0709.1561/MY.tex
1: \documentclass[letter,oldversion]{aa}
2: \usepackage{graphicx}
3: \usepackage{txfonts}
4: \usepackage{natbib}
5: \bibpunct{ (}{)}{;}{a}{}{,}
6: 
7: \begin{document}
8: %
9: 
10: \title{Calibration of the galaxy cluster $M_{\rm 500}$--$Y_{\rm X}$
11:   relation with XMM-Newton} 
12: \author{M. Arnaud \inst{1},
13:               E. Pointecouteau \inst{2} and
14:                       G.W. Pratt \inst{3}}
15: \offprints{M. Arnaud, \email{Monique.Arnaud@cea.fr}}
16: 
17:  \institute{
18:  $^1$ Laboratoire AIM, DAPNIA/Service d'Astrophysique - CEA/DSM - CNRS
19:  - Universit\'{e} Paris Diderot, B\^{a}t. 709, CEA-Saclay, F-91191
20:  Gif-sur- Yvette Cedex, France \\ 
21:  $^2$ CESR, 9 Av du colonel Roche, BP 44346, 31028 Toulouse Cedex 4,
22:  France\\ 
23:  $^3$ Max-Planck-Institut f\"ur extraterrestriche Physik, Giessenbachstra{\ss}e, 85748 Garching, Germany}
24: 
25:   \date{Received 24 August 2007; accepted 9 September 2007}
26:   \abstract 
27:    {The quantity $Y_{\rm X}$, the product of the
28:      X-ray temperature $T_{\rm X}$ and gas mass $M_{\rm g}$, has
29:      recently been proposed as a robust low-scatter mass indicator for
30:      galaxy clusters. 
31:     Using precise measurements from XMM-Newton data of a sample of 10
32:     relaxed nearby clusters, spanning a $Y_{\rm X}$ range of
33:      $10^{13}$--$10^{15}$\,M$_\odot$\,keV, we investigate  the
34:      $M_{500}$--$Y_{\rm X}$ relation.
35:      The $M_{500}$ -- $Y_{\rm
36:        X}$ data exhibit a power law relation with slope
37:      $\alpha=0.548 \pm 0.027$, close to the self-similar value (3/5) and
38:      independent of the mass range considered. However, the 
39:      normalisation is $\sim 20\%$  below the prediction from numerical
40:      simulations including cooling and galaxy feedback.  We discuss two effects that could contribute to the normalisation offset: an
41:      underestimate of the true mass due to  
42:      the HE assumption used in X-ray mass estimates, and an underestimate of
43:      the hot gas mass fraction in the simulations. A comparison of
44:      the functional form and scatter of the relations between
45:      various observables and the mass suggest that $Y_{\rm X}$
46:      may indeed be a better mass proxy than $T_{\rm X}$ or $M_{\rm
47:        g,500}$.}  
48:  \keywords{Cosmology: observations,  Cosmology: dark
49:       matter, Galaxies: cluster: general, (Galaxies) Intergalactic  
50: medium, X-rays: galaxies: clusters}
51: 
52:    \maketitle
53: %
54: %________________________________________________________________
55: \def\etal{et al.}
56: 
57: \def\Mgv{M_{\rm g,500}}
58: \def\Mg{M_{\rm g}}
59: \def\YX {Y_{\rm X}}
60: \def\TX {T_{\rm X}}
61: \def\fgv {f_{\rm g,500}}
62: \def\fg  {f_{\rm g}}
63: \def\kT {{\rm k}T}
64: \def\Mv {M_{\rm 500}}
65: \def \Rv {R_{500}}
66: \def\keV {\rm keV}
67: 
68: 
69: \def\MT {$M_{500}$--$T_{\rm X}$}
70: \def\MY {$M_{500}$--$Y_{\rm X}$}
71: \def\MMg {$M_{500}$--$M_{\rm g,500}$}
72: \def\MgT {$M_{\rm g,500}$--$T_{\rm X}$}
73: \def\MgY {$M_{\rm g,500}$--$Y_{\rm X}$}
74: 
75: \def\msol {{\rm M_{\odot}}}
76: 
77: \def\lesssim{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$<$}}}}
78: \def\gtrsim{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$>$}}}}
79: 
80: % satellites
81: \def \xmm {\hbox{\it XMM-Newton}}
82: \def \chandra {\hbox{\it Chandra }}
83: 
84: \section{Introduction}
85: 
86: All theoretical approaches characterise galaxy clusters in terms of
87: their mass.  Models  of
88: structure formation predict the space density, distribution
89: and physical properties of clusters as a function of mass and redshift
90: \citep[e.g. ][]{bertschinger98}. However, the mass is not easily measured.  
91: X-ray estimates from the hydrostatic equilibrium (HE) equation are
92: valid only for reasonably relaxed clusters and require temperature
93: profiles of high statistical quality; furthermore, the available
94: precision rapidly degrades with redshift.  
95: Based on the regularity of the cluster population, other X-ray
96: observables such as luminosity, temperature $\TX$, or gas mass $\Mg$,
97: have been used as proxies for the mass, e.g. to constrain cosmological
98: parameters using cluster surveys \citep{voi05}.  Studies of cluster
99: formation physics must also rely on mass proxies when considering
100: unbiased (i.e., covering a variety of dynamical 
101: states) or distant cluster samples \citep[e.g.][]{mau07}.  
102: The identification of the best mass proxy, and knowledge
103: of its exact relation to the mass, are therefore important.
104: 
105: The most commonly used mass proxy, $\TX$, is expected to
106: be closely related to the mass via the virial theorem.  
107: Significant progress on the calibration of the local $M$--$\TX$
108: relation for {\it relaxed} clusters has recently been made, with
109: excellent agreement now achieved between various observations
110: \citep{app05,vik06}, and comparison between observations and numerical
111: models including cooling and galaxy feedback showing agreement to the
112: $\sim10\%$  level \citep[e.g.][]{nag07b,app05}.  \citet{kra06}
113: recently proposed a new mass proxy,  $\YX=\TX\Mgv$, where
114: $\Mgv$ is the gas mass within $\Rv$, the radius corresponding to a
115: density contrast of  $\delta=500$. $\YX$ is related to the thermal
116: energy of the gas and is the X-ray analogue of the integrated SZ
117: Comptonisation parameter, $Y_{\rm SZ}$. The numerical simulations of \citeauthor{kra06} showed
118: that, as compared to $\TX$ or $\Mgv$,  $\YX$ is a better mass proxy,
119: in the sense that the intrinsic scatter was lower than for any other
120: mass indicator, regardless of cluster dynamical state (similar to
121: previous results for the $M$--$Y_{\rm SZ}$ relation, e.g.,  
122: \citealt{das04,mot05,nag06}). Furthermore, its evolution appears to be
123: close to the standard self-similar 
124: expectation.   
125: 
126: In this Letter, we present  the \MY\ relation derived from precise
127: \xmm\ data and compare it to the \MMg\ and \MT\ relations. The
128: \MY\  relation is discussed with respect to previous \chandra\ results
129: and theoretical expectations \citep{nag07b}.  Other relations between
130: observables,  such as the variation of the gas mass fraction $\fgv$
131: with mass,  are also investigated in order to shed new light on the
132: scatter and slope of the various mass-proxy relations.   
133: 
134: \begin{table*}
135:      \caption[]{Physical cluster parameters.  $\Mv$, $M_{\rm g, 500}$
136:        and  $f_{\rm g,500}$ are the total mass, gas mass and gas mass
137:        fraction respectively, within the radius $\Rv$, inside which
138:        the mean mass density is 500 times the critical
139:        density at the cluster redshift. $\TX$ is the
140:        spectroscopic temperature within $[0.15-0.75]R_{500}$ and
141:        $Y_{\rm X}= M_{\rm g, 500} \TX$. Values  are given for a
142:        $\Lambda$CDM cosmology  with $\Omega_{\rm m}=0.3$,
143:        $\Omega_{\Lambda}=0.7$, $H_0=70$~km s$^{-1}$ Mpc$^{-1}$. Errors
144:        are $1\,\sigma$.} 
145:      \label{tab:clu}
146:      \begin{center}
147: %    $$
148:     \begin{tabular}{lccccccccc}
149:     \hline
150:     \hline
151:     Cluster & $z$ & $\TX (\keV)$&  $M_{500}~(10^{14}\, \msol)$& $M_{\rm g,500}~(10^{13}\, \msol)$&$Y_{\rm X}~(10^{13}\, \msol\,\keV)$& $f_{\rm g,500}$ \\
152:     \hline
153: \object{A 1983}& $0.0442$ & $2.18\pm0.09$ & $1.09_{-0.29}^{+0.45}$ & $ 0.64_{- 0.08}^{+ 0.10}$ & $ 1.39_{-0.18}^{+0.23}$ & $0.058_{-0.017}^{+0.026}$\\
154: \object{MKW9}& $0.0382$ & $2.43\pm0.24$ & $0.88_{-0.18}^{+0.23}$ & $ 0.49_{- 0.05}^{+ 0.06}$ & $ 1.19_{-0.17}^{+0.18}$ & $0.055_{-0.012}^{+0.016}$\\
155:  \object{A 2717}& $0.0498$ & $2.56\pm0.06$ & $1.10_{-0.11}^{+0.13}$ & $ 1.02_{- 0.05}^{+ 0.06}$ & $ 2.60_{-0.15}^{+0.16}$ & $0.093_{-0.011}^{+0.012}$\\
156: \object{A 1991}& $0.0586$ & $2.71\pm0.07$ & $1.20_{-0.12}^{+0.13}$ & $ 1.25_{- 0.06}^{+ 0.06}$ & $ 3.39_{-0.19}^{+0.19}$ & $0.104_{-0.011}^{+0.012}$\\
157:  \object{A 2597}& $0.0852$ & $3.67\pm0.09$ & $2.22_{-0.21}^{+0.23}$ & $ 2.51_{- 0.08}^{+ 0.09}$ & $ 9.21_{-0.38}^{+0.39}$ & $0.113_{-0.011}^{+0.012}$\\
158:  \object{A 1068}& $0.1375$ & $4.67\pm0.11$ & $3.87_{-0.27}^{+0.29}$ & $ 3.77_{- 0.10}^{+ 0.10}$ & $17.6_{-0.62}^{+0.63}$ & $0.097_{-0.007}^{+0.008}$\\
159: \object{A 1413}& $0.1430$ & $6.62\pm0.14$ & $4.82_{-0.40}^{+0.44}$ & $ 7.55_{- 0.27}^{+ 0.28}$ & $50.0_{-2.1}^{+2.1}$ & $0.157_{-0.014}^{+0.016}$\\
160: \object{A 478}& $0.0881$ & $7.05\pm0.12$ & $7.57_{-1.02}^{+1.20}$ & $ 9.33_{- 0.43}^{+ 0.46}$ & $65.8_{-3.2}^{+3.4}$ & $0.123_{-0.017}^{+0.020}$\\
161: \object{PKS 0745-191}& $0.1028$ & $7.97\pm0.28$ & $7.27_{-0.70}^{+0.80}$ & $10.71_{- 0.47}^{+ 0.50}$ & $85.3_{-4.8}^{+5.0}$ & $0.147_{-0.016}^{+0.018}$\\
162: \object{A 2204}& $0.1523$ & $8.26\pm0.22$ & $8.39_{-0.77}^{+0.86}$ & $10.55_{- 0.39}^{+ 0.40}$ & $87.2_{-4.0}^{+4.1}$ & $0.126_{-0.012}^{+0.014}$\\
163:     \hline
164:     \end{tabular}
165: \end{center}
166: \end{table*}
167: 
168: \section{The data}
169: 
170: \subsection{The sample}
171: \label{sample}
172: The sample comprises ten nearby 
173: morphologically
174: relaxed clusters in the temperature range $[2-9]\,\keV$. We have
175: previously used \xmm\ data to study  the structural and
176: scaling properties of the total mass \citep{pap05b,app05} and of the
177: entropy \citep{pap06}; the  $\TX$, $\Mv$ and
178: $\Rv$ values derived in these papers are used in
179: the present Letter (Table~\ref{tab:clu}). The observations and data
180: reduction steps are 
181: fully described in \citet{pap05b}. $\Mv$ values were derived from NFW
182: model fits to mass profiles measured  down to 
183: $\delta_{\rm  obs}=600-700$,  
184: except for the two lowest mass clusters
185: ($\delta_{\rm obs}\sim1400$), 
186: thus the $\Mv$ estimates involve some
187: data extrapolation. However, as discussed  in  \citet{app05}, the
188: $\Mv$ estimates rely solely on the physically and
189: observationally-motivated assumption that the best fitting NFW model
190: remains valid between $\delta_{\rm obs}$ and $\delta=500$, and not on
191: a less reliable extrapolation of density and temperature profiles.
192: The temperature $\TX$ was derived from a single-temperature fit to the
193: integrated spectrum in the $[0.1$--$0.5]\,R_{200}$ aperture,   the
194: inner radius defined to exclude the 
195: cooling core region and the outer radius chosen to ensure a 
196: sufficiently
197: precise $\TX$ estimate over the whole mass range.   This aperture corresponds to  $[0.15$--$0.75]\,\Rv$, while an aperture of
198: $[0.15$--$1]\,\Rv$ is used for the definition of $\TX$ in numerical
199: simulations and in the Chandra analysis \citep{nag07b}.  For typical decreasing temperature profiles, these $\TX\ $ values are expected to be 
200: slightly 
201: smaller  by $3$--$6\%$
202:   \footnote{The difference is $3\%$ for A1413 \citep[][]{app05}, a  cluster for which the temperature profiles measured up to $\Rv$  both  with \xmm\ and \chandra\ are in excellent agreement
203:   \citep{pa02,vik05}.
204:   In the simulations of \citet{nag07a}, the $[0.15$--$0.5]\,\Rv$ temperature is $6\%$ higher than that in $[0.15$--$1]\,\Rv$. A smaller difference is expected for the aperture used here.}.
205: 
206: 
207: \begin{table}[t]
208:      \caption[]{Observed scaling relations.  For each observable
209:        set $(B,A)$, we fitted a power law relation of the form $B
210:        = C(A/A_0)^\alpha$, with $A_0 = 5\,\keV;  4\times10^{13}\msol;
211:        2\times10^{14}\,\msol\,\keV$ for $\TX$,  $M_{\rm g,500}$ and
212:        $Y_{\rm X}$ respectively.   $\sigma_{\rm log,r}$ and
213:        $\sigma_{\rm log,i}$ are the raw and intrinsic scatter about
214:        the best fitting relation in the $\log$--$\log$ plane. The \MT\
215:        relation is the same as that given in \cite{app05}. } 
216:      \label{tab:rel}
217:      \begin{center}
218:     \begin{tabular}{lccccc}
219:     \hline
220:     \hline
221:     Relation &  $\log_{10} C $ & $\alpha$ & $\sigma_{\rm log,r}$&$\sigma_{\rm log,i}$ \\% & $\chi^2$\\
222:     \hline
223:  $h(z) M_{500}$--$\TX$   &  $14.580 \pm 0.016$ &$1.71 \pm 0.09$ & $0.064$ & $0.039$ \\%& 12.7\\
224:  $h(z)^{2/5} M_{500}$--$Y_{\rm X}$   &  $14.556 \pm 0.015$ &$0.548 \pm 0.027$ & $0.062$ & $0.039$ \\%& 14.4\\
225: $M_{500}$--$M_{\rm g,500}$  &$14.542 \pm 0.015$ &$0.803 \pm 0.040$ & $0.065$ & $0.044$\\% & 16.6  \\
226:  $h(z) M_{g,500}$--$\TX$   &  $13.651 \pm 0.010$ &$2.10 \pm 0.05$ & $0.048$ & $0.036$ \\%& 12.7\\
227: $h(z)^{2/5} M_{\rm g,500}$--$Y_{\rm X}$  &   $13.619 \pm 0.008$ &$0.678 \pm 0.014$ & $0.017$ & - \\%& 5.7 \\
228: $f_{\rm g,500}$--$Y_{\rm X}$  &   $-0.939 \pm 0.016$ &$0.133 \pm 0.028$ & $0.067$ & 0.044 \\%& 5.7 \\
229:      \hline
230:     \end{tabular}
231:  \end{center}
232: \end{table}
233: 
234:  \begin{figure}[t]
235: \centering
236: \includegraphics[ width=0.85\columnwidth, keepaspectratio]{Fig_MYXcomp.eps} 
237:  \caption{\label{my} The \MY\ relation as seen by \xmm\ from a sample
238:    of 10 local relaxed clusters. The red solid line is the best fitting
239:    power law and  the shaded orange area corresponds to the $1\,\sigma$
240:    uncertainty. The predicted relation from numerical simulations
241:    including cooling and galaxy feedback \citep{nag07b} is over-plotted
242:    as a green dot-dashed line (true mass) and as a green solid
243:    line (mass estimated from mock X-ray observations and the HE
244:    equation). The dotted and dashed blue lines are the observed
245:    relations derived from Chandra data by \citet{nag07b} and
246:    \citet{mau07} respectively (see text). } 
247: \end{figure}
248: 
249: %===============================================================
250: \begin{figure*}[th]
251: \begin{center}
252: \begin{minipage}[t]{0.9\hsize}
253: \resizebox{\hsize}{!} {
254: \includegraphics{Fig_MT.eps}
255: \hspace{8mm}
256: \includegraphics{Fig_MMg.eps}
257: \hspace{8mm}
258: \includegraphics{Fig_MYX.eps}
259: } 
260: \end{minipage}\\[2mm]
261: \begin{minipage}[t]{0.9\hsize}
262: \resizebox{\hsize}{!} {
263: \includegraphics{Fig_MgYX.eps}
264: \hspace{8mm}
265: \includegraphics{Fig_fgMb.eps}
266: \hspace{8mm}
267: \includegraphics{Fig_fgYX.eps}
268: } 
269: \end{minipage}
270: \caption{\label{rel}   Correlations between X-ray
271:   observables. Solid lines: best fitting power law relations.  Dashed lines: standard self-similar
272:   relation (slope indicated in each figure) normalised to data from the three
273:   most massive clusters. Green
274:   dotted line in top-left panel:  best fitting power law  \MT\  relation
275:   for the hot cluster sub-sample \citep{app05}.  Green  lines in bottom-middle
276:   panel: gas mass fraction from the \MMg\ relation in the numerical simulations of \citet{nag07b},  using the true  mass (dash-dotted line) and the HE mass (full line). } 
277: \end{center}
278: \end{figure*}
279: %===============================================================
280: 
281: 
282: The integrated gas mass depends sensitively on the gas density at
283: large radius.  To compute $\Mgv$, we re-derived the gas density
284: profile from the emissivity corrected surface brightness profiles
285: using the deprojection and PSF-deconvolution technique recently
286: developed by \citet{cro06}. This derivation is free of any
287: assumption on profile shape, such as power law behaviour at large
288: radius, a feature common to all analytical fitting models used thus far
289: \citep[e.g.][]{pa02,pap05b,vik06}. Furthermore, the statistical errors
290: are readily estimated from a built-in Monte-Carlo procedure
291: \citep[see][for details]{cro06}. For the present sample, there is
292: excellent agreement  between the deprojected density profiles and the 
293: analytical model profiles derived in our previous work
294: \citep{pap05b,pap06}. The significant differences are in the very
295: central regions of some clusters  \citep[e.g.][Fig 12]{cro06} and for
296: A2597 at large radii, where the deprojected profile is slightly
297: steeper than the model profile. The gas mass estimated
298: with the two methods differs by less than $3\%$, except for  A2597
299: ($8\%$ difference). For all clusters, except for A1983 and MKW9, the
300: surface brightness profiles extend at least up to $\Rv$, or very close to
301: it, so that extrapolation uncertainty is not an issue. For A1983 and MKW9, the $\Mgv$
302: estimated from extrapolation in the $\log$--$\log$ plane are $31\%$
303: (A1983) and $67\%$ (MKW9) larger than the gas mass measured at
304: $\delta_{\rm obs}$
305: ; as shown below, these points do not however have a
306: significant effect on the results. 
307: 
308: The  resulting  $\Mgv$, $\fgv$ and $\YX$ values are listed in
309: Table~\ref{tab:clu}. Errors  on $\Mgv$ include both statistical errors
310: and errors due to uncertainties in $\Rv$, which are summed
311: quadratically, with the latter dominating the error budget. 
312: 
313: \subsection{Scaling relations}
314: 
315: For each observable set $(B,A)$ we fitted a power law relation of the form
316: $h(z)^nB = C(A/A_0)^\alpha$, where 
317: $h(z)$  is the Hubble constant
318: normalised to its present value 
319: and $n$ is fixed to the expected
320: scaling with $z$.  The fit was performed using linear
321: regression in the $\log$--$ \log$ plane, taking into account  the
322: errors on both variables
323: \citep[\texttt{FITEXY};][]{numrec}.  The pivot
324: point $A_{0}$ is chosen so that the normalisation and slope are nearly
325: independent parameters. For the $M_{500}$--$Y_{X}$ relation for
326: instance, the covariance in $\log(C)$ and $\alpha$ normalised to the
327: product of their standard errors is 0.042. The resulting values are
328: given in Table~\ref{tab:rel}, and the various correlations are plotted
329: in Fig.~\ref{my} and   Fig.~\ref{rel}. 
330: Table~\ref{tab:rel} also shows the raw and intrinsic scatter about 
331: the best fitting  relations in the  $\log$--$ \log$ plane.
332: The raw scatter was estimated using the vertical distances to the  
333: regression line, weighted by the error. The  
334: intrinsic scatter was computed from the quadratic difference between the  
335: raw scatter and the scatter expected from the statistical errors.
336: 
337: The regression method is strictly valid only if the intrinsic
338: scatter is negligible as compared to the statistical scatter; in fact
339: they are of the same order (Table~\ref{tab:rel}). We verified that the
340: results are unchanged using the variation of the method discussed in
341: \citet{pap06}. 
342: Finally, the \MY\ relation is robust to exclusion of
343: A1983 and MKW9, for which data extrapolations were required (see above): the differences are at the $+0.4\,\sigma$  and $-0.2\,\sigma$ levels for the slope and normalisation, respectively.
344: 
345: \section{Discussion and conclusions}
346: \subsection{Comparison with theoretical predictions}
347: The slope of the observed relation:
348: \begin{equation}
349: h(z)^{2/5}\Mv = 10^{14.556 \pm 0.015} \left[\frac{\YX}{2\times10^{14}\,{\msol}\,\keV}\right]^{0.548 \pm 0.027}{\rm h_{70}^{-1}\,\msol}
350: \end{equation}
351: is slightly smaller than the standard self-similar  value
352: ($\alpha=3/5$),  at the $1.9\,\sigma$ significance level, consistent
353: with the  \MMg\ and \MT\  relations  (Table \ref{tab:rel} and
354: Fig.~\ref{rel} top panel). The  \MMg\ relation is shallower than
355: expected, reflecting the increase in gas mass fraction with mass
356: (Fig.~\ref{rel} bottom-middle panel), while the  \MT\ is steeper.
357: At a given mass the gas mass is smaller and the temperature is
358: higher, leading to a partial cancellation in the product $\YX= \Mg
359: \TX$ (see also below).    
360: 
361: The observed normalisation is $\sim 20\%$ smaller over the whole $\YX$ range
362: than that derived from numerical simulations including cooling and galaxy feedback \citep{nag07b}, while the observed slope is
363: consistent with the predicted slope, $\alpha = 0.568\pm 0.006$, within
364: the $1\,\sigma$ error (Fig.~\ref{my}).  
365: Better agreement is obtained
366: with the  simulated $M_{500}^{HE}$--$\YX$ relation,  where $M_{500}^{HE}$ is the mass estimated from mock X--ray
367: observations and the HE equation.  Although
368: the predicted slope, $\alpha = 0.596\pm 0.010$,
369: %(close to the self-similar value), 
370: is slightly higher,  the difference in normalisation drops
371: to  $\sim\,8\%$ ($2.4\,\sigma$) at $\YX = 2\times10^{14}\,{\msol}\,\keV$.   As
372: discussed by \citet{nag07b}, the offset in normalization, also observed
373: with \chandra\ data,  may arise from an underestimate of the true mass
374: by the HE equation, perhaps due to residual non-thermal pressure
375: support.  These numerical simulations also
376: predict a hot gas mass fraction systematically smaller than observed
377: (Fig.~\ref{rel} bottom-middle panel). The difference is smaller for simulated  $\fgv$ using $M_{500}^{HE}$ and again could be due, in part, to  biases in X--ray mass estimates. Nevertheless, there may also be an underestimate of $\fgv$ in the simulations, possibly due in part  to  over-condensation of hot gas into the cold dense phase \citep[][]{nag07b}. This would  contribute to the offset, by shifting the \MY\ relation to the left in the $\log$--$\log$ plane. Finally, as the normalization  depends on $\TX^{0.6}$,  the difference in the exact definition of $\TX$ (see Sec.~\ref{sample}) could contribute by $\lesssim 4\%$ to the offset.
378:   
379: \subsection{Comparison with \chandra\ results}
380: 
381: Our \MY\ relation is very similar to that derived by \citet{nag07b}
382: from the \chandra\ data presented in \citet[][see our
383: Fig.~\ref{my}]{vik06}. The slope $\alpha =0.526  
384: \pm0.038$ is consistent with our value, $\alpha = 0.548\pm0.027$,
385: and the normalisation at $\YX = 2\times10^{14}\,\msol\,\keV$, $\Mv =
386: 3.82\times10^{14}\,{\rm h_{70}^{-1}}\,\msol$, is higher than our value,
387: $(3.60\pm0.13)\times10^{14}\msol$, at only the $1.6\,\sigma$ level. Even
388: better agreement is obtained with the best fitting relation quoted  by
389: \citet[][dashed line in Fig.~\ref{my}]{mau07}, derived from the same
390: data excluding the lowest mass cluster
391: (A. Vikhlinin, priv. communication). Here the  slope ($\alpha
392: =0.564$) is closer to the self-similar value, as we have found, and
393: the difference in normalisation is less than $5\%$ over the whole mass
394: range.  
395: 
396: 
397: \subsection{Comparison of  mass proxies for relaxed clusters}
398: 
399: For {\it relaxed} clusters, \citet{kra06} found similar scatter in the
400: \MT\  and \MMg\ relations ($\sigma_{\rm log} = 0.055$ and 0.047
401: respectively), but two times less scatter in the \MY\ relation (0.022).
402: We can compare with the present data, the statistical quality allowing us to estimate the intrinsic scatter for the
403: first time. %%In fact 
404: The scatter 
405: (Table~\ref{tab:rel}) 
406: is the same for the \MY\ and \MT\  relations
407: ($\sigma_{\rm log,i} = 0.039$) and slightly larger for the \MMg\
408: relation ($\sigma_{\rm log,i} = 0.044$). The latter may reflect that
409: the \MMg\ relation is not actually a power law: 
410: the gas mass fraction appears constant at $\Mv
411: \gtrsim 2-3\times10^{14} \msol$, with a progressive drop at lower mass (Fig.~\ref{rel} bottom-middle panel). 
412: 
413: In fact the behaviour of $\fgv$ appears to be the primary factor
414: driving the scatter in the \MY\ relation. The \MgY\ relation is
415: extremely tight (Fig.~\ref{rel}
416: and Table~\ref{tab:rel}), being well fitted by a power law 
417: %($\chi^2_{red}=0.7$),
418: with no measurable scatter, in spite of the precision of the
419: data. Since $\Mv=\Mgv /\fgv$, the scatter in the \MY\ relation simply
420: reflects the scatter in the $\fgv$--$\YX$ relation (cf. top and
421: bottom left panels of Fig.~\ref{rel}). This scatter could arise from  true scatter in $\fgv$ and/or scatter in the X--ray mass  to true mass ratio, e.g., due to variations in the magnitude of nonthermal pressure support.
422: Note that a low-scatter
423: correlation between $\Mgv$ and $\YX$ is expected: it is
424: straightforward to show that the logarithmic scatter in the \MgY\
425: relation is 1/3 of the scatter in the \MgT\ relation for
426: $\Mgv \propto \TX^{\sim 2}$ (Table~\ref{tab:rel}).
427: 
428: In terms of observed scatter in the relation with mass,  $\YX$
429: thus does not appear to be a better proxy than $\TX$, and is only slightly
430: better than $\Mgv$.  However we caution against over-inerpretation. Firstly, the present results are for relaxed
431: clusters only: with the current data we cannot check if the scatter is
432: insensitive to dynamical state \citep{kra06,poo07}. Secondly, the 
433: scatter estimates should be
434: confirmed using larger cluster samples with stricter
435: selection criteria.
436: 
437: However, in terms of functional dependence with mass, $\YX$ is clearly
438: a better proxy than $\Mgv$: it is better fitted by  a simple
439: power-law, and has a slope closer to the standard self-similar value
440: (Table~\ref{tab:rel}).  Furthermore, although the  quality of the
441: power law fits to \MT\ and \MY\ are formally similar ($\chi^2/{\rm
442:   d.o.f}\sim13/8$), with similar ($\sim2\,\sigma$) deviations from the standard slope, there is some indication that $\YX$ is
443: also a better proxy than $\TX$ in this regard. The slope of the \MT\
444: relation may depend on mass range \citep{app05}, reaching the standard value  when cool clusters are excluded,  but the slope of the \MY\ relation remains stable in that case ($0.7\,\sigma$ difference). 
445: 
446: \subsection{Concluding remarks}
447: 
448: Our results suggest that the various mass scaling relations
449: might be better understood by considering the gas thermal energy
450: ($\YX$) and mass ($\fgv$) as its most fundamental properties.  
451: Let us suppose that the thermal energy content of the gas is the quantity most closely related to the mass (i.e. the best mass proxy is indeed $\YX$), and
452: that its relation with mass has a quasi-standard slope. Let us further note that the gas mass fraction appears constant at high
453: mass, with a progressive decrease below 
454: a 'break' mass (reflecting gas loss or incomplete accretion in low
455: mass systems due to non gravitational effects). Since
456: $M/T^{3/2}$ varies as $(M/\YX^{3/5})^{5/2} f_{\rm g}^{3/2}$, one
457: then expects a steepening of the \MT\ relation at low mass, with a
458: standard slope at high mass.
459: 
460: A deeper understanding of the mass scaling relations will come from
461: the X-ray study of larger unbiased samples of local clusters,
462: such as REXCESS \citep{boh07}, combined with lensing data.
463: This is necessary
464: to ascertain the dependence of the \MY\ relation on dynamical
465: state, and to calibrate its normalisation and slope. This step is essential because the use of $\YX$ as a mass proxy, as in the case of $\TX$, requires a detailed understanding of non-gravitational effects, in particular of the impact of cooling and feedback on the fraction of primordial gas that remains in the gravitationally bound hot phase. Precise measurements at $z=0$ are
466: needed to constrain models, on which one must rely for high z
467: studies.  
468: Significant progress is also expected from forthcoming SZ data (e.g from the Planck Surveyor all sky survey),
469: especially if combined with \xmm\ or Chandra data, which will allow a
470: full study of the $M$--$Y_{\rm SZ}$ relation. 
471: 
472: 
473: \begin{acknowledgements}
474:     We thank A. Kravtsov and D. Nagai  for useful comments on the manuscript, and the referee for a speedy and pertinent response. 
475: \end{acknowledgements}
476: 
477: \begin{thebibliography}{}
478: 
479: \bibitem[{{Arnaud \etal}(2005)}]{app05}
480:     Arnaud, M., Pointecouteau, E. \& Pratt, G.W. 2005, \aap, 441, 893
481: 
482: \bibitem[{{Bertschinger}(1998)}]{bertschinger98}
483: {Bertschinger}, E. 1998, \araa, 36, 599
484: 
485: \bibitem[{{B{\"o}hringer \etal}(2007)}]{boh07}
486: {B{\"o}hringer}, H., {Schuecker}, P., {Pratt}, G.~W.  \etal, 2007,
487: \aap, 469, 363 
488: 
489: \bibitem[{{Croston \etal}(2006)}]{cro06}
490:     Croston, J.H., Arnaud, M., Pointecouteau, E. \& Pratt, G.W. 2006,
491:     \aap, 459, 1007 
492: 
493: \bibitem[{{da Silva \etal}(2004)}]{das04}
494: da Silva, A.C., Kay, S.T., Liddle, A.R \& Thomas, P. 2004, \mnras,
495: 348, 1401 
496: 
497: \bibitem[{{Kravtsov \etal}(2006)}]{kra06}
498:     Kravtsov, A.V., Vikhlinin, A. \& Nagai, D.  2006, \apj, 650, 128
499: 
500: \bibitem[{{Maughan}(2007)}]{mau07}
501:     Maughan, B.J.  2007, \apj, in press, astro-ph/0703504
502: 
503: \bibitem[{{Motl \etal }(2005)}]{mot05}
504:    Motl, P.M., Hallman, E.J., Burns, J.O. \& Norman, M.L. 2005, \apj,
505:    623, L63 
506: 
507: \bibitem[{{Nagai }(2006)}]{nag06} 
508:         Nagai, D.  2006,  \apj, 650, 538
509:         
510: \bibitem[{{Nagai \etal}(2007a)}]{nag07a} 
511:         Nagai, D., Vikhlinin, A. \& Kravtsov, A.  2007a,  \apj,
512:         655, 98
513:         
514: \bibitem[{{Nagai \etal}(2007b)}]{nag07b} 
515:         Nagai, D.,  Kravtsov, A. \& Vikhlinin, A.  2007b,  \apj,
516:         submitted, astro-ph/0703661 
517: 
518: \bibitem[{{Pointecouteau \etal}(2005)}]{pap05b}
519:     Pointecouteau, E.,  Arnaud, M. \& Pratt, G.W.  2005, \aap, 435, 1
520: 
521: \bibitem[{{Pratt  \&  Arnaud}(2002)}]{pa02}
522:          Pratt, G.W. \&  Arnaud, M. 2002, \aap, 394, 375
523: 
524: 
525: \bibitem[{{Pratt \etal}(2006)}]{pap06}
526:     Pratt, G.W., Arnaud, M. \& Pointecouteau, E.  2006, \aap, 446, 429
527: 
528: \bibitem[{{Poole  \etal}(2007)}]{poo07}
529:  Poole, G..P., Babul, A., McCarthy, I.G., Fardal, M.A., Bildfell, C.J., Quinn, T., Mahdavi, A.,  2007, \mnras, submitted, astro-ph/0701586
530: 
531: \bibitem[{{Press} {\etal}(1992)}]{numrec}
532: {Press}, W.H., {Teukolsky}, S.A., {Vetterling}, S.A. \& {Flannery}  
533: B.P.  1992, Numerical Recipes in Fortran 77, Second Edition, p.660
534: 
535: \bibitem[{{Vikhlinin \etal}(2005)}]{vik05} 
536:         Vikhlinin, A.,  Markevitch, M.,  Murray, S, Jones, C., Forman,
537:         W., \& Van Speybroeck, L. 2005,  \apj, 628, 655 
538: 
539: \bibitem[{{Vikhlinin \etal}(2006)}]{vik06} 
540:         Vikhlinin~A.,  Kravtsov, A., Forman, W., Jones, C.,
541:         Markevitch, M., Murray, S. \& Van Speybroeck, L. 2006,  \apj,
542:         640, 691 
543: 
544: \bibitem[Voit(2005)]{voi05}Voit, G.M.~2005, Rev. Mod. Phys., 77, 207
545: 
546: 
547: 
548: \end{thebibliography}
549: 
550: \end{document}
551: 
552: