1: \documentstyle [12pt,aaspp4]{article}
2: %\documentstyle [12pt,aasms4]{article}
3: %\documentclass{aastex}
4: %\documentclass[preprint]{aastex}
5: %\documentclass[preprint2]{aastex}
6: %\documentclass{article}
7: %\usepackage{emulateapj}
8: %\usepackage{epsf}
9: %\usepackage{amssymb}
10: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
11: \newcommand{\simgt}{\lower.5ex\hbox{$\; \buildrel > \over \sim \;$}}
12: \newcommand{\simlt}{\lower.5ex\hbox{$\; \buildrel < \over \sim \;$}}
13: \newcommand{\himpc}{{\hbox {$h^{-1}$}{\rm Mpc}} }
14: \def\pppm{\rm P^3M}
15: \newcommand{\citet}[1] {\cite{#1}}
16: \newcommand{\citep}[1] {(\cite{#1})}
17: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
18: \slugcomment{SHAO/MPA-003}
19: %
20: %
21: \begin{document}
22: %
23: \title{A new test for the stable clustering hypothesis}
24:
25: \author{Y.P. Jing$^{1,2}$} %and Yasushi Suto$^{3}$}
26: \affil{$ ^1$
27: Shanghai Astronomical Observatory, the Partner Group of MPI f\"ur
28: Astrophysik, Nandan Road 80, Shanghai 200030, China}
29: \affil {$^2$ National Astronomical Observatories, Chinese Academy of
30: Sciences, Beijing 100012, China}
31: %\affil{$^3$ Department of Physics and Research Center for
32: % the Early Universe (RESCEU) \\ School of Science, University of
33: % Tokyo, Tokyo 113-0033, Japan.}
34: %\affil{ypjing@center.shao.ac.cn, suto@phys.s.u-tokyo.ac.jp}
35: \affil{ypjing@center.shao.ac.cn}
36:
37: \received{2001 January}
38: \accepted{2001 ???}
39:
40: \begin{abstract}
41: The stable clustering hypothesis is a fundamental assumption about the
42: nonlinear clustering of matter in cosmology. It states that the mean
43: physical separation of particles is a constant on sufficiently small
44: scales. While many authors have attempted to test the hypothesis with
45: cosmological N-body simulations, no consensus has been reached on
46: whether and where the hypothesis is valid, because of the limited
47: dynamical range that this type of simulations can achieve. In this {\it
48: Letter}, we propose to test the hypothesis with high resolution halo
49: simulations, since the individual halo simulations can resolve much
50: better the fine structures of the halos and since almost all pairs of
51: particles with small separations are presumed to be inside virialized
52: halos. We calculated the mean pair velocity for 14 high resolution
53: halos of $\sim 1$ million particles in a low-density flat cold dark
54: matter model. The result agrees very well with the stable clustering
55: prediction within the measurement uncertainty $\sim 30\%$ over a large
56: range of scales where the overdensity is $10^3$ to $10^6$. The
57: accuracy of the test can be improved to $\sim 10\%$ if some 100 halos
58: with a similar resolution are analyzed.
59:
60: \end{abstract}
61:
62: \keywords{galaxies: clusters: general -- cosmology}
63:
64: \section{Introduction}
65: The stable clustering hypothesis is a basic assumption in cosmology
66: about the nonlinear clustering of cosmic matter
67: (\cite{Peebles1980}). It states that the mean physical separations $r$
68: of particles in highly non-linear regions do not change with the
69: expansion of the Universe. In terms of the scale factor $a$ and the
70: comoving coordinates ${\bf x}$, the hypothesis equivalently states
71: that the relative pair velocity in the comoving coordinates $v=a
72: \dot{x}$ cancels out the Hubble flow $\dot{a}x$, i.e.
73: \begin{equation}
74: v=-\dot{a}x=-Hr\,.
75: \label{eq1}
76: \end{equation}
77: Based on this assumption plus self-similarity, it is
78: predicted that 1) the power-law index
79: of the two-point correlation function $\xi(r)$ on sufficiently small
80: scales is related to the power-law index $n$ of the linear power spectrum $P(k) \propto
81: k^{n}$ as $\xi(r) \propto r^{-3(3+n)/(5+n)}$, and 2) the N-point
82: correlation functions $\zeta_{N}$ scale with $\xi$ as
83: $\zeta_{N}\propto \xi^{N-1}$ (\cite{P74}, \cite{DP77},
84: \cite{GR75}, \cite{Peebles1980}, \cite{bs89}, \cite{Suginohara1991}, \cite{Suto1993},
85: \cite{Jain1997}). Based on the
86: first prediction for clustering on small scale and the linear theory
87: prediction for large scale, a phenomenological fitting formula has
88: been found for the nonlinear power spectrum (\cite{Hamilton1991},
89: \cite{JMW1995}, \cite{PD1996}, \cite{Ma1998}, \cite{Suginohara2001}).
90: This fitting formula has been applied to study a wide range of
91: non-linear clustering problems, notably the gravitational lensing and
92: the non-linear clustering of galaxies.
93:
94: There have been many attempts to verify the stable clustering
95: hypothesis with numerical N-body simulations. In their seminal work,
96: Efstathiou et al. (\cite{efs}) tested the hypothesis [Eq.(\ref{eq1})]
97: and its two predictions with a set of scale-free N-body simulations of
98: $32^3$ particles. They concluded that their simulation results are
99: consistent with the hypothesis, but it is not difficult to imagine
100: that the accuracy they could achieve is quite limited for the limited
101: resolution of this first generation of simulations. Subsequently,
102: many improved simulations with $64^3$ to $128^3$ particles have been
103: used to solve this problem (e.g. \cite{Suginohara1991},
104: \cite{Suto1993}, \cite{cbh1995}, \cite{munshietal1997}, and
105: \cite{Jain1997} and references therein). As shown by Jain (1997; his
106: Fig.2, Fig.3 and Fig.8), his simulation results of the mean pair
107: velocity have just approached the stable clustering condition
108: [Eq.(\ref{eq1})] at the smallest scale that his simulations of $\sim
109: 128^3$ particles can resolve. Thus, it remains an open question if the
110: stable clustering holds for a range of scales
111: (\cite{munshietal1997}). One point clear from these previous studies
112: is that only in high overdensity regions $\Delta \rho/\rho\gg 100$ is
113: the stable clustering assumption possibly valid. One needs simulations
114: which have much higher resolutions than those Jain analyzed, in order
115: to convincingly test the stable clustering.
116:
117: It is well known that close pairs of particles reside inside virialized
118: halos (e.g. \cite{mcs77}, \cite{Peebles1980}). This is
119: the reason why many people have recently constructed
120: successful models for non-linear statistics based on the
121: simulation result of the halo density profile and
122: the halo mass function (e.g. \cite{mjb},
123: \cite{sj97}, Ma \& Fry 2000a,b,c, \cite{uros},
124: \cite{CHM00}, \cite{jp},
125: \cite{halo}, \cite{shethetal2000}). In particular,
126: Ma \& Fry (2000b) checked the stable clustering assumption using such
127: an approach. They derived the mean pair velocity by inserting the halo
128: model two-point correlation function into the pair conservation
129: equation (e.g. \cite{Peebles1980}), and found that the stable clustering is
130: valid only when the halo mass function and the halo density profile
131: obey specific relations. Considering the fact that both models of the
132: halo density profile and the halo mass function have
133: considerable uncertainties and scatters
134: (\cite{jing2000}, \cite{JS00},
135: \cite{st99}, \cite{J00}), they concluded
136: that it is yet unclear if the stable clustering can be realized in
137: practice.
138:
139: In this {\it Letter}, we propose to use high-resolution halo
140: simulations to test the stable clustering hypothesis by examining the
141: mean pair velocity within virialized halos. The validity of this
142: approach underlies the above-mentioned fact that almost all close
143: pairs of galaxies reside inside virialized halos. If the mean pair
144: velocity within halos agrees with the stable clustering hypothesis,
145: one can sufficiently conclude that the stable clustering is
146: established, since the mean pair velocity in general is a
147: $N_{pair}^h(r,M) n(M)$ weighted average of the mean pair velocity
148: within halos, where $N_{pair}^h(r,M)$ is the mean pair number of
149: separation $r$ inside a halo of mass $M$, and $n(M)$ is the halo
150: number density per unit halo mass. The advantage of this approach over the
151: conventional approach which uses cosmological N-body simulations
152: (e.g. Jain 1997) is that the individual halos can be simulated with a
153: much better resolution. Here we use fourteen high-resolution halos to
154: make such an analysis. Twelve of these halos were used for studying
155: density profiles of dark matter halos (Jing \& Suto 2000) and the
156: other two were simulated after the work of Jing \& Suto (2000).
157:
158:
159: \section{Simulation data}
160:
161: Fourteen dark matter halos in total were selected from our one
162: cosmological P$^3$M N-body simulation of $256^3$ particles in a
163: $(100h^{-1}{\rm Mpc})^3$ cube (\cite{JS1998}, \cite{Jing1998}). The
164: model for this simulation is the standard Lambda cold dark matter
165: model which has the density parameter $\Omega_0=0.3$, the cosmological
166: constant $\lambda_0=0.7$, the Hubble constant $h=0.7$ and the power
167: spectrum normalization $\sigma_8=1.0$. These model parameters are
168: consistent with the current observations of the cluster abundance, of
169: the cosmic background radiation anisotropies, of high-redshift
170: supernova luminosity, and of many others. Of the fourteen halos, five,
171: five, and four have mass scales of clusters, groups and galaxies
172: respectively.
173:
174: These halos are re-simulated with high resolution using the multiple
175: mass method (e.g. \cite{kw93}) and our nested-grid $\pppm$ code
176: (\cite{JS00}). To minimize the contamination of the coarse particles
177: on the halo properties within the virial radius at $z=0$, $r_{\rm
178: vir}$, we trace back the particles within $3r_{\rm vir}$ of each halo
179: to their initial conditions at redshift $z=72$. The virial radius
180: $r_{\rm vir}$ is defined such that the spherical overdensity inside is
181: $\sim 18\pi^2 \Omega_0^{0.4} \sim 110$ times the critical density
182: $\rho_{\rm crit}(z=0)$ (\cite{Kitayama1997}, \cite{bn1998}). Typically
183: $ 1.5\times 10^6$ fine particles are placed in the high resolution
184: region, and $ 0.7\times 10^6$ coarse particles, whose mass increases
185: monotonically with the distance from the high-resolution region, are
186: placed outside the high resolution region to model the tidal force
187: effect. The initial conditions for these particles are produced with
188: the same random initial field as the cosmological N-body simulation
189: but with the shorter wavelength perturbations added that the
190: cosmological N-body simulation missed. The simulations are evolved by
191: our nested-grid $\pppm$ code which was designed to simulate
192: high-resolution halos. The force resolution is typically $0.004
193: r_{vir}$. At last about $(0.5 - 1)\times 10^6$ particles end up
194: within $r_{\rm vir}$ of each halo. For more details about the
195: simulation, we refer readers to Jing \& Suto (2000).
196:
197: \section{The mean pair velocity}
198: We calculate the mean pair velocity $v_r(r)$ by averaging the relative
199: velocity of all pairs of particles within the virial
200: radius. Separation bins are taken from $0.01r_{vir}$ to $r_{vir}$ with
201: $\Delta \lg r =0.05$. Since one halo has $(0.5 - 1)\times 10^6$
202: particles, there are more than $10^{11}$ pairs in total in each halo,
203: and more than $10^{6}$ pairs within the single innermost bin. The
204: high number of pairs is very crucial for investigating $v_r(r)$ at the
205: smallest separation. This is because relative to the stable clustering
206: prediction $v_r=-0.01H_0r_{vir}$ at $r=0.01r_{vir}$, the random
207: motion of particles (i.e. the noise for determining $v_r$) is very
208: large which, in terms of the one-dimensional velocity
209: dispersion, is about $5H_0r_{vir}$. In order to suppress the noise,
210: more than $500^2=2.5\times 10^5$ particle pairs are needed for the
211: innermost bin. Only high-resolution halo simulations like those used
212: in this paper can easily meet this requirement. In contrast,
213: cosmological N-body simulations would need many more than $128^3$
214: particles to study $v_r(r)$ at a similar scale.
215:
216: We find that the mean pair velocity $v_r(r)$ varies significantly from
217: one halo to another. In Figure 1, we show $v_r(r)/H_0 r$ for the five
218: cluster halos as a function of $r$. The halos are labeled in the same
219: way as Jing \& Suto (2000) with the newly added halo labeled as
220: CL5. Although the measured values of $v_r(r)/H_0 r$ are around the
221: stable clustering prediction $-1$, there is significant difference in
222: this quantity between one halo and another. Some halo has $v_r(r)/H_0
223: r$ systematically higher than $-1$ (CL3), and some lower than $-1$ (CL1 \&
224: CL2). We have examined whether $v_r(r)/H_0 r$ is correlated with the
225: dynamical state and the density profiles of halos. From Jing \& Suto
226: (2000), we know that CL1 has the smallest concentration parameter and
227: CL2 has the largest among the first four halos, while they have
228: similar $v_r(r)/H_0 r$. Therefore there seems to be no apparent
229: correlation between the density profiles and $v_r(r)/H_0 r$, which is
230: also supported by the results of galaxy and group halos (Figure
231: 2). The large variation of $v_r(r)/H_0 r$ from halo to halo might
232: reflect the fact that the merging and relaxation processes are
233: still taking place within the halos, which is not unexpected in the
234: hierarchical clustering scenario, and the value of $v_r(r)/H_0 r$
235: depends on a fine balance of the both processes within a halo.
236:
237: For testing the stable clustering hypothesis, the important quantity
238: is not the $v_r(r)/H_0 r$ of an individual halo, but the averaged
239: value of $v_r(r)/H_0 r$ over halos of a given mass. Figure 2 shows the
240: averaged $v_r(r)/H_0 r$ for the cluster (CL), group (GR) and galaxy
241: (GX) halos (note that the vertical axis is different from that of
242: Figure 1). The measurement has a typical uncertainty $0.6\sim 1.0$
243: (For clarity, the error bars are plotted for the cluster halos
244: only). As the figure shows, the result of each set of halos is
245: consistent with the stable clustering hypothesis within the
246: measurement uncertainty.
247:
248: Similarities have been found for the halos of different mass after the
249: length (and correspondingly other quantities) is scaled to the halo
250: size. Navarro et al. (1996) proposed a universal density profile for
251: halos, and Klypin et al (1999) and Moore et al. (1999) showed that the
252: number distributions of subhalos (after scaling to the circular
253: velocity) are very similar between galaxy and cluster halos. Although
254: there is some systematic difference between halos of different
255: mass even after scaling (e.g. the density profiles, Jing \& Suto
256: 2000), the small difference might not be important for the current
257: work. This is reflected in Figure 2 which shows no dependence of
258: $v_r(r)/H_0 r$ on the halo mass. As a further step, we average
259: $v_r(r)/H_0 r$ over all 14 halos at each $r/r_{vir}$-bin, and the
260: results are presented in Figure 3. It shows that the averaged result
261: is well consistent with the stable clustering prediction within a
262: typical uncertainty $\sim 30\%$. According to the density profiles,
263: the typical density contrast at $r=(0.01\sim 0.1)r_{vir}$ is $10^3$ to
264: $10^6$.
265:
266:
267: \section{Conclusion and future work}
268: We have made a first attempt to investigate the validity of the
269: stable clustering hypothesis using high-resolution halo
270: simulations. We have obtained the following conclusions.
271: \begin{itemize}
272: \item The mean pair velocity within a single halo fluctuates strongly
273: around the stable clustering prediction. This may indicate that the
274: merging and relaxation processes inside halos have not fully
275: completed, as should be expected for hierarchical clustering scenarios.
276: \item The mean pair velocity averaged over four or five halos of
277: a similar mass is consistent with the stable clustering prediction
278: with an uncertainty of $\sim 60\%$. No dependence on halo mass
279: has been seen for the quantity $v_r(r)/H_0 r$.
280: \item Based on the similarities exhibited by halos of different mass,
281: we averaged the $v_r(r)/H_0 r$ over the fourteen halos at each
282: $r/r_{vir}$-bin. The result is in very good agreement with the stable
283: clustering assumption with an uncertainty of $\sim 30\%$ over a large
284: range of the scales $r$, thus lending strong support for the stable clustering
285: hypothesis.
286: \end{itemize}
287:
288: It is important to emphasize again that the stable clustering
289: condition (1) is not generally satisfied within one single
290: virialized halo, but it can be achieved through averaging over many
291: halos. Although there is clear evidence for the similarities among
292: halos of different mass, we are still concerned of any dependence of
293: $v_r(r)/H_0 r$ on the halo mass. We have started a program to simulate
294: 50 halos at each mass range of clusters, groups, and galaxies, 150
295: halos in total. This will reduce the uncertainty of $v_r(r)/H_0 r$ to
296: $\sim 15\%$ for each mass range and will show the dependence of
297: $v_r(r)/H_0 r$ on the halo mass, if any.
298:
299:
300:
301: \acknowledgments
302: The author would like to thank Gerhard B\"orner and
303: Houjun Mo for helpful comments, and Yasushi Suto for his collaboration
304: on the simulation sample. The research is supported in part by the
305: One-Hundred-Talent Program, by NKBRSF(G19990754) and by NSFC.
306: Numerical computations were carried out on VPP300/16R and VX/4R at
307: ADAC (the Astronomical Data Analysis Center) of the National
308: Astronomical Observatory, Japan, as well as at RESCEU (Research Center
309: for the Early Universe, University of Tokyo) and KEK (High Energy
310: Accelerator Research Organization, Japan). Several helpful comments
311: from the referee, S. Colombi, are acknowledged.
312: %This
313: %research was supported in part by the Grant-in-Aid by the Ministry of
314: %Education, Science, Sports and Culture of Japan (07CE2002) to RESCEU,
315: %and by the Supercomputer Project (No.99-52) of KEK.
316:
317:
318: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
319: \begin{thebibliography}{}
320: \parskip=0pt
321: \baselineskip=7pt
322: \bibitem[Balian \& Schaeffer 1989]{bs89} Balian, R., Schaeffer, R. 1989,
323: \aap, 220, 1
324: \bibitem[Bryan \& Norman 1998]{bn1998} Bryan, G.L., Norman, M.L. 1998, \apj, 495, 80
325: \bibitem[Colombi et al. 1996]{cbh1995} Colombi S., Bouchet F.R., Hernquist L., 1996, ApJ, 465, 14
326: \bibitem[Cooray, Hu, \& Miralda-Escude 2000]{CHM00}Cooray, A., Hu, W., \& Miralda-Escude, J. 2000, astro-ph/0003205
327: \bibitem[Davis \& Peebles 1977]{DP77}Davis, M., \& Peebles, P. J. E. 1977, \apjs, 34, 425
328: \bibitem[1988]{efs} Efstathiou G., Frenk C. S., White S. D. M., Davis M. 1988, MNRAS, 235, 715
329: \bibitem[Gott \& Rees 1975]{GR75} Gott III J. R., Rees M., 1975, A\&A, 45, 365
330: \bibitem[Hamilton et al. 1991]{Hamilton1991} Hamilton, A.J.S., Kumar, P., Lu, E., \& Matthews, A. 1991, ApJ, 374, L1
331: \bibitem[Jain 1997]{Jain1997} Jain, B. 1997, \mnras, 287, 687
332: \bibitem[Jain, Mo, \& White 1995]{JMW1995} Jain, B., Mo, H.J., \& White, S.D.M. 1995, MNRAS, 276, 25
333: \bibitem[Jenkins etal. 2000]{J00}Jenkins, A., Frenk, C. S., White, S. D. M., Colberg, J. M., Cole, S.,Evrard, A. E., \& Yoshida, N. 2000, astro-ph/0005260
334: \bibitem[Jing 1998]{Jing1998} Jing, Y. P. 1998, ApJ, 503, L9
335: \bibitem[Jing 2000]{jing2000} Jing, Y.P. 2000, ApJ, 535, 30
336: \bibitem[Jing \& Suto 1998]{JS1998} Jing, Y.P. \& Suto, Y. 1998, ApJ, 494, L5
337: \bibitem[Jing \& Suto 2000]{JS00}Jing, Y. P., \& Suto, Y. 2000, \apj, 529, L69
338: \bibitem[Katz \& White 1993]{kw93}Katz, N., White, S.D.M. 1993, \apj, 412, 455
339: \bibitem[Kitayama \& Suto 1997]{Kitayama1997} Kitayama, T. \& Suto, Y. 1997, ApJ, 490, 557
340: \bibitem[Klypin et al. 1999] {Klypin} Klypin, A.A., Gottl\"{o}ber, S., Kravtsov, A.V. \& Khokholov, A.M. 1999, \apj, 516 530
341: \bibitem[Ma 1998] {Ma1998} Ma, C.-P. 1998, ApJ, 508, L5
342: \bibitem[Ma \& Fry 2000a]{mafry2000a} Ma, C.\ \& Fry, J.\ N.\ 2000, \apj, 543, 503
343: \bibitem[Ma \& Fry 2000b]{mafry2000b} Ma, C.\ \& Fry, J.\ N.\ 2000, \apjl, 538, L107
344: \bibitem[Ma \& Fry 2000c]{mafry2000c} Ma, C.\ \& Fry, J.\ N.\ 2000, \apjl, 531, L87
345: \bibitem[McClelland \& Silk 1977]{mcs77} McClelland J., Silk J., 1977, ApJ, 217, 331
346: \bibitem[Mo, Jing, \& B\"orner 1997]{mjb} Mo H. J., Jing Y. P., B\"orner G., 1997, MNRAS, 286, 979
347: \bibitem[Moore et al. 1999] {moore} Moore, B., Ghigna, S., Governato, F., Lake, G., Quinn,T., Stadel, J. \& Tozzi, P. 1999a, \apjl, 524, L19
348: \bibitem[Munshi et al. 1998] {munshietal1997} Munshi, D., Chiang, L.-Y., Coles, P., Melott, A.L. 1998, \mnras, /astro-ph/9707259
349: \bibitem[Navarro, Frenk, \& White 1996]{NFW1996} Navarro, J.F., Frenk, C.S., \& White, S.D.M. 1996, ApJ, 462, 563
350: \bibitem[Padmanabhan 1996]{paddy} Padmanabhan T., 1996, MNRAS, 278, 29P
351: \bibitem[Peacock \& Dodds 1996]{PD1996} Peacock, J.A. \& Dodds, S.J. 1996, MNRAS, 280, L19
352: \bibitem[Peacock \& Smith 2000]{jp} Peacock J., Smith R., 2000, MNRAS, submitted, astro-ph/0005010
353: \bibitem[Peebles 1974]{P74}Peebles, P. J. E. 1974, \apj, 189, L51
354: \bibitem[Peebles 1980]{Peebles1980} Peebles, P.J.E. 1980, The Large Scale Structure of the Universe (Princeton University Press: Princeton)
355: \bibitem[Scoccimarro et al. 2000]{halo} Scoccimarro R., Sheth R. K., Hui L., Jain B., 2000, ApJ,accepted, astro-ph/0006319
356: \bibitem[Seljak 2000]{uros} Seljak U., 2000, MNRAS, in press, astro-ph/0001493
357: \bibitem[Sheth \& Jain 1997] {sj97} Sheth R. K., Jain B., 1997, MNRAS, 285, 231
358: \bibitem[Sheth \& Tormen 1999]{st99} Sheth R. K., Tormen G., 1999, MNRAS, 308, 119
359: \bibitem[Sheth et al. 2000]{shethetal2000} Sheth, R.K., Hui, L., Diaferio, A., Scoccimarro, R., 2000, \mnras (submitted),
360: astro-ph/0009167
361: \bibitem[Suginohara et al. 1991]{Suginohara1991} Suginohara, T., Suto, Y., Bouchet, F.R., \& Hernquist, L. 1991, ApJS, 75, 631
362: \bibitem[Suginohara et al. 2001]{Suginohara2001} Suginohara, T., Taruya, A., Jing, Y.P., \& Suto, Y. 2001, ApJ (submitted)
363: \bibitem[Suto 1993]{Suto1993} Suto, Y. 1993, Prog.Theor.Phys., 90, 1173
364: (Princeton: Princeton Univ. Press)
365:
366: \end{thebibliography}
367: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
368:
369: \newpage
370:
371: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
372: \begin{figure}
373: \epsscale{1.0} \plotone{fig1.cps}
374: \caption{The mean peculiar velocity, in units of $H_0r$, of particles within cluster halos. }
375: \end{figure}
376: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
377: \begin{figure}
378: \epsscale{1.0} \plotone{fig2.cps}
379: \caption{The mean peculiar velocity, in units of $H_0r$, of particles averaged over four to five halos at cluster, group or galaxy masses. For clarity, errorbars are plotted for the cluster halos only. }
380: \end{figure}
381:
382: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
383: \begin{figure}
384: \epsscale{1.0} \plotone{fig3.cps}
385: \caption{The mean peculiar velocity, in units of $H_0r$, of particles
386: averaged over the 14 halos. }
387: \end{figure}
388: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
389:
390:
391:
392: \end{document}
393:
394:
395:
396: