0707.1690/ms.tex
1: \documentclass[12pt,preprint]{aastex}
2: \begin{document}
3: \title{The Nonlinear Evolution of Galaxy Intrinsic Alignments}
4: \author{Jounghun Lee}
5: \affil{Department of Physics and Astronomy, FPRD, Seoul National University, 
6: Seoul 151-747, Korea} 
7: \email{jounghun@astro.snu.ac.kr}
8: \author{Ue-Li Pen}
9: \affil{Canadian Institute for Theoretical Astrophysics, Toronto, ON M5S 3H8, 
10: Canada} 
11: \email{pen@cita.utoronto.ca}
12: 
13: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
14: \begin{abstract}
15: The non-Gaussian contribution to the intrinsic halo spin alignments 
16: is analytically modeled and numerically detected. Assuming that the growth 
17: of non-Gaussianity in the density fluctuations caused the tidal field to have 
18: nonlinear-order effect on the orientations of the halo angular momentum, 
19: we model the intrinsic halo spin alignments as a linear scaling of the 
20: density correlations on large scales, which is different from the previous 
21: quadratic-scaling model based on the linear tidal torque theory.
22: Then, we analyze the halo catalogs from the recent high-resolution 
23: Millennium Run simulation at four different redshifts ($z=0,0.5,1$ and $2$) 
24: and measure quantitatively the degree of the nonlinear effect on the halo 
25: spin alignments and its changes with redshifts. A clear signal of spin 
26: correlations is found on scales as large as $10h^{-1}$Mpc at $z=0$, which 
27: marks a detection of the nonlinear tidal effect on the intrinsic halo 
28: alignments. We also investigate how the nonlinear effect depends on the 
29: intrinsic properties of the halos. It is found that the degree of the 
30: nonlinear tidal effect increases as the halo mass scale decreases, the 
31: halo specific angular momentum increases, and the halo peculiar velocity 
32: decreases. We discuss implication of our result on the weak gravitational 
33: lensing. 
34: \end{abstract}
35: \keywords{cosmology:theory --- large-scale structure of universe}
36: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
37: 
38: \section{INTRODUCTION}
39: 
40: The intrinsic galaxy alignment refers to a cosmic phenomenon that the 
41: orientations of the galaxy spin axes (or galaxy ellipticities) are not 
42: random but locally aligned between neighbors due to the spatial correlations 
43: of the initial tidal field. The topic of the intrinsic galaxy alignments has 
44: recently drawn considerable attention in the field of the large-scale 
45: structure for three reasons. Firstly, since it is believed to be induced by 
46: the spatial correlations of the initial tidal field, it will be possible in 
47: principle by measuring the galaxy intrinsic alignments to reconstruct the 
48: initial density field on galactic scales 
49: \citep{lee-pen00,lee-pen01,lee-pen02,lee-erd07}.
50: 
51: Secondly, the initial tidal correlations are also responsible for the web-like 
52: distribution of the large scale structures \citep{bon-etal96}. Thus, the galaxy 
53: intrinsic alignments will be useful to characterize and quantify the influence 
54: of the large-scale structures on the orientations of the galaxies distributed 
55: in a cosmic web \citep{nav-etal04,tru-etal06,pat-etal06,ara-etal07,hah-etal07}.
56:  
57: Thirdly and most importantly, the galaxy intrinsic alignments could cause 
58: non-negligible contamination on weak lensing signals. Plenty of efforts have 
59: been made so far to distinguish and separate the intrinsic contaminations 
60: from the galaxy extrinsic alignments caused by the weak gravitational lensing 
61: effect \citep{cro-met00,hea-etal00,cat-etal01,cri-etal01,jin02,hui-zha02,
62: hey-hea03,tak-whi04,hir-sel04,kin05,man-etal06}.
63: 
64: For these reasons, it is of fundamental importance to predict the degree 
65: of galaxy intrinsic alignments with a valid theory and measure quantitatively 
66: their signals in practice. In fact, \citet{pen-etal00} have provided an 
67: analytic model for the intrinsic galaxy spin alignments in the frame of the 
68: linear tidal torque theory \citep{dor70,whi84,cat-the96}. According to their  
69: model, the galaxy intrinsic alignments can be expressed in terms of a 
70: quadratic scaling of the density correlation function and thus 
71: are expected to exist only at small distances of order of a few Mpc. 
72: A crucial implication of their analytic model is that due to the quadratic 
73: scaling nature of the galaxy intrinsic alignments the cross-correlations 
74: between the gravitational lensing shears from large scale density field 
75: and the galaxy intrinsic alignments with local tidal shears will be zero.
76: 
77: \citet{pen-etal00} tested their analytic predictions against low-redshift 
78: samples of spiral galaxies and found that the observed signals are 
79: consistent with the analytic predictions. Yet, their detections 
80: suffered from poor-number statistics and thus were still tentative. Later, 
81: several other authors have confirmed the existence of galaxy intrinsic 
82: alignments at small scales either in numerical simulations or in low-redshift 
83: observational surveys \citep{hea-etal00,jin02,bro-etal02,hey-etal04}.
84: 
85: However, \citet{hui-zha02} has pointed out that it might be invalid to 
86: describe the galaxy intrinsic alignments on large scales as a quadratic 
87: function of the density correlation. Their logic is as follows: 
88: Since the quadratic scaling is based on the linear tidal torque theory 
89: which adopts a somewhat oversimplified  assumption that the tidal field is 
90: Gaussian in the subsequent evolutionary stages, it should not be a good 
91: approximation to describe the galaxy intrinsic alignments  on large scales. 
92: In reality, the density fluctuations will develop non-Gaussianity via 
93: gravity which would in turn lead to non-negligible contributions of the 
94: nonlinear-order of the tidal tensors to the generation of the galaxy 
95: angular momentum. Due to this nonlinear effect on the galaxy angular 
96: momentum, their intrinsic spin alignments should be better approximated 
97: as linear scaling with the density correlation function. Because the 
98: linear scaling of the density correlation drops much slowly than the 
99: quadratic scaling, the intrinsic spin alignments would not be completely 
100: negligible even on large scales. If their claims turn out to be true, then 
101: it will have a significant impact not only on the weak lensing analysis but 
102: also on our fundamental understanding of the evolution of the tidal alignments. 
103: 
104: In the light of their claims, the following questions naturally arise; 
105: Does the nonlinear tidal effect on the galaxy intrinsic alignments really 
106: exist to a non negligible degree? If so, at what epochs and on which scales 
107: does its contribution begin to be significant? Does it depend on the 
108: intrinsic properties of the galaxies or dark halos? Our goal here is to 
109: answer the above questions using both analytical and numerical methods.
110: 
111: The outline of this paper is as follows: \S 2, we overview briefly the 
112: previous analytic model for the galaxy spin-spin alignments based on the 
113: linear tidal torque theory and propose a new model to account for the 
114: nonlinear tidal effect. \S 3, we report a detection of the signals 
115: of the nonlinear tidal effect on the intrinsic spin alignments of dark 
116: matter halos simulated in N-body experiments and show how the signals 
117: depend on redshift, scale, and the halo intrinsic property. \S 4, we 
118: summarize our results and discuss the implications of our work on 
119: the weak lensing effect.
120: 
121: \section{ANALYTIC MODEL}
122: 
123: \subsection{Overview of the Linear Tidal Torque Theory}
124: 
125: The linear tidal torque theory explains that unless a proto-halo has a 
126: perfectly spherical shape, it can acquire spin angular momentum at first 
127: order through its tidal interaction with the surrounding matter 
128: \citep{dor70,whi84,cat-the96}. The main prediction of the linear tidal 
129: torque theory is that the proto-halo angular momentum vector, ($J_{i}$), 
130: is proportional to the anti-symmetric product of the two tensors, the 
131: inertial momentum tensor $(I_{ij})$ and the local tidal shear tensor,
132: $(T_{ij})$, as
133: \begin{equation}
134: \label{eqn:ang}
135: L_{i} \propto \epsilon_{ijk}T_{jl}I_{lk}. 
136: \end{equation}
137: Here the defintions of $I_{ij}$ and $T_{ij}$ are given as
138: \begin{equation}
139: \label{eqn:int}
140: I_{ij} \equiv \int_{V}\! d^{3}{\bf q}\rho({\bf q})q_{i}q_{j}, \qquad
141: T_{ij}\equiv \frac{\partial^{2}}{\partial q_{i}\partial q_{j}}\Phi({\bf q}).
142: \end{equation}
143: where ${\bf q}$ is the Lagrangian position of the particles that reside in 
144: the proto-halo regions, $V$ and $\rho({\bf q})$ are the Lagrangian volume 
145: and the density of the proto-halo region, respectively, and $\Phi({\bf q})$ 
146: is the velocity perturbation potential.
147: 
148: Given the property of the perfectly anti-symmetric tensor 
149: $\epsilon_{ijk}$ in equation (\ref{eqn:ang}), an additional 
150: condition has to be satisfied for the first-order generation of the 
151: proto-halo angular momentum: the principal axes of $I_{ij}$ and $T_{ij}$ 
152: have to be misaligned with each other \citep{cat-the96}. 
153: Furthermore, a crucial implication of equation (\ref{eqn:ang}) is 
154: that if this additional condition is satisfied and thus the angular 
155: momentum of a proto-halo is generated at first order, then the direction 
156: of the proto-halo angular momentum is not random but preferentially 
157: aligned with the principal axes of the local tidal tensor \citep{lee-pen00}. 
158: 
159: Numerical experiments have revealed that the principal axes of $(I_{ij})$ 
160: and $(T_{ij})$ are correlated strongly but not perfectly 
161: \citep{lee-pen00,por-etal02}, which indicates that the first-order generation
162: of the proto-halo angular momentum is not so efficient and the degree of the 
163: alignments between the proto-halo spin directions and the principal axes of 
164: the local tidal shear tensors would not be so high.
165: 
166: Motivated by this numerical clue, \citet{lee-pen00} suggested the following 
167: quadratic formula to quantify the expected degree of the alignments between 
168: the halo spins and the local tidal shears, generalizing the linear tidal 
169: torque theory:
170: \begin{equation}
171: \label{eqn:stc}
172: \langle\hat{L}_{i}\hat{L}_{j}\rangle = \frac{1+a}{3}\delta_{ij} -
173: a\hat{T}_{ik}\hat{T}_{kj}.
174: \end{equation}
175: where $(\hat{L}_{i})$ is the unit spin vector of a halo, $(\hat{T}_{ij})$ is 
176: the unit traceless tidal tensor smoothed on the halo mass scale, and $a$ is 
177: a correlation parameter in the range of $[0,3/5]$ which measures the strength 
178: of the correlation between $(\hat{L}_{i})$ and $(\hat{T}_{ij})$. It is 
179: worth recalling the fact that the ensemble average  
180: $\langle\hat{L}_{i}\hat{L}_{j}\rangle$ in the left hand side of equation 
181: (\ref{eqn:stc}) is obtained by taking the average of $\hat{L}_{i}\hat{L}_{j}$ 
182: from the sample halos having all different mass, while the tidal shear field 
183: in the right hand side of equation (\ref{eqn:stc}) is smoothed on the single 
184: mass scale which amounts to the mean halo mass (i.e., the mean value of the 
185: mass of the sample halos).\citet{lee-pen00} derived equation (\ref{eqn:stc}) 
186: empirically, relating the angular momenta of the halos with different mass 
187: to the tidal shear field smoothed on the mean halo mass.
188: 
189: The correlation parameter $a$ in equation (\ref{eqn:stc}) which characterizes 
190: equation (\ref{eqn:stc}) is introduced to take into account strong 
191: correlations between $(I_{ij})$ and $(T_{ij})$ and any modification in the 
192: subsequent evolution after the moment of the turn-around. 
193: The two extreme cases of $a=0$ and $a=3/5$ represent no correlation and 
194: maximum correlation between $(\hat{L}_{i})$ and $(\hat{T}_{ij})$, 
195: respectively. It is expected that the true value of $a$ would be between 
196: these two extremes.
197: 
198: The linear tidal torque theory itself cannot make any prediction on the 
199: value of $a$. Thus, the true value of this free parameter $a$ has to be 
200: determined empirically. So far, several attempts have been made to measure 
201: the value of $a$ either from numerical simulations or from observations 
202: \citep{lee-pen00,por-etal02,nav-etal04,tru-etal06,lee-erd07}. Although no 
203: general consensus has been reached on the true value of $a$, most of the 
204: numerical and observational evidences indicated that the true value of 
205: $a$ deviates from zero.
206: 
207: \subsection{The First Order Approximation to the Halo Spin Correlations}
208: 
209: A difficulty in constraining the value of $a$ lies in the fact that 
210: it is hard to measure the principal axes of the linear tidal field in 
211: real space.  An alternative approach to the true value of $a$ is to 
212: measure the intrinsic spin-spin correlations of dark halos \citep{pen-etal00}.
213: The local alignments between the halo spin vectors and the tidal shear tensors 
214: will lead to spatial spin alignments between the neighbor halos. 
215: Using equation (\ref{eqn:stc}), \citet{pen-etal00} derived the following 
216: analytic model for the halo spin-spin correlations:
217: \begin{equation}
218: \label{eqn:1st}
219: \eta(r) \equiv \langle\vert\hat{\bf J}({\bf x})\cdot
220: \hat{\bf J}({\bf x + r})\vert^{2}\rangle - \frac{1}{3}
221: \approx \frac{a^2}{6}\frac{\xi^{2}(r;R)}{\xi^{2}(0;R)}.
222: \end{equation}
223: Here, the constant $1/3$ represents the value of 
224: $\langle\vert\hat{\bf J}({\bf x})\cdot\hat{\bf J}({\bf x}+{\bf r})
225: \vert^{2}\rangle$ for the case of no alignment, and  $\xi(r;R)$ is the linear 
226: two-point correlation function defined as
227: \begin{equation}
228: \label{eqn:xi}
229: \xi(r;R) \equiv \int_{-\infty}^{\infty}\Delta^{2}(k)\frac{\sin kr}{kr}
230: W^{2}(kR)d\ln k,
231: \end{equation}
232: where $\Delta^{2}(k)$ is the dimensionless power spectrum and $W(kR)$ 
233: is the top-hat spherical filter of scale radius, $R$, which is related 
234: to the halo mass scale $M$ as $R \equiv [3M/(4\pi\bar{\rho})]^{1/3}$ with 
235: the mean density $\bar{\rho}$. 
236: 
237: Note that equation (\ref{eqn:1st}) is proportional to the square of the 
238: two point density correlation $\xi(r)$, which means that the halo spin-spin 
239: correlations would decrease very rapidly. For instance, on a galactic scale 
240: ($R \sim 1 h^{-1}$Mpc), it is expected that the halo spin-spin 
241: correlations exist only at distances of order of a few Mpcs and 
242: effectively vanishes at larger distances.
243: 
244: Since the tidal field on different scales are cross-correlated, equation 
245: (\ref{eqn:stc}) leads to the existence of the cross-correlations of the 
246: spin axes of halos on different mass scales. In accordance with equation 
247: (\ref{eqn:1st}) the halo spin cross-correlations on different 
248: scales ($R_{1}$ and $R_{2}$) are modeled as
249: \begin{equation}
250: \label{eqn:crr}
251: \eta_{C}(r) \approx \frac{a^2}{6}\frac{\xi^{2}_{C}(r;R_{1},R_{2})}
252: {\xi^{2}_{C}(0;R_{1},R_{2})}.
253: \end{equation}
254: where $\xi^{2}_{C}(r;R_{1},R_{2})$ is the density cross-correlation 
255: function defined as 
256: \begin{equation}
257: \label{eqn:cxi}
258: \xi_{C}(r;R_{1},R_{2}) \equiv 
259: \int_{-\infty}^{\infty}\Delta^{2}(k)\frac{\sin kr}{kr}
260: W(kR_{1})W(kR_{2})d\ln k.
261: \end{equation} 
262: It is worth mentioning that the value of $a$ for the cross correlation 
263: $\eta_{C}(r)$ would not be same as its value for the auto-correlation 
264: $\eta(r)$ in equation (\ref{eqn:1st}) due to the difference in the smoothing 
265: scale for the tidal field. Note that equation (\ref{eqn:crr}) is a modified 
266: version of equation (\ref{eqn:1st}), both of which are based on equation 
267: (\ref{eqn:stc}). The correlation parameter $a$ in equation (\ref{eqn:1st}) 
268: represents the spin-spin correlations caused by the spatial correlations of 
269: the tidal fields smoothed on the same scale (one mean halo mass scale, $R$). 
270: While the correlation parameter $a$ in equation (6) represents the spin-spin 
271: correlations caused by the spatial correlations of the tidal fields on two 
272: different scales (two mean halo mass scales, $R_{1}$ and $R_{2}$). 
273: Since the spatial correlations of the two tidal fields smoothed on two 
274: different scales, $R_{1}$ and $R_{2}$ (say, $R_{1} < R_{2}$), are lower 
275: than the case that the two tidal fields are smoothed on the same scale 
276: of $R_{2}$ and higher than the case that they are smoothed on the same scale 
277: of $R_{1}$, it is naturally expected that the value of $a$ for the 
278: cross-correlation $\eta_{C}(r)$ should be different from that for the 
279: correlation $\eta(r)$.
280: 
281: \subsection{The Nonlinear Effect and Large Scale Correlations of Halo Spins}
282: 
283: As mentioned in \S 1, it was \citet{hui-zha02} who first pointed out that 
284: equation (\ref{eqn:1st}) is valid only if the gravitational tidal 
285: field is Gaussian. They argued that the non-Gaussianity in the tidal field
286: should not be negligible in the nonlinear regime and suggested that the 
287: halo spin-spin alignments should be better described as a linear scaling 
288: with the density correlation function on sufficiently large scales. 
289: 
290: Inspired by the work of \citet{hui-zha02}, here we suggest a new formula 
291: to model the halo spin-spin correlations:
292: \begin{equation}
293: \label{eqn:2nd}
294: \eta(r) \approx \frac{a^2_{\rm l}}{6}\frac{\xi^{2}(r;R)}{\xi^{2}(0;R)} + 
295: \varepsilon_{\rm nl}\frac{\xi(r;R)}{\xi(0;R)},
296: \end{equation}
297: where a nonlinear correlation parameter $\varepsilon_{\rm nl}$ is introduced 
298: to measure the strength of the nonlinear tidal effect on the halo 
299: spin-spin correlations. In this model the linear correlation parameter is 
300: notated as $a_{\rm l}$ since its value could be different between the 
301: two cases that the nonlinear effect is ignored as in equation 
302: (\ref{eqn:1st}) and taken into account as in (\ref{eqn:2nd}).
303:  
304: Equation (\ref{eqn:2nd}) implies that if the nonlinear contribution 
305: exists, then the halo spin-spin correlations must exist on larger scale 
306: than predicted by the linear model (eq.[\ref{eqn:1st}]) since it scales 
307: linearly with the density correlation function in the nonlinear model. 
308: That is, the nonlinear tidal effect would generate large-scale correlations 
309: of the halo spin axes. Of course, the values of the nonlinear correlation 
310: parameter $\varepsilon_{\rm nl}$ as well as the linear correlation parameter 
311: have to be determined empirically.
312: 
313: The cross-correlation, $\eta_{C}(r)$ can be also modeled in accordance 
314: with (\ref{eqn:2nd}) as
315: \begin{equation}
316: \label{eqn:crn}
317: \eta_{C}(r) \approx 
318: \frac{a^2_{\rm l}}{6}\frac{\xi^{2}_{C}(r;R_{1},R_{2})}
319: {\xi^{2}_{C}(0;R_{1},R_{2})} + 
320: \varepsilon_{\rm nl}\frac{\xi_{C}(r;R_{1},R_{2})}{\xi_{C}(0;R_{1},R_{2})}.
321: \end{equation}
322: It is worth mentioning that the value of $\epsilon_{\rm nl}$ for the cross 
323: correlation $\eta_{C}(r)$ would not be necessarily lower than its value 
324: for the auto-correlation $\eta(r)$ in equation (\ref{eqn:2nd}), unlike the 
325: case of the linear tidal torque model in \S 2.2. In the nonlinear regime 
326: where the non-Gaussianity in the density field grows, the occurrence of the 
327: halo merging is quite frequent. This occurrence of the halo merging plays a 
328: role of transferring the orbital angular momentum generated by the external 
329: tidal field into the spin angular momentum of a merged halo. In other words, 
330: the orbital angular momentum of a system composed of small individual halos 
331: become the spin angular momentum of a large halo formed through merging of 
332: the small halos. Given that the orbital angular momentum of the system 
333: before merging is generated by the external tidal field on larger scale 
334: \citep{vit-etal02}, the transfer of the orbital angular momentum into the 
335: spin angular momentum creates cross-correlations of the tidal fields between 
336: different scales, which in turn generates large-scale cross-correlations 
337: $\eta_{C}$ in the nonlinear regime.
338: 
339: \section{SIGNALS FROM NUMERICAL SIMULATIONS}
340: 
341: \subsection{Numerical Data}
342: 
343: We use the halo catalogs from the Millennium Run Simulation
344: \footnote{It is now available at http://www.mpa-garching.mpg.de/millennium} 
345: of $10^{10}$ dark matter particles for a $\Lambda$CDM universe with the 
346: cosmological parameters given as $\Omega_{m}=0.25$, $h=0.73$, 
347: $\sigma_{8}=0.9$ and $n_{s}=1$ \citep{spr-etal05}. 
348: The simulation was performed in a periodic box of linear size of  
349: $500 h^{-1}$Mpc to follow the evolution of all particles from $z=200$ 
350: to $z=0$, each of which has a mass of $8.6\times 10^{8}h^{-1}M_{\odot}$.
351: 
352: We analyze the halo catalogs at four different redshifts $z=0$, $0.5$, $1$ and 
353: $2$ which contain information on the halo mass, comoving position, comoving 
354: peculiar velocity and the spin vector. At each redshift, we restrict our 
355: attention only to those halos which have more than $200$ particles for 
356: better accuracy in the measurement of the halo spin vector. It amounts to 
357: selecting only halos which are more massive than 
358: $17.2\times 10^{9}h^{-1}M_{\odot}$.
359: 
360: Basically, we measure the spatial correlations of the spin axes between 
361: neighbor halos as a function of separation distance and investigate how 
362: the correlation strength changes with redshift (z), halo mass ($M$), 
363: velocity ($v$) and specific angular momentum ($l$, angular momentum per mass). 
364: Then, we compare the numerical results with the linear and the nonlinear 
365: analytic models described in \S 2 to determine the best-fit values of the 
366: correlation parameters, $a$, $a_{l}$, and $\varepsilon_{\rm nl}$.
367: 
368: \subsection{Dependence on the Redshift}
369: 
370: Figure \ref{fig:z} plots the halo spin-spin correlation function $\eta(r)$ at 
371: $z=0$,$0.5$,$1$ and $2$ in the top-left, top-right, bottom-left, and 
372: bottom-right panel, respectively. In each panel, the solid dots correspond 
373: to the numerical results from the Millennium data while the solid and the 
374: dashed lines represent the nonlinear model (eq.[\ref{eqn:2nd}]) and the 
375: linear model (eq.[\ref{eqn:1st}]), respectively. The case of no alignment 
376: is also shown as a dotted line in each panel for comparison. 
377: 
378: The errors of the numerical results are calculated as one standard deviation
379: for the case of no alignment. Basically, it is computed by the formula of
380: $2/\sqrt{45n_{h}}$ where $n_{h}$ is the number of halos belonging to each bin 
381: \citep{lee-pen01}. Note that the sizes of the vertical errors in Figure 
382: \ref{fig:z}  depend on the value of $n_{h}$ which in turn depends on the 
383: bin size of the spatial separation $r$. Here we choose the bin size of $r$ 
384: as $2h^{-1}$Mpc for all plots. We have tested whether or not the final 
385: results depend on the bin size of $r$ and confirmed the robustness 
386: of the final results.
387: 
388: For the analytic models, the correlation parameters are 
389: determined through fitting to the numerical results by means of the 
390: $\chi^{2}$-minimization. For the evaluation of the analytic models, we use 
391: the formula for the linear power spectrum given by \citet{bar-etal86} 
392: using the same values of the cosmological parameters that are used for the 
393: Millennium Run simulations and  the shape parameter $\Gamma = \Omega_{m}h$ 
394: (private communication with V. Springel).
395: 
396: Table \ref{tab:z} lists the number of halos ($N_{h}$), the mean mass 
397: ($\bar{M}$), the best-fit values of $a$, $a_{nl}$ and $\varepsilon_{\rm nl}$ 
398: at $z=0-2$. Here, the mean mass $\bar{M}$ for each bin is used to find the 
399: Lagrangian smoothing scale, $R$ which has to be implemented into 
400: the analytic models (eqs.\ref{eqn:1st} -\ref{eqn:2nd}).
401: As shown in Figure \ref{fig:z}, we detect clear signals of the halo spin-spin 
402: alignments within a few $h^{-1}$Mpc distances at all four redshifts. Note, 
403: however, a conspicuous difference between the results at low redshifts 
404: ($z=0$ and $0.5$) and at higher redshifts ($z=1$ and $2$). At $z=1$ and $2$, 
405: the correlation signals rapidly decrease with distance and disappear  
406: at distances around $10h^{-1}$Mpc. Thus, for the low redshifts, the linear 
407: model (dashed line) for itself fits the  numerical data pretty well. 
408: Whereas, at $z=0$ and $0.5$ the correlation signals decrease rather slowly, 
409: still existent to a nennegligible level at large separations beyond 
410: $10h^{-1}$Mpc.  Thus, at low redshifts the nonlinear model (solid line) 
411: fits the numerical data much better than the linear model, predicting the 
412: existence of the large-scale correlations.
413: 
414: This phenomena can be quantified in terms of the bestfit-values of 
415: the correlation parameters listed in Table \ref{tab:z}. At $2$,  
416: the best-fit value of $\varepsilon_{\rm nl}$ is effectively zero within 
417: one standard deviation ($\sigma$) while the best-fit value of $a_{\rm l}$ is 
418: higher than $10\sigma$ (not marginalized level).  In contrast, at $z=0$ the 
419: value of $\varepsilon_{\rm nl}$ deviates from zero as significantly as 
420: $12\sigma$ while the value of $a_{\rm l}$ decreases significantly to a 
421: $3\sigma$ level. It is, however, worthing mentioning here that the 
422: standard deviation, $\sigma$, here is not marginalized.
423: Figure \ref{fig:cz} plots the $68\%$, $95\%$ and $99\%$ contours for 
424: $a_{\rm l}$ and $\varepsilon_{\rm nl}$ at $z=0-2$, showing the degeneracy 
425: in putting the constraints on the two parameters. As can be seen, at $z=0$ and 
426: $0.5$, the values of $\varepsilon_{\rm nl}$ deviate from zero at higher than 
427: $99\%$ level.
428: 
429: Our numerical detection implies that the linear model for the halo spin-spin 
430: correlations (eq.[\ref{eqn:1st}]) works well at $z > 1$ but the nonlinear 
431: effect indeed dominates at low redshifts ($z < 1$), producing large-scale 
432: correlations of the halo spins, which can be well described by our nonlinear 
433: model (eq.[\ref{eqn:2nd}).
434: 
435: \subsection{Dependence on the Specific Angular Momentum Magnitude}
436: 
437: We define the magnitude of the rescaled specific angular momentum (angular 
438: momentum per unit mass) as $l \equiv L/M_{*}$ where $L$ and $M_{*}$ 
439: are the halo angular momentum magnitude and the halo mass in unit of 
440: $10^{10}h^{-1}M_{\odot}$, respectively. 
441: To determine quantitatively how the halo spin-spin correlation changes 
442: with $l$, we classify the halos at  $z=0$ into four logarithmic bins 
443: of $l$ and measure the correlations separately for each bin. 
444: Figure \ref{fig:a} plots $\eta(r)$ at four logarithmic bins of $l$ in 
445: the top-left, top-right, bottom-left, and bottom-right panels, respectively. 
446: Table \ref{tab:a} also lists the number of halos, the mean mass, the best-fit 
447: values of $a$, $a_{\rm l}$ and $\varepsilon_{\rm nl}$ for each bin, and 
448: Figure \ref{fig:ca} plots the $68\%$, $95\%$ and $99\%$ contours for 
449: $a_{\rm l}$ and$\varepsilon_{\rm nl}$ at the four bins. 
450: 
451: As shown in Figure \ref{fig:a}, in the bins of low-$l$ halos (top two panels), 
452: the correlations are rather weak and rapidly diminish to zero at separations 
453: greater than a few $h^{-1}$Mpc. Thus, for this case the linear model with 
454: the single parameter $a$ fits the data pretty well. In other words, 
455: for the spin-spin correlations of low-$l$ halos, the values of 
456: $\varepsilon_{\rm nl}$ are effectively zero as shown in Table \ref{tab:a}. 
457: While in the bins of high-$l$ halos, the correlations are stronger, still 
458: existent even at separations as large as $10h^{-1}$Mpc. Therefore for the 
459: spin-spin correlations of high-$l$ halos, the nonlinear model with non-zero 
460: value of $\varepsilon_{\rm nl}$ fits the data much better than the linear 
461: model. Note that for the case of highest-$l$ halos, the value of 
462: $\varepsilon_{\rm nl}$ is approximately $0.005$, which is five times 
463: greater than the average value of $0.001$ at $z=0$.
464: 
465: This phenomena that the halos with higher specific angular momentum tend to 
466: have stronger non-linear (and linear) spin-spin correlations may be explained 
467: as follows. If a halo is located in a region where the effect of the tidal 
468: field is very strong, then the halo will possess high specific angular 
469: momentum. Furthermore, its spin direction will be strongly correlated with 
470: that of its neighbors since the strong effect of the tidal field tends to 
471: diminish any randomization of the spin orientations that can occur in the 
472: subsequent stages.Therefore, those halos which have higher specific angular 
473: momentum are likely to show strong spin-spin correlations under the strong 
474: effect of the tidal field. We have also explored whether this phenomena 
475: occur at higher redshifts $z=0.5$,$1$ and $2$ and found that this is still 
476: the case.
477: 
478: \subsection{Dependence on the Mass Scale}
479: 
480: The halos at  $z=0$ are classified into four bins according to the value 
481: of the rescaled mass $M_{*}$ and their spin-spin alignments are measured 
482: separately for each bin. Figure \ref{fig:m} plots $\eta(r)$, 
483: Table \ref{tab:m} lists the number of halos, the mean mass, the best-fit 
484: values of $a$, $a_{\rm l}$ and $\varepsilon_{\rm nl}$, and Figure \ref{fig:cz} 
485: plots the $68\%$, $95\%$ and $99\%$ contours for $a_{\rm l}$ and 
486: $\varepsilon_{\rm nl}$ for the four bins.
487: 
488: As shown in Figure \ref{fig:m}, in the third bin which includes the galactic 
489: halos with mass of order of $(1-2)\times 10^{12}h^{-1}M_{\odot}$ 
490: (bottom-left panel), the correlation signal is strongest,  existent at 
491: separations as large as $10h^{-1}$Mpc. For this bin, the nonlinear model 
492: with nonzero value of $\varepsilon_{\rm nl}$ gives a better fit the numerical 
493: results than the linear with model single parameter $a$. The best-fit value 
494: of $\varepsilon_{\rm nl}$ deviates from zero by a factor of $3\sigma$ 
495: as shown in Table \ref{tab:m}. Whereas in the first and the fourth bins 
496: which include dwarf halos and galaxy-group halos, respectively, the signals 
497: are rather weak and the correlations disappear at large distances. For 
498: these bins, the linear model for itself fits the numerical data quite well 
499: and the values of $\varepsilon_{\rm nl}$ are effectively zero.
500: This result implies that the strength of the nonlinear effect depends on mass
501: scales, being strongest on the galactic mass scale. Note, however, here that 
502: the value of the characteristic mass scale changes with $z$. Thus, the 
503: results given in \S 3.2 and \S 3.4 are related to each other.
504: 
505: \subsection{Scale Bias and Cross-Correlations}
506: 
507: As mentioned in \S 2.3, if the nonlinear effect dominates on the halo 
508: spin-spin correlations, then it is likely to produce stronger 
509: cross-correlations between different mass scales (eq.[\ref{eqn:crn}]). 
510: We measure the halo spin cross-correlations between the halos of 
511: different mass bins that are constructed in \S 3.2 (the first bin 
512: corresponds to the lowest mass scales. see Table \ref{tab:m}), 
513: and find the best-fit values of $a$, $a_{nl}$ and $\varepsilon_{\rm nl}$ 
514: by fitting the numerical data to the analytic model (eq.[\ref{eqn:crn}]). 
515: 
516: Figure \ref{fig:b} plots the numerical results of $\eta_{C}$ for the six 
517: different cases: $\eta_{C12}$, $\eta_{C13}$, $\eta_{C14}$, $\eta_{C23}$, 
518: $\eta_{C24}$,$\eta_{C34}$, where $\eta_{Cij}$ represents the 
519: cross-correlations between the $i$-th and the $j$-th mass bins. 
520: As can be seen, the nonlinear model basically gives better fit to the 
521: numerical result for all cases, indicating that the nonlinear effect 
522: increases indeed increase spin cross-correlations between different 
523: mass scales, as presumed in \S 2.3. Figure \ref{fig:cb} plots the $68\%$, 
524: $95\%$ and $99\%$ contours for $a_{\rm l}$ and $\varepsilon_{\rm nl}$ 
525: for these two cases of $\eta_{13}$ and $\eta_{23}$ where the strong signals 
526: of the cross-correlations are detected. Obviously, the values of 
527: $\varepsilon_{\rm nl}$ for each case deviates from zero at $95\%$ 
528: confidence level.
529: 
530: \subsection{Dependence on the Velocity Magnitude}
531: 
532: We also investigate the dependence of the halo spin-spin correlations 
533: on the magnitude of halo peculiar velocity $v$ in unit of km/s. The 
534: correlations $\eta(r)$ are plotted in Fig.\ref{fig:v} at four logarithmic 
535: bins of $v$, and the $68\%$, $95\%$ and $99\%$ contours for the correlation 
536: parameters of the nonlinear models are shown in Fig.\ref{fig:cv}.
537: 
538: As shown in Figure \ref{fig:v}, the dependence of the halo spin-spin 
539: correlations on the velocity magnitude turns out not to be so strong.
540: The significant nonlinear effect on the halo-halo correlation is found  
541: only in the first bin of low-$v$ halos (top-left panel), For this bin, 
542: the linear model fails in fitting the numerical data while the nonlinear 
543: model gives a better fit, predicting the large-scale correlations. 
544: As shown in the top-left panel of Fig.\ref{fig:cv}, the value of 
545: $\varepsilon_{\rm nl}$ for the case of lowest-$v$ halos deviates 
546: from zero at $95\%$ confidence level.
547: 
548: \section{SUMMARY AND DISCUSSION}
549: 
550: We summarize our results in the following:
551: \begin{itemize}
552: \item
553: To account for the possible nonlinear tidal effect on the intrinsic 
554: halo alignments, we suggest a new model characterized by two free 
555: parameters, $a_{\rm l}$ and $\varepsilon_{\rm nl}$. The two parameters
556: $a_{\rm l}$ and $\varepsilon_{\rm nl}$ represents the strength of the 
557: linear and the nonlinear tidal effect on the orientations of the halo spins, 
558: respectively. If $\varepsilon_{\rm nl}=0$, it corresponds to the case 
559: that the nonlinear tidal effect is negligible and the halo spin-spin 
560: correlations is a quadratic scaling of the density correlation function 
561: $\xi(r)$, existent on small scales. While if $\varepsilon_{\rm nl}$ deviates 
562: from zero, the nonlinear tidal effect is nennegligible and the halo 
563: spin-spin correlations is a linear scaling of $\xi(r)$, present on large 
564: distance scales.
565: \item
566: Using the halo catalogs from numerical simulations at $z=0,1,0.5$ and $2$, 
567: we have measured the halo spin-spin correlations and determined the best-fit 
568: values of $a_{\rm l}$ and $\varepsilon_{\rm nl}$. It is found that the 
569: values of $a_{\rm l}$ increases with $z$ while $\varepsilon_{\rm nl}$ 
570: decreases with $z$. 
571: At $z=1$ and $2$, the values of $\varepsilon_{\rm nl}$ are found to be 
572: negligibly small, while at $z=0.5$ and $z=0$ $\varepsilon_{\rm nl}$ has 
573: an order of $10^{-3}$ higher than $5\sigma$. Especially at $z=0$, 
574: $a_{\rm l}$ is effectively zero and $\varepsilon_{\rm nl}$ deviates from 
575: zero by a factor more than $10\sigma$, which implies that at present epoch 
576: the nonlinear tidal effect is dominant inducing halo spin correlations 
577: on scales as large as $10h^{-1}$Mpc.
578: \item 
579: We have investigated the dependence of $a_{\rm l}$ and 
580: $\varepsilon_{\rm nl}$ on the halo intrinsic properties such as mass ($M$), 
581: specific angular momentum magnitude ($l$), and velocity magnitude ($v$) at 
582: $z=0$. It is found that the value of $\varepsilon_{\rm nl}$ depends most 
583: sensitively on the value of $l$. For those halos with highest $l$, the 
584: value of $\varepsilon_{\rm nl}$ reaches as high as $5\times 10^{-3}$ 
585: and the value of $a_{\rm l}$ is effectively zero. 
586: Regarding the dependence on $M$ and $v$, it is found that 
587: $\varepsilon_{\rm l}$ is less sensitive. However, for those halos with  
588: $M \sim (1-2)\times 10^{12}h^{-1}M_{\odot}$ $\log v < 2.5$km/s, the 
589: value of $\varepsilon_{\rm nl}$ tend to be higher.
590: \item
591: We have measured the cross-correlations of the spin axes between halos 
592: on different mass scales. It is found that on two different galactic mass 
593: scales ($M_{1} \sim 10^{11}h^{-1}M_{\odot}$ and 
594: $M_{s} \sim 10^{12}h^{-1}M_{\odot}$, the cross-correlations exist at 
595: scales as large as $10h^{-1}$Mpc, due to the growth of the non-Gaussianity 
596: increases cross-correlations of the tidal field on different scales. 
597: \end{itemize}
598: 
599: Our results have shown that the galaxy intrinsic alignments can exist not only 
600: for the case that the separation distance between the two galaxies is small 
601: but also for the case that any galaxy pair separated by distances larger than 
602: $10h^{-1}$Mpc can still have aligned axes. It also implies preferential E-mode 
603: contamination of the galaxy intrinsic alignments in weak lensing signals 
604: \citep{cri-etal02}.
605: 
606: In addition, we have also found quantitatively that the galaxy intrinsic 
607: alignments could be cross-correlated with the larger-scale tidal field. These 
608: results imply that the intrinsic galaxy correlations are indeed correlated 
609: with the surrounding density field on large scales, which will in turn lead 
610: to correlation between the gravitational lensing shears and the intrinsic 
611: galaxy alignments, (often called, GI correlations). 
612: 
613: Recently, \citet{hir-etal07} claimed that a signal of the GI correlations has 
614: been detected from the large low-redshift galaxy surveys 
615: \citep[see also,][]{man-etal06}. 
616: According to their claim, the GI correlations exist to a nennegligible level 
617: for the bright Luminous Red Galaxies (LRGs) and it is zero for the faint blue 
618: galaxies, mentioning an uncertainty in the estimation of the GI correlation for 
619: the faint blue galaxies. According to our numerical detection, the nonlinear 
620: tidal effect which causes the large-scale intrinsic alignments is  strongest 
621: for the halos with high specific angular momentum magnitude and low peculiar 
622: velocity, which usually correspond to the galaxies with low surface brightness 
623: \citep[][and references therein]{jim-etal98} and blue color \citep{dav-etal97}.
624: Thus, our results indicate that the GI correlation amplitude for the faint blue 
625: galaxies, may exceed the estimation upper limit, as noted by 
626: \citet{hir-etal07}. It will be of interest to compare the cross-correlations 
627: between the spin axes of the blue galaxies and the shapes of the red 
628: galaxies for the comparison of their correlation scales.
629: 
630: \acknowledgments
631: 
632: The Millennium Run simulation used in this paper was carried out by the Virgo 
633: Supercomputing Consortium at the Computing Centre of the Max-Planck Society 
634: in Garching. We thank V. Springel and G. Lemson for useful comments. We also 
635: thank an anonymous referee for helpful suggestions. J.L. is very grateful to 
636: the warm hospitality of L. Kofman and the Canadian Institute for Theoretical 
637: Astrophysics (CITA) where this work was initiated and performed. 
638: J.L. acknowledges the financial support from the Korea Science and Engineering 
639: Foundation (KOSEF) grant funded by the Korean Government 
640: (MOST, NO. R01-2007-000-10246-0).
641: 
642: 
643: \begin{thebibliography}{}
644: \bibitem[Aragon-Calvo et al.(2007)]{ara-etal07}
645: Aragon-Calvo, M. A., van de Weygaert, R., Jones, B. J. T., \& van der Hulst, 
646: J. M. 2007, \apj, 655, 5
647: \bibitem[Bardeen et al.(1986)]{bar-etal86}
648: Bardeen, J. M., Bond, J. R., Kaiser, N., \& Szalay, A. S. 1986, \apj, 304, 15
649: \bibitem[Bond, Kofman, \& Pogosyan(1996)]{bon-etal96}
650: Bond, J., R., Kofman, L., \& Pogosyan, D. 1996, Nature, 380, 603
651: \bibitem[Brown et al.(2002)]{bro-etal02}
652: Brown, M. L., Taylor, A. N., Hambly, N. C., \& Dye, S. 2002, \mnras, 333, 501
653: \bibitem[Catelan et al.(2001)]{cat-etal01}
654: Catelan, P., Kamionkowski, M., \& Blandford, R. D. 2001, \mnras, 320, L7
655: \bibitem[Catelan \& Theuns(1996)]{cat-the96}
656: Catelan, P., \& Theuns, T. 1996, \mnras, 282, 436
657: \bibitem[Crittenden et al.(2001)]{cri-etal01}
658: Crittenden, R. G., Natarajan, P., Pen, U. L \& Theuns, T. 2001, \apj, 559, 552
659: \bibitem[Crittenden et al.(2002)]{cri-etal02}
660: Crittenden, R. G., Natarajan, P., Pen, U. L \& Theuns, T. 2001, \apj, 568, 20
661: \bibitem[Croft \& Metzler(2000)]{cro-met00}
662: Croft, R. A. C. \& Metzler, C. A. 2000, \apj, 545, 561
663: \bibitem[Davis et al.(1997)]{dav-etal97}
664: Davis, M., Miller, A., \& White, S.D.M. 1997, \apj, 490, 63
665: \bibitem[Doroshkevich(1970)]{dor70}
666: Doroshkevich, A. G. 1970, Astrofizika, 6, 581
667: \bibitem[Hahn et al.(2007)]{hah-etal07}
668: Hahn, O., Porciani, C., Carollo, C. M., \& Dekel, A. 2007, \mnras, 375, 489
669: \bibitem[Heavens et al.(2000)]{hea-etal00}
670: Heavens, A., Refregier, A., \& Heymans, C. 2000, \mnras, 319, 649
671: \bibitem[Heymans \& Heavens(2003)]{hey-hea03}
672: Heymans, C. \& Heavens, A. 2003, \mnras, 339, 711
673: \bibitem[Heymans et al.(2004)]{hey-etal04}
674: Heymans, C., Brown, M., Heavens, A., Meisenheimer, K., Taylor, A., \& 
675: Wolf, C. 2004, \mnras, 361, 160
676: \bibitem[Hirata \& Seljak(2004)]{hir-sel04}
677: Hirata, C. M. \& Seljak, U. 2004, \prd, 70, 063526
678: \bibitem[Hirata et al.(2004)]{hir-etal04}
679: Hirata, C. M., et al. 2004, \mnras, 353, 529
680: \bibitem[Hirata et al.(2007)]{hir-etal07}
681: Hirata, C. M., et al. 2007, \mnras, 381, 1197
682: \bibitem[Hui \& Zhang(2002)]{hui-zha02}
683: Hui, L. \& Zhang Z. 2002, preprint [astro-ph/0205512]
684: \bibitem[Jimenez et al(1998)]{jim-etal98}
685: Jimenez, R., Padoan, P., Matteucci, F. \& Heavens, A. F.  1998, \mnras, 
686: 299, 123
687: \bibitem[Jing(2002)]{jin02}
688: Jing, Y.  2002, \mnras, 335, L89
689: \bibitem[King(2005)]{kin05}
690: King, L. 2005, \aap, 441, 47
691: \bibitem[Lee et al.(2005)]{lee-etal05}
692: Lee, J., Kang, X., \& Jing, Y. 2005, \apj, 629, L5
693: \bibitem[Lee \& Pen(2000)]{lee-pen00}
694: Lee, J. \& Pen, U. L. 2000, \apj, 532, L5
695: \bibitem[Lee \& Pen(2001)]{lee-pen01}
696: Lee, J. \& Pen, U. L. 2001, \apj, 555, 106
697: \bibitem[Lee \& Pen(2002)]{lee-pen02}
698: Lee, J. \& Pen, U. L.  2002, \apj, 567, 111
699: \bibitem[Lee \& Erdogdu(2007)]{lee-erd07}
700: Lee, J. \& Erdogdu, P.  2007, 671, 1248
701: \bibitem[Mandelbaum et al.(2006)]{man-etal06}
702: Mandelbaum, R., Hirata, C. M., Ishak, M., Seljak, U., \& Brinkmann, J. 
703: 2006, \mnras, 367, 611
704: \bibitem[Mo et al.(1998)]{mo-etal98} 
705: Mo, H. J., Mao, S. \& White, S. D. M. 1998, \mnras, 295, 319
706: \bibitem[Navarro et. al.(2004)]{nav-etal04}
707: Navarro, J.F., Abadi, M.G., \& Steinmetz, M. 2004, \apj, 613, L41
708: \bibitem[Patiri et al.(2006)]{pat-etal06}
709: Patiri, S. G., Cuesta, A. J., Prada, F., Betancort-Rijo, J., \& Klypin, A. 
710: 2006, \apj, 652, 75
711: \bibitem[Peebles(1969)]{pee69}
712: Peebles, P. J. E. 1969, \apj, 155, 393
713: \bibitem[Pen et al.(2000)]{pen-etal00}
714: Pen, U. L., Lee, J., \& Seljak, U. 2000, 543, L107
715: \bibitem[Porciani et al.(2002)]{por-etal02}
716: Porciani, C., Dekel, A., \& Hoffman, Y. 2002, \mnras, 332, 339
717: \bibitem[Springel et al.(2005)]{spr-etal05}
718: Springel, V. et al. 2005, \nat , 435, 629
719: \bibitem[Takada \& White(2004)]{tak-whi04}
720: Takada, M., \& White, S. D. M. 2004, \apj, 601, L1
721: \bibitem[Trujillo et. al.(2006)]{tru-etal06} 
722: Trujillo, I., Carretero, C., \& Patiri, S. 2006, \apj, 610, L111
723: \bibitem[Vitvitska et. al.(2002)]{vit-etal02} 
724: Vitvitska, M., Klypin, A. A., Kravtsov, A. V., Wechsler, R. H., 
725: Primack, J. R., \& Bullock, J. S. 2002, \apj, 581, 799
726: \bibitem[White(1984)]{whi84}
727: White, S. D. M. 1984, \apj, 286, 38
728: \end{thebibliography}
729: 
730: 
731: \clearpage
732: \begin{deluxetable}{cccccc}
733: \tablewidth{0pt}
734: \setlength{\tabcolsep}{5mm}
735: \tablehead{
736: $z$ & $N_{h}$ & $\bar{M}_{*}$ & $a\times 10^{2}$ & $a_{\rm l}\times 10^{2}$ & 
737: $\varepsilon_{\rm nl}\times 10^{3}$ \\ &  &  ($10^{10}h^{-1}M_{\odot}$) & &  & }
738: \tablecaption{The number of halos, the mean mass, the best-fit values of the 
739: linear and the nonlinear correlations parameters for the intrinsic halo spin 
740: correlations at four redshifts from the Millennium Run data.}
741: \startdata   
742: $0$ & $1846776$ & $169.7$ & $11.3\pm 0.6$ & $3.9\pm 1.6$  & $1.2\pm 0.1$  \\ 
743: $0.5$  & $1942263$ & $133.3$ & $9.8\pm 0.7$ & $6.1\pm 1.1$ & $0.7\pm 0.1$  \\ 
744: $1$ & $1928690$   & $106.7$ & $8.6\pm 0.8$ & $7.9\pm 0.9$& $0.1\pm 0.1$ \\ 
745: $2$& $1558504$   & $69.6$  &  $10.6\pm 0.9$ & $10.4\pm 0.9$ &$0.0\pm 0.2$  \\ 
746: \enddata
747: \label{tab:z}
748: \end{deluxetable}
749: %\end{tabular}
750: %\end{table}
751: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
752: \clearpage
753: \begin{deluxetable}{cccccc}
754: \tablewidth{0pt}
755: \setlength{\tabcolsep}{5mm}
756: \tablehead{$\log (l)$ & $N_{h}$  & $\bar{M}_{*}$ & $a\times 10^{2}$
757: & $a_{\rm l}\times 10^{2}$ & $\varepsilon_{\rm nl}\times 10^{3}$ \\ &  & 
758: ($10^{10}h^{-1}M_{\odot}$) &  &  &}
759: \tablecaption{Numerical results for the intrinsic halo spin correlations 
760: at four bins of the specific angular momentum magnitude $l$.}
761: \startdata   
762: $(-4.0, -2.0)$ & $350541$   & $607.1$ & $9.4\pm 2.4$ & $9.4\pm 2.4$  &
763: $0.0\pm 0.4$  \\ 
764: $(-2.0, -1.7)$  & $604356$   & $102.7$ & $10.1\pm 2.2$ & $6.6\pm 3.4$ &
765: $0.7\pm 0.4$ \\ 
766: $(-1.7,-1.5)$    & $475842$   & $51.0$ & $16.8\pm 2.1$ & $5.4\pm 6.3$   &
767: $2.7\pm 0.6$ \\ 
768: $(-1.5, 0.0)$    & $416032$   & $32.5$  &  $20.2\pm 2.2$ & $0.0\pm 9.7$  &
769: $4.9\pm 0.8$  \\ 
770: \enddata
771: \label{tab:a}
772: \end{deluxetable}
773: %\end{tabular}
774: %\end{table}
775: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
776: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
777: \clearpage
778: \begin{deluxetable}{cccccc}
779: \tablewidth{0pt}
780: \setlength{\tabcolsep}{5mm}
781: \tablehead{$M_{*}$ & $N_{h}$  & $\bar{M}_{*}$ & $a\times 10^{2}$   
782: & $a_{\rm l}\times 10^{2}$ & $\varepsilon_{\rm nl}\times 10^{3}$ \\ 
783: ($10^{10}h^{-1}M_{\odot}$) & &  ($10^{10}h^{-1}M_{\odot}$) &  &  &}
784: \tablecaption{Numerical results for the intrinsic halo spin correlations 
785: at four different mass ranges.}
786: \startdata   
787: $(17.2,38.0)$    & $925011$   & $25.0$ & $14.1\pm 1.6$ & $8.5\pm 2.6$  &
788: $1.3\pm 0.4$  \\ 
789: $(38.0, 85.0)$  & $469647$   & $55.7$ & $14.1\pm 2.3$ & $10.6\pm 3.2$ &
790: $1.0\pm 0.6$   \\ 
791: $(85.0,190.0)$    & $233724$   & $124.6$ & $20.8\pm 2.5$ & $15.0\pm 3.5$   &
792: $2.4\pm 0.9$ \\ 
793: $(190.0,10^{5})$    & $218248$   & $1022.5$  &  $15.5\pm 1.9$ & 
794: $13.0\pm 2.2$  & $0.8\pm 0.6$  \\ 
795: \enddata
796: \label{tab:m}
797: \end{deluxetable}
798: %\end{tabular}
799: %\end{table}
800: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
801: 
802: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
803: \clearpage
804: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
805:  \begin{figure}
806:   \begin{center}
807:    \plotone{f1.eps}
808: \caption{Intrinsic halo spin-spin correlations at four different redshifts: 
809: $z=0,0.5,1$ and $2$ (top-left, top-right, bottom-left, and bottom-right, 
810: respectively). In each panel, the solid dots represent the numerical 
811: results, while the solid and dashed lines correspond to the nonlinear 
812: (eq.[\ref{eqn:1st}]) and linear (eq.[\ref{eqn:1st}]) analytic model, 
813: respectively.The errors are calculated as the standard deviation for
814: the case of no correlation. The dotted line represents the case of
815: no correlation.}
816: \label{fig:z}
817:  \end{center}
818: \end{figure}
819: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
820: \clearpage
821: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
822:  \begin{figure}
823:   \begin{center}
824:    \plotone{f2.eps}
825: \caption{$68\%$, $95\%$ and $99\%$ contour plots for the linear and the 
826: nonlinear correlation parameters, $a_{\rm l}$ and $\varepsilon_{\rm nl}$ 
827: at $z=0,0.5,1$ and $2$ (top-left,top-right,bottom-left, and bottom-right,
828: respectively). In each panel, the cross mark represents the best-fit value.}
829: \label{fig:cz}
830:  \end{center}
831: \end{figure}
832: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
833: \clearpage
834: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
835:  \begin{figure}
836:   \begin{center}
837:    \plotone{f3.eps}
838: \caption{Intrinsic halo spin-spin correlations at $z=0$ for the four 
839: different logarithmic bins of the specific angular momentum magnitude 
840: $l$ at $z=0$.}
841: \label{fig:a}
842:  \end{center}
843: \end{figure}
844: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
845: \clearpage
846: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
847:  \begin{figure}
848:   \begin{center}
849:    \plotone{f4.eps}
850: \caption{$68\%$, $95\%$ and $99\%$ contour plots for $a_{\rm l}$ and 
851: $\varepsilon_{\rm nl}$at the four logarithmic bins of $l$.}
852: \label{fig:ca}
853:  \end{center}
854: \end{figure}
855: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
856: \clearpage
857: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
858:  \begin{figure}
859:   \begin{center}
860:    \plotone{f5.eps}
861: \caption{The dark halo spin-spin correlations at $z=0$ for the four 
862: different bins of the halo mass $M_{*}$ in unit of $10^{10}h^{-1}M_{\odot}$.}
863: \label{fig:m}
864:  \end{center}
865: \end{figure}
866: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
867: \clearpage
868: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
869:  \begin{figure}
870:   \begin{center}
871:    \plotone{f6.eps}
872: \caption{$68\%$, $95\%$ and $99\%$ contour plots for  $a_{\rm l}$ and 
873: $\varepsilon_{\rm nl}$ at the four bins of halo mass $M_{*}$.}
874: \label{fig:cm}
875:  \end{center}
876: \end{figure}
877: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
878: \clearpage
879: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
880:  \begin{figure}
881:   \begin{center}
882:    \plotone{f7.eps}
883: \caption{Intrinsic halo spin cross correlations between different mass scales 
884: at $z=0$. In each panel, $\eta_{Cij}$ represents the cross-correlation 
885: between the $i$-th and the $j$-th mass bin, given in Table \ref{tab:m}. }
886: \label{fig:b}
887:  \end{center}
888: \end{figure}
889: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
890: \clearpage
891: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
892:  \begin{figure}
893:   \begin{center}
894:    \plotone{f8.eps}
895: \caption{Comparison of the spin correlation $\eta_{ii}(r)$ between halos 
896: in the same mass bins with the spin cross-correlation $\eta_{Cij}(r)$ 
897: between halos in different mass bins. }
898: \label{fig:tb}
899:  \end{center}
900: \end{figure}
901: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
902: \clearpage
903: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
904:  \begin{figure}
905:   \begin{center}
906:    \plotone{f9.eps}
907: \caption{$68\%$, $95\%$ and $99\%$ contour plots for $a_{\rm l}$ and 
908: $\varepsilon_{\rm nl}$ for the spin cross-correlations between halos 
909: in the $1$st mass bin and the $2nd$ mass bin (top) and between halos 
910: in the $2$nd mass bin and the $3rd$ mass bin (bottom). }
911: \label{fig:cb}
912:  \end{center}
913: \end{figure}
914: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
915: \clearpage
916: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
917:  \begin{figure}
918:   \begin{center}
919:    \plotone{f10.eps}
920: \caption{Intrinsic halo spin correlations at the four logarithmic bins of 
921: the velocity magnitude $v$ in unit of km/s.}
922: \label{fig:v}
923:  \end{center}
924: \end{figure}
925: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
926: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
927:  \begin{figure}
928:   \begin{center}
929:    \plotone{f11.eps}
930: \caption{$68\%$, $95\%$ and $99\%$ contour plots for $a_{\rm l}$ and 
931: $\varepsilon_{\rm nl}$ at the four logarithmic bins of $v$.}
932: \label{fig:cv}
933:  \end{center}
934: \end{figure}
935: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
936: 
937: \end{document}
938: