_{500}$ (0.114$^{+0.021}_{-0.020}$, extrapolated)
1878: is comparable to those of 5 - 7 keV clusters. The \chandra\
1879: exposure is short but the best-fit values of our temperature profile
1880: agree well with the \xmm\ results by Snowden et al. (2008).
1881: The derived $c_{500}$ in this work (5.05$^{+2.37}_{-1.34}$) is consistent with
1882: the result by Piffaretti et al. (2005), 4.33$\pm$0.51.
1883:
1884: \paragraph{A1991} was also studied by V05 and V06. Our derived system properties
1885: at $r_{500}$ and $r_{2500}$ agree very well
1886: with those of V05 and V06. The temperature decline at large radii in this system
1887: was also found from the \xmm\ data by Snowden et al. (2008) with consistent values.
1888:
1889: \paragraph{A2462} hosts a small corona (with a radius of $\lsim$ 4 kpc)
1890: at the center, without a large cool core. This is common for BCGs
1891: (Sun et al. 2007; also see $\S$8.2). The \chandra\ surface brightness
1892: profile shows a significant break at $\sim$ 180 kpc, which is about the size
1893: of the central radio source from the NVSS image. The central radio source in A2462 is the most
1894: luminous one in our sample. It may have heated the group core, as shown by the
1895: high temperature and entropy beyond the central corona.
1896:
1897: \paragraph{RXJ~1159+5531} has been studied by V05, V06 and G07. Our temperature
1898: profile agrees well with both V05 and G07. The derived gas fraction and total mass
1899: at $r_{2500}$ are well consistent with those from V06. $r_{500}$ is also consistent.
1900: Our $c_{500}$ (2.95$^{+1.16}_{-0.90}$) lies between the results from V06 (1.70$\pm$0.29)
1901: and from G07 (5.6$\pm$1.5). Our $r_{500}$ and $M_{500}$ are consistent with G07's.
1902:
1903: \paragraph{A2550} is in a large filamentary structure that connects with A2554
1904: ($z=0.111$, 17.5$'$ on the northeast), while A2556 ($z=0.087$) is 21$'$ to the east.
1905: Based on the derived surface brightness profiles centered on each system, we exclude
1906: regions within 13.7$'$ and 14.1$'$ (in radius) of A2554 and A2556 respectively,
1907: which are about 1.4 times $r_{500}$ of each system.
1908: There is also an X-ray clump south of A2550's core that is excised.
1909:
1910: \end{appendix}
1911:
1912: \begin{references}
1913:
1914: \reference{} Anders, E., \& Grevesse N. 1989, Geochimica et Cosmochimica Acta, 53, 197
1915: \reference{} Akritas, M. G., \& Bershady, M. A. 1996, ApJ, 470, 706
1916: \reference{} Arabadjis, J. S., \& Bregman, J. N. 1999, ApJ, 510, 806
1917: \reference{} Arnaud, M., Pointecouteau, E., Pratt, G. W. 2005, A\&A, 441, 893
1918: \reference{} Arnaud, M., Pointecouteau, E., Pratt, G. W. 2007, A\&A, 474, L37
1919: \reference{} B\^{i}rzan, L., Rafferty, D. A., McNamara, B. R., Wise, M. W., Nulsen, P. E. J. 2004, ApJ, 607, 800
1920: \reference{} Borgani, S. et al. 2004, MNRAS, 348, 1078
1921: \reference{} Borgani, S. et al. 2005, MNRAS, 361, 233
1922: \reference{} Borgani, S., \& Viel, M. 2008, MNRAS, submitted
1923: \reference{} Bullock, J. S. et al. 2001, MNRAS, 321, 559
1924: \reference{} Buote, D. A., Lewis, A. D.; Brighenti, F., Mathews, W. G. 2003, ApJ, 595, 151
1925: \reference{} Buote, D. A. et al. 2007, ApJ, 664, 123
1926: \reference{} Croton, D. J. et al. 2006, MNRAS, 365, 11
1927: \reference{} De Grandi, S., \& Molendi, S. 2002, ApJ, 567, 163
1928: \reference{} Dickey, J. M., \& Lockman, F. J. 1990, ARA\&A, 28, 215
1929: \reference{} Ettori, S. 2000, MNRAS, 318, 1041
1930: \reference{} Finoguenov, A., Reiprich, T. H., B\"{o}hringer, H. 2001, A\&A, 368, 749
1931: \reference{} Finoguenov, A., Jones, C., B\"{o}hringer, H., Ponman, T. 2002, ApJ, 578, 74
1932: \reference{} Finoguenov, A., Ponman, T. J., Osmond, J. P. F., Zimer, M. 2007, MNRAS, 374, 737
1933: \reference{} Fujimoto, R. et al. 2007, PASJ, 59, 133
1934: \reference{} Gastaldello, F. et al. 2007, ApJ, 669, 158 (G07)
1935: \reference{} Gilfanov, M. 2004, MNRAS, 349, 146
1936: \reference{} Gonzalez, A. H., Zaritsky, D., Zabludoff, A. I. 2007, ApJ, 666, 147
1937: \reference{} Hickox, R. C., \& Markevitch, M. 2006, ApJ, 645, 95 (HM06)
1938: \reference{} Humphrey, P. J. et al. 2006, ApJ, 646, 899
1939: \reference{} Jones, L. R., Ponman, T. J., Horton, A., Babul, A., Ebeling, H., Burke, D. J. 2003, MNRAS, 343, 627
1940: \reference{} Kalberla, P. M. W. et al. 2005, A\&A, 440, 775
1941: \reference{} Kay, S. T. 2004, MNRAS, 347, 13L
1942: \reference{} Kay, S. T., Thomas, P. A., Jenkins, A., Pearce, F. R. 2004, MNRAS, 355, 1091
1943: \reference{} Khosroshahi, H. G., Jones, L. R., Ponman, T. J. 2004, MNRAS, 349, 1240
1944: \reference{} Kim, D. W., \& Fabbiano, G. 2004, ApJ, 611, 846
1945: \reference{} Kim, M. et al. 2007, ApJ, 659, 29 (K07)
1946: \reference{} Komatsu, E. et al. 2008, ApJS, in press, arXiv:0803.0547
1947: \reference{} Koutroumpa, D., Acero, F., Lallement, R., Ballet, J., Kharchenko, V. 2007, A\&A, 475, 901
1948: \reference{} Kravtsov, A. V. et al. 2005, ApJ, 625, 588
1949: \reference{} Kravtsov, A., Vikhlinin, A., Nagai, D. 2006, ApJ, 650, 128
1950: \reference{} Lapi, A., Cavaliere, A., Menci, N. 2005, ApJ, 619, 60L
1951: \reference{} Leccardi, A., \& Molendi, S. 2008, A\&A, 486, 359 (LM08)
1952: \reference{} Lin, Y. T., Mohr, J. J., \& Stanford, S. A. 2003, ApJ, 591, 749
1953: \reference{} Loken, C. et al. 2002, ApJ, 579, 571
1954: \reference{} Mahdavi, A., B\"{o}hringer, H., Geller, M. J., Ramella, M. 2000, ApJ, 534, 114
1955: \reference{} Mahdavi, A., Finoguenov, A., B\"{o}hringer, H., Geller, M. J., Henry, J. P. 2005, ApJ, 622, 187
1956: \reference{} Markevitch, M., Forman, W. R., Sarazin, C. L., Vikhlinin, A. 1998, ApJ, 503, 77
1957: \reference{} Markevitch, M. et al. 2003, ApJ, 583, 70
1958: \reference{} Maughan, B. J. 2007, ApJ, 668, 772
1959: \reference{} McLaughlin, D. E. 1999, AJ, 117, 2398
1960: \reference{} Miller, E. et al. 2008, PASJ, 60, 95
1961: \reference{} Mohr, J. J., Mathiesen, B., Evrard, A. E. 1999, ApJ, 517, 627
1962: \reference{} Mulchaey, J. S., \& Zabludoff, A. I. 1999, ApJ, 514, 133
1963: \reference{} Mulchaey, J. S., Davis, D. S., Mushotzky, R. F., Burstein, D. 2003, ApJS, 145, 39
1964: \reference{} Mushotzky, R., Figueroa-Feliciano, E., Loewenstein, M., Snowden, S. L. 2003, astro-ph/0302267
1965: \reference{} Nagai, D., Vikhlinin, A., Kravtsov, A. V. 2007a, ApJ, 655, 98
1966: \reference{} Nagai, D., Kravtsov, A. V., Vikhlinin, A. 2007b, ApJ, 668, 1 (NKV07)
1967: \reference{} Osmond, J. P. F., Ponman, T. J. 2004, MNRAS, 350, 1511
1968: \reference{} Ostriker, J. P., \& McKee, C. F. 1988, Rev. Mod. Phys., 60, 1
1969: \reference{} Piffaretti, R., Jetzer, Ph., Kaastra, J. S., Tamura, T. 2005, A\&A, 433, 101
1970: \reference{} Pizzolato, F., Molendi, S., Ghizzardi, S., De Grandi, S. 2003, ApJ, 592, 62
1971: \reference{} Ponman, T. J., Cannon, Damian B., Navarro, J. F. 1999, Nature, 397, 135
1972: \reference{} Ponman, T. J., Sanderson, A. J. R., \& Finoguenov, A. 2003, MNRAS, 343, 331
1973: \reference{} Pratt, G. M., Arnaud, M. 2005, 429, 791
1974: \reference{} Pratt, G. M., Arnaud, M., \& Pointecouteau, E. 2006, A\&A, 429, 438
1975: \reference{} Pratt, G. W. et al. 2007 A\&A, 461, 71
1976: \reference{} Puchwein, E., Sijacki, D., Springel, V. 2008, ApJL, submitted, arXiv:0808.0494
1977: \reference{} Rasia, E., Tormen, G., Moscardini, L., 2004, MNRAS, 351, 237
1978: \reference{} Rasia, E. et al. 2008, ApJ, 674, 728
1979: \reference{} Rasmussen, J., Ponman, T. J. 2007, MNRAS, 380, 1554
1980: \reference{} Sanderson, A. J. R., Ponman, T. J., Finoguenov, A., Lloyd-Davies, E. J., \& Markevitch, M. 2003, MNRAS, 340, 989
1981: \reference{} Scannapieco, E. et al. 2005, ApJ, 635, L13
1982: \reference{} Snowden, S. L., McCammon, D., Burrows, D. N., Mendenhall, J. A. 1994, ApJ, 424, 714
1983: \reference{} Snowden, S. L. et al. 1997, ApJ, 485, 125
1984: \reference{} Snowden, S. L., Collier, M. R., \& Kuntz, K. D. 2004, ApJ, 610, 1182
1985: \reference{} Snowden, S. L., Mushotzky, R. F., Kuntz, K. D., Davis, D. S. 2008, A\&A, 478, 615
1986: \reference{} Sun, M., Forman, W., Vikhlinin, A., Hornstrup, A., Jones, C., Murray, S. S. 2003, ApJ, 598, 250
1987: \reference{} Sun, M., Forman, W., Vikhlinin, A., Hornstrup, A., Jones, C., Murray, S. S. 2004, ApJ, 612, 805
1988: \reference{} Sun, M. et al. 2007, ApJ, 657, 197
1989: \reference{} Trinchieri, G., Breitschwerdt, D., Pietsch, W., Sulentic, J., Wolter, A. 2007, A\&A, 463, 153
1990: \reference{} Vikhlinin, A., Forman, W., Jones, C. 1999, ApJ, 525, 47
1991: \reference{} Vikhlinin, A., McNamara, B. R., Hornstrup, A., Quintana, H., Forman, W., Jones, C., Way, M. 1999, ApJ, 520, L1
1992: \reference{} Vikhlinin, A., Markevitch, M., Murray, S. S., Jones, C., Forman, W., Van Speybroeck, L. 2005, ApJ, 628, 655 (V05)
1993: \reference{} Vikhlinin, A., Kravtsov, A., Forman, W., Jones, C., Markevitch, M., Murray, S. S., Van Speybroeck, L. 2006, ApJ, 640, 691 (V06)
1994: \reference{} Vikhlinin, A. 2006, ApJ, 640, 710
1995: \reference{} Vikhlinin, A. et al. 2008, ApJ, submitted, arXiv:0805.2207 (V08)
1996: \reference{} Voit, G. M., \& Bryan, G. L. 2001, Nature, 414, 425
1997: \reference{} Voit, G. M., Balogh, M. L., Bower, R. G., Lacey, C. G., Bryan, G. L. 2003, ApJ, 593, 272
1998: \reference{} Voit, G. M., \& Ponman, T. J. 2003, ApJ, 594, L75
1999: \reference{} Voit, G. M. 2005, Rev. Mod. Phys., 77, 207
2000: \reference{} Voit, G. M., Kay, S. T., \& Bryan, G. L. 2005, MNRAS, 364, 909 (VKB05)
2001: \reference{} Wargelin, B. J., Markevitch, M., Juda, M., Kharchenko, V., Edgar, R., Dalgarno, A. 2004, ApJ, 607, 596
2002:
2003: \end{references}
2004:
2005: \clearpage
2006:
2007: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2008: \begin{table}
2009: \begin{center}
2010: \caption{The group sample and the \chandra\ observations}
2011: {\scriptsize
2012: \begin{tabular}{ccccccccc} \hline \hline
2013: Group & $z$\tablenotemark{a} & D\tablenotemark{b} & ObsID\tablenotemark{c} & Date & Exposure\tablenotemark{d} & PSPC\tablenotemark{e} & $L_{\rm Ks}$\tablenotemark{f} & $L_{\rm 1.4 GHz}$\tablenotemark{g} \\ \hline
2014:
2015: NGC~1550 & 0.0124 & 51.4 & 3186 & 2002-01-08 & 9.7 & & 11.29 & 21.72 \\
2016: & & & 3187 & 2002-01-08 & 9.5 & & & \\
2017: & & & 5800 & 2005-10-22 & 43.0 (44.0) & & & \\
2018: & & & 5801 & 2005-10-24 & 44.0 (44.3) & & & \\
2019: NGC~3402 & 0.0153 & 63.5 & 3243 & 2002-11-05 & 22.9 (29.5) & & 11.40 & 22.18 \\
2020: A262 & 0.0163 & 67.8 & 2215 & 2001-08-03 & 28.7 (28.7) & 7.6 & 11.60 & 22.56 \\
2021: & & & 7921 & 2006-11-20 & 110.5 (110.5) & & & \\
2022: NGC~383 (3C~31) & 0.0170 & 70.8 & 2147 & 2000-11-06 & 41.0 (44.3) & 24.0 & 11.67 & 24.46 \\
2023: 3C~449 & 0.0171 & 71.2 & 4057 & 2003-09-18 & 14.3 (26.4) & 8.9 & 11.13 & 24.36 \\
2024: NGC~533 & 0.0185 & 77.1 & 2880 & 2002-07-28 & 29.7 (37.1) & 11.5 & 11.76 & 22.31 \\
2025: NGC~741 & 0.0185 & 77.3 & 2223 & 2001-01-28 & 28.7 (30.2) & 12.4 & 11.82 & 23.85 \\
2026: MKW4 & 0.0200 & 83.4 & 3234 & 2002-11-24 & 27.6 (29.7) & 9.4 & 11.81 & 22.15 \\
2027: A3581 & 0.0230 & 96.2 & 1650 & 2001-06-07 & 7.2 (7.2) & & 11.51 & 23.85 \\
2028: NGC~5129 & 0.0230 & 96.2 & 6944 & 2006-04-13 & 20.5 (20.5) & 5.5 & 11.63 & 21.90 \\
2029: & & & 7325 & 2006-05-14 & 25.6 (25.6) & & & \\
2030: NGC~1132 & 0.0233 & 97.5 & 3576 & 2003-11-16 & 20.5 (37.1) & & 11.64 & 21.79 \\
2031: UGC~2755 & 0.0245 & 102 & 2189 & 2001-02-07 & 7.2 (15.6) & 16.4 & 11.49 & 24.26 \\
2032: NGC~4325 & 0.0257 & 108 & 3232 & 2003-02-04 & 25.6 (29.9) & 4.8 & 11.31 & $<$21.32 \\
2033: HCG~51 & 0.0258 & 108 & 4989 & 2004-02-15 & 18.4 (18.9) & & 11.47 & $<$21.32 \\
2034: & & & 5304 & 2004-02-16 & 12.2 (12.7) & & & \\
2035: 3C~442A & 0.0263 & 110 & 5635 & 2005-07-27 & 26.6 & & 11.51, 11.40 & 24.70 \\
2036: & & & 6353 & 2005-07-28 & 13.6 & & & \\
2037: & & & 6359 & 2005-10-07 & 19.4 & & & \\
2038: & & & 6392 & 2006-01-12 & 31.7 & & & \\
2039: UGC~5088 & 0.0274 & 115 & 3227 & 2002-03-10 & 33.8 & & 11.26 & 21.04 \\
2040: NGC~6338 & 0.0274 & 115 & 4194 & 2003-09-17 & 39.9 & 3.5 & 11.73 & 22.95 \\
2041: NGC~4104 & 0.0282 & 118 & 6939 & 2006-02-16 & 35.8 (35.8) & 13.6 & 11.85 & 21.60 \\
2042: RBS~461 & 0.0296 & 124 & 4182 & 2003-03-11 & 22.0 & & 11.41 & 22.57 \\
2043: ESO~552-020 & 0.0314 & 132 & 3206 & 2002-10-14 & 18.7 & & 11.92 & $<$21.72 \\
2044: A1177 & 0.0316 & 133 & 6940 & 2006-12-27 & 32.8 (33.5) & & 11.73 & $<$21.50 \\
2045: IC~1262 & 0.0326 & 138 & 2018 & 2001-08-23 & 25.6 (30.5) & & 11.49 & 23.22 \\
2046: & & & 6949 & 2006-04-17 & 38.1 & & & \\
2047: & & & 7321 & 2006-04-19 & 36.8 & & & \\
2048: & & & 7322 & 2006-04-22 & 37.4 & & & \\
2049: NGC~6269 & 0.0348 & 147 & 4972 & 2003-12-29 & 38.6 & 10.1 & 11.94 & 23.11 \\
2050: ESO~306-017 & 0.0358 & 151 & 3188 & 2002-03-08 & 13.6 & & 11.91 & 22.58 \\
2051: & & & 3189 & 2002-03-09 & 13.8 & & & \\
2052: & & & 914* & 2000-07-26 & 50.4 & & & \\
2053: NGC~5098 & 0.0368 & 156 & 2231 & 2001-08-04 & 10.0 & & 11.44, 11.36 & 23.38 \\
2054: & & & 6941 & 2005-11-01 & 37.9 (38.6) & & & \\
2055: A1139 & 0.0398 & 169 & 9387 & 2008-03-28 & 10.0 & & 11.50 & 23.16 \\
2056: A160 & 0.0447 & 190 & 3219 & 2002-10-18 & 54.5 & & 11.69 & 24.65 \\
2057: UGC~842 & 0.0452 & 192 & 4963 & 2005-02-13 & 38.9 (39.2) & & 11.77 & 22.24 \\
2058: A2717 & 0.0498 & 213 & 6973 & 2006-08-17 & 46.1 & 8.5 & 11.82 & 24.52 \\
2059: & & & 6974 & 2006-04-10 & 18.9 & & & \\
2060: RXJ~1022+3830 & 0.0543 & 233 & 6942 & 2006-10-14 & 40.9 (41.4) & 10.3 & 11.69, - & $<$21.98 \\
2061: AS1101 & 0.0564 & 242 & 1668 & 2001-08-13 & 9.2 (9.4) & 11.9 & 11.89 & 24.27 \\
2062: ESO~351-021 & 0.0571 & 245 & 5784 & 2005-04-24 & 34.8 (35.1) & & 11.95 & 22.71 \\
2063: A3880 & 0.0581 & 250 & 5798 & 2004-12-23 & 19.7 & & 11.93 & 24.22 \\
2064: A1991 & 0.0587 & 253 & 3193 & 2002-12-16 & 36.8 (36.3) & 20.0 & 11.86 & 23.47 \\
2065: A1275 & 0.0637 & 275 & 6945 & 2006-02-05 & 49.1 (48.4) & & 11.31 & $<$22.13 \\
2066: A2092 & 0.0669 & 290 & 9384 & 2007-11-13 & 9.7 & & 11.44 & $<$22.17 \\
2067: RXJ~1206-0744 & 0.0680 & 295 & 9388 & 2007-11-15 & 10.0 & & 11.75 & $<$22.19 \\
2068: A1238 & 0.0720 & 313 & 4991 & 2004-03-28 & 18.2 & & 11.70 & 24.44 \\
2069: A744 & 0.0729 & 317 & 6947 & 2006-10-22 & 36.3 & & 11.80 & 22.24 \\
2070: A2462 & 0.0733 & 319 & 4159 & 2002-11-19 & 37.9 (38.6) & 3.4 & 11.82 & 25.36 \\
2071: RXJ~1159+5531 & 0.0808 & 354 & 4964 & 2004-02-11 & 69.6 (74.2) & & 11.97 & $<$22.34 \\
2072: A1692 & 0.0848 & 372 & 4990 & 2004-08-12 & 21.5 & & 11.88 & 22.33 \\
2073: & & & 6930* & 2006-03-06 & 76.0 & & & \\
2074: & & & 7289* & 2006-03-09 & 75.0 & & & \\
2075: A2550 & 0.122 & 550 & 2225 & 2001-09-03 & 57.3 (58.8) & & 11.62 & 23.18 \\
2076:
2077: \hline \hline
2078: \end{tabular}}
2079: \begin{flushleft}
2080: \leftskip 35pt
2081: \tablenotetext{a}{The group redshift is extracted from NASA/IPAC Extragalactic Database (NED)}
2082: \tablenotetext{b}{The luminosity distance of the group derived from its redshift}
2083: \tablenotetext{c}{The ObsIDs with * are observations that happened to be close to the interested groups. We used these observations to constrain the local soft CXB.}
2084: \tablenotetext{d}{Effective exposure after excluding time intervals of background flares. For observations with ACIS-S3 at the optical axis, two exposure values are listed, with the one in the brackets as the exposure for the FI chips.} \tablenotetext{e}{\rosat\ PSPC clean exposure time (in ksec) if the pointed PSPC observations exist.}
2085: \tablenotetext{f}{2MASS $K_{\rm s}$ band luminosity of the cD galaxy as shown as log(L$_{\rm Ks}$/L$_{\odot}$), $M_{K\odot}$ = 3.39 mag. There are three groups (3C~442A, NGC~5098 and RXJ~1022+3830)
2086: with two BCGs at the center. The $K_{\rm s}$ band magnitude of one central galaxy in RXJ~1022+3830
2087: with two BCGs at the center. The $K_{\rm s}$ band magnitude of one central galaxy in RXJ~1022+3830
2088: is unknown.}
2089: \tablenotetext{g}{1.4 GHz luminosity of the cD galaxy as shown as log(L$_{\rm 1.4 GHz}$/W Hz$^{-1}$) from the NRAO VLA Sky Survey (NVSS) or the Sydney University Molonglo Sky Survey (SUMSS), assuming a spectral index of -0.8, unless it can be derived from NED.}
2090: \end{flushleft}
2091: \end{center}
2092: \end{table}
2093: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2094:
2095: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2096: \begin{table}
2097: \begin{center}
2098: \caption{Absorption, radial range of the analysis and the local CXB}
2099: %\vspace{0.3cm}
2100: {\tiny
2101: \begin{tabular}{cccccccccc} \hline \hline
2102: Group & $N_{H}$\tablenotemark{a} & R45\tablenotemark{b} & outermost bin\tablenotemark{c} & r$_{\rm det, spe}$\tablenotemark{d} & r$_{\rm det, sur}$\tablenotemark{e} & $kT_{\rm hot}$\tablenotemark{f} & $f_{\rm 0.47 - 1.21 keV}$\tablenotemark{g} & $f_{\rm 2 - 8 keV}$\tablenotemark{h} & $f_{\rm 2 - 8 keV, expected}$ ($f_{\rm limit}$)\tablenotemark{i} \\
2103: & (10$^{20}$ cm$^{-2}$) & & (\%) & (kpc) & (kpc) & (keV) & & & \\ \hline
2104:
2105: NGC~1550 & 12.5 (10.0) & 125 & 11 & 355 & 364 & 0.20$^{+0.04}_{-0.07}$ & 4.5$^{+0.2}_{-0.4}$ & 13.5$^{+4.2}_{-4.5}$, 7.4$^{+4.5}_{-4.0}$ & $\sim$12.6 ($\sim$2.1), $\sim$8.5 ($\sim$0.56) \\
2106: NGC~3402 & 11.0 (4.0) & 119 & 31 & 239 & 239 & (0.25) & 5.0$^{+0.2}_{-0.6}$ & 6.5$^{+2.5}_{-3.5}$ & $\sim$6.8 ($\sim$ 0.27) \\
2107: A262 & 8.1 (5.8) & 152 & 19 & 363 & 387 (800) & 0.27$\pm$0.03 & 2.7$^{+0.6}_{-0.5}$ & 10$^{+4.0}_{-4.5}$, 8.4$^{+3.4}_{-4.0}$ & $\sim$9.4 (0.51-1.1), $\sim$7.7 (0.31-0.65) \\
2108: NGC~383 & 5.3 & 100 & 21 & 289 & 350 (500) & (0.25) & 3.2$^{+0.4}_{-0.9}$ & 9.3$^{+3.2}_{-2.6}$ & $\sim$8.4 (0.47 - 0.61) \\
2109: 3C 449 & 13.3 (9.0) & 161 & 32 & 230 & 230 (380) & (0.25) & 6.7$^{+0.8}_{-0.3}$ & 10.0$^{+5.0}_{-2.6}$ & $\sim$8.4 ($\sim$0.47) \\
2110: NGC~533 & 5.7 (3.0) & 103 & 26 & 238 & 340 (380) & 0.31$\pm$0.05 & 3.9$^{+0.9}_{-0.5}$ & 7.1$^{+2.5}_{-2.2}$ & $\sim$ 7.2 ($\sim$ 0.33) \\
2111: NGC~741 & 5.9 (4.3) & 100 & 33 & 271 & 360 (420) & 0.23$^{+0.05}_{-0.03}$ & 3.7$^{+0.5}_{-0.3}$ & 6.7$^{+3.0}_{-3.5}$ & $\sim$ 7.5 ($\sim$ 0.36) \\
2112: MKW4 & 3.1 (1.8) & 117 & 10 & 490 & 490 (720) & (0.25) & 4.7$^{+0.6}_{-0.3}$ & 7.5$^{+3.7}_{-3.9}$ & $\sim$8.4 (0.34 - 0.78) \\
2113: A3581 & 5.8 (4.5) & 296 & 25 & 322 & 450 & 0.27$^{+0.03}_{-0.02}$ & 12.5$^{+0.4}_{-0.2}$ & 10.6$^{+2.3}_{-5.5}$ & $\sim$10.1 ($\sim$0.98)\\
2114: NGC~5129 & 5.5 (1.7) & 298 & 45 & 214 & 222 (270) & 0.29$^{+0.02}_{-0.01}$ & 12.3$^{+0.7}_{-0.3}$ & 7.0$^{+3.9}_{-4.8}$ & $\sim$7.0 (0.17-0.41) \\
2115: NGC~1132 & 7.8 (5.5) & 92 & 33 & 284 & 310 & 0.28$\pm$0.04 & 2.5$^{+0.7}_{-0.3}$ & 7.3$^{+2.9}_{-2.5}$ & $\sim$6.5 ($\sim$0.22) \\
2116: UGC~2755 & 13.6 & 93 & 52 & 193 & 193 (205) & 0.26$^{+0.06}_{-0.04}$ & 3.3$^{+0.6}_{-0.5}$ & 10.4$^{+3.0}_{-3.7}$ & $\sim$ 8.7 (0.55-0.73) \\
2117: NGC~4325 & 2.4 & 143 & 23 & 232 & 232 (250) & (0.25) & 4.5$^{+1.2}_{-1.3}$ & 8.1$^{+5.4}_{-5.0}$ & $\sim$ 7.0 ($\sim$ 0.30) \\
2118: HCG~51 & 5.3 (1.1) & 118 & 30 & 329 & 360 & (0.25) & 2.2$^{+0.6}_{-0.3}$ & 6.8$^{+2.5}_{-3.0}$ & $\sim$ 7.2 ($\sim$ 0.33) \\
2119: 3C~442A & 6.1 (4.8) & 85 & 68 & 396 & 530 & (0.25) & 2.0$^{+0.5}_{-0.6}$ & 6.6$\pm$3.4 & $\sim$7.6 (0.25-0.7) \\
2120: UGC~5088 & 1.2 & 90 & 63 & 317 & 320 & (0.25) & 2.5$^{+0.3}_{-0.4}$ & 6.5$^{+2.4}_{-2.2}$ & $\sim$7.0 ($\sim$0.27) \\
2121: NGC~6338 & 2.3 & 133 & 43 & 349 & 360 (510) & (0.25) & 2.0$^{+0.5}_{-0.6}$ & 9.2$^{+4.6}_{-2.8}$ & $\sim$7.0 ($\sim$0.31) \\
2122: NGC~4104 & 4.4 (1.8) & 113 & 33 & 407 & 420 (550) & 0.23$^{+0.07}_{-0.05}$ & 2.7$^{+0.4}_{-0.2}$ & 8.1$^{+5.0}_{-5.8}$ & $\sim$7.0 ($\sim$0.30) \\
2123: RBS~461 & 17.7 (15.0) & 65 & 54 & 359 & 515 & (0.25) & 3.6$^{+0.3}_{-0.6}$ & 10.0$^{+6.0}_{-3.7}$ & $\sim$8.6 ($\sim$ 0.56) \\
2124: ESO~552-020 & 3.9 & 119 & 51 & 380 & 480 & (0.25) & 2.4$^{+0.4}_{-0.7}$ & 8.8$^{+3.4}_{-3.5}$ & $\sim$7.8 ($\sim$ 0.46) \\
2125: A1177 & 4.6 (1.1) & 103 & 35 & 418 & 420 & 0.15$\pm$0.02\tablenotemark{j} & 2.0$^{+0.4}_{-0.2}$ & 6.5$^{+2.8}_{-3.4}$ & $\sim$ 7.3 ($\sim$ 0.34) \\
2126: IC~1262 & 3.4 (1.8) & 162 & 74 & 375 & 390 & 0.32$\pm$0.04 & 5.4$^{+0.4}_{-1.0}$ & 8.1$^{+3.1}_{-2.2}$ & 6.0 - 7.3 (0.15 - 0.31) \\
2127: NGC~6269 & 5.2 & 106 & 48 & 439 & 510 (740) & (0.25) & 2.5$\pm$0.4 & 8.2$^{+2.3}_{-3.3}$ & $\sim$ 7.0 ($\sim$ 0.30) \\
2128: ESO~306-017 & 3.0 & 105 & 77 & 484 & 490 & 0.23$^{+0.04}_{-0.03}$ & 4.9$^{+0.6}_{-0.8}$ & 11.0$^{+4.9}_{-2.6}$ & $\sim$ 9.7 ($\sim$ 0.85) \\
2129: NGC~5098 & 6.8 (1.3) & 120 & 81 & 421 & 480 & 0.25$^{+0.05}_{-0.04}$ & 2.6$^{+0.4}_{-0.5}$ & 5.9$^{+3.7}_{-3.9}$, 9.5$^{+4.8}_{-4.2}$ & $\sim$ 10.1 ($\sim$ 0.95), $\sim$ 6.9 ($\sim$ 0.26) \\
2130: A1139 & 3.2 & 93 & 94 & 318 & 370 & (0.25) & 2.6$^{+0.4}_{-0.5}$ & 10.8$^{+5.8}_{-4.3}$ & $\sim$ 9.6 ($\sim$ 0.87) \\
2131: A160 & 4.0 & 100 & 62 & 507 & 640 & (0.25) & 3.2$\pm$0.5 & 7.8$^{+3.7}_{-2.9}$ & $\sim$ 6.7 ($\sim$ 0.22) \\
2132: UGC~842 & 4.0 & 99 & 25 & 410 & 510 & 0.26$\pm$0.03 & 4.4$^{+0.5}_{-0.6}$ & 7.5$^{+3.5}_{-2.6}$ & $\sim$6.8 ($\sim$ 0.22)\\
2133: A2717 & 1.2 & 122 & 37 & 730 & 800 (760) & 0.31$^{+0.07}_{-0.06}$ & 3.6$\pm$0.6 & 7.1$^{+3.2}_{-2.9}$, 9.4$^{+4.1}_{-3.2}$ & $\sim$ 6.3 ($\sim$ 0.20), $\sim$ 8.6 ($\sim$ 0.58) \\
2134: RXJ~1022+3830 & 3.3 (1.5) & 116 & 32 & 548 & 610 (670) & 0.30$\pm$0.06 & 2.5$\pm$0.6 & 8.8$^{+5.2}_{-3.8}$ & $\sim$ 7.0 ($\sim$ 0.27)\\
2135: AS1101 & 1.1 & 130 & 23 & 568 & 650 (850) & 0.28$\pm$0.07 & 3.5$\pm$0.9 & 13.0$^{+6.5}_{-5.0}$ & $\sim$10.1 ($\sim$ 0.95) \\
2136: ESO~351-021 & 2.4 & 129 & 34 & 511 & 515 & 0.28$^{+0.07}_{-0.04}$ & 2.5$^{+0.4}_{-0.6}$ & 7.2$^{+3.7}_{-2.9}$ & $\sim$6.8 ($\sim$ 0.26)\\
2137: A3880 & 1.2 & 149 & 37 & 779 & 810 & (0.25) & 3.5$^{+1.0}_{-0.6}$ & 9.2$^{+3.5}_{-3.9}$ & $\sim$8.0 ($\sim$ 0.43)\\
2138: A1991 & 2.5 & 280 & 31 & 655 & 820 (750) & 0.33$\pm$0.02 & 14.4$\pm$0.6 & 6.7$^{+2.9}_{-2.5}$ & $\sim$7.0 ($\sim$ 0.29)\\
2139: A1275 & 6.2 (2.0) & 91 & 42 & 495 & 530 & 0.21$^{+0.06}_{-0.09}$ & 2.1$^{+0.5}_{-0.4}$ & 7.7$^{+3.8}_{-3.4}$ & $\sim$6.3 ($\sim$ 0.16) \\
2140: A2092 & 2.2 & 130 & 79 & 555 & 600 & (0.25) & 3.1$^{+0.3}_{-0.4}$ & 6.3$^{+1.9}_{-1.8}$ & $\sim$ 9.2 ($\sim$ 0.79) \\
2141: RXJ~1206-0744 & 2.3 & 134 & 92 & 451 & 580 & (0.25) & 2.1$^{+0.6}_{-1.2}$ & 9.2$^{+4.3}_{-3.0}$ & $\sim$ 9.4 ($\sim$ 0.84) \\
2142: A1238 & 3.5 & 103 & 81 & 673 & 690 & 0.22$^{+0.09}_{-0.05}$ & 2.4$^{+0.4}_{-0.7}$ & 8.3$^{+4.2}_{-3.3}$ & $\sim$ 8.2 ($\sim$ 0.49) \\
2143: A744 & 3.5 & 104 & 92 & 433 & 600 & 0.20$^{+0.30}_{-0.06}$ & 2.7$^{+0.4}_{-0.5}$ & 6.4$^{+3.3}_{-2.6}$ & $\sim$ 6.8 ($\sim$ 0.24) \\
2144: A2462 & 3.1 & 158 & 50 & 564 & 650 (800) & 0.26$\pm$0.02 & 7.8$\pm$0.6 & 7.6$^{+5.0}_{-2.1}$ & $\sim$6.4 ($\sim$ 0.19) \\
2145: RXJ~1159 & 1.2 & 134 & 56 & 529 & 560 & 0.27$\pm$0.03 & 2.7$\pm$0.5 & 6.0$^{+3.1}_{-2.5}$ & $\sim$5.5 (0.08-0.14) \\
2146: A1692 & 1.8 & 158 & 80 & 766 & 770 & 0.26$^{+0.05}_{-0.04}$ & 4.2$\pm$0.5 & 7.3$^{+3.9}_{-3.0}$ & $\sim$8.0 ($\sim$ 0.39)\\
2147: A2550 & 1.9 & 130 & 75 & 597 & 740 & 0.28$\pm$0.03 & 4.8$\pm$0.7 & 7.8$^{+3.4}_{-3.1}$ & $\sim$ 5.9 ($\sim$ 0.13) \\
2148:
2149: \hline \hline
2150: \end{tabular}}
2151: \vspace{-1.2cm}
2152: \begin{flushleft}
2153: \leftskip 35pt
2154: \tablenotetext{a}{The absorption column density in our analysis. If the value from
2155: our spectral analysis ($\S$3.3) is consistent with the Galactic value from the
2156: Leiden/Argentine/Bonn (LAB) HI survey (Kalberla et al. 2005), the LAB value is used.
2157: Both values are listed if they are significantly different, with the number in
2158: brackets is the LAB value.}
2159: \tablenotetext{b}{The \rosat\ All-Sky Survey R45 flux (Snowden et al. 1997), in a unit of
2160: 10$^{-6}$ cts/s/arcmin$^{2}$, measured from an annulus centered on the source. The inner
2161: radius of the annulus is 0.4 - 0.8 deg (depending on the source size), while the outer
2162: radius is the inner radius + 0.4 deg. In a few cases, we have to use partial apertures
2163: to exclude the bright sources near our targets (e.g., A1692 and A2550).}
2164: \tablenotetext{c}{The fraction of the outermost radial bin for the spectral analysis covered
2165: by the \chandra\ data, compared with the full annulus. The median is 37\%. Note
2166: this fraction is always less than one because of point sources and chip gaps.}
2167: \tablenotetext{d}{The outermost radius for the \chandra\ spectral analysis}
2168: \tablenotetext{e}{The radius where X-ray surface brightness is detected at $> 2\sigma$.
2169: $r_{out}/r_{in}$ = 1.06-1.1. Note that our estimate of local background is conservative so our 2$\sigma$ range is smaller than V06's 3$\sigma$ range for A1991 and RXJ~1159+5531. The value in brackets is for PSPC if available.}
2170: \tablenotetext{f}{The temperature of the hotter component of the local soft CXB}
2171: \tablenotetext{g}{The 0.47 - 1.21 keV observed flux of the local soft CXB
2172: (in unit of 10$^{-12}$ ergs cm$^{-2}$ s$^{-1}$ deg$^{-2}$). The energy band is
2173: chosen to match that of the RASS R45 band.}
2174: \tablenotetext{h}{The 2 - 8 keV unabsorbed flux of the unresolved hard CXB
2175: (in unit of 10$^{-12}$ ergs cm$^{-2}$ s$^{-1}$ deg$^{-2}$)}
2176: \tablenotetext{i}{The expected 2 - 8 keV unabsorbed flux of the unresolved hard X-ray
2177: CXB (in unit of 10$^{-12}$ ergs cm$^{-2}$ s$^{-1}$ deg$^{-2}$), estimated
2178: based on the limiting flux (shown in brackets, in unit of 10$^{-14}$ ergs
2179: cm$^{-2}$ s$^{-1}$) and from the derived average relation in K07.}
2180: \tablenotetext{j}{The 0.1 keV component has zero normalization in this case.}
2181: \end{flushleft}
2182: \end{center}
2183: \end{table}
2184: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2185:
2186: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2187: \begin{table}
2188: \begin{center}
2189: \caption{Derived properties of groups (I: temperature, mass and gas fraction)}
2190: \vspace{0.3cm}
2191: {\tiny
2192: \begin{tabular}{cccccccccc} \hline \hline
2193: Group & $T_{500}$\tablenotemark{a} & $T_{2500}$ & $r_{500}$\tablenotemark{a} & $r_{2500}$ & $M_{500}$\tablenotemark{a} & $f_{\rm gas, 500}$\tablenotemark{a} & $f_{\rm gas, 2500}$ & $f_{\rm gas, 2500-500}$\tablenotemark{a} & $c_{500}$\tablenotemark{b} \\
2194: & (keV) & (keV) & (kpc) & (kpc) & (10$^{13}$ M$_{\odot}$) & & & & \\ \hline
2195:
2196: NGC~1550 & 1.06$\pm0.02$* & 1.18$\pm0.02$ & 465$^{+15}_{-19}$* & 222$\pm6$ & 3.18$^{+0.32}_{-0.37}$* & 0.097$^{+0.012}_{-0.009}$* & 0.056$\pm$0.003 & 0.149$^{+0.033}_{-0.024}$* & 4.93$^{+0.50}_{-0.46}$ (10) \\
2197: NGC~3402 & (0.74$\pm0.03$)& 0.80$\pm0.02$ & (380) & 205$^{+62}_{-22}$ & & & 0.032$^{+0.008}_{-0.014}$ & & \\
2198: A262 & (1.94$^{+0.11}_{-0.15}$) & 2.18$^{+0.06}_{-0.07}$ & (644) & 288$^{+17}_{-16}$ & & & 0.064$^{+0.005}_{-0.007}$ & & 3.48$^{+0.49}_{-0.45}$ (9) \\
2199: NGC~383 & (1.67$^{+0.13}_{-0.11}$) & 1.89$^{+0.17}_{-0.10}$ & (593) & 252$^{+35}_{-20}$ & & & 0.031$^{+0.003}_{-0.004}$ & & 3.09$^{+1.84}_{-1.30}$ (6) \\
2200: 3C 449 & (0.97$^{+0.04}_{-0.05}$) & 1.08$\pm0.04$ & (437) & 211$^{+11}_{-21}$ & & & 0.041$^{+0.006}_{-0.004}$ & & 2.97$^{+2.60}_{-1.64}$ (5) \\
2201: NGC~533 & (1.06$^{+0.08}_{-0.04}$) & 1.21$^{+0.06}_{-0.07}$ & (461) & 207$^{+18}_{-29}$ & & & 0.031$^{+0.005}_{-0.004}$ & & 4.58$^{+3.90}_{-2.34}$ (5) \\
2202: NGC~741 & (1.27$^{+0.08}_{-0.12}$) & 1.37$^{+0.08}_{-0.12}$ & (510) & 214$^{+20}_{-14}$ & & & 0.026$^{+0.004}_{-0.003}$ & & 3.05$^{+1.68}_{-1.22}$ (6) \\
2203: MKW4 & 1.58$\pm$0.09 & 1.75$^{+0.05}_{-0.04}$ & 538$^{+24}_{-29}$ & 259$^{+12}_{-8}$ & 4.85$^{+0.71}_{-0.68}$ & 0.086$\pm$0.009 & 0.047$^{+0.002}_{-0.003}$ & 0.134$^{+0.38}_{-0.26}$ & 3.93$^{+1.16}_{-0.78}$ (7) \\
2204: A3581 & (1.68$^{+0.10}_{-0.09}$) & 1.85$^{+0.12}_{-0.07}$ & (593) & 259$^{+22}_{-17}$ & & & 0.067$^{+0.006}_{-0.007}$ & & 7.43$^{+2.06}_{-1.40}$ (6) \\
2205: NGC~5129 & (0.76$\pm0.03$) & 0.83$\pm0.02$ & (384) & 174$^{+7}_{-9}$ & & & 0.035$^{+0.004}_{-0.003}$ & & 3.43$^{+1.72}_{-1.22}$ (5) \\
2206: NGC~1132 & (0.99$\pm0.04$) & 1.08$\pm0.03$ & (442) & 215$\pm$11 & & & 0.039$^{+0.005}_{-0.002}$ & & 1.77$^{+0.70}_{-0.58}$ (6) \\
2207: UGC~2755 & (0.76$\pm0.05$) & 0.83$^{+0.05}_{-0.05}$ & (384) & 188$^{+11}_{-12}$ & & & 0.031$\pm$0.005 & & \\
2208: NGC~4325 & (0.89$\pm0.03$) & 0.97$\pm0.03$ & (418) & 212$\pm$31 & & & 0.037$^{+0.014}_{-0.007}$ & & 5.19$^{+3.68}_{-2.36}$ (7) \\
2209: HCG~51 & (1.06$^{+0.04}_{-0.03}$) & 1.15$^{+0.04}_{-0.03}$ & (460) & 275$^{+13}_{-14}$ & & & 0.028$\pm$0.003 & & 2.11$^{+0.89}_{-0.56}$ (6) \\
2210: 3C~442A & 1.34$\pm0.04$* & 1.61$^{+0.05}_{-0.06}$ & 495$^{+12}_{-16}$* & 277$^{+16}_{-15}$ & 3.90$^{+0.22}_{-0.40}$* & 0.068$^{+0.006}_{-0.003}$* & 0.028$^{+0.003}_{-0.002}$ & 0.230$^{+0.210}_{-0.060}$* & \\
2211: UGC~5088 & 0.81$\pm0.03$* & 0.83$^{+0.02}_{-0.03}$ & 364$^{+27}_{-19}$* & 163$^{+10}_{-8}$ & 1.48$^{+0.36}_{-0.24}$* & 0.049$^{+0.008}_{-0.010}$* & 0.029$\pm$0.003 & 0.067$^{+0.021}_{-0.013}$* & 4.15$^{+1.66}_{-1.16}$ (5) \\
2212: NGC~6338 & (1.92$^{+0.06}_{-0.07}$) & 2.14$\pm$0.05 & (636) & 288$^{+13}_{-11}$ & & & 0.053$\pm$0.003 & & 5.27$^{+2.66}_{-1.80}$ (8) \\
2213: NGC~4104 & 1.41$^{+0.09}_{-0.06}$* & 1.64$^{+0.05}_{-0.08}$ & 535$^{+19}_{-20}$* & 274$\pm$12 & 4.85$^{+0.55}_{-0.53}$* & 0.069$^{+0.009}_{-0.006}$* & 0.036$\pm$0.003 & 0.137$^{+0.043}_{-0.025}$* & 4.29$^{+1.58}_{-1.12}$ (7) \\
2214: RBS~461 & (1.93$\pm0.10$) & 2.17$\pm0.06$ & (637) & 318$^{+25}_{-27}$ & & & 0.059$\pm$0.006 & & 5.45$^{+1.32}_{-1.16}$ (9) \\
2215: ESO~552-020 & (1.72$^{+0.13}_{-0.09}$) & 1.97$\pm0.09$ & (598) & 297$^{+27}_{-24}$ & & & 0.039$\pm$0.005 & & 4.27$^{+1.84}_{-1.22}$ (5) \\
2216: A1177 & 1.37$^{+0.06}_{-0.07}$* & 1.48$^{+0.06}_{-0.07}$ & 550$^{+29}_{-27}$* & 264$^{+20}_{-25}$ & 5.28$^{+0.84}_{-0.73}$* & 0.060$^{+0.009}_{-0.007}$* & 0.037$\pm$0.004 & 0.091$^{+0.033}_{-0.020}$* & 5.26$^{+4.68}_{-2.61}$ (6) \\
2217: IC~1262 & (1.73$^{+0.06}_{-0.05}$) & 1.91$^{+0.06}_{-0.05}$ & (600) & 330$^{+20}_{-40}$ & & & 0.045$^{+0.004}_{-0.010}$ & & \\
2218: NGC~6269 & 1.72$^{+0.12}_{-0.11}$* & 2.16$^{+0.08}_{-0.10}$ & 645$^{+50}_{-52}$* & 232$^{+27}_{-21}$ & 8.49$^{+1.97}_{-2.01}$* & 0.076$^{+0.011}_{-0.010}$* & 0.044$^{+0.003}_{-0.004}$ & 0.087$^{+0.027}_{-0.015}$* & \\
2219: ESO~306-017 & 2.37$^{+0.12}_{-0.14}$* & 2.54$^{+0.08}_{-0.10}$ & 690$^{+44}_{-30}$* & 337$^{+16}_{-14}$ & 10.3$^{+2.1}_{-1.3}$* & 0.081$^{+0.010}_{-0.011}$* & 0.052$^{+0.003}_{-0.004}$ & 0.119$^{+0.031}_{-0.026}$* & \\
2220: NGC~5098 & 0.96$\pm0.04$ & 1.05$^{+0.02}_{-0.03}$ & 398$^{+17}_{-33}$ & 195$^{+17}_{-13}$ & 2.00$^{+0.28}_{-0.46}$ & 0.108$^{+0.021}_{-0.012}$ & 0.048$^{+0.006}_{-0.004}$ & 0.205$^{+0.131}_{-0.060}$ & 4.29$^{+1.46}_{-1.06}$ (7) \\
2221: A1139 & (2.01$^{+0.33}_{-0.34}$) & 2.20$^{+0.35}_{-0.33}$ & (650) & 298$^{+37}_{-36}$ & & & 0.026$^{+0.016}_{-0.008}$ & & \\
2222: A160 & 1.68$^{+0.10}_{-0.10}$ & 2.05$^{+0.07}_{-0.06}$ & 626$^{+27}_{-31}$ & 286$^{+25}_{-23}$ & 7.90$^{+1.06}_{-1.10}$ & 0.085$^{+0.009}_{-0.008}$ & 0.043$^{+0.003}_{-0.004}$ & 0.121$^{+0.038}_{-0.023}$ & 2.73$^{+0.70}_{-0.60}$ (8) \\
2223: UGC~842 & 1.54$^{+0.14}_{-0.12}$* & 1.78$\pm0.09$ & 570$^{+84}_{-45}$* & 276$^{+23}_{-29}$ & 5.60$^{+2.60}_{-1.10}$* & 0.056$^{+0.012}_{-0.014}$* & 0.032$\pm$0.003 & 0.089$^{+0.064}_{-0.041}$* & 6.06$^{+3.31}_{-2.19}$ (6) \\
2224: A2717 & 2.43$^{+0.13}_{-0.12}$ & 2.60$^{+0.08}_{-0.09}$ & 732$^{+49}_{-32}$ & 342$^{+13}_{-11}$ & 12.9$^{+2.7}_{-1.7}$ & 0.076$\pm$0.010 & 0.053$\pm$0.003 & 0.098$^{+0.029}_{-0.025}$ & 2.15$^{+0.36}_{-0.32}$ (9) \\
2225: RXJ~1022+3830 & 1.94$^{+0.20}_{-0.14}$ & 2.36$^{+0.17}_{-0.12}$ & 631$^{+32}_{-41}$ & 326$^{+25}_{-23}$ & 8.00$^{+1.31}_{-1.40}$ & 0.075$^{+0.007}_{-0.013}$ & 0.038$^{+0.004}_{-0.005}$ & 0.134$^{+0.074}_{-0.046}$ & 4.03$^{+1.24}_{-0.80}$ (6) \\
2226: AS1101 & 2.57$^{+0.14}_{-0.11}$* & 2.69$^{+0.10}_{-0.09}$ & 768$^{+90}_{-65}$* & 362$^{+21}_{-18}$ & 14.1$^{+5.0}_{-3.4}$* & 0.114$^{+0.021}_{-0.020}$* & 0.085$\pm$0.008 & 0.129$^{+0.088}_{-0.041}$* & 5.05$^{+2.37}_{-1.34}$ (10) \\
2227: ESO~351-021 & 1.14$^{+0.07}_{-0.04}$ & 1.34$^{+0.07}_{-0.03}$ & 437$^{+78}_{-38}$ & 194$^{+28}_{-35}$ & 3.22$^{+1.80}_{-0.90}$ & 0.074$^{+0.013}_{-0.018}$ & 0.036$^{+0.007}_{-0.005}$ & 0.119$^{+0.066}_{-0.050}$ & 2.42$^{+1.20}_{-0.76}$ (6) \\
2228: A3880 & 2.49$^{+0.14}_{-0.12}$ & 2.75$\pm0.11$ & 799$^{+89}_{-68}$ & 309$^{+24}_{-12}$ & 14.9$^{+5.0}_{-3.5}$ & 0.088$^{+0.016}_{-0.021}$ & 0.074$^{+0.005}_{-0.007}$ & 0.090$^{+0.047}_{-0.022}$ & 4.03$^{+1.54}_{-1.26}$ (8) \\
2229: A1991 & 2.68$^{+0.10}_{-0.08}$ & 2.86$\pm0.07$ & 749$^{+45}_{-35}$ & 348$^{+13}_{-10}$ & 13.4$^{+2.5}_{-1.9}$ & 0.094$^{+0.010}_{-0.012}$ & 0.066$\pm$0.004 & 0.115$^{+0.038}_{-0.022}$ & 4.69$^{+0.76}_{-0.70}$ (9) \\
2230: A1275 & 1.46$^{+0.08}_{-0.07}$* & 1.63$^{+0.06}_{-0.07}$ & 592$^{+87}_{-50}$* & 202$^{+12}_{-9}$ & 6.90$^{+3.00}_{-1.67}$* & 0.094$^{+0.018}_{-0.031}$* & 0.069$\pm$0.004 & 0.100$^{+0.032}_{-0.035}$* & 4.10$^{+1.69}_{-1.27}$ (7) \\
2231: A2092 & 1.67$^{+0.13}_{-0.12}$* & 2.14$^{+0.17}_{-0.21}$ & 659$^{+45}_{-40}$* & 283$^{+34}_{-24}$ & 8.95$^{+1.81}_{-1.62}$* & 0.078$\pm$0.013* & 0.050$^{+0.006}_{-0.007}$ & 0.096$^{+0.027}_{-0.020}$* & \\
2232: RXJ~1206-0744 & (1.91$^{+0.20}_{-0.21}$) & 2.14$^{+0.19}_{-0.21}$ & (624) & 248$^{+28}_{-27}$ & & & 0.053$^{+0.009}_{-0.007}$ & & \\
2233: A1238 & (2.51$^{+0.27}_{-0.32}$)& 2.88$^{+0.35}_{-0.36}$ & (725) & 272$^{+63}_{-59}$ & & & 0.035$\pm$0.008 & & \\
2234: A744 & (2.24$^{+0.19}_{-0.16}$) & 2.49$\pm$0.16 & (681) & 325$^{+30}_{-24}$ & & & 0.046$^{+0.005}_{-0.006}$ & & 3.33$^{+1.66}_{-1.04}$ (6) \\
2235: A2462 & 2.32$^{+0.12}_{-0.10}$ & 2.62$\pm$0.09 & 646$\pm$30 & 327$^{+13}_{-12}$ & 8.80$^{+1.29}_{-1.19}$ & 0.099$^{+0.011}_{-0.009}$ & 0.052$\pm$0.003 & 0.180$^{+0.058}_{-0.034}$ & 3.41$^{+1.56}_{-0.84}$ (6) \\
2236: RXJ~1159 & 1.84$^{+0.14}_{-0.08}$* & 2.12$^{+0.09}_{-0.10}$ & 630$^{+77}_{-30}$* & 273$^{+15}_{-13}$ & 8.30$^{+3.10}_{-1.12}$* & 0.065$^{+0.007}_{-0.012}$* & 0.042$\pm$0.003 & 0.080$^{+0.019}_{-0.023}$* & 2.95$^{+1.16}_{-0.90}$ (6) \\
2237: A1692 & 2.61$^{+0.16}_{-0.24}$ & 3.06$^{+0.25}_{-0.22}$ & 658$^{+64}_{-47}$ & 356$^{+42}_{-32}$ & 9.70$^{+2.99}_{-1.91}$ & 0.090$^{+0.014}_{-0.020}$ & 0.043$\pm$0.006 & 0.216$^{+0.155}_{-0.087}$ & 5.46$^{+1.77}_{-1.19}$ (7) \\
2238: A2550 & 1.95$\pm$0.10 & 2.05$\pm$0.08 & 617$^{+74}_{-27}$ & 286$\pm$9 & 7.90$^{+2.90}_{-1.00}$ & 0.093$^{+0.011}_{-0.016}$ & 0.067$^{+0.003}_{-0.004}$ & 0.119$^{+0.027}_{-0.038}$ & 4.57$^{+0.92}_{-0.76}$ (7) \\
2239: \hline \hline
2240: \end{tabular}}
2241: \vspace{-1.2cm}
2242: \begin{flushleft}
2243: \leftskip 35pt
2244: \tablenotetext{a}{The $T_{500}$ in parenthesis are from the empirical relation between $T_{1500}$
2245: and $T_{500}$ ($\S$4). The $r_{500}$ in parenthesis are estimated from the $M_{500} - T_{500}$
2246: relation in this work. The values with an asterisk are for tier 2 groups from extrapolation.}
2247: \tablenotetext{b}{The value in parenthesis is the number of radial points between 40 kpc
2248: and $r_{\rm det, spe}$ (including 40 kpc and $r_{\rm det, spe}$) are used in the NFW fit ($\S$7.3).}
2249: \end{flushleft}
2250: \end{center}
2251: \end{table}
2252: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2253:
2254: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2255: \begin{table}
2256: \begin{center}
2257: \caption{Derived properties of groups (II: entropy)}
2258: %\vspace{0.3cm}
2259: {\scriptsize
2260: \begin{tabular}{ccccccc} \hline \hline
2261: Group & $K_{500}$\tablenotemark{a} & $K_{1000}$\tablenotemark{a} & $K_{1500}$\tablenotemark{a} & $K_{2500}$ & $K_{\rm 0.15 r500}$ & $K_{\rm 30 kpc}$ \\ \hline
2262:
2263: NGC~1550 & (297$^{+46}_{-42}$) & 253$\pm$23 & 228$\pm$15 & 198$^{+9}_{-8}$ & 94$\pm$3 & 49$\pm$1 \\
2264: NGC~3402 & & & (278$^{+218}_{-88}$) & 226$^{+100}_{-52}$ & 66$\pm$4 & 34$\pm$2 \\
2265: A262 & & (385$^{+87}_{-67}$) & 383$^{+73}_{-46}$ & 354$^{+43}_{-29}$ & 161$^{+9}_{-7}$ & 69$\pm$2 \\
2266: NGC~383 & & & (495$^{+83}_{-46}$) & 448$^{+71}_{-39}$ & 327$^{+23}_{-24}$ & 214$^{+28}_{-25}$ \\
2267: 3C 449 & & & (217$^{+28}_{-37}$) & 199$^{+22}_{-27}$ & 158$^{+11}_{-12}$ & 96$\pm$6 \\
2268: NGC~533 & & (336$^{+52}_{-76}$) & 317$^{+45}_{-66}$ & 289$^{+44}_{-50}$ & 171$^{+13}_{-11}$ & 103$\pm$6 \\
2269: NGC~741 & & (460$^{+64}_{-114}$) & 428$^{+57}_{-86}$ & 388$^{+53}_{-66}$ & 243$^{+36}_{-29}$ & 146$^{+15}_{-17}$ \\
2270: MKW4 & 574$^{+54}_{-96}$ & 454$^{+44}_{-59}$ & 388$^{+42}_{-39}$ & 332$^{+31}_{-24}$ & 161$\pm$6 & 80$^{+3}_{-4}$ \\
2271: A3581 & & (384$^{+84}_{-78}$) & 344$^{+72}_{-59}$ & 298$^{+49}_{-41}$ & 124$\pm$7 & 38$\pm$1 \\
2272: NGC~5129 & & (262$^{+32}_{-30}$) & 229$^{+23}_{-22}$ & 193$^{+18}_{-17}$ & 100$\pm$6 & 75$^{+6}_{-4}$ \\
2273: NGC~1132 & & & 336$^{+49}_{-45}$ & 238$\pm$26 & 97$^{+7}_{-6}$ & 68$^{+9}_{-7}$ \\
2274: UGC~2755 & & & (260$^{+65}_{-72}$) & 211$^{+50}_{-45}$ & 120$^{+15}_{-14}$ & 113$^{+24}_{-18}$ \\
2275: NGC~4325 & & & (195$^{+86}_{-70}$) & 179$^{+46}_{-36}$ & 84$^{+7}_{-6}$ & 39$\pm$2 \\
2276: HCG~51 & & (371$^{+196}_{-166}$) & 348$^{+120}_{-101}$ & 303$^{+49}_{-47}$ & 126$\pm$5 & 82$\pm$4 \\
2277: 3C~442A & (455$^{+49}_{-77}$) & 438$^{+34}_{-49}$ & 431$^{+29}_{-40}$ & 419$^{+33}_{-37}$ & 210$^{+27}_{-26}$ & 72$\pm$8 \\
2278: UGC~5088 & (414$^{+91}_{-191}$) & 326$^{+48}_{-49}$ & 288$^{+32}_{-33}$ & 230$\pm$23 & 84$^{+8}_{-7}$ & 51$^{+6}_{-4}$ \\
2279: NGC~6338 & & (401$^{+46}_{-45}$) & 371$^{+36}_{-35}$ & 334$^{+28}_{-23}$ & 239$^{+14}_{-11}$ & 108$\pm$5 \\
2280: NGC~4104 & (571$^{+74}_{-85}$) & 471$^{+42}_{-50}$ & 421$^{+32}_{-36}$ & 360$^{+26}_{-29}$ & 221$^{+16}_{-15}$ & 148$^{+12}_{-11}$ \\
2281: RBS~461 & & & (446$^{+59}_{-56}$) & 370$^{+43}_{-41}$ & 165$\pm$8 & 93$\pm$5 \\
2282: ESO~552-020 & & (580$^{+107}_{-124}$) & 550$^{+79}_{-87}$ & 460$^{+58}_{-61}$ & 190$\pm$9 & 97$^{+6}_{-5}$ \\
2283: A1177 & (648$^{+130}_{-111}$) & 480$^{+74}_{-70}$ & 397$^{+48}_{-47}$ & 317$^{+28}_{-45}$ & 179$^{+11}_{-12}$ & 138$^{+13}_{-11}$ \\
2284: IC~1262 & & & (489$^{+63}_{-99}$) & 411$^{+70}_{-63}$ & 134$\pm$4 & 47$\pm$2 \\
2285: NGC~6269 & (390$^{+189}_{-183}$) & 440$^{+56}_{-68}$ & 434$^{+54}_{-50}$ & 406$^{+38}_{-40}$ & 272$^{+18}_{-21}$ & 127$^{+7}_{-8}$ \\
2286: ESO~306-017 & (914$^{+203}_{-192}$) & 700$^{+108}_{-81}$ & 596$^{+74}_{-59}$ & 475$^{+44}_{-38}$ & 191$\pm$12 & 86$\pm$6 \\
2287: NGC~5098 & 217$\pm$43 & 210$^{+28}_{-26}$ & 205$\pm$21 & 194$^{+16}_{-17}$ & 84$^{+7}_{-8}$ & 45$\pm$2 \\
2288: A1139 & & & (486$^{+429}_{-206}$) & 431$^{+112}_{-100}$ & 368$^{+69}_{-48}$ & 261$^{+69}_{-43}$ \\
2289: A160 & 477$^{+86}_{-90}$ & 429$^{+34}_{-36}$ & 407$^{+29}_{-28}$ & 379$^{+33}_{-25}$ & 268$^{+17}_{-16}$ & 172$^{+31}_{-26}$ \\
2290: UGC~842 & (652$^{+231}_{-149}$) & 522$^{+90}_{-77}$ & 459$^{+60}_{-58}$ & 406$^{+48}_{-39}$ & 238$^{+22}_{-20}$ & 122$^{+11}_{-9}$ \\
2291: A2717 & 1022$^{+163}_{-181}$ & 817$^{+106}_{-102}$ & 673$^{+71}_{-77}$ & 502$^{+40}_{-41}$ & 186$\pm$11 & 75$\pm$5 \\
2292: RXJ~1022+3830 & 655$^{+89}_{-94}$ & 587$^{+69}_{-64}$ & 548$^{+62}_{-51}$ & 488$^{+72}_{-46}$ & 223$^{+33}_{-30}$ & 77$^{+6}_{-7}$ \\
2293: AS1101 & (476$^{+457}_{-218}$) & 465$^{+142}_{-91}$ & 441$^{+73}_{-53}$ & 386$^{+39}_{-36}$ & 120$^{+15}_{-13}$ & 29$\pm$1 \\
2294: ESO~351-021 & 309$^{+105}_{-84}$ & 315$^{+85}_{-40}$ & 310$^{+58}_{-34}$ & 289$^{+35}_{-38}$ & 175$^{+19}_{-20}$ & 57$\pm$4 \\
2295: A3880 & 624$^{+198}_{-236}$ & 681$^{+103}_{-98}$ & 560$^{+88}_{-52}$ & 420$^{+44}_{-35}$ & 176$\pm$20 & 50$\pm$3 \\
2296: A1991 & 752$^{+252}_{-275}$ & 653$^{+95}_{-142}$ & 568$^{+68}_{-94}$ & 463$^{+43}_{-63}$ & 166$^{+11}_{-23}$ & 39$\pm$1 \\
2297: A1275 & (464$^{+278}_{-98}$) & 324$^{+66}_{-44}$ & 284$^{+23}_{-30}$ & 231$^{+18}_{-17}$ & 130$^{+16}_{-14}$ & 59$\pm$3 \\
2298: A2092 & (423$^{+192}_{-140}$) & 449$^{+90}_{-66}$ & 439$^{+85}_{-69}$ & 401$^{+69}_{-57}$ & 202$^{+24}_{-21}$ & 200$^{+57}_{-34}$ \\
2299: RXJ~1206-0744 & & 432$^{+97}_{-104}$ & 402$^{+88}_{-80}$ & 354$^{+68}_{-67}$ & 201$^{+34}_{-41}$ & 102$^{+32}_{-35}$ \\
2300: A1238 & & 805$^{+347}_{-165}$ & 676$^{+167}_{-113}$ & 584$^{+114}_{-91}$ & 464$^{+129}_{-76}$ & 341$^{+200}_{-78}$ \\
2301: A744 & & (695$^{+249}_{-220}$) & 580$^{+134}_{-115}$ & 470$^{+72}_{-63}$ & 215$^{+16}_{-15}$ & 115$^{+12}_{-11}$ \\
2302: A2462 & 627$^{+102}_{-118}$ & 537$^{+79}_{-77}$ & 488$^{+62}_{-60}$ & 422$^{+48}_{-40}$ & 247$^{+12}_{-13}$ & 172$^{+20}_{-21}$ \\
2303: RXJ~1159 & (800$^{+200}_{-121}$) & 580$^{+78}_{-56}$ & 491$^{+49}_{-40}$ & 410$^{+35}_{-28}$ & 230$^{+16}_{-17}$ & 92$\pm$6 \\
2304: A1692 & 622$^{+202}_{-176}$ & 644$^{+186}_{-167}$ & 636$^{+174}_{-156}$ & 576$^{+126}_{-120}$ & 252$\pm$29 & 149$^{+30}_{-24}$ \\
2305: A2550 & 704$^{+132}_{-128}$ & 530$^{+65}_{-67}$ & 431$^{+46}_{-45}$ & 331$^{+27}_{-26}$ & 114$^{+11}_{-8}$ & 46$\pm$3 \\
2306:
2307: \hline \hline
2308: \end{tabular}}
2309: \vspace{-1.2cm}
2310: \begin{flushleft}
2311: \leftskip 35pt
2312: \tablenotetext{a}{The values in parentheses are from extrapolation (see $\S$4).}
2313: \end{flushleft}
2314: \end{center}
2315: \end{table}
2316: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2317:
2318:
2319: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2320: \begin{table}
2321: \begin{center}
2322: \caption{$K - T$ relations\tablenotemark{a}}
2323: \begin{tabular}{lcc} \hline \hline
2324: Relation & $K_{1}$ & $\alpha$ \\
2325: & (keV cm$^{2}$) & \\ \hline
2326:
2327: $K_{500} - T_{500}$ (tier 1) & 286$\pm$60 & 1.12$\pm$0.28 \\
2328: $K_{500} - T_{500}$ (tier 1+2) & 356$\pm$45 & 0.825$\pm$0.200 \\
2329: $K_{1000} - T_{500}$ (tier 1+2+3) & 310$\pm$19 & 0.832$\pm$0.100 \\
2330: $K_{1500} - T_{500}$ (all groups) & 288$\pm$14 & 0.790$\pm$0.078 \\
2331: $K_{2500} - T_{2500}$ (all groups) & 230$\pm$10 & 0.760$\pm$0.061 \\
2332: $K_{0.15 r500} - T_{2500}$ (all groups) & 114$\pm$9 & 0.778$\pm$0.124 \\
2333: $K_{30 \rm kpc} - T_{2500}$ (all groups) & 68$\pm$9 & 0.459$\pm$0.229 \\
2334: \hline
2335: $K_{500} - T_{500}$ (tier 1 + clusters) & 286$\pm$23 & 1.08$\pm$0.05 \\
2336: $K_{500} - T_{500}$ (tier 1+2 + clusters) & 329$\pm$25 & 0.994$\pm$0.054 \\
2337: $K_{1000} - T_{500}$ (tier 1+2+3 + clusters) & 303$\pm$13 & 0.887$\pm$0.034 \\
2338: $K_{2500} - T_{500}$ (all groups + clusters) & 252$\pm$8 & 0.740$\pm$0.027 \\
2339: $K_{0.15 r500} - T_{500}$ (all groups + clusters) & 137$\pm$9 & 0.494$\pm$0.047 \\
2340:
2341: \hline \hline
2342: \end{tabular}
2343: \vspace{-1cm}
2344: \tablenotetext{a}{$E(z)^{4/3} K = K_{1}$ ($T$ / 1 keV)$^{\alpha}$,
2345: where $K_{1}$ is the corresponding entropy at 1 keV. The last four relations
2346: include 14 clusters from V06 and V08. There is little correction between $K_{30 \rm kpc}$
2347: and $T_{2500}$, but we still list it for the studies of intrinsic scatter (Fig. 11).
2348: We used the BCES (Y$|$X) regression (Akritas \& Bershady 1996) as the temperature
2349: errors are smaller than the entropy errors.}
2350: \end{center}
2351: \end{table}
2352: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2353:
2354: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2355: \begin{table}
2356: \begin{center}
2357: \caption{$M_{500} - T_{500}$ relation\tablenotemark{a}}
2358: \begin{tabular}{cccc} \hline \hline
2359: Sample & $M_{3}$ & $\alpha$ & $r_{3}$ \\
2360: & (10$^{14} h^{-1} M_{\odot}$) & & ($h^{-1}$ Mpc) \\ \hline
2361:
2362: Tier 1 & 1.17$\pm$0.21 & 1.64$\pm$0.21 & 0.587$\pm$0.035 \\
2363: Tier 1 + clusters & 1.21$\pm$0.08 & 1.68$\pm$0.04 & 0.593$\pm$0.014 \\
2364: Tier 1+2 & 1.27$\pm$0.12 & 1.67$\pm$0.15 & 0.602$\pm$0.020 \\
2365: Tier 1+2 + clusters & 1.26$\pm$0.07 & 1.65$\pm$0.04 & 0.600$\pm$0.011 \\
2366: Tier 1+2 ($>$ 1 keV) + clusters & 1.27$\pm$0.06 & 1.60$\pm$0.03 & 0.602$\pm$0.010 \\
2367:
2368: \hline \hline
2369: \end{tabular}
2370: \vspace{-1cm}
2371: \tablenotetext{a}{$E(z) M_{500} = M_{3}$ ($T_{500}$ / 3 keV)$^{\alpha}$,
2372: where $T_{500}$ is spectroscopic temperature. $r_{3}$ is the corresponding
2373: scale in the $r_{500} - T_{500}$ relation ($E(z) r_{500} = r_{3}$ ($T_{500}$ /
2374: 3 keV)$^{\alpha/3}$). The cluster sample includes fourteen
2375: $T_{500} > 3.7$ keV systems from V08. We used the BCES orthogonal regression
2376: (Akritas \& Bershady 1996).}
2377: \end{center}
2378: \end{table}
2379: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2380:
2381: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2382: \begin{table}
2383: \begin{center}
2384: \caption{$M_{500} - Y_{\rm X, 500}$ relation\tablenotemark{a}}
2385: \begin{tabular}{ccc} \hline \hline
2386: Sample & $M_{\rm C}$ & $\alpha$ \\
2387: & (10$^{14} h^{-1} M_{\odot}$) & \\ \hline
2388:
2389: Tier 1 & 1.12$\pm$0.16 & 0.605$\pm$0.050 \\
2390: Tier 1 + clusters & 1.10$\pm$0.07 & 0.588$\pm$0.012 \\
2391: Tier 1+2 & 1.12$\pm$0.10 & 0.564$\pm$0.031 \\
2392: Tier 1+2 + clusters & 1.14$\pm$0.05 & 0.571$\pm$0.010 \\
2393:
2394: \hline \hline
2395: \end{tabular}
2396: \vspace{-1cm}
2397: \tablenotetext{a}{$E(z)^{2/5} M_{500} = M_{\rm C}$ ($Y_{\rm X, 500}$ / 4$\times10^{13}$ keV $M_{\odot}$)$^{\alpha}$.
2398: The cluster sample includes fourteen $T_{500} > 3.7$ keV systems from V08.
2399: We used the BCES orthogonal regression (Akritas \& Bershady 1996).}
2400: \end{center}
2401: \end{table}
2402: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2403:
2404: \clearpage
2405:
2406: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2407: \begin{figure}
2408: \centerline{\includegraphics[height=0.4\linewidth]{f1.eps}}
2409: %\vspace{-0.5cm}
2410: \caption{The spectra of the NGC~1550 in the outermost radial bin that is to the edge
2411: of the \chandra\ FOV (one from the S1 chip of the ObsID 5800 in the solid line and
2412: another from the S2 chip of the ObsID 3187 in the dashed line). Besides the still
2413: significant iron L hump from the group emission, the O hump from the soft CXB is also
2414: strong in the S1 spectrum, which allows a robust separation of these two components.
2415: }
2416: \end{figure}
2417: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2418:
2419: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2420: \begin{figure}
2421: \centerline{\includegraphics[height=0.5\linewidth,angle=270]{f2.ps}}
2422: %\vspace{-0.5cm}
2423: \caption{The 0.47 - 1.21 keV observed flux surface density of the soft
2424: CXB from the \chandra\ data vs. the RASS R45 flux (in RASS channels of 13 - 18)
2425: measured in an annulus around the target (inner radius of 0.4-0.8 deg and the
2426: outer radius is the inner radius + 0.4 deg). The three lines are the expected
2427: conversions between two fluxes, with the assumed two thermal components for
2428: the soft CXB (all with $T_{\rm cool}$ = 0.1 keV). The solid line is for:
2429: $T_{\rm hot}$ = 0.25 keV and NORM$_{\rm hot}$ / NORM$_{\rm cool}$ = 0.5. The
2430: dotted line is for the $T_{\rm hot}$ = 0.2 keV and NORM$_{\rm hot}$ /
2431: NORM$_{\rm cool}$ = 0.5. The dashed line is for: $T_{\rm hot}$ = 0.3 keV and
2432: NORM$_{\rm hot}$ / NORM$_{\rm cool}$ = 1. The
2433: assumed absorption is 4$\times10^{20}$ cm$^{-2}$. The total hard CXB
2434: (1.74$\times10^{-11}$ ergs s$^{-1}$ cm$^{-2}$ deg$^{-2}$ in the 2-8 keV band
2435: from K07) is added in the conversion. The good agreement on average can be seen.
2436: One should be aware that the RASS R45 flux is extracted in a much larger area
2437: surrounding the interested group range and some of the soft CXB may come from
2438: the time-variable SWCX emission. We also include 3C~296 ($T_{\rm hot} = 0.38 \pm 0.03$)
2439: in the high CXB flux end (star). It is an 1 keV group but was left out
2440: as the high local soft CXB (on NPS) prohibits deriving gas properties to $r_{2500}$.
2441: }
2442: \end{figure}
2443: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2444:
2445: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2446: \begin{figure}
2447: %\vspace{-0.2cm}
2448: \centerline{\includegraphics[height=1.25\linewidth]{f3.eps}}
2449: %\vspace{-0.5cm}
2450: \caption{The temperature profiles of 43 groups in our sample. The red lines
2451: are the reconstructed 3D temperature profile with the 1 $\sigma$ errors,
2452: derived from 1000 simulations. The black line is the best-fit projected
2453: temperature profile. The effective radius of each bin for the projected temperature
2454: profile is derived by weighting the projected emissivity profile with the actual
2455: spatial coverage of the each bin. $r_{2500}$ and $r_{500}$ are marked. The $r_{500}$
2456: in parentheses is estimated from the $M_{500} - T_{500}$ relation.
2457: }
2458: \end{figure}
2459: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2460: \clearpage
2461:
2462: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2463: \begin{figure}
2464: \centerline{\includegraphics[height=1.25\linewidth]{f4.eps}}
2465: %\vspace{-0.5cm}
2466: \caption{Continue from Fig. 3.
2467: }
2468: \end{figure}
2469: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2470: \clearpage
2471:
2472: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2473: \begin{figure}
2474: \centerline{\includegraphics[height=1.25\linewidth]{f5.eps}}
2475: %\vspace{-0.5cm}
2476: \caption{Continue from Fig. 4.
2477: }
2478: \end{figure}
2479: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2480: \clearpage
2481:
2482: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2483: \begin{figure}
2484: \centerline{\includegraphics[height=0.5\linewidth,angle=270]{f6.ps}}
2485: \caption{The best constrained 3D abundance profile in this sample with the best
2486: parametric fit ($\S$3.3) and 1 $\sigma$ errors from 1000 simulations.
2487: The 1000 simulated abundance profiles are all
2488: used to derive 1000 simulated temperature and density profiles.
2489: As errors of other abundance profiles in this sample are larger and there are
2490: fewer bins, our simple 3D abundance model ($\S$3.3) always fits well
2491: beyond the central 10 kpc.
2492: }
2493: \end{figure}
2494: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2495:
2496: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2497: \begin{figure}
2498: \centerline{\includegraphics[height=0.5\linewidth,angle=270]{f7.ps}}
2499: \caption{One example of the surface brightness profiles (\chandra\ +
2500: PSPC) with the best fits derived from the best-fit 3D temperature and abundance profiles.
2501: The \chandra\ profile is the upper one, while the PSPC profile is the lower one.
2502: We generated response files for each \chandra\ radial bin.
2503: Note that the density errors are derived from 1000 Monte Carlo simulations
2504: with 1000 simulated 3D temperature and abundance profiles folded in.
2505: The dashed and dotted lines mark $r_{2500}$ and $r_{500}$ (see Fig. 4 for NGC~6269).
2506: }
2507: \end{figure}
2508: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2509: \clearpage
2510:
2511: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2512: \begin{figure}
2513: \centerline{\includegraphics[height=0.47\linewidth]{f8.eps}}
2514: %\vspace{-0.5cm}
2515: \caption{Temperature profiles in the logarithmic and linear scales of $r_{500}$.
2516: Temperatures are scaled by $T_{2500}$. Despite the large scatter at small radii,
2517: the temperature profiles outside of 0.2 $r_{500}$ are generally similar. The thick
2518: solid line in the linear plot is the universal temperature profile (also projected)
2519: derived from the simulations in Loken et al. (2002). We simply used $T_{2500}$ to replace
2520: $T_{0}$ in Loken et al. (2002). Good agreement can be seen even though the normalization
2521: is not adjusted. The thick dashed line is a simple linear fit to the data (see $\S$5).
2522: }
2523: \end{figure}
2524: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2525:
2526: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2527: \begin{figure}
2528: \vspace{-1.2cm}
2529: \centerline{\includegraphics[height=0.5\linewidth,angle=270]{f9.ps}}
2530: \caption{Mean temperature profile of groups (black circles) and the 1-$\sigma$ scatter
2531: in dotted lines. The solid line is the best-fit from equ. 6. The dashed
2532: line is the mean temperature profile of 1-3 keV systems from Borgani et al.
2533: (2004) simulations. The data points in red triangles are the mean temperature
2534: profile from LM08 on 48 $kT >$ 3.3 keV clusters
2535: at $z$ = 0.1 - 0.3. Note the mean temperature $T_{\rm M}$ defined in
2536: LM08 is computed by fitting the profile with a constant after excluding the
2537: central 0.1 $r_{180}$ region. It should be smaller than $T_{2500}$ as
2538: $T_{2500}$ is emission-weighted within $r_{2500}$.
2539: Nevertheless, it is clear that the group temperature profiles are more
2540: peaky than those of clusters around the center.
2541: }
2542: \end{figure}
2543: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2544: \clearpage
2545:
2546: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2547: \begin{figure}
2548: %\vspace{-1cm}
2549: \centerline{\includegraphics[height=0.4\linewidth,angle=270]{f10.ps}}
2550: %\vspace{-0.5cm}
2551: \caption{The entropy profiles scaled by $K_{\rm 500, adi}$ ($\S$6).
2552: The dashed line represents the baseline entropy profile derived by VKB05
2553: with a power index of 1.1.
2554: The observed entropy profiles all lie above the baseline.
2555: }
2556: \end{figure}
2557: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2558:
2559: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2560: \begin{figure}
2561: \vspace{-5cm}
2562: \centerline{\includegraphics[height=1.27\linewidth]{f11.ps}}
2563: \vspace{-4.8cm}
2564: \caption{Entropy values at $r_{500}$, $r_{1000}$, $r_{1500}$, $r_{2500}$,
2565: 0.15 $r_{500}$ and 30 kpc vs. system temperature ($T_{500}$ or $T_{2500}$).
2566: The open data points (in the upper three panels) are based on extrapolation of
2567: the temperature and density profiles (see $\S$4 for details).
2568: The solid lines are the best-fits to our data from the BCES (Y$|$X) estimator (Table 5),
2569: while the dashed lines are the best-fits from the NKV07 simulations.
2570: The entropy excess above the NKV07 simulations (with cooling and SF) is
2571: significant at $r_{2500}$, while the agreement is better at larger radii.
2572: The dotted line in the $K_{500}$ plot represents the base-line entropy by
2573: VKB05 (or 1.40 $K_{\rm 500, adi}$, $\S$6). Note that the baseline has a slope
2574: of 1.1 rather than 1.0 as the $M_{500} - T_{500}$ relation used (the fourth row of
2575: Table 6) has a slope of 1.65 (rather than 1.5).
2576: It is also clear that the entropy values at 0.15 $r_{500}$ and 30 kpc
2577: show large intrinsic scatter. The measured intrinsic scatter in the $K - T$
2578: relations decreases with radius and stays the same from $r_{2500}$ to $r_{500}$ at
2579: $\sim$ 10\%.
2580: }
2581: \end{figure}
2582: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2583:
2584: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2585: \begin{figure}
2586: \centerline{\includegraphics[height=0.7\linewidth,angle=270]{f12.ps}}
2587: %\vspace{-5.1cm}
2588: \caption{$K - T_{500}$ relations at $r_{500}$, $r_{1000}$, $r_{2500}$ and 0.15 $r_{500}$
2589: for groups in our sample and 14 clusters from V08. The solid lines are the best-fits of
2590: the data from the BCES (Y$|$X) estimator (Table 5), while the dashed lines are
2591: the best-fits from the NKV07 simulations. The dotted lines in the $K_{500}$ and $K_{2500}$
2592: plots represent the base-line entropy by VKB05.
2593: The best-fit slopes from observations are
2594: also shown. The agreement between observations and the NKV07 simulations becomes better
2595: with increasing radius. At $r_{2500}$, the observed entropy values are on average 56\% - 22\%
2596: higher than those from the NKV07 simulations at 0.8 - 2.5 keV.
2597: At $r_{1000}$, the difference is 18\% - 8\% from 0.8 - 10 keV.
2598: At $r_{500}$, the NKV07 line is basically the same as our best-fit, which also
2599: has a slope expected from the self-similar relation (1.0).
2600: The $K_{0.15 r500} - T_{500}$ relation has significant scatter.
2601: Almost all clusters in the V08 sample have dense cool cores. Inclusion of
2602: non-cool-core clusters may steepen the $K - T$ relation at 0.15 $r_{500}$.
2603: }
2604: \end{figure}
2605: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2606:
2607: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2608: \vspace{-1cm}
2609: \begin{figure}
2610: \centerline{\includegraphics[height=0.96\linewidth,angle=270]{f13.ps}}
2611: %\vspace{-0.5cm}
2612: \caption{The ratios of the observed entropy values to the entropy baseline from
2613: VKB05 at $r_{500}$, $r_{1000}$ and $r_{2500}$. Fourteen clusters from V06 and V08 are included
2614: for comparison. The open data points are from extrapolation. We also show weighted
2615: means at each radius for the group and cluster samples. The observed entropy values are
2616: always larger than or comparable to the baseline at all radii, but the average ratios
2617: decrease with radius for both clusters and groups. The decrease
2618: is more dramatic in groups.
2619: }
2620: \end{figure}
2621: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2622:
2623: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2624: \begin{figure}
2625: \centerline{\includegraphics[height=0.8\linewidth,angle=270]{f14.ps}}
2626: %\vspace{-0.5cm}
2627: \caption{The entropy slopes between $r_{1500}$ - $r_{\rm det, spe}$, $r_{2500}$ - $r_{1500}$
2628: (or $r_{\rm det, spe}$), 0.15 $r_{500}$ - $r_{2500}$ and 30 kpc - 0.15 $r_{500}$ vs. $T_{500}$.
2629: Beyond 0.15 $r_{500}$, the slopes are always shallower than 1.1 and the weighted averages
2630: are all around 0.7.
2631: }
2632: \end{figure}
2633: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2634:
2635: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2636: \begin{figure}
2637: \centerline{\includegraphics[height=0.42\linewidth]{f15.eps}}
2638: %\vspace{-0.5cm}
2639: \caption{$M_{500} - T_{500}$ (left panel) and $M_{500} - Y_{\rm X, 500}$ (right panel) relations, combining the
2640: results from this work and V08. The open data points are tier 2 groups. The solid
2641: lines are the relations for the real mass in the NKV07 simulations, while the dotted
2642: lines are the relations for the mass derived under the assumption of hydrostatic
2643: equilibrium in the NKV07 simulations. The dashed lines are the best-fit relations from
2644: V08. The $M_{500} - T_{500}$ relation can be well described by a power law down to at
2645: least $M_{500}$ of 2 $\times 10^{13}$ h$^{-1}$ M$_{\odot}$, although the HSE mass
2646: may be systematically lower than the real mass. The $M_{500} - Y_{\rm X, 500}$ relation
2647: has a smaller scatter and the agreement with the NKV07 simulations is much better.
2648: Note as $Y_{\rm X, 500} \propto h^{-2.5}$, the $h$-dependence of $M_{500}$ should be
2649: $h^{1/2}$ for the self-similar relation, $M_{500} \propto Y_{\rm X, 500}^{3/5}$.
2650: We still use the $h^{-1}$ dependence to directly compare with the left panel
2651: }
2652: \end{figure}
2653: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2654:
2655: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2656: \begin{figure}
2657: \centerline{\includegraphics[height=0.45\linewidth,angle=270]{f16.ps}}
2658: %\vspace{-0.5cm}
2659: \caption{The density gradient at $r_{500}$ in terms of $\beta_{\rm eff}$
2660: (see the appendix of V06) for tier 1 and 2 groups that $M_{500}$ is derived.
2661: Most groups have $\beta_{\rm eff}$ of 0.55 - 0.75.
2662: }
2663: \end{figure}
2664: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2665:
2666: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2667: \begin{figure}
2668: \centerline{\includegraphics[height=0.45\linewidth,angle=270]{f17.ps}}
2669: \caption{The enclosed gas fraction within $r_{2500}$ vs. $T_{500}$ (groups +
2670: 14 clusters from V06 and V08). The solid line is the BCES fit to the group
2671: sample (0.0355$\pm$0.0018 ($T_{500}$/1 keV)$^{0.449\pm0.096}$),
2672: while the dotted line is the BCES fit to the group + cluster sample
2673: (0.0347$\pm$0.0016 ($T_{500}$/1 keV)$^{0.509\pm0.034}$). The intrinsic scatter on the
2674: $f_{\rm gas, 2500} - T_{500}$ relation is 22\%.
2675: Two dashed lines enclose the 1 $\sigma$ region of the universal baryon fraction
2676: derived from the \wmap\ 5-year data combined with the data of the Type Ia supernovae
2677: and the Baryon Acoustic Oscillations (0.1669$\pm$0.0063, Komatsu et al. 2008).
2678: }
2679: \end{figure}
2680: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2681:
2682: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2683: \begin{figure}
2684: \centerline{\includegraphics[height=0.95\linewidth,angle=270]{f18.ps}}
2685: \caption{The enclosed gas fraction within $r_{2500}$ is correlated with
2686: the entropy at 0.15 $r_{500}$, after the temperature dependence on both variables
2687: are removed (right panel). The slope is about -0.7. The intrinsic scatter on
2688: the scaled $f_{\rm gas, 2500}$ and the scaled
2689: $K_{0.15 r500}$ is 11\% and 14\% respectively, compared with 22\% and
2690: 29\% intrinsic scatter in their relations to temperature.
2691: This is primarily driven by density - density correlation, but quantitatively
2692: shows their connection.
2693: }
2694: \end{figure}
2695: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2696:
2697: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2698: \begin{figure}
2699: \centerline{\includegraphics[height=1.0\linewidth,angle=270]{f19.ps}}
2700: \caption{The enclosed gas fraction within $r_{500}$ (left panel) and between
2701: $r_{2500}$ - $r_{500}$ (right panel) vs. $T_{500}$. The open data points are tier 2 groups.
2702: The solid lines are fits to all 23 groups and 14 V08 clusters, 0.0708$\pm$0.0046
2703: ($T_{500}$/1 keV)$^{0.220\pm0.046}$, while the dashed line in the left panel is
2704: the fit excluding 12 tier 2 groups, 0.0776$\pm$0.0057 ($T_{500}$/1 keV)$^{0.158\pm0.047}$.
2705: The listed power law index can be compared with that in NKV07 simulations
2706: (0.152). Two horizontal dashed lines enclose the 1 $\sigma$ region of the
2707: universal baryon fraction (see the caption of Fig. 17). We also estimate
2708: the total baryon fraction within $r_{500}$ by adding the $f_{\rm gas, 500} - T_{500}$
2709: relation from 23 groups and 14 V08 clusters and the relation for stellar fraction
2710: from Lin et al. (2003) and Gonzalez et al. (2007), as shown by the two dotted
2711: lines. The upper one uses the Gonzalez et al. relation, while the lower one
2712: uses the Lin et al. relation (see $\S$7.4, note that the Gonzalez et al. relation
2713: includes the intracluster light). It appears that gas fraction between
2714: $r_{2500}$ - $r_{500}$ has no temperature dependence ($f_{\rm gas} \sim$ 0.12)
2715: on average, although there is still scatter. The solid line is the fit to all
2716: 23 groups and 14 clusters with a slope of 0.018$\pm$0.068. The dashed-dotted line
2717: is the expected average $f_{\rm gas, 2500-500}$ expected from the best-fit
2718: $f_{\rm gas, 500} - T_{500}$ and $f_{\rm gas, 2500} - T_{500}$
2719: scaling relations in this work (see $\S$7.2).
2720: }
2721: \end{figure}
2722: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2723:
2724: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2725: \begin{figure}
2726: \centerline{\includegraphics[height=0.4\linewidth,angle=270]{f20.ps}}
2727: \caption{The enclosed fraction of the count flux for 17 groups in our
2728: sample that $r_{500}$ is reached by \chandra\ or (and) PSPC. The dotted line
2729: marks the position of 0.15 $r_{500}$. There is large scatter for the
2730: enclosed fractions within 0.15 $r_{500}$ and $r_{2500}$ ($\sim$ 0.465 $r_{500}$ on
2731: average in this sample), mainly depending on the existence of a central cool core.
2732: }
2733: \end{figure}
2734: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2735:
2736: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2737: \begin{figure}
2738: \centerline{\includegraphics[height=0.45\linewidth,angle=270]{f21.ps}}
2739: %\vspace{-0.5cm}
2740: \caption{$c_{500} - M_{500}$ relation from this work. The filled and open
2741: circles are for tier 1 and 2 groups respectively. The open triangles are
2742: for tier 3 and 4 groups with mass estimated from the $M_{500} - T_{500}$
2743: relation derived in this work. The solid line is the median relation from the
2744: model of Bullock et al. (2001) with parameters F=0.001 and K=2.8, while
2745: the two dotted lines enclose the 1$\sigma$ region. We use:
2746: $\sigma_{8}$=0.817, $\Omega$$_{\rm M}$=0.279, $n_{s}$=0.96 (tilt) from
2747: Komatsu et al. (2008). Note that $\sigma_{8}$ affects the predicted $c_{500}$
2748: significantly (see Buote et al. 2007). The dashed line is the best-fit from
2749: G07. Our results show no significant mass dependence of $c_{500}$ in this
2750: narrow mass range and are generally consistent with the prediction under the current
2751: value of $\sigma_{8}$. The weighted mean of $c_{500}$ is 4.2.
2752: }
2753: \end{figure}
2754: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2755:
2756: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2757: \begin{figure}
2758: \centerline{\includegraphics[height=0.45\linewidth,angle=270]{f22.ps}}
2759: \caption{2MASS $K_{\rm s}$ band luminosity of the cD galaxy vs. $M_{500}$.
2760: The filled and open circles are for tier 1 and 2 groups respectively. The
2761: open triangles are for tier 3 and 4 groups with mass estimated from the
2762: $M_{500} - T_{500}$ relation derived in this work. The red points are fossil
2763: groups identified in this work. There is a general trend that more massive groups
2764: host more massive central galaxies. The dotted line represents a constant
2765: $M_{500} / L_{\rm Ks, cD}$ line. For the same system mass, the fossil groups
2766: host more luminous (or more massive) cDs than non-fossil groups.
2767: }
2768: \end{figure}
2769: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2770:
2771: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2772: \begin{figure}
2773: \centerline{\includegraphics[height=0.9\linewidth,angle=270]{f23.ps}}
2774: %\vspace{-0.5cm}
2775: \caption{Two groups with a significant entropy bump: UGC~2755 and 3C~449.
2776: Both hosts an active FR I radio source (see $\S$8.1).
2777: The dotted lines enclose the 1$\sigma$ error region, while the dashed lines
2778: represent the baseline entropy profile (estimated from the group's mass)
2779: with a slope of 1.1.
2780: }
2781: \end{figure}
2782: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2783:
2784: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2785: \begin{figure}
2786: \centerline{\includegraphics[height=0.45\linewidth,angle=270]{f24.ps}}
2787: %\vspace{-0.5cm}
2788: \caption{The spectral residual after the period D stowed background subtracted by the
2789: period E stowed background (in black). Their 9.5 - 12 keV fluxes are scaled to be the same.
2790: The residual is flat and is only $\sim$ 6\%. The red data points are
2791: the residual ($\sim$ 0\%) if the period D stowed background is scaled up by 5.7\%.
2792: Clearly after this special scaling, the 9.5 - 12 keV fluxes of two background are not the same,
2793: which implies a small spectral change for the BI PB from the period D to E.
2794: }
2795: \end{figure}
2796: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2797:
2798: \end{document}
2799: