1: \documentclass[12pt,preprint]{aastex}
2:
3: \newcommand{\1}{\'\i}
4: \newcommand{\hmpc}{$h^{-1}$\,Mpc}
5: \newcommand{\be}{\begin{equation}}
6: \newcommand{\e}{\end{equation}}
7: \newcommand{\f}{\frac}
8:
9: \shorttitle{KSZ anisotropies generated by gas in cosmic structures}
10:
11: %\received{/\the\month/\the\day/\the\year}
12: \begin{document}
13: \title{Kinematic Sunyaev-Zeldovich Cosmic Microwave Background Temperature
14: Anisotropies Generated by Gas in Cosmic Structures.}
15:
16: \author{F. Atrio-Barandela\altaffilmark{1},
17: J.P. M\"ucket\altaffilmark{2}, R. G\'enova-Santos\altaffilmark{3}}
18: \altaffiltext{1}{F\'{\i}sica Te\'orica, Universidad de Salamanca, 37008 Salamanca, Spain;
19: email: atrio@usal.es}
20: \altaffiltext{2}{Astrophysikalisches Institut Potsdam. D-14482 Potsdam, Germany;
21: email: jpmuecket@aip.de}
22: \altaffiltext{3}{Cavendish Laboratory. University of Cambridge. Cambridge CB3OHE, U.K;
23: email: rgs@mrao.cam.ac.uk}
24:
25: \begin{abstract}
26: If the gas in filaments and halos shares the same velocity field
27: than the luminous matter, it will generate measurable temperature anisotropies
28: due to the Kinematic Sunyaev-Zeldovich effect. We compute the distribution
29: function of the KSZ signal produced by a typical filament and show it is highly
30: non-gaussian. The combined contribution of the Thermal and
31: Kinematic SZ effects of a filament of size $L\simeq 5$Mpc and electron
32: density $n_e\simeq 10^3m^{-3}$ could explain the cold spots of $\delta\sim -200\mu$K
33: on scales of $30'$ found in the Corona Borealis Supercluster by the VSA experiment.
34: PLANCK, with its large resolution and frequency coverage,
35: could provide the first evidence of the existence of filaments
36: in this region. The KSZ contribution of the network of filaments and halo
37: structures to the radiation
38: power spectrum peaks around $l\sim 400$, a scale very different from that of
39: clusters of galaxies, with a maximum amplitude
40: $l(l+1)C_l/2\pi\sim 10-25 (\mu K)^2$, depending on model parameters, i.e.,
41: $\sigma_8$ and the Jeans length. About 80\% of the signal comes from filaments
42: with redshift $z\le 0.1$. Adding this component to the intrinsic
43: Cosmic Microwave Background temperature anisotropies of the concordance model
44: improves the fit to WMAP 3yr data by $\Delta\chi^2\simeq 1$. The improvement is
45: not statistically significant but a more systematic study could demonstrate
46: that gas could significantly contribute to the anisotropies measured by WMAP.
47: \end{abstract}
48:
49: \keywords{Cosmic Microwave Background. Cosmology: theory. Cosmology: observations.}
50:
51: \section{Introduction}
52: At low redshift, about 50 \% of all baryons have not been identified
53: (Fukugita \& Peebles, 2004). The highly ionized intergalactic gas, that has evolved
54: from the initial density perturbations into a complex
55: network of mildly non-linear structures in the redshift interval $0<z<6$
56: (Rauch 1998; Stocke, Shull \& Penton 2004), could contain most of the
57: baryons in the Universe (Rauch et al. 1997, Schaye 2001, Richter et al 2005).
58: With cosmic evolution, the fraction
59: of baryons in these structures decreases as more matter is concentrated within
60: compact virialized objects. The Ly$\alpha$
61: forest absorbers at low redshifts are filaments (with low HI column densities)
62: containing about 30\% of all baryons
63: (Stocke et al. 2004). Hydrodynamical simulations predict that
64: another large fraction of all baryons resides within mildly-nonlinear
65: structures with temperatures $0.01-1$KeV, called Warm-Hot Intergalactic
66: Medium (WHIM). The amount of baryons within this WHIM is estimated to reach
67: 20 - 40\% (Cen \& Ostriker 1999; Dave et al. 1999, 2001).
68:
69: Being highly ionized the WHIM could generate temperature anisotropies
70: by means of the thermal or kinematic Sunyaev-Zeldovich effect
71: (TSZ, KSZ; Sunyaev \& Zeldovich 1972, Sunyaev \& Zeldovich 1980),
72: adding to the contribution coming from gas in halos.
73: Hallman et al (2007) carried out a numerical simulation
74: and found that about a third of the SZ flux would be generated by unbound gas.
75: In Atrio-Barandela \& M\"ucket (2006) -hereafter ABM- we estimated the CMB anisotropies
76: generated by the WHIM due to the TSZ effect.
77: In this letter we compute the imprint on CMB temperature anisotropies
78: due to the peculiar motions of gas in filaments and halos.
79: The relative KSZ contribution with respect to the TSZ increases with decreasing
80: temperature and the anisotropies generated could be used to trace
81: the gas distribution on large scales. Briefly, in Sec. 2 we show
82: the effect of individual filaments on the CMB temperature anisotropies.
83: We extend an early calculation of the KSZ probability distribution function due to halos
84: by Yoshida, Sheth \& Diaferio (2001) by including gas not bound in halos
85: and generalize the work of Juszkiewicz, Fisher \& Szapudi (1998).
86: In Sec. 3 we compute the radiation power spectrum, showing that
87: it reaches its maximum at $l\sim 400$, an angular scale much larger than that of
88: clusters of galaxies; we also show that a statistical analysis
89: including the anisotropies generated by the peculiar motions in the WHIM
90: improves the fit to WMAP 3yr data. Finally, we present our main conclusions.
91:
92:
93: \section{The effect of single filament.}
94:
95: The temperature anisotropy due to the KSZ effect,
96: along a given line of sight ($los$), is the number density
97: weighted average of the radial component of the peculiar velocity of electrons:
98: $\Delta T=T_o\sigma_T\int n_e(\hat{l})(v_{e,r}(\hat{l})/c) dl=\tau v_r/c$.
99: In this expression, $n_e$ is the electron number density,
100: $v_{e,r}$ is the projected peculiar velocity along the
101: line of sight $\hat{l}$, $\sigma_T$ is the Thomson cross section
102: and $c$ the speed of light. The last expression follows when
103: the KSZ signal is dominated by a single filament of size $L$ and radial component
104: of its peculiar velocity $v_r$. In this case, $\tau=\sigma_Tn_eL$ is the optical
105: depth along the filament of average electron density $n_e$.
106:
107: To compute the probability distribution that a filament produces
108: a given KSZ anisotropy, we assumed that the matter distribution is
109: well described by a log-normal random field.
110: Coles \& Jones (1990) showed this approximation to be adequate when matter
111: density perturbations are weakly non-linear and the velocity
112: field is still in the linear regime.
113: The probability that at any given point
114: and at redshift $z$ there is an overdensity $\xi=n_e(\vec{x},z)/\bar{n}_e(z)$,
115: with $\bar{n}_e(z)$ the average electron density, is:
116: $f_1(\xi)=(\xi\sigma_\delta\sqrt{2\pi})^{-1}
117: \exp\left[-(\log\xi-\mu_\delta)^2/(2\sigma_\delta^2)\right]$
118: where $\mu_\delta$ and $\sigma_\delta$ are the mean and rms of the matter density contrast
119: in the linear regime (i.e., $\mu_\delta=0$).
120: For filaments of characteristic scale length $L$, this distribution
121: is also the distribution of the optical depth $\tau$ at each location.
122: The KSZ contribution includes the effect of
123: all baryons along the $los$, independently of whether they are in halos
124: or in the form of WHIM.
125:
126: The velocity field
127: is accurately described as Gaussian Random Field with mean and rms $\mu_v,\sigma_v$.
128: Its coherence scale is much larger than that of the density field;
129: the motion of galaxies, clusters and the IGM filaments is not
130: determined solely by the matter distribution in their vicinity,
131: but from a much larger scale. Under this assumption, we can assume
132: that density and velocity fields are independent. At each location, the
133: probability distribution function (PDF) that an ionized gas cloud
134: would produce a KSZ of a given
135: amplitude is simply the product of the distribution of the optical
136: depth (log-normal) and of the velocity (gaussian).
137: If $X=\tau$ and $Y=v_r$ are two continuous random variables with
138: joint probability $f_{X,Y}(x,y)$, then the PDF of $U=XY=\tau v_r$ is:
139: $f_U(u)=\int f_{X,Y}(x,u/x)|x|^{-1}dx$ (Rohatgi, 1976).
140: When both random variables are independent, then $f_{X,Y}=f_Xf_Y$.
141: The resulting PDF is symmetric with respect to the y-axis.
142: In Fig.~\ref{fig:pdf} we plot only the region $u>0$; the thick solid line
143: represents the PDF of the KSZ effect due to the IGM.
144: For comparison, we also plot as a thin solid line the gaussian distribution.
145: Both distributions were normalized such that the velocity and linear density
146: fields have zero mean and unit variance ($\sigma_v=\sigma_\delta=1$).
147: Then, u is expressed in units of $\langle\tau\rangle\sigma_v$, where $\langle\tau\rangle$
148: is the average optical depth of filaments of a size $L$.
149: Let us remark that the KSZ distribution has much larger tails than the
150: normal distribution. The same PDF was computed by Yoshida et al. (2001) for the halo model.
151: In their approach, it was unlikely that a given $los$ will contain more than one halo and
152: their PDF was dominated by the more abundant and less massive halos,
153: and only in the tails of the distribution is the contribution of the more
154: massive halos significant. In ours, the optical depth along each line of sight can be
155: much larger since it includes the effect of baryons in halos together
156: with the WHIM in low dense regions, and this effect will raise the
157: tails of the PDF.
158:
159: More informative is the Distribution Function $F(u)=\int_{-\infty}^{u} f_U(u')du'$.
160: In Fig.~\ref{fig:pdf} we plot the quantity $P(u>u_0)=1-F(u_0)$ (dashed line)
161: that represents the probability that a given filament produces a KSZ signal $u$ equal
162: or larger than $u_0=\langle\tau\rangle v_r$.
163: The shaded regions $[0,1.7]$ and $[0,6.2]$ correspond to percentiles 0.9 and 0.99.
164: For a gaussian distribution, the probability that a random variable has an
165: amplitude 2.7 times the rms deviation is 1\%.
166: For the KSZ distribution, at the same percentage, the amplitude is 6.2 times the average;
167: if a filament of size $L$ and electron density $n_e$ has
168: an optical depth $\tau\simeq 2\times 10^{-3}(n_e/10^3m^{-3})(L/1Mpc)$ and
169: moves with a velocity of $300km/s$, it will typically give rise to
170: $\Delta T_{KSZ}\sim 5.5 \mu$K. About 9\% of the filaments in this
171: population will have enough velocity and density to
172: produce an effect in the range $[9.3,34]\mu$K and 1\% will
173: have values larger than $34\mu$K. The same region will produce, on average, a
174: thermal SZ signal of $\delta_{TSZ}\sim 11.5 G(\nu)\mu$K$(T_x/1keV)$,
175: where $G(\nu)\sim -2$ in the Rayleigh-Jeans regime.
176:
177: Observations with the Very Small Array
178: of the Corona Borealis Supercluster (G\'enova-Santos et al. 2005) have detected
179: an extended ($\sim 30$ arcmin) cold spot without significant X-ray emission on the
180: ROSAT-R6 map, termed decrement H. Those authors described
181: the likelihood that it could be partly generated by an unknown massive
182: cluster in the background or by a filament. They did not include
183: the significant KSZ contribution in their study. For example, 1\% of all the filaments
184: of size $\sim 5$Mpc, aligned with the $los$ and with a coherent flow at similar scale
185: will produce a $\delta_{TSZ}+\delta_{KSZ}\sim (-150,-200)\mu$K,
186: enough to explain the central temperature of $-230\pm 23\mu$K of decrement H.
187: Further observations, like those to be carry out by the PLANCK satellite, at various frequencies
188: and with better angular resolution, are required to quantify the KSZ contribution.
189:
190:
191: \section{Contribution to CMB anisotropies.}
192:
193: The KSZ effect is independent of redshift; along a $los$
194: contributions from individual filaments will average out since their
195: peculiar velocities will have random directions.
196: Since the non-linear density and velocity fields
197: evolve with redshift, so does the contribution of the filament network.
198: To estimate the temperature anisotropies generated by the
199: peculiar motions of filaments and halos
200: we adapted the formalism developed by Choudhury, Padmanabhan \& Srianand (2001)
201: and ABM. The correlation function of the KSZ signal is
202: \begin{equation}
203: C(\theta)=\langle\frac{\Delta T_{KSZ}(\hat{x})}{T_o}\frac{\Delta T_{KSZ}(\hat{x}')}{T_o}\rangle=
204: \frac{\sigma_T^2}{c^2}\int_o^{l(z_f)}\int_o^{l(z_f)} dl dl' \langle n_e(l\hat{x},z)v_{e,r}(l\hat{x},z)
205: n_e(l\hat{x}',z')v_{e,r}(l\hat{x}',z')\rangle
206: \label{eq:definition}
207: \end{equation}
208: where $\theta$ is the angular separation between the two $los$, $\hat{x}$ and $\hat{x}'$.
209: The double integration is carried out from the origin to the highest redshift $z_f$ where the
210: IGM is still fully ionized.
211:
212: The main difficulty in making an analytical estimate
213: is to relate density and velocity fields at each location
214: since both random fields have very different scale lengths.
215: If we assume they are independent, like in the previous section, we can
216: decompose the velocity field into two contributions: the random motion
217: of matter with respect to the center of mass $v_R(z)$,
218: and the velocity of the center of mass with respect to
219: a comoving observer, also termed Bulk Flow $V_B$: $v_e(l\hat{x},z)=V_B(z)+v_R(l\hat{x},z)$.
220: With this decomposition, the integrals containing the random component
221: in eq.~(\ref{eq:definition}) average out and
222: \begin{equation}
223: C(\theta)=\frac{\sigma_T^2}{c^2}
224: \int_o^{l(z_f)}\int_o^{l'(z_f)} dl dl' V_B(R(z))D_v(z)V_B(R(z'))D_v(z')
225: \langle n_e(l\hat{x},z)n_e(l\hat{x}',z')\rangle
226: \label{eq:corr}
227: \end{equation}
228: where $V_B(R(z))$ denotes the mean bulk velocity of a sphere with radius
229: $R(z)$, the comoving distance from the observer to a filament at redshift $z$ and
230: $D_v(z)$ is the velocity linear growth factor.
231:
232: Assuming a log-normal distribution for matter and baryons,
233: in the non-linear regime, the electron number
234: density $n_e({\bf x},z)$ in the IGM is:
235: $n_B({\bf x},z)=n_0(z)\exp[\delta_B({\bf x},z)-\Delta_B^2(z)/2]$,
236: where ${\bf x}$ denotes the spatial position at redshift $z$
237: and $|{\bf x}(z)|$ is the proper distance;
238: $\delta_B({\bf x},z)$ is the linear baryon density contrast, $n_0(z)=\rho_B(1+z)^3/\mu_B m_p$
239: and $\rho_B$, $m_p$ are the baryon density and proton mass, respectively. Finally,
240: \be
241: \Delta_B^2(z)={\langle {\delta_B^2({\bf x},z)}\rangle} =
242: D^2(z)\int{\f{d^3k}{(2\pi)^3}\f{P_{DM}(k)}{[1+x_b^2(z)k^2]^2}} ,
243: \e
244: where $D(z)=D(z,\Omega_{\Lambda},\Omega_m)$ is the matter linear growth factor.
245: The spatial average in eq.~(\ref{eq:corr}) is ABM:
246: \begin{equation}
247: C(\theta)=\frac{\sigma_T^2}{c^2}
248: \int_o^{l(z_f)}\int_o^{l'(z_f)} dl dl' V_B(R(z))D_v(z)V_B(R(z'))D_v(z')
249: n_e(z)n_e(z')F(\theta,z,z'),
250: \label{eq:corr_simple}
251: \end{equation}
252: with
253: \be
254: F(\theta,z,z')=-1 + \exp\left[\frac{D(z)D(z')}{2\pi^2}
255: \int^\infty_0{{P(k)k^2 dk\over [1 + x_b^2(z)k^2][1 + x_b^2(z')k^2]}{\sin
256: (k|{\bf x}-{\bf x'}|)\over k|{\bf x}-{\bf x'}|}}\right] ,
257: \e
258: where $|{\bf x}-{\bf x'}|$ denotes the proper distance between two filaments at
259: positions ${\bf x}(z)$ and ${\bf x'}(z')$ separated by the angle $\theta$,
260: $P(k)$ is the matter
261: power spectrum and $x_b$ is the Jeans length at each redshift, that relates
262: the matter power spectrum to that of baryons (Fang et al. 1993):
263: $x_b(z)=H_0^{-1}\left[(2 \gamma k_{\rm B} T_m(z))/(3 \mu m_p \Omega_m (1+z))\right]^{1/2}$.
264: In this last expression, $T_m$ is the mean
265: temperature of the IGM, $\gamma\sim(1.2-1.6)$ is the polytropic index,
266: $\Omega_m$ is the cosmological fraction of matter density and
267: $\mu$ is the mean molecular weight of the IGM. The helium weight fraction
268: was $Y=0.24$. The factor $\gamma T_m$ defines the Jeans length and fixes
269: the smallest scale that contributes to the spectrum. We will take
270: $\gamma T_m \approx 2\times 10^4$K (Schaye et al. 2000).
271: The number density of electrons $n_e$ in the IGM can be obtained by assuming ionization
272: equilibrium between recombination, photo-ionization and collisional ionization. At the
273: conditions valid for the IGM (temperature in the range $10^4-10^7$K,
274: and density contrast $\delta < 100$) the degree of ionization is very high.
275: Commonly $n_e = \epsilon n_B$, with $0.9<\epsilon\le 1$ depending on the
276: degree of ionization. For future reference, $x_b=0.3-0.5h^{-1}$Mpc, depending
277: on model parameters (see ABM).
278:
279: In eq.~\ref{eq:corr_simple}, for each $\theta$ and $z$ the function
280: $F(\theta,z,z')$ is strongly peaked at $z = z'$ and is different from zero
281: only within a very narrow range $z' \in [z-\Delta z, z+\Delta z]$.
282: The effective peak width $2\Delta z$ increases and its maximum amplitude
283: decreases with decreasing angle and increasing redshift,
284: simplifying the numerical evaluation of eq.~\ref{eq:corr_simple};
285: in our numerical integration we took $\Delta z=0.1$.
286: Integration around the peak width was carried out with high accuracy, using $z,z'$ intervals of
287: $\delta z=5\times 10^{-4}$ and angle separations of 30 arcseconds.
288: Further integration (with respect to $z$ only) is carried out with a much smoother function
289: and can be performed by a relatively rough spacing in redshift.
290:
291: In Fig.~\ref{fig:cl}a, we show the radiation power spectrum obtained using
292: the formalism developed before. Since some of the parameters in our model are poorly
293: determined, we considered the effect on the radiation power spectrum of:
294: (a) the amplitude of the matter power spectrum measured by $\sigma_8$,
295: (b) the averaged temperature of the IGM, $T_m$ and
296: (c) the maximum overdensity $\xi$ for which the log-normal
297: distribution is a good description of the IGM.
298: We took $\sigma_8=(0.8;0.9),
299: T_m = (1.4\times 10^4; 5.\times 10^4)$ K and $\gamma=1.3$ throughout. We
300: restricted all integrations to overdensities $\xi \le 500$.
301: The overall shape of the power spectra is very similar for
302: a large range of the parameter space:
303: the spectrum peaks around $l\sim 200-600$ (shaded area)
304: with a maximum amplitude of $10-25\mu$K.
305: This signal is much larger than the one considered by Hajian et al (2007),
306: associated to the free electrons in the halo of the Galaxy.
307: The amplitude and angular scale depend on $x_b$ or, equivalently, on
308: $T_m$. The Jeans length fixes the scale where the power in the baryon density
309: field falls below that of the DM. Decreasing $x_b$ decreases the scale
310: where baryons evolve non-linearly and increases the fraction of baryons
311: sharing the same coherent motion, increasing the KSZ power spectrum
312: and shifting the maximum to smaller angular scales.
313: The power also increases with increasing either $\sigma_8$ or $\xi$.
314: If we vary parameters within the range $\sigma_8\in (0.7-0.9)$ and $\xi\in(50-500)$
315: the amplitude of the power spectrum could change up to a factor $\approx 5$.
316: $T_m$ has the largest effect: the amplitude scales as $\sim T_m^{-6}$ similarly
317: to the Thermal Sunyaev-Zeldovich contribution (see ABM).
318:
319: In Fig.~\ref{fig:cl}b we plot the contributions to the power spectrum from
320: different redshift bins ($0< z < 0.15; dz = 0.01$). As
321: remarked above, the differential contribution strongly decreases with redshift.
322: The signal is dominated by: (1) the bulk motion on scales
323: of $300h^{-1}$Mpc and below and (2) by the structures
324: with largest density contrast. Contributions from higher
325: redshifts and lower densities are negligible due to the superposition
326: of filaments with randomly oriented velocities with decreasing contributions.
327: Like in our earlier results for the TSZ effect,
328: the scale of the maximum is fixed by the Jeans length; at z=0.01, $x_b\sim 400h^{-1}$kpc
329: corresponds to an angular scale $l\sim 300$, as shown.
330:
331: Comparing the spectrum of Fig.~\ref{fig:cl} with the results of
332: the KSZ effect obtained using the halo model (Molnar \& Birkinshaw 2000,
333: Cooray \& Sheth 2002) the former shows more power (factor 3-10, depending on model
334: parameters) and at different angular scale. While the halo model
335: only accounts for the gas in halos, our formalism also incorporates the
336: gas in low dense regions. Those regions with intermediate density contrast
337: include the largest gas fraction at low redshift (see ABM)
338: when the bulk flow velocities are largest, which boost the power
339: with respect to the halo model result.
340:
341: Let us consider whether the KSZ contribution described above
342: could explain the excess of power in the range $l\sim 200-600$ measured
343: by WMAP 3yr data with respect to the best fit model
344: (see Hinshaw et al 2007, Fig. 17). We explored the parameter
345: space of $\Lambda$CDM models using a Monte Carlo Markov Chain (MCMC);
346: we added a KSZ contribution with an amplitude at the maximum of $25\mu$K
347: to the radiation power spectrum of each cosmological model.
348: This approach is not self consistent since when $\sigma_8$ changes so does
349: the amplitude of the KSZ contribution. Even though, the best fit model improved
350: the fit in one unit: $\Delta\chi^2=1$. This improvement is not statistically
351: significative from a Bayesian point of view (Liddle 2004): we introduced two parameters
352: $\gamma T_m$ and the maximum overdensity $\xi$, to obtain only a mild improvement on the fit.
353: Comparison with the concordance model, parameters deduced from
354: our MCMC differ by less than 1\%, being the variation on $\Omega_bh^2$ the most
355: significative. This exercise does show that ionized gas could originate measurable CMB
356: temperature anisotropies and CMB observations could provide evidence of the
357: gas distribution at large scales.
358:
359:
360:
361: \section{Discussion.}
362:
363: If the 'missing baryons' are in the form of a highly ionized WHIM,
364: and this medium shares the same velocity field than galaxies and dark matter,
365: then it will produce a contribution to the CMB temperature anisotropies
366: that could be detectable. Gas in halos and filaments with bulk flows aligned with
367: the $los$ could explain a significant fraction of the signal measured in the decrement H
368: of the Corona Borealis supercluster observed by the VSA interferometer.
369: Also, we computed the effect of the overall population of filaments on
370: CMB temperature anisotropies. Their contribution has a caracteristic
371: angular scale at $l\sim 400$, very different from that of clusters
372: of galaxies. Uncertainties in
373: our model parameters translate into an uncertainty of a factor $5$
374: in the amplitude of the power spectrum.
375:
376: Intrinsic and KSZ temperature anisotropies have the same frequency dependence, they
377: can not be separated directly from the data. Using a MCMC analysis
378: we have shown that this extra component improves the fit of the
379: concordance model to WMAP 3yr data, but the improvement can not
380: be considered significant. WMAP 5yr data or the future
381: PLANCK observations could have enough statistical power to indicate the existence
382: of this component. Further, PLANCK has enough angular resolution and frequency coverage
383: as to measure the KSZ contribution of structures such as decrement H in the VSA data of
384: the Corona Borealis Supercluster.
385:
386: \acknowledgments
387: F.A.B. acknowledges financial support from the Spanish Ministerio de Educaci\'on y Ciencia
388: (project FIS2006-05319) and Junta de Castilla y Le\'on (project SA010C05).
389: We thank R. Sheth for discussions. J.P.M. thanks the University of Salamanca for
390: its support and hospitality.
391:
392:
393: \begin{thebibliography}{99}
394: \bibitem{2} Atrio-Barandela, F., M\"ucket, J. 2006, ApJ, 643, 1 (ABM)
395: \bibitem{5} Cen, R., Ostriker, J.P. 1999, ApJ, 519, L109
396: \bibitem{6} Choudhury, T.R., Padmanabhan, T., Srianad, R. 2001, MNRAS, 322, 561
397: \bibitem{7} Coles, P., Jones, B. 1991, MNRAS 248, 1
398: \bibitem{7b} Cooray, A. \& Sheth, R. (2002) Physics Reports, 372, 1
399: \bibitem{8} Dave, R., Hernquist, L., Katz, N., Weinberg, D.H. 1999, ApJ, 511, 521
400: \bibitem{9} Dave, R., et al. 2001, ApJ, 552, 473
401: \bibitem{10} Fang, L.Z., Bi, H., Xiang, S., B\"orner, G. 1993, ApJ, 413, 477
402: \bibitem{101} Fukugita, M., Peebles, P.J.E. 2004, ApJ, 616, 643
403: \bibitem{102} Hajian, A., Hernandez-Monteagudo, C, Jim\'enez, R., Spergel, D., Verde, L.
404: (2007). Preprint, arXiv:0705.3245
405: \bibitem{11} Hallman, E. J. et al (2007) astro-ph 0704.2607
406: \bibitem{12} Hern{\'a}ndez-Monteagudo, C., Verde, L., Jim\'enez, R., Spergel, D.N. 2006, ApJ, 643, 598
407: \bibitem{121} Hinshaw, G. et al. (2007) ApJSS, 170, 288
408: \bibitem{122} Juskiewicz, R., Fisher, K.B. \& Szapudi, I. (1998) ApJ, 504, L1
409: \bibitem{13} Liddle, A.R. (2004) MNRAS, 351, L49
410: \bibitem{14} Molnar, S.M. \& Birkinshaw, M. (2000) ApJ, 537, 542
411: \bibitem{15} Rauch, M., et al. 1997, ApJ, 489, 7
412: \bibitem{16} Rauch, M. 1998, ARA\&A, 36, 267
413: \bibitem{17} Richter, P. et al. (2005). Preprint, astro-ph/0509539
414: \bibitem{18} Rohagti, B.K. (1976) ''An Introduction to Probability Theory Mathematical
415: Statistics''', Wiley, New York
416: \bibitem{20} Schaye, J., Theuns, T., Rauch, M., Efstathiou, G., Sargent, W.L.W. 2000, MNRAS, 318, 817
417: \bibitem{21} Schaye, J. 2001, ApJ, 559, 507
418: \bibitem{22} Stocke, J.T., Shull, J.M., Penton, S.V. 2004, astro-ph/0407352
419: \bibitem{23} Sunyaev, R.A. \& Zel'dovich, Ya. B. 1972, Comments Astrophys. Space Phys., 4, 173
420: \bibitem{24} Sunyaev, R.A. \& Zel'dovich, Ya. B. 1980, MNRAS, 190, 413
421: \bibitem{25} Yoshida, N., Sheth, R. \& Diaferio, A. (2001) MNRAS, 328, 669
422: \end{thebibliography}
423:
424: \clearpage
425:
426: \begin{figure}[ht]
427: \plotone{f1.eps}
428: \caption[]{KSZ (thick solid) and Gaussian (thin
429: solid) probability distribution functions. The dashed line represents
430: $P(u>u_0)$, the probability that the a given filament produces a KSZ anisotropy
431: $u$ equal or larger than a given value $u_0$. Shaded areas correspond limit the regions of these
432: probabilities being 10\% and 1\%.
433: }
434: \label{fig:pdf}
435: \end{figure}
436:
437: \begin{figure}[ht]
438: \plotone{f2.eps}
439: \caption[]{
440: (a) KSZ Radiation power spectra for different parameter sets $\sigma_8,
441: T_m, \xi$ (see labels). (b) Differential Power spectrum contribution from
442: subsequent redshift bins of width $dz=0.01$. The shaded area indicates the
443: interval where power spectra with different parameters reach
444: their maxima.}
445: \label{fig:cl}
446: \end{figure}
447: \end{document}
448:
449: