0811.1339/ms.tex
1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass[apjl]{emulateapj}
3: \voffset=-0.6in
4: \font\bften=cmbx10 scaled 1000 \font\tenrm=cmr10 scaled 1000
5: \font\eightrm=cmr8 scaled 1000 \font\sevenrm=cmr7 scaled 1000
6: %\slugcomment{ApJL, ***, 2008 ***, Accepted 2008 ***} 
7: \shorttitle{Abundance of Void Groups}
8: \shortauthors{Song \& Lee}
9: \begin{document}
10: \title{The Mass Function of Void Groups as a Probe of the Primordial 
11: non-Gaussianity}
12: \author{Hyunmi Song and Jounghun Lee}
13: \affil{Department of Physics and Astronomy, FPRD, Seoul National University, 
14: Seoul 151-747, Korea} 
15: \email{yesuane@gmail.com}
16: \email{jounghun@astro.snu.ac.kr}
17: \begin{abstract}
18: The primordial non-Gaussianity signal, if measured accurately, will allow us
19: to distinguish between different candidate models for the cosmic inflation.
20: Since the galaxy groups located in void regions are rare events, their 
21: abundance may be a sensitive probe of the primordial non-Gaussianity.
22: We construct an analytic model for the mass function of void groups in the 
23: framework of the extended Press-Schechter theory with non-Gaussian initial 
24: conditions and investigate how it depends on the primordial non-Gaussianity 
25: parameter. A feasibility study is conducted by fitting the analytic mass 
26: function of void groups to the observational results from the galaxy 
27: group catalog of the Sloan Digital Sky Survey Data Release 4 with adjusting 
28: the primordial non-Gaussianity parameter.
29: \end{abstract}
30: \keywords{cosmology:theory --- large-scale structure of universe}
31: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
32: 
33: \section{INTRODUCTION}
34: 
35: In precision cosmology, a fundamental task is to distinguish between various 
36: candidate models for the cosmic inflation. One way to perform this task is to 
37: measure the degree of non-Gaussianity in the primordial density field. 
38: Although the primordial density field is regarded as nearly Gaussian in all 
39: inflationary scenarios \citep{GP82}, the degree of its non-Gaussianity 
40: differs between the models. For instance, in the single-field slow-roll 
41: inflationary model, the deviation from Gaussianity is small enough 
42: to be unobservable \citep{ver-etal01}. Meanwhile in some multi-field 
43: inflationary models the non-Gaussianity can be generated to a detectable 
44: level \citep[][and references therein]{BE07}. For a thorough 
45: review on the predictions of various inflationary models for the 
46: primordial non-Gaussianity, see \citet{bar-etal04}
47: 
48: The cosmic microwave background radiation (CMB) is a useful probe of the 
49: primordial non-Gaussianity as it reflects the linear density perturbations. 
50: Yet, if the primordial non-Gaussianity is scale-dependent, the constraint 
51: from the CMB analysis is rather restricted to the very large-scale 
52: \citep{ver-etal01,KS01}.  
53: The large-scale structure is a powerful alternative probe of the primordial 
54: non-Gaussianity on the sub-CMB scale. It has been well known that the 
55: abundance of high-$z$ clusters are rare enough to constrain the primordial 
56: non-Gaussianity \citep{LM88,chi-etal98,mat-etal00,WK03,car-etal08,lov-etal08,
57: gro-etal09}. This probe, however, is likely to suffer from large systematics 
58: involved in the inaccurate measurement of the masses of high-$z$ clusters. 
59: The abundance of cosmic voids is another probe of the primordial 
60: non-Gaussianity based on the large-scale structure \citep{kam-etal09}. 
61: One difficulty in using this probe lies in the fact that there is no unique 
62: way to define voids \citep{car-etal08}. Recently, it has been claimed that 
63: using the clustering properties of highly biased large-scale structures 
64: the primordial non-Gaussianity parameter can be measured with accuracy 
65: as high as the one obtained from the CMB analysis 
66: \citep{car-etal08,slo-etal08,jeo-kom09}. 
67: 
68: Here, we propose the mass function of present galaxy groups embedded in 
69: void regions as a new probe of the primordial non-Gaussianity. It is not only 
70: the high-$z$ clusters but also the low-$z$ void groups that are so rare that 
71: their abundance may depend sensitively on the initial conditions. 
72: Furthermore, the mass estimation of low-$z$ galaxy groups should be much 
73: more reliable than that of high-$z$ clusters \citep{yan-etal07}. 
74: Throughout this Letter, we assume a WMAP 5 cosmology \citep{wmap5}.
75: 
76: \section{AN ANALYTIC MODEL}
77: 
78: The Press-Schechter theory \citep[][PS hereafter]{PS74} provides an analytic 
79: framework within which the number density of bound objects as a function of 
80: mass, $dN_{\rm PS}/dM$, can be obtained:
81: \begin{equation}
82: \label{eqn:ps}
83: \frac{dN}{dM} = A\frac{\bar{\rho}}{M}\frac{d}{dM}\left\vert
84: \int_{\delta_c(z)}^{\infty}p(\delta_{M})d\delta\right\vert,
85: \end{equation} 
86: where $\bar{\rho}$ is the mean background density, $\delta_{c}(z)$ is the 
87: critical density contrast for gravitational collapse at redshift $z$, 
88: $p(\delta_{M})$ is the probability density distribution of the density field 
89: $\delta_{M}$ smoothed on the mass scale of $M$, and $A$ is the normalization 
90: constant. Basically, equation (\ref{eqn:ps}) states that the number density 
91: of bound objects can be inferred from the differential volume fraction 
92: occupied by those regions in the linear density field whose average density 
93: contrast $\delta_{M}$ reaches a certain threshold, $\delta_{c}$.
94: In the original PS theory, the initial density field is assumed to be 
95: Gaussian as $p(\delta_{M})=\exp\left[-\delta^{2}_{M}/(2\sigma^{2}_{M})\right]
96: /(\sqrt{2\pi}\sigma_{M})$. 
97: 
98: Here $\sigma_{M}$ is the rms density fluctuation 
99: smoothed on the mass scale $M$ and $\delta_{c}\equiv\delta_{c0}/D_{+}(z)$ 
100: where $\delta_{c0}$ is the critical density contrast at $z=0$ and $D_{+}(z)$ 
101: is the linear growth factor. For a WMAP 5 cosmology, we find 
102: $\delta_{c0}\approx 1.62$. Note that the mass function of bound objects 
103: depends on the initial conditions through its dependence on $\sigma_{M}$ 
104: which is a function of the density parameter, $\Omega_{m}$ and the 
105: amplitude of the linear power spectrum, $\sigma_{8}$. 
106: 
107: Now, let us consider the case of non-Gaussian initial conditions that is 
108: often characterized by the primordial non-Gaussianity parameter $f_{\rm NL}$ 
109: as  $\psi({\bf x})=\phi({\bf x}) + f_{NL}\left(\phi^{2}({\bf x})-
110: \langle\phi^{2}({\bf x})\rangle\right)$ where $\phi$ is a Gaussian random 
111: field and $\psi({\bf x})$ is the linearly extrapolated gravitational 
112: potential at $z=0$ \citep[e.g.,][]{lov-etal08,gro-etal09}. 
113: The functional form of $p(\delta_{M})$ for the 
114: non-Gaussian case will be in general different from the Gaussian case. 
115: However, provided that the degree of the non-Gaussianity is very small and 
116: scale independent, $p(\delta_{M})$ has the same functional form as the 
117: Gaussian case at first order \citep{LM88,mat-etal00,ver-etal01}. 
118: The only difference is the value of the critical density contrast, 
119: $\delta_{c*}$, which is related to that of the Gaussian case, $\delta_{c}$ as 
120: \begin{equation}
121: \label{eqn:nldc}
122: \delta_{c*}(z)=\delta_{c}(z)\left[1-\frac{S_3}{3}\delta_{c}(z)\right]^{1/2}.
123: \end{equation}
124: Here $S_3$ is a skewness parameter, related to the primordial non-Gaussianity 
125: parameter $f_{\rm NL}$ as $S_{3}=\frac{3}{5}f_{\rm NL}\mu_{3}^{(1)}/
126: (\mu_{2}^{(1)})^2$ where $\mu_{2}^{(1)}$ and $\mu_{3}^{(1)}$ denote the 
127: variance and skewness of the smoothed non-Gaussian density field at first 
128: order, respectively  \citep[see eqs.43-45 in][]{mat-etal00}. 
129: Therefore, the PS mass function with non-Gaussian initial conditions is a 
130: function of $M$ and $f_{\rm NL}$: $dN_{\rm PS}(f_{\rm NL},M)/dM$. The 
131: case of $f_{\rm NL}=0$ corresponds to the original one with Gaussian initial 
132: conditions.
133: 
134: On the group scale, $dN_{\rm PS}(f_{\rm NL},M)/dM$ does not change sensitively 
135: with the initial conditions since the galaxy groups are not rare events. 
136: However, those groups located in voids should be so rare that their 
137: abundance may depend sensitively on the initial conditions. Before deriving 
138: the abundance of void groups, we clarify the meaning of a {\it void region}. 
139: Following \citet{hah-etal07}, we define a void region on mass scale $M$ as a 
140: region where the three eigenvalues of the tidal tensor at a given region on 
141: mass scale $M$ are less than zero. Then, we replace $p(\delta_{M})$ in 
142: eq.~(\ref{eqn:ps}) by $p(\delta_{M}|\lambda^{\prime}_{1M^{\prime}}<0)$ which 
143: represents the conditional probability density distribution that the density 
144: contrast has a certain value on the mass scale $M$ provided that the largest 
145: eigenvalue of the tidal tensor is negative on some larger mass scale 
146: $M^{\prime}>M$. From here on, $\delta$ denotes the density contrast on the 
147: mass scale $M$, while $\lambda^{\prime}_{1},\ \lambda^{\prime}_{2},\ 
148: \lambda^{\prime}_{3}$ are the three eigenvalues of the tidal field 
149: on some larger mass scale $M^{\prime}$.
150: 
151: The joint probability distribution 
152: $p(\delta,\lambda^{\prime}_{1},\lambda^{\prime}_{2},\lambda^{\prime}_{3})$ 
153: for the case of Gaussian initial conditions has been already derived by 
154: \cite{lee06} as  
155: \begin{eqnarray}
156: \label{eqn:ess}
157: p(\delta,\lambda^{\prime}_{1},\lambda^{\prime}_{2},
158: \lambda^{\prime}_{3}) = 
159: \frac{1}{\sqrt{2\pi}\sigma_{\Delta}}\frac{3375}{8\sqrt{5}\pi\sigma^{\prime6}}
160: \exp\left[-\frac{(\delta-I^{\prime}_{1})^{2}}
161: {2\sigma^{2}_{\Delta}}\right]\times \nonumber \\
162: \exp\left(-\frac{3I^{\prime 2}_{1}}{\sigma^{\prime 2}} + 
163: \frac{15I^{\prime}_{2}}{2\sigma^{\prime 2}}\right)
164: (\lambda^{\prime}_{1}-\lambda^{\prime}_{2})
165: (\lambda^{\prime}_{2}-\lambda^{\prime}_{3})
166: (\lambda^{\prime}_{1}-\lambda^{\prime}_{3}), 
167: \end{eqnarray}
168: with $\sigma^{2}_{\Delta} \equiv \sigma^{2} - \sigma^{\prime 2}$, 
169: $I^{\prime}_{1}=\lambda^{\prime}_{1} + \lambda^{\prime}_{2} + 
170: \lambda^{\prime}_{3}$, and $I^{\prime}_{1}=\lambda^{\prime}_{1}
171: \lambda^{\prime}_{2}+\lambda^{\prime}_{2}\lambda^{\prime}_{3}+ 
172: \lambda^{\prime}_{1}\lambda^{\prime}_{3}$. Here $\sigma$ and 
173: $\sigma^{\prime}$ represents the rms density fluctuations on the mass 
174: scale $M$ and $M^{\prime}$, respectively. 
175: The conditional probability density, $p(\delta| \lambda^{\prime}_{1}<0)$, is 
176: now written as 
177: \begin{equation}
178: \label{eqn:void}
179: p(\delta| \lambda^{\prime}_{1}<0) = 
180: \frac{\int^{0}_{-\infty}d\lambda_{1}
181: \int^{\lambda^{\prime}_{1}}_{-\infty}d\lambda_{2}
182: \int^{\lambda^{\prime}_{2}}_{-\infty}d\lambda_{3}
183: p(\delta,\lambda^{\prime}_{1},\lambda^{\prime}_{2},\lambda^{\prime}_{3})}
184: {\int^{0}_{-\infty}d\lambda_{1}
185: \int^{\lambda^{\prime}_{1}}_{-\infty}d\lambda_{2}
186: \int^{\lambda^{\prime}_{2}}_{-\infty}d\lambda_{3}
187: p(\lambda^{\prime}_{1},\lambda^{\prime}_{2},\lambda^{\prime}_{3})}.
188: \end{equation} 
189: Putting $p(\delta|\lambda^{\prime}_{1}<0)$ in eq.~(\ref{eqn:ps}), we evaluate 
190: the mass function of void groups with Gaussian initial conditions, 
191: $dN_{\rm V}^{\rm G}/dM$.
192: \begin{figure} 
193: \plotone{f1.eps}
194: \caption{Mass function of void groups (solid line) and of all 
195: groups (dashed line) in the extended Press-Schechter formalism.}
196: \label{fig:analytic}
197: \end{figure}
198: 
199: Using the methodology suggested by \citet{lov-etal08}, we model the mass 
200: function of void groups with non-Gaussian initial conditions, 
201: $dN^{\rm NG}_{\rm V}/dM$, as
202: \begin{equation}
203: \label{eqn:ratio}
204: \frac{dN^{\rm NG}_{\rm V}}{dM} =
205: \frac{dN^{\rm G}_{\rm V}}{dM}\frac{dN_{\rm PS}(f_{\rm NL},M)/dM}
206: {dN_{\rm PS}(f_{\rm NL}=0,M)/dM}, 
207: \end{equation}
208: where $dN_{\rm PS}(f_{\rm NL}=0,M)/dM$ and $dN_{\rm PS}(f_{\rm NL},M)/dM$ 
209: represents the PS mass function  with Gaussian and non-Gaussian initial 
210: conditions, respectively (see eq.[4.20] in Loverde et al. 2009). 
211: Figure \ref{fig:analytic} plots the mass function of void groups (solid line), 
212: which normalized as $\int dN_{V}/d\log M_{i}=N_{vg}$ where $N_{vg}$ is the 
213: total number of void groups found in the observational data (see \S 3.)
214: The mass function of all groups (dashed line) are also plotted for comparison. 
215: As can be seen, the mass function of void groups decreases very rapidly with 
216: mass, which indicates that it must depend sensitively on the initial 
217: conditions. 
218: 
219: It has to be mentioned here that equation (\ref{eqn:ratio}) has yet to be 
220: validated against numerical results. The methodology of \citet{lov-etal08} 
221: that equation (\ref{eqn:ratio}) is based on has recently been tested against 
222: N-body simulations and found to be valid in the high-mass sections 
223: \citep{gro-etal09}. Yet, to fully justify the use of equation 
224: (\ref{eqn:ratio}) for the evaluation of the abundance of void groups with 
225: non-Gaussian initial conditions, it will be required to test equation 
226: (\ref{eqn:ratio}) numerically on the group-mass scale. 
227: Furthermore, for the case of scale-dependent non-Gaussianity the mass 
228: function of void groups would have much more complicated even at first order 
229: \citep{lov-etal08}. The focus of this work, however, is on the proof of a 
230: concept that the abundance of present galaxy groups in voids can be used 
231: as a probe of the primordial non-Gaussianity. Henceforth, here we use 
232: equation (\ref{eqn:ratio}) as an ansatz and consider only the 
233: scale-independent case for simplicity.
234: \begin{figure} 
235: \plotone{f2.eps}
236: \caption{Mass function of void groups with the best-fit value of 
237: $f_{\rm NL}$ (solid line) determined by fitting to the observational data 
238: from the SDSS DR4 (dots). The errors include both the Poisson noise and the 
239: cosmic variance. The mass function of void groups for the Gaussian 
240: case ($f_{\rm NL}=0$) is also plotted for comparison (dashed).}
241: \label{fig:void}
242: \end{figure}
243: 
244: \section{A FEASIBILITY STUDY}
245: 
246: We conduct a feasibility study by comparing the analytic mass function of 
247: void groups obtained in \S 2 with the observational result from the 
248: galaxy group catalog of the Sloan Digital Sky Survey Data Release 4 
249: provided by \citep{yan-etal07} who measured the masses of the SDSS galaxy 
250: groups were estimated from the characteristic luminosity (or stellar masses) 
251: using the WMAP 3 cosmology \citep{wmap3}.
252: To identify void groups from the SDSS group catalog, we use the real-space 
253: tidal field reconstructed by \citet{LE07} on $64^{3}$ pixels in a box of 
254: linear length $400h^{-1}$Mpc from the Two Mass Redshift survey (2MRS) 
255: \citep{huc-etal05}. We smooth the 2MRS tidal field with a Gaussian filter of 
256: scale radius $8h^{-1}$Mpc on which scale the density field is still in the 
257: quasi-linear regime, calculate the three eigenvalues of the smoothed tidal 
258: field at each pixel and mark as voids those pixels in which all three 
259: eigenvalues are less than zero. 
260: A total of $550$ galaxy groups at redshifts $0.01\le z<0.04$ in mass range of 
261: $11.7\le\log M/(h^{-1}M_{\odot})\le 13.4$ are found in the void pixels. 
262: Binning their mass range in the logarithmic scale, we measure their abundance,
263: $dN_{V}/d\log M$, which is renormalized to be $\int dN_{V}=1$. 
264: To account for the cosmic variance as well as the Poisson noise in the 
265: measurement of $dN_{V}/d\log M$, we divide the selected void groups into 
266: $6$ subsamples and measure  $dN_{V}/d\log M$ for each subsample separately. 
267: We calculate the jackknife errors as the standard deviation in the measurement 
268: of the mean $dN_{V}/d\log M$ averaged over the $6$ subsamples at each mass bin.
269: 
270: We fit the observational result to the analytic model by adjusting the 
271: value of $f_{\rm NL}$. To account for the correlations between the mass bins, 
272: we employ the {\it generalized} $\chi^{2}$-statistics to determine the 
273: best-fit value of $f_{\rm NL}$: $\chi^{2} = [n_{i}-n(\log M_{i};f_{\rm NL})]
274: C^{-1}_{ij}[n_{i}-n(\log M_{i};f_{\rm NL})]$ where $n_{i}\equiv dN/d\log M_{i}$
275: and $n(\log M_{i};f_{\rm NL})$ represents the observational and analytical 
276: results evaluated at the $i$-th logarithmic mass bin, $\log M_{i}$, 
277: respectively. And $(C_{ij})$ is the covariance matrix defined as 
278: $C_{ij}=\langle(n_{i}-n_{0i})(n_{j}-n_{0j})\rangle$ where $n_{0i}$ 
279: represents the mean of $n_{i}$ averaged over all samples. 
280: Finally, the uncertainty in the measurement of $f_{\rm NL}$ is calculated as 
281: the curvature of the $\chi^{2}$ function at the minimum. Through this fitting 
282: procedure, we find $f_{\rm NL}=36\pm 1$. Figure \ref{fig:void} plots the 
283: observational results (dots) and the analytic model with the best-fit value 
284: of $f_{\rm NL}=36$ (solid line). The median redshift of the SDSS void 
285: groups, $z=0.03$, is used for the value of $z$ in the analytic model.  
286: For comparison, the analytic model with $f_{\rm NL}=0$ is also plotted 
287: (dotted line). As can be seen, the observational results agree better with 
288: the analytic model with non-Gaussian initial conditions. 
289: 
290: It is, however, expected that there is a degeneracy between $f_{\rm NL}$ and 
291: the other key cosmological parameters on which the mass function of void 
292: groups depend. Here, we investigate the degeneracy between $\Omega_{m}$ 
293: and $f_{\rm NL}$, setting $\sigma_{8}$ at the value determined by 
294: WMAP 5 cosmology. Varying the values of $\Omega_{m}$ and $f_{\rm NL}$, 
295: we recalculate $dN_{V}/d\log M$ at a typical group mass scale of 
296: $M=10^{13}h^{-1}M_{\odot}$.  Figure \ref{fig:con} plots a family of the 
297: degeneracy curves in the $\Omega_m$-$f_{\rm NL}$ plane with the value 
298: of $\sigma_{8}$ set at $0.76$ (solid) and $0.83$ (dashed). As can be seen, 
299: a strong degeneracy exists between the two parameters. For a fixed value 
300: of $dN_{V}/d\log M$, the value of $f_{\rm NL}$ increases as the value of 
301: $\Omega_{m}$ decreases. The comparison between the solid and the dashed lines 
302: indicate that the degree of the degeneracy between $\Omega_{m}$ and 
303: $f_{\rm NL}$ increases as the value of $\sigma_{8}$ increases. 
304: It also indicates that if $\Omega_{m}$ is fixed, the value of 
305: $f_{\rm NL}$ increases as the value of $\sigma_{8}$ decreases.
306: To break this parameter degeneracy, it will be necessary to combine 
307: our analysis with other analyses.
308: \begin{figure} 
309: \plotone{f3.eps}
310: \caption{Degeneracy curves in the $f_{\rm NL}$-$\Omega_{\rm m}$ plane with 
311: the value of $\sigma_{8}$ set at $0.76$ (solid)and at $0.83$ (dashed).}
312: \label{fig:con}
313: \end{figure}
314: 
315: The preliminary results of this feasibility study, however, are subject to 
316: several caveats. First, the analytic model assumes the scale-independent 
317: Gaussianity. To be more realistic, it is necessary to account for the 
318: scale-dependent non-Gaussianity. The second caveat lies in the limitation of 
319: the Press-Schechter formalism. As shown by several authors 
320: \citep{LS98,ST99,jen-etal01}, the real gravitational process deviates from 
321: the spherical dynamics on which the PS mass function is based. It has to 
322: be tested how significantly the deviation of gravitational collapse process 
323: from the spherical dynamics affects the abundance of void groups.  
324: The third caveat comes from the fact that the validity of equation 
325: (\ref{eqn:ratio}) has yet to be confirmed.  Although \citet{gro-etal09} have 
326: shown that this methodology suggested by \citet{lov-etal08} to model 
327: departures from non-Gaussianity leads to an excellent approximation on the 
328: cluster scale, it has to be confirmed by N-body simulations whether the 
329: same methodology can be used to count the number of void groups with 
330: non-Gaussian initial conditions. Fourth, the different mass-to-light ratios 
331: of the void galaxies from that of the wall galaxies has to be taken into 
332: account. According to \citet{roj-etal05}, the specific star formation rate 
333: in void galaxies is higher than that in wall galaxies, suggesting that the 
334: mass of the void groups in the SDSS Galaxy group catalog are likely to be 
335: overestimated. Fifth, \citet{yan-etal07} measured the masses of SDSS galaxy 
336: groups assuming the old WMAP 3 cosmology \citep{wmap3}. It will be necessary 
337: to use the values of the most updated WMAP 5 parameters for the more accurate 
338: calculation of the masses of void groups.
339: 
340: As a final conclusion, we have proved that the abundance of void groups 
341: can in principle be a useful probe of the primordial non-Gaussianity 
342: parameter. For a robust probe, however, it will be required to refine 
343: further the analytic model of the abundance of void groups and 
344: to improve the mass estimation of galaxy groups in void regions, 
345: which is the direction of our future work.
346: 
347: \acknowledgments
348: 
349: We thank an anonymous referee for many helpful suggestions.
350: This work is financially supported by the Korea Science and Engineering 
351: Foundation (KOSEF) grant funded by the Korean Government 
352: (MOST, NO. R01-2007-000-10246-0).
353: 
354: Funding for  the SDSS and SDSS-II  has been provided by  the Alfred P.
355: Sloan Foundation, the Participating Institutions, the National Science
356: Foundation, the  U.S.  Department of Energy,  the National Aeronautics
357: and Space Administration, the  Japanese Monbukagakusho, the Max Planck
358: Society, and  the Higher Education  Funding Council for  England.  The
359: SDSS Web  Site is  http://www.sdss.org/.  The SDSS  is managed  by the
360: Astrophysical    Research    Consortium    for    the    Participating
361: Institutions. The  Participating Institutions are  the American Museum
362: of  Natural History,  Astrophysical Institute  Potsdam,  University of
363: Basel,   Cambridge  University,   Case  Western   Reserve  University,
364: University of Chicago, Drexel  University, Fermilab, the Institute for
365: Advanced   Study,  the  Japan   Participation  Group,   Johns  Hopkins
366: University, the  Joint Institute  for Nuclear Astrophysics,  the Kavli
367: Institute  for   Particle  Astrophysics  and   Cosmology,  the  Korean
368: Scientist Group, the Chinese  Academy of Sciences (LAMOST), Los Alamos
369: National  Laboratory, the  Max-Planck-Institute for  Astronomy (MPIA),
370: the  Max-Planck-Institute  for Astrophysics  (MPA),  New Mexico  State
371: University,   Ohio  State   University,   University  of   Pittsburgh,
372: University  of  Portsmouth, Princeton  University,  the United  States
373: Naval Observatory, and the University of Washington.
374: 
375: \begin{thebibliography}{}
376: \bibitem[Battefeld \& Easther(2007)]{BE07} 
377: Battefeld, T., \& Easther, R.\ 2007, Journal of Cosmology and Astro-Particle 
378: Physics, 3, 20 
379: \bibitem[Bartolo et al.(2004)]{bar-etal04} 
380: Bartolo, N., Komatsu, E., Matarrese, S., \& Riotto, A.\ 2004, 
381: \physrep, 402, 103 
382: \bibitem[Chiu et al.(2998)]{chi-etal98} 
383: Chiu, W.~A., Ostriker, J. P., \& Strauss, M.~A.\ 1998, \apj, 494, 479 
384: \bibitem[Carbone et al.(2008)]{car-etal08} 
385: Carbone, C., Verde, L., \& Matarrese, S.\ 2008, \apjl, 684, L1 
386: \bibitem[Colberg et al.(2008)]{col-etal08} 
387: Colberg, J.~M., et al.\ 2008, \mnras, 387, 933 
388: \bibitem[Dunkley et al.(2009)]{wmap5} 
389: Dunkley, J., et al.\ 2009, \apjs, 180, 306 
390: \bibitem[Eke et al.(1996)]{eke-etal96} 
391: Eke, V.~R., Cole, S., \& Frenk, C.~S.\ 1996, \mnras, 282, 263 
392: \bibitem[Erdo{\u g}du et al.(2006)]{erd-etal06} Erdo{\u g}du, P., et al.
393: \ 2006, \mnras, 373, 45 
394: \bibitem[Grossi et al.(2009)]{gro-etal09} 
395: Grossi, M., Verde, L., Carbone, C., Dolag, K., Branchini, E., Iannuzzi, F., 
396: Matarrese, S., \& Moscardini, L.\ 2009, arXiv:0902.2013 
397: \bibitem[Guth \& Pi(1982)]{GP82} 
398: Guth, A.~H., \& Pi, S.-Y.\ 1982, \prl, 49, 1110 
399: \bibitem[Hahn et al.(2007)]{hah-etal07} 
400: Hahn, O., Carollo, C.~M., Porciani, C., \& Dekel, A.\ 2007, \mnras, 381, 41 
401: \bibitem[Huchra et al.(2005)]{huc-etal05} Huchra, J., et al.\ 
402: 2005, Maps of the Cosmos, 216, 170 
403: \bibitem[Jenkins et al.(2001)]{jen-etal01}
404: Jenkins, A., Frenk, C.~S., White, S.~D.~M., Colberg, J.~M., Cole, S., 
405: Evrard, A.~E., Couchman, H.~M.~P., \& Yoshida, N.\ 2001, \mnras, 321, 372
406: \bibitem[Jeong \& Komatsu(2009)]{jeo-kom09} 
407: Jeong, D., \& Komatsu, E.\ 2009, arXiv:0904.0497 
408: \bibitem[Kamionkowski et al.(2009)]{kam-etal09} 
409: Kamionkowski, M., Verde, L., \& Jimenez, R.\ 2009, Journal of Cosmology 
410: and Astro-Particle Physics, 1, 10 
411: \bibitem[Komatsu \& Spergel(2001)]{KS01} 
412: Komatsu, E., \& Spergel, D.~N.\ 2001, \prd, 63, 063002 
413: \bibitem[Lee(2006)]{lee06}
414: Lee, J. 2006, preprint (arXiv:0605697v1)
415: \bibitem[Lee \& Erdogdu(2007)]{LE07} 
416: Lee, J., \& Erdogdu, P.\ 2007, \apj, 671, 1248 
417: \bibitem[Lee \& Shandarin(1998)]{LS98}
418: Lee, J. \& Shandarin, S.~F.\ 1998, ApJ, 500, 14
419: \bibitem[LoVerde et al.(2008)]{lov-etal08} 
420: Lo Verde, M., Miller, A., Shandera, S., \& Verde, L.\ 2008, 
421: Journal of Cosmology and Astro-Particle Physics, 4, 14 
422: \bibitem[Lucchin \& Matarrese(1988)]{LM88} 
423: Lucchin, F., \& Matarrese, S.\ 1988, \apj, 330, 535 
424: \bibitem[Matarrese et al.(2000)]{mat-etal00} 
425: Matarrese, S., Verde, L., \& Jimenez, R.\ 2000, \apj, 541, 10 
426: \bibitem[Press \& Schechter(1974)]{PS74} 
427: Press, W.~H., \& Schechter, P.\ 1974, \apj, 187, 425 
428: \bibitem[Rojas et al.(2005)]{roj-etal05}
429: Rojas, R.~R., Vogeley, M.~S., Hoyle, F., \& Brinkmann, J.\ 2005, 624, 571
430: \bibitem[Sheth \& Tormen(1999)]{ST99}
431: Sheth, R.~K., \& Tormen, G.\ 1999, \mnras, 308, 119 
432: \bibitem[Slosar et al.(2008)]{slo-etal08} 
433: Slosar, A., Hirata, C., Seljak, U., Ho, S., \& Padmanabhan, N.\ 2008, 
434: Journal of Cosmology and Astro-Particle Physics, 8, 31 
435: \bibitem[Spergel et al.(2007)]{wmap3}
436: Spergel, D.~N., et al.\ 2007, \apjs, 170, 377 
437: \bibitem[Verde et al.(2000)]{ver-etal00} 
438: Verde, L., Wang, L., Heavens, A.~F., \& Kamionkowski, M.\ 2000, \mnras, 
439: 313, 141 
440: \bibitem[Verde et al.(2001)]{ver-etal01} 
441: Verde, L., Jimenez, R., Kamionkowski, M., \& Matarrese, S.\ 2001, 
442: \mnras, 325, 412 
443: \bibitem[Weinberg \& Kamionkowski(2003)]{WK03} 
444: Weinberg, N.~N., \& Kamionkowski, M.\ 2003, \mnras, 341, 251 
445: \bibitem[Yang et al.(2007)]{yan-etal07} 
446: Yang, X., Mo, H.~J., van den Bosch, F.~C., Pasquali, A., Li, C., 
447: \& Barden, M.\ 2007, \apj, 671, 153 
448: 
449: \end{thebibliography}
450: \end{document}
451: