1: \documentclass[12pt,preprint]{aastex}
2: \begin{document}
3: \title{Reconstructing the Intrinsic Triaxial Shape of the Virgo Cluster}
4: \author{Bomee Lee and Jounghun Lee}
5: \affil{Department of Physics and Astronomy, FPRD, Seoul National University,
6: Seoul 151-747, Korea}
7: \email{bmlee@astro.snu.ac.kr, jounghun@astro.snu.ac.kr}
8: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
9: \begin{abstract}
10: To use galaxy clusters as a cosmological probe, it is important to account
11: for their triaxiality. Assuming that the triaxial shapes of galaxy clusters
12: are induced by the tidal interaction with the surrounding matter, Lee and
13: Kang recently developed a reconstruction algorithm for the measurement of
14: the axial ratio of a triaxial cluster. We examine the validity of this
15: reconstruction algorithm by performing an observational test of it with
16: the Virgo cluster as a target. We first modify the LK06 algorithm by
17: incorporating the two dimensional projection effect. Then, we analyze
18: the $1275$ member galaxies from the Virgo Cluster Catalogue and find
19: the projected direction of the Virgo cluster major axis by measuring the
20: anisotropy in the spatial distribution of the member galaxies in the two
21: dimensional projected plane. Applying the modified reconstruction algorithm
22: to the analyzed data, we find that the axial ratio of the triaxial
23: Virgo cluster is ($1$: $0.54$ : $0.73$). This result is consistent with the
24: recent observational report from the Virgo Cluster Survey, proving the
25: robustness of the reconstruction algorithm. It is also found that at the
26: inner radii the shape tends to be more like prolate. We discuss the
27: possible effect of the Virgo cluster triaxiality on the mass estimation.
28: \end{abstract}
29: \keywords{cosmology:theory --- large-scale structure of universe}
30: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
31:
32: \section{INTRODUCTION}
33:
34: Galaxy clusters provide one of the most powerful tools to constrain the key
35: cosmological parameters. In the era of precision cosmology, it is
36: important to determine their mass as accurately as possible before using
37: them as a cosmological probe. Any kind of simplified assumption about
38: the properties of galaxy clusters could cause substantial systematics
39: in the mass estimation. The triaxial shapes of galaxy clusters are one
40: of such properties.
41:
42: It has been long known both observationally and numerically that the galaxy
43: clusters are noticeably triaxial \citep[e.g.,][]{fre-etal88,wes89,plio91,
44: war-etal92}. While plenty of efforts have been already made to take into
45: account the triaxial shapes of galaxy clusters \citep{jin02,fox-pen01,
46: suw-etal03,lee04,kas-evr05,hop-etal05,lee05,smi-wat05,hay07}, previous
47: studies have been largely focused on the statistical treatment of
48: cluster triaxiality. For the measurement of the gas mass fraction of galaxy
49: clusters that can provide powerful constraints on the density parameter and
50: the dark energy equation of state \citep{whi-etal93,lub-etal96,cen97,evr97,
51: coo98,lar06,fer-bla07}, however, it is necessary to deal with the individual
52: clusters and their triaxial shapes.
53:
54: The standard picture based on the cosmic web theory \citep{bar-etal86,
55: bon-etal96} explains that galaxy clusters are rare events corresponding
56: to the local maxima of the initial density field and form at the dense
57: nodes of the local filaments in the cosmic web through tidal interactions
58: with the surrounding matter. The tidal effect from the surrounding matter
59: results in the deviation of cluster shapes from spherical symmetry as well
60: as the preferential alignments of cluster galaxies (or halos) with the
61: major axes of their host clusters \citep{bin82,str-pee85,hop-etal05,kas-evr05,
62: lee05,alg-etal06,alt-etal06,paz06}.
63:
64: In the frame of this standard scenario, \citet[][hereafter, LK06]{lee-kan06}
65: have recently developed an analytic algorithm to reconstruct the triaxial
66: shapes of individual clusters. The key concept of the LK06 algorithm is that
67: the two axial ratios of a triaxial cluster are related to the eigenvalues of
68: the local tidal shear tensor. By measuring the spatial alignment of the
69: cluster galaxies with the major axis of their host cluster, one can
70: determine the eigenvalues of the local tidal tensor, which will in turn
71: yields the two axial ratios of a triaxial cluster.
72:
73: Testing their analytic model against high-resolution N-body simulations,
74: LK06 have shown that their algorithm works well within $20\%$ errors.
75: Now that the LK06 algorithm is known to work in principle, it is time to
76: test the algorithm against observations. Our goal here is to apply the
77: LK06 algorithm to real observational data and examine its validity in
78: practice. Here, we use the Virgo cluster as a target, whose triaxial
79: shape has been very recently measured observationally \citep{mei-etal07}.
80:
81: The organization of this paper is as follows. In \S 2, a brief overview of
82: the LK06 algorithm is provided and how to incorporate the two dimensional
83: projection effect into the algorithm is explained. In \S 3, the Virgo cluster
84: data are analyzed and its triaxial shapes are reconstructed using the LK06
85: algorithm. In \S 4, the results are summarized, and the advantages and the
86: caveats of our model are discussed.
87:
88: \section{THEORETICAL MODEL}
89:
90: \subsection{\it Overview of the LK06 Algorithm}
91:
92: According to the LK06 algorithm, the cluster triaxial shape originates from
93: its tidal interaction with the surrounding matter distribution.
94: This assumption has been verified from high-resolution N-body simulation which
95: demonstrated clearly that the tidal field elongates the cluster shapes
96: \citep[e.g.,][and references therein]{alt-etal06}. LK06 has shown that the
97: two axial ratios of a triaxial cluster are related to the three eigenvalues
98: of the local tidal tensor defined as the second derivative of the
99: gravitational potential:
100: \begin{equation}
101: \label{eqn:axi}
102: \frac{b}{c}=\left (\frac{1-\lambda_{2}}{1-\lambda_{3}}\right)^{1/2} , \qquad
103: \frac{a}{c}=\left (\frac{1-\lambda_{1}}{1-\lambda_{3}}\right)^{1/2} ,
104: \end{equation}
105: where $\{a,b,c\}$ (with $a \le b \le c$)are the three principal axis lengths
106: of a cluster and $\{\lambda_{1},\lambda_{2},\lambda_{3}\}$ (with
107: $\lambda_{1} \ge \lambda_{2} \ge \lambda_{3}$) are the three eigenvalues of
108: the local tidal tensor, ${\bf T}$. According to this formula, one can
109: estimate the cluster axial ratios if $\lambda_{1},\lambda_{2},\lambda_{3}$
110: is fixed by fitting the probability density distribution, $p(\cos\theta)$
111: analytically to the observational data. LK06 suggested that the preferential
112: locations of the cluster galaxies near the cluster major axes given the local
113: tidal tensor be described as
114: \begin{equation}
115: \label{eqn:corr}
116: \langle \hat{x}_i\hat{x}_j\vert\hat{T}\rangle =
117: \frac{1-s}{3}\delta_{ij}+s\hat{T}_{ik}\hat{T}_{kj}.
118: \end{equation}
119: where $\hat{\bf x} \equiv (\hat{x}_{i})$ is the unit position vector of a
120: cluster galaxy, $\hat{\bf T}\equiv {\bf T}/\vert{\bf T}\vert$ is the
121: unit tidal shear tensor, and $s \in [-1,1]$ is the correlation parameter
122: that represents the correlation strength between $\hat{\bf x}$ and
123: $\hat{\bf T}$. Under the assumption that the minor axis of the tidal shear
124: tensor is in the direction of the cluster major axis,
125: equation (\ref{eqn:corr}) basically describes the alignment between the
126: position of a cluster galaxy and the major axis of its host cluster.
127: If $s =-1$, there is a maximum alignment. If $s=1$, there is a maximum
128: anti-alignment. The case of $s=0$ corresponds to no alignment.
129:
130: Let $\theta_{3d}$ be the angle between the host cluster major axis and the
131: galaxy position vector. Under the assumption that the cluster major axis is
132: in the direction of the minor principal axis of the tidal shear tensor,
133: The probability density distribution of $\cos\theta_{3d}$ was derived by
134: LK6 as
135: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
136: \begin{equation}
137: \label{eqn:gen}
138: p(\cos\theta_{3d})=\frac{\vert M\vert}{2\pi}
139: (\hat{x}_{i}\cdot M^{-1}_{ij}\cdot\hat{x}_{j})^{-\frac{3}{2}},
140: \end{equation}
141: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
142: where $\phi$ is an azimuthal angle of $\hat{\bf x}$ measured in an arbitrary
143: coordinate system. Here, covariance matrix $\bf{M}$ is defined as
144: ${\bf M}\equiv \langle\hat{x}_i\hat{x}_j\vert\hat{T}\rangle$. Note that
145: equation (\ref{eqn:gen}) holds good for any arbitrary coordinate
146: system in which the tidal shear tensor is not necessarily diagonal.
147:
148: In the principal axis frame of the tidal tensor, equation (\ref{eqn:pri})
149: can be expressed only in terms of the three eigenvalues of the tidal shear
150: tensor as
151: \begin{eqnarray}
152: \label{eqn:pri}
153: p(\cos\theta)&=&\frac{1}{2\pi} \prod_{i=1}^{3}
154: (1-s+3s\hat{\lambda}^{2}_{i})^{-1/2}\nonumber \\
155: &&\times \int_{0}^{2\pi}\left(\frac{\sin^2\theta \cos^2\phi}{1-s+3s
156: \hat{\lambda}^{2}_{1}} +\frac{\sin^2\theta \sin^2\phi}{1-s+3s
157: \hat{\lambda}^{2}_{2}}+\frac{\cos^2\theta}{1-s+3s\hat{\lambda}^{2}_{3}}
158: \right)^{-3/2} d\phi,
159: \end{eqnarray}
160: where $\{\hat{\lambda}_{i}\}^{3}_{i=1}$ are the eigenvalues of $\hat{\bf T}$,
161: related to $\{\lambda_{i}\}^{3}_{i=1}$ as
162: \begin{equation}
163: \label{lambda}
164: \lambda_{1}=\frac{\delta_{c}\hat{\lambda}_{1}}{\hat{\lambda}_{1}
165: +\hat{\lambda}_{2}+\hat{\lambda}_{3}} ,\quad \lambda_{2}=
166: \frac{\delta_{c}
167: \hat{\lambda}_{2}}{\hat{\lambda}_{1}+\hat{\lambda}_{2}+
168: \hat{\lambda}_{3}},\quad \lambda_{3}=\frac{\delta_{c}\hat{\lambda}_{3}}
169: {\hat{\lambda}_{1}+\hat{\lambda}_{2}+\hat{\lambda}_{3}},
170: \end{equation}
171: where $\delta_{c} \approx 1.68 $ is the linear density threshold for a
172: dark halo \citep{eke96} satisfying the following constraint
173: of $\delta_{c}=\sum_{i=1}^{3}\lambda_{i}$.
174:
175: The key concept of LK06 algorithm is that by fitting equation (\ref{eqn:gen})
176: to the observed probability distribution, one can find the best-fit values
177: of $\lambda_{1},\lambda_{2}$ and $s$, and then determine the cluster axial
178: ratios using equation (\ref{eqn:axi}). Although LK06 algorithm allows us
179: in principle to reconstruct the three dimensional intrinsic triaxial
180: shapes of individual clusters, it is restricted to the cases where the
181: informations on the three dimensional positions of the cluster galaxies
182: in the cluster principal axis frame are given. Unfortunately, for most
183: clusters, these informations are not available but only two dimensional
184: projected images of clusters.
185:
186: In $\S 2.2$, we attempt to modify the LK06 algorithm in order to incorporate
187: the two dimensional projection effect.
188:
189: \subsection{\it Projection Effect}
190:
191: Let us suppose that the position vectors of the cluster galaxies
192: are all projected along the line of sight direction onto the plane of a sky.
193: Unless the major axis of the host cluster is perfectly aligned with the
194: line-of-sight direction to the cluster center, one would expect that the
195: projected position vectors of the cluster galaxies should show a tendency
196: to be aligned with the projected major principal axes.
197:
198: Let $\theta_{2d}$ be the angle between the projected cluster major axis and
199: galaxy position vector in the plane of sky. The probability distribution can
200: be calculated by integrating equation (\ref{eqn:gen}) along the line of sight
201: as
202: \begin{equation}
203: \label{eqn:2dp}
204: p(\cos\theta_{2d})=\int_{-1}^{1}\frac{\vert M\vert}{2\pi}
205: (\hat{x}_{i}\cdot M^{-1}_{ij}\cdot\hat{x}_{j})^{-\frac{3}{2}}d
206: \hat{x}_{3},
207: \end{equation}
208: where $(\hat{x}_{3})$ is now chosen to be in the direction of the line of
209: sight. In other words, we consider a certain Cartesian coordinate system
210: in which the $\hat{x}_{3}$ direction is parallel to the line of the sight
211: to the cluster center of mass. Note that in this coordinate system the
212: tidal tensor is not necessarily diagonal.
213:
214: Through the similarity transformation
215: \begin{equation}
216: \label{eqn:rot}
217: \hat{\bf T} =\bf{R}^{t}\cdot\hat{\bf \Lambda}_{\rm T}
218: \cdot {\bf {R}},
219: \end{equation}
220: with
221: \begin{equation}
222: \label{eqn:dia}
223: \mathbf{\Lambda_{T}} =
224: \left( \begin{array}{ccc}
225: \hat{\lambda_1} & 0 & 0 \\
226: 0 & \hat{\lambda_2} & 0 \\
227: 0 & 0 & \hat{\lambda_3}
228: \end{array} \right),
229: \end{equation}
230: one can express the nondiagonal unit tidal tensor $\hat{\bf T}$ in
231: terms of its eigenvalues.Here, the rotation matrix ${\bf R}$ has the form
232: of \citep{bin85}
233: \begin{equation}
234: \label{eqn:bin}
235: \mathbf{R} =
236: \left( \begin{array}{ccc}
237: -\sin\psi & -\cos\psi\cos\xi & \cos\psi\sin\xi \\
238: \cos\psi & -\sin\psi\cos\xi & \sin\psi\sin\xi \\
239: 0 & \sin\xi & \cos\xi
240: \end{array} \right),
241: \end{equation}
242: where $(\xi,\psi) $ is the polar coordinate of the line-of-sight
243: direction in the principal axis frame of the cluster.
244:
245: Through equations (\ref{eqn:2dp})-(\ref{eqn:bin}), we finally final
246: an analytic expression for the probability density distribution of
247: $\cos\theta{2d}$ in terms of $\{\hat{\lambda}_{i}\}_{i=1}^{3}$ and
248: the correlation parameter $s$:
249: \begin{equation}
250: \label{real}
251: p(\cos\theta_{2d})=\frac{(1+s)^2(1-2s)}{2\pi}\int_{-1}^{1}Q^{-3/2}d\hat{x}_3,
252: \end{equation}
253: with the factor $Q$ defined as
254: \begin{eqnarray}
255: \label{xi}
256: Q &\equiv & [(1 + s)^2 - 3s(1 + s)](A_1\hat{\lambda}_{1}+A_2\hat{\lambda}_{2}
257: +A_3\hat{\lambda}_{3})(1 - \hat{x}^{2}_{3}), \nonumber \\
258: && + [(1 + s)^2 - 3s(1 + s)](C_1\hat{\lambda}_{1} + C_2\hat{\lambda}_{2}
259: + C_3\hat{\lambda}_{3})\hat{x}^{2}_{3},
260: \nonumber \\
261: && + 6s(1 + s)(B_1\hat{\lambda}_{1}+ B_2\hat{\lambda}_{2} +
262: B_3\hat{\lambda}_{3})\hat{x}_{3}\sqrt{1 - \hat{x}^{2}_{3}},
263: \end{eqnarray}
264: where the coefficients $\{A_{i}\}^{3}_{i=1},\{B_{i}\}^{3}_{i=1},
265: \{C_{i}\}^{3}_{i=1}$ are given as
266: \begin{eqnarray}
267: \label{a}
268: A_1 &=&
269: \cos\psi\cos\theta_{2d}(\cos\psi\cos\theta_{2d} -
270: 2\sin\psi\cos\xi\sin\theta_{2d}) +
271: \sin^2\theta_{2d}(\sin^2\psi + \sin^2\xi\cos^2\psi), \cr
272: A_2 &=&
273: \sin\psi\cos\theta_{2d}(\sin\psi\cos\theta_{2d} +
274: 2\cos\psi\cos\xi\sin\theta_{2d}) +
275: \sin^2\theta_{2d}(\cos^2\psi + \sin^2\xi\sin^2\psi), \cr
276: A_3 &=& \cos^2\theta_{2d} + \cos^2\xi\sin^2\theta_{2d}, \cr
277: B_1 &=& -\sin\xi\cos\psi(\cos\xi\sin\theta_{2d}\cos\psi +
278: \sin\psi\cos\theta_{2d}), \cr
279: B_2 &=& -\sin\xi\sin\psi(\cos\xi\sin\theta_{2d}\sin\psi -
280: \cos\psi\cos\theta_{2d}), \cr
281: B_3 &=& \cos\xi\sin\xi\sin\theta_{2d}, \cr
282: C_1 &=& \sin^2\psi + \cos^2\xi\cos^2\psi, \nonumber \cr
283: C_2 &=& \cos^2\psi + \cos^2\xi\sin^2\psi, \nonumber \cr
284: C_3 &=& \sin^2\xi .
285: \end{eqnarray}
286: With this new modified algorithm in the two dimensional space, we can
287: reconstruct the intrinsic shape of a triaxial cluster halo from the
288: observed two dimensional image.
289:
290: \section{APPLICATION TO THE VIRGO CLUSTER HALO}
291:
292: \subsection{Observational Data and Analysis}
293:
294: The Virgo cluster is the nearest richly populated cluster of galaxies
295: whose properties has been studied fruitfully for long \citep[e.g.,]
296: [and references therein]{boh94,wes00}. It is known to have
297: approximately $1275$ member galaxies \citep{bin-etal85} and
298: located at a distance of approximately $16.1$ Mpc from us \citep{ton01}
299: with the major axis inclined at an angle of approximately $10^{o}$ with
300: respect to the line of sight \citep{wes00}.
301:
302: Near the center of the Virgo cluster is located the large ellipticity galaxy
303: $M87$ (or Virgo A). The major axis of the Virgo cluster is found to be
304: inclined approximately $10^{\circ}$ with respect to the line of sight
305: direction to $M87$ \citep{wes00}. We use data from the Virgo Cluster
306: Catalogue \citep{bin-etal85} which compiles the equatorial coordinates of
307: total $1275$ member galaxies.
308:
309: \subsection{Coordinate Transformation and Projection Effect}
310:
311: Let $r$ and $(\alpha,\delta)$ represent the three dimensional distance to a
312: member galaxy and its equatorial coordinates, respectively.
313: Under the assumption that the position of $M87$ is the center of mass of the
314: Virgo cluster, the Cartesian coordinate of a member galaxy in the center of
315: mass frame can be written as
316: \begin{eqnarray}
317: \label{eqn:flat}
318: x_1 &=&r\cos\alpha\cos\delta - x_{\rm 1va} \nonumber \\
319: x_2 &=&r\sin\alpha\cos\delta - x_{\rm 2va} \nonumber \\
320: x_3 &=&r\sin\delta - x_{\rm 3va}
321: \end{eqnarray}
322: where $(x_{\rm 1va},x_{\rm 2va},x_{\rm 3va})$ represents the position of
323: $M87$. The equatorial coordinates of $M87$ is measured to be
324: $\alpha_{\rm va}=187.71^{\circ}$ and $\delta_{\rm va}=12.39^{\circ}$.
325: The distance to Virgo A, $r_{va}$, from us is also known to be
326: approximately $16.1$Mpc (\citealt{ton01}; SBF survey).
327: Thus,we have a full information on $(x_{\rm 1va},x_{\rm 2va},x_{\rm 3va})$.
328:
329: Now let us consider a coordinate system where the third axis is in
330: in the direction of the line of sight to $M87$. Let $(u,\vartheta,\varphi)$
331: be the spherical polar coordinate of the member galaxy in this coordinate
332: system. It can be found through coordinate transformation as
333: \begin{equation}
334: \label{eqn:tra}
335: \left(
336: \begin{array}{c} u \\ \vartheta \\ \varphi
337: \end{array}\right)=
338: \left(
339: \begin{array}{ccc}
340: \sin\xi_{\rm va}\cos\psi_{\rm va} & \cos\xi_{\rm va}\cos\psi_{\rm va}
341: & -\sin\psi_{\rm va} \\
342: \sin\xi_{\rm va}\sin\psi_{\rm va} & \cos\xi_{\rm va}\sin\psi_{\rm va}
343: & \cos\psi_{\rm va} \\
344: \cos\xi_{\rm va} & - \sin\xi_{\rm va} & 0
345: \end{array} \right)
346: \left(
347: \begin{array}{c} x_{1}\\ x_{2} \\ x_{3}
348: \end{array}\right )
349: \end{equation}
350: where $(\xi_{\rm va},~\psi_{\rm va})$ is the polar coordinate of M87,
351: so that $\xi_{\rm va}\equiv \pi/2-\delta_{\rm va}$.
352:
353: In the plane of sky projected along the line of sight direction, the
354: spherical polar coordinates of the member galaxy can be regarded as
355: the Cartesian coordinates as $r \to 0$, $\vartheta \to x_{l}$, and
356: $\phi \to y_{l}$. Basically, it represents a two dimensional projected
357: position of a Virgo cluster member galaxy in the plane of sky with the
358: position of $M87$ as a center.
359:
360: For each member galaxy from the Virgo Cluster Catalogue, we have determined
361: the two dimensional projected position $(x_{l},y_{l})$ using the given
362: equatorial coordinates.
363:
364: \subsection{Virgo Cluster Reconstruction}
365:
366: To measure the alignments between the positions of the Virgo cluster galaxies
367: and the projected major axis of the Virgo cluster and compare the distribution
368: of the alignment angles with the analytic model (eq.[\ref{eqn:2dp}]), it is
369: necessary to find the direction of the projected major axis in the coordinate
370: system of $(x_{1l},x_{2l})$. Equivalently, it is necessary to have an
371: information on the polar coordinates of the line of sight, $(\xi,\psi)$
372: with respect to the three dimensional principal axis of the Virgo cluster.
373:
374: The seminal paper of \citet{wes00} provides us with the information on
375: $\xi$ i.e., the angle between the three dimensional major axis of the Virgo
376: cluster and the line of sight direction as approximately $10^{\circ}$.
377: However, we still need the azimuthal angle $\psi$ of the major axis of the
378: Virgo cluster.
379:
380: To find the azimuthal angle $\psi$, we first let $\xi=10^{\circ}$ and
381: $\psi=0$ in the analytic model (eq.[\ref{eqn:2dp}]). It implies that we
382: choose a certain Cartesian coordinates $(x_{1p},x_{2p})$ where the angle
383: $\psi$ vanishes. Then, we transform the coordinate system
384: of $(x_{1l},x_{2l})$ into this new coordinate system as
385: \begin{equation}
386: \label{eqn:new}
387: \left(\begin{array}{cc} x_{1p} \\ x_{2p} \end{array}\right)
388: =\left(\begin{array}{cc} \cos\psi & \sin\psi \\ -\sin\psi &
389: \cos\psi \end{array}\right) \left(\begin{array}{cc} x_{1l} \\ x_{2l}
390: \end{array}\right).
391: \end{equation}
392: Here, note that $(\cos\psi, \sin\psi)$ corresponds to the projected
393: major axis of the Virgo cluster in the $(x_{1l},x_{2l})$ coordinate system.
394: It is expected that in this new $(x_{1p},x_{2p})$-coordinate system the
395: the observationally measured distribution should fit the analytic model
396: (eq.[\ref{eqn:2dp}] best.
397:
398: For a given $\psi$, we measure the alignment angle between the projected
399: major axis and position vector of each Virgo cluster galaxy as
400: \begin{eqnarray}
401: \label{cos}
402: \cos\theta_{2d} = \hat{x}_{1p}\cos\psi + \hat{x}_{2p}\sin\psi,
403: \end{eqnarray}
404: where $\hat{\bf x}_{p}\equiv {\bf x}_{p}/\vert{\bf x}_{p}\vert$. Then, by
405: counting the number of galaxies galaxy's number density as a function of
406: $\cos\theta_{d}$, we can derive the probability distribution,
407: $p(\cos\theta_{d})$. We fit this observational distribution with the analytic
408: model, adjusting the values of $\lambda_{1},\lambda_{2}$ and $s$ through
409: $\chi^{2}$-minimization. We repeart the whole process varying the value
410: of $\psi$, and seek for the value of $\psi$ which yields the smallest
411: $\chi^{2}$ value. As a final step, we determine the corresponding best-fit
412: values of $\lambda_{1}$ and $\lambda_{2}$.
413:
414: Finally, we find the axial ratios of the Virgo cluster to be $a/c=0.54$
415: and $b/c=0.73$ by using equation (\ref{eqn:axi}) with the constraint of
416: $\delta_{c}=\sum_{i=1}^{3}\lambda_{i}$. The best-fit value of $s$ is also
417: determined to be $-0.25$, indicating that the positions of the cluster
418: galaxies are indeed aligned with the major axes of the Virgo cluster.
419: Figure [\ref{fig:plane}] plots the two dimensional projected positions of
420: the Virgo cluster galaxies in the $(x_{1p},x_{2p})$-coordinate system.
421: The arrow represents the direction of the projected major axis determined
422: from the chosen value of $\psi$.
423:
424: Figure \ref{fig:fit} plots the probability distribution of the alignment
425: angles for four different cases of $\psi$. In each panel, the histogram
426: with Poisson errors is the observational distribution while the solid line
427: represents the analytic fitting model. The dotted line stands for the
428: case of no alignment at all. The top left panel corresponds to the
429: finally chosen value of $\psi$ for which the analytic fitting
430: model and the observational result agree with each other best. The other
431: three panels show the three exemplary cases of $\psi$ for which the
432: analytic fitting model and the observational result do not agree with
433: each other well with relatively high $\chi^{2}$ value.
434:
435: To investigate whether the reconstructed axial ratios of the Virgo cluster
436: changes with the radius from the center, we introduce a cut-off radius,
437: $R_{\rm cut}$, and remeasure the axial ratios using only those galaxies
438: located within $R_{\rm cut}$ from the center in the $(x_{1p},x_{2p})$-
439: corrodinate system. We repeat the same process but using four different
440: values of $R_{\rm cut}$. Table \ref{tab:fit} lists the values of the
441: resulting best-fit axial ratios $a/c$ and $b/c$, and the best-fit correlation
442: parameter for the four different cut-off radii of $R_{cut}$. As can be seen,
443: at the inner radius smaller than the maximum one, the shape tends to be more
444: prolate-like, consistent with recent numerical report \citep[e.g.,][]{hay07}.
445: Note also that the value of $s$ is consistently $-0.25$, which implies the
446: strength of the tidal interaction with surrounding matter is consistent.
447:
448: Figure \ref{fig:radii} plots the probability distributions for the four
449: different cases of $R_{\rm cut}$. The top left panel shows the case of maximum
450: cut-off radius. As can be seen, the agreements between the analytic fitting
451: model and the observational result are quite good even at inner radii.
452:
453: \section{SUMMARY AND DISCUSSION}
454:
455: We have modified the cluster reconstruction algorithm which was originally
456: developed by \citet{lee-kan06} to apply it to the two dimensional projected
457: images of galaxy clusters in practice. Assuming that unless the cluster major
458: axes are in the line-of-sight direction, we have found that the alignments
459: between the galaxy positions and the projected major axes can be used to
460: reconstruct the two axial ratios of the triaxial clusters. We have applied
461: the modified algorithm to the observational data of the Virgo cluster and
462: shown that the reconstructed axial ratios of ($1$: $0.54$ : $0.73$) are in
463: good agreement with the recent report from the Virgo Cluster survey
464: \citep{mei-etal07}, which proves the validity and usefulness of our
465: method in practice.
466:
467: Now that the Virgo cluster is found to be triaxial, let us discuss on
468: the triaxiality effect on the mass estimation. For simplification, let us
469: assume that the Virgo cluster has a uniform density. Then, the mass of
470: the triaxial Virgo cluster with the axial ratios given as $(1:0.54:0.73)$
471: is estimated to be $2.6$ times larger than the spherical case since the
472: spherical radius is close to the minor axis length of the Virgo cluster
473: since the major axis of the Virgo cluster is very well aligned with the
474: line of sight direction. Therefore, it can cause maximum $\sim 50\%$ errors
475: to neglect the Virgo cluster triaxiality.
476:
477: The most prominent merit of our method over the previous one is that
478: it reconstructs directly the {\it intrinsic three dimensional structures}
479: of the underlying triaxial dark matter halo using the fact that the cluster
480: triaxiality originated from the tidal interaction.
481: Conventionally, the triaxial shape of a cluster is found through calculating
482: its inertia momentum tensor. This conventional method, however, is unlikely
483: to yield the {\it intrinsic} shape of the underlying dark matter halo unless
484: the target cluster is a well relaxed system. In addition, our method does not
485: resort to any simplified assumption like the axis-symmetry or the alignment
486: with the line-of-sight and etc.
487:
488: Yet, it is worth mentioning here a couple of limitations of our method.
489: First, it assumes that the major axes of the clusters are not aligned with
490: the line-of-sight direction so that the alignments between the galaxy
491: positions and the projected major axes can be measured.
492:
493: Second, for the Virgo cluster, the crucial information on the angle
494: $\xi$ between the three dimensional major axis and the line of sight
495: has been already given \cite{wes00}. Which simplifies the whole of
496: our method since it was only $\psi$ that has to be determined.
497: For most clusters, however, this information is not given, so that
498: both the values of $\xi$ and $\psi$ have to be determined through
499: fitting before finding the axial ratios.
500:
501: Third, its success is subject to the validity of the LK06 algorithm.
502: According to the numerical test, the LK06 algorithm suffers approximately
503: $20\%$ errors for the case that the number of the member galaxies is not
504: high enough. It implies that the LK06 algorithm is definitely restricted
505: to the rich clusters with large numbers of galaxies.
506: Therefore, it may be necessary to refine and improve the LK06 algorithm
507: itself for the application to poor cluster samples. Our future work is in
508: this direction.
509:
510: \acknowledgments
511:
512: This work is supported by the research grant No. R01-2005-000-10610-0 from
513: the Basic Research Program of the Korea Science and Engineering Foundation.
514:
515: \clearpage
516: \begin{thebibliography}{100}
517: \bibitem[Algood et al.(2006)]{alg-etal06}
518: Algood, B. et al. 2006, \mnras, 367, 1781
519: \bibitem[Altay et al.(2006)]{alt-etal06}
520: Altay, G., Colberg, J. M., \& Croft, R. A. C. 2006, \mnras, 370, 1422
521: \bibitem[Anninos \& Norman(1996)]{ann96}
522: Anninos, P., \& Norman, M. L. 1996, \apj, 459, 12
523: \bibitem[Bardeen et al.(1986)]{bar-etal86}
524: Bardeen, B., Bond, R. J., Kaiser, N., \& Szalay, H. 1986, \apj, 23, 567
525: \bibitem[Basilakos et al.(2000)]{bas00}
526: Basilakos, S., Plionis, M., \& Maddox, S.J. 2000, \mnras, 316, 779
527: \bibitem[Binggeli(1982)]{bin82}
528: Binggeli, B. 1982, \aap, 107, 338
529: \bibitem[Binggeli et al.(1985)]{bin-etal85}
530: Binggeli, B., Sandage, A., \& Tammann, G., A. 1985, \aj, 90, 1681
531: \bibitem[Binney(1985)]{bin85}
532: Binney, J. 1985, \mnras, 212, 767
533: \bibitem[Bohringer et al.(1994)]{boh94}
534: Bohringer, H. et al. 1994, Nature, 368, 828
535: \bibitem[Bond et al.(1996)]{bon-etal96}
536: Bond, J. R., Kofman, L., \& Pogosyan, D. 1996, Nature, 380, 603
537: \bibitem[Cen(1997)]{cen97}
538: Cen, R. 1997, \apj, 485, 39
539: \bibitem[Cooray(1998)]{coo98}
540: Cooray, A. R. 1998, \aap, 339, 623
541: \bibitem[Eke et al.(1996)]{eke96}
542: Eke, V. R., Cole, S., \& Frenk, C. S. 1996, \mnras, 282, 263
543: \bibitem[Evrard(1997)]{evr97}
544: Evrard, A. E. 1997, \mnras, 292, 289
545: \bibitem[Fabricant \& Gorenstein(1983)]{fab83}
546: Fabricant, D., Gorenstein, P. 1983, \apj, 267, 535
547: \bibitem[Ferramacho \& Blanchard(2007)]{fer-bla07}
548: Ferramacho, L. D., \& Blanchard, A. 2007, \aap, 463, 423
549: \bibitem[Fouque et al.(2001)]{fou01}
550: Fouque, P. S., Solanes, J. M., Sanchis, T., \& Balowski, C. 2001,
551: \aap, 375, 770
552: \bibitem[Fox \& Pen(2001)]{fox-pen01}
553: Fox, D. C., \& Pen U. L. 2002, \apj, 574, 38
554: \bibitem[Frenk et al.(1988)]{fre-etal88}
555: Frenk, C. S., White, S. D. M., Davis, M., Efstathiou, G. 1988, \apj, 327, 507
556: \bibitem[Hayashi et al.(2007)]{hay07}
557: Hayashi, E., Navarro, J. F., \& Springel, V. 2007, accepted \mnras
558: \bibitem[Hopkins et al.(2005)]{hop-etal05}
559: Hopkins, P. F., Bahcall, N., \& Bode, P. 2005, \apj, 618, 1
560: \bibitem[Jing \& Suto(2002)]{jin02}
561: Jing, Y.P., \& Suto, Y. 2002, \apj, 574, 538
562: \bibitem[Kasun \& Evrard(2005)]{kas-evr05}
563: Kasun, S. F., Evrard, A. E. 2005, \apj, 629, 781
564: \bibitem[Klypin et al.(1999)]{kly99}
565: Klypin, A., Cottlober, S., \& Kravtsov, A. V. 1999, \apj, 516, 530
566: \bibitem[Laroque et al.(2006)]{lar06}
567: Laroque, S. J. 2006, \apj, 652, 917
568: \bibitem[Lee \& Pen(2001)]{lee01}
569: Lee, J., Pen, U. L. 2001, \apj , 555, 106
570: \bibitem[Lee \& Suto(2003)]{lee03}
571: Lee, J., Suto, Y. 2003, \apj, 585, 151
572: \bibitem[Lee \& Suto(2004)]{lee04}
573: Lee, J., Suto, Y. 2004, \apj , 601, 599
574: \bibitem[Lee et al.(2005)]{lee05}
575: Lee, J., Kang, Xi, \& Jing, Y. P. 2005, \apj, 639, L5
576: \bibitem[Lee \& Kang (2006)]{lee-kan06}
577: Lee, J., Kang, Xi 2006, \apj, 637, 561
578: \bibitem[Lokas \& Mamon(2003)]{lok03}
579: Lokas, E. L., Mamon, G. A. 2003, \mnras, 343, 401
580: \bibitem[Lubin et al.(1996)]{lub-etal96}
581: Lubin, L. M., Cen, R., Bahcall, N. A., \& Ostriker, J. P. 1996, ApJ, 460, 10
582: \bibitem[Mamon et al.(2004)]{mam04}
583: Mamon, G. A., Sanchis, T. S., \& Solanes, J. M. 2004,
584: \aap, 414, 445
585: \bibitem[Mclaughlin et al.(1999)]{mcl99}
586: Mclaughlin, D. E. 1999, \apj, 512, L9
587: \bibitem[Mei et al.(2007)]{mei-etal07}
588: Mei, S. et al. 2007, \apj , 655, 144
589: \bibitem[Mohr et al.(1999)]{moh99}
590: Mohr, J. J., Mathiesen, B., \& Evrard,A. E. 1999, \apj, 517, 627
591: \bibitem[Nulsen \& Bohringer(1995)]{nul95}
592: Nulsen, P. E. J., Bohringer, H. 1995, \mnras, 274, 1093
593: \bibitem[Paz et al.(2006)]{paz06}
594: Paz, D. J., Lambas, D. G., Padilla, N., \& Merchan, M. 2006, \mnras, 366, 1503
595: \bibitem[Plionis et al.(1991)]{plio91}
596: Plionis, M., Barrow, J. D., \& Frenk, C. S. 1991, \mnras, 249, 662
597: \bibitem[Sanderson et al.(2003)]{san03}
598: Sanderson, A. J. R. 2003, \mnras, 340, 989
599: \bibitem[Schindler et al.(1999)]{sch99}
600: Schindler, S., Binggeli, B., \& Bohringer, H. 1999, \aap , 343, 420
601: \bibitem[Smith \& Watts(2005)]{smi-wat05}
602: Smith, R. E., \& Watts, P. I. R. 2005, \mnras, 360, 203
603: \bibitem[Solanes et al.(2002)]{sol02}
604: Solanes, J. M. 2002, \apj, 124, 2440
605: \bibitem[Struble \& Peebles(1985)]{str-pee85}
606: Struble, M. F., Peebles, P. J. E. 1985, \aj, 90, 582
607: \bibitem[Suto et al.(1998)]{sut98}
608: Suto, Y., Sasaki, S., \& Makino, N. 1998, \apj, 509, 544
609: \bibitem[Suwa et al.(2003)]{suw-etal03}
610: Suwa, T., Habe, A., Yoshikawa, K., \& Okamoto, T. 2003, \apj, 588, 7
611: \bibitem[Tonry et al.(2001)]{ton01}
612: Tonry, J. L. et al. 2001, \apj, 546, 681
613: \bibitem[Warren et al.(1992)]{war-etal92}
614: Warren, M. S., Quinn, P. J., Salmon, J. K., Zurek, W. H., \apj, 399, 405
615: \bibitem[West(1989)]{wes89}
616: West, M. J. 1989, \apj, 347, 610
617: \bibitem[West \& Blakeslee(2000)]{wes00}
618: West, M. J., Blakeslee, J. P. 2000, \apj, 543, L27
619: \bibitem[White et al.(1993)]{whi-etal93}
620: White, S. D. M., Navarro, J., \& Evrard, A. E., \& Frenk, C. S. 1993,
621: Nature, 366, 429
622: \bibitem[Zentner et al.(2005)]{zen05}
623: Zentner, A. R. et al. 2005, \apj, 629, 219
624: \end{thebibliography}
625: \clearpage
626: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
627: \begin{figure}
628: \begin{center}
629: \plotone{f1.eps}
630: \caption{Positions of the Virgo member galaxies in the two dimensional
631: projected space. The arrow represents the direction of the projected major
632: axes of the Virgo cluster.}
633: \label{fig:plane}
634: \end{center}
635: \end{figure}
636: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
637: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
638: \begin{figure}
639: \begin{center}
640: \plotone{f2.eps}
641: \caption{Probability density distributions of the cosines of the angles
642: between the position vectors of the Virgo cluster galaxies and four different
643: choices of the Virgo cluster major axis in the two dimensional projected
644: plane of sky. In each panel, the histogram with Poisson errors represents the
645: observational data points from the Virgo Catalog, the solid line is the
646: analytic fitting function based on the LK06 reconstruction algorithm, and
647: the dotted line corresponds to the case of no correlation.
648: The top left panel corresponds to the best-fit result according to which
649: the major axes of the Virgo cluster in the two dimensional projected plane
650: is determined, while the other three panels show how the agreements between
651: the observational and the analytic results change if different directions
652: other than the major axes are used.}
653: \label{fig:fit}
654: \end{center}
655: \end{figure}
656: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
657:
658: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
659: \begin{figure}
660: \begin{center}
661: \plotone{f3.eps}
662: \caption{Comparison between the observational and the analytic results for
663: the alignments at different two dimensional cut-off radii ($R_{\rm cut}$).}
664: \label{fig:radii}
665: \end{center}
666: \end{figure}
667: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
668: \clearpage
669: \begin{deluxetable}{ccccc}
670: \tablewidth{0pt}
671: \setlength{\tabcolsep}{5mm}
672: \tablehead{$R_{\rm cut}$ & $N_{g}$ & $a/c$ & $ b/c $ & $s$ \\
673: (Mpc) & & & & }
674: \tablecaption{The cut-off radius ($R_{\rm cut}$) in the two dimensional
675: projected space, the number of the member galaxies enclosed within
676: $R_{\rm cut}$, the reconstructed two axial ratios, and the best-fit value
677: of the correlation parameter.}
678: \startdata
679: $2.24$ & $1275$ & $0.53$ & $0.73$ & $-0.25$ \\
680: $1.68$ & $1221$ & $0.64$ & $0.69$ & $-0.25$ \\
681: $1.49$ & $1154$ & $0.64$ & $0.69$ & $-0.25$ \\
682: $1.12$ & $902$ & $0.64$ & $0.69$ & $-0.25$ \\
683: \enddata
684: \label{tab:fit}
685: \end{deluxetable}
686: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
687: \end{document}
688: