0705.3245/ms.tex
1: %\documentclass[aps,prl,twocolumn,showpacs,groupedaddress]{revtex4}
2: %\documentclass[aps,preprint,superscriptaddress]{revtex4}
3: %\documentclass[preprint2]{aastex}
4: \documentclass[usegraphicx,usenatbib,useapjfonts,apj]{emulateapj}
5: \usepackage{graphicx}  % needed for figures
6: \usepackage{dcolumn}   % needed for some tables
7: \usepackage{bm}        % for math
8: \usepackage{amssymb}   % for math
9: 
10: 
11: % My definitions
12: \def\be{\begin{equation}}
13:   \def\ee{\end{equation}}
14: \def\bea{\begin{eqnarray}}
15:   \def\eea{\end{eqnarray}}
16: \def\fun#1#2{\lower3.6pt\vbox{\baselineskip0pt\lineskip.9pt
17: \ialign{$\mathsurround=0pt#1\hfil##\hfil$\crcr#2\crcr\sim\crcr}}}
18: \def\lap{\mathrel{\mathpalette\fun <}}
19: \def\gap{\mathrel{\mathpalette\fun >}}
20: 
21: \newcommand{\vdag}{(v)^\dagger}
22: \newcommand{\myemail}{ahajian@PRINCETON.edu}
23: 
24: 
25: \shorttitle{Galactic kSZ}
26: \shortauthors{Hajian et al.}
27: 
28: \begin{document}
29: 
30: 
31: 
32: \title{The Kinetic Sunyaev-Zel'dovich Effect Due to the Electrons of Our Galaxy}
33: \author{Amir Hajian\altaffilmark{1,2}, Carlos Hern\'andez-Monteagudo\altaffilmark{3}, Raul Jimenez\altaffilmark{3}, David Spergel\altaffilmark{2}, Licia Verde\altaffilmark{3}} 
34: \altaffiltext{1}{Department of Physics, Jadwin Hall, Princeton University,
35:     Princeton, NJ 08542; {\it ahajian@princeton.edu}.}
36: \altaffiltext{2}{Department of Astrophysical Sciences, Peyton Hall,
37:    Princeton University,
38:     Princeton, NJ 08544; {\it dns@astro.Princeton.edu}. }
39: \altaffiltext{3}{Department of  Physics and Astronomy, University of Pennsylvania, 209s 33rd str, Philadelphia, PA 19104; {\it carloshm@astro.upenn.edu; 
40: raulj@physics.upenn.edu;lverde@physics.upenn.edu}.}
41: %\email{ahajian@princeton.edu; carloshm@astro.physics.upenn.edu; 
42: %raulj@physics.upenn.edu; dns@astro.Princeton.edu;
43: %lverde@physics.upenn.edu}
44: 
45: 
46: \begin{abstract}
47: We compute the effect of local
48: electrons on the CMB temperature anisotropies.
49: The number density and distribution of free electrons in our Galaxy
50: has been accurately measured from pulsar dispersion measurements.
51: Because of their distribution, the dynamics of our Galaxy and the
52: Galaxy peculiar velocity with respect to the Hubble flow, these free
53: electrons leave a frequency-independent imprint on the cosmic
54: microwave background (CMB). In particular, the coherent motion of the
55: free electrons respect to us and to the CMB rest frame produce a
56: kinetic Sunyaev-Zeldovich signal. We compute this effect and 
57: we note that the large-scale antisymmetry of the signal gives 
58: it an angular power spectrum with a sawtooth pattern where 
59: even multipoles are suppressed with respect to the odd ones.  
60: We find the signal to
61: be small ($\sim 2 \mu K$) and sub-dominant compared to the primary CMB
62: and other foreground signals. However, since there are no free
63: parameters in the modeling of this signal, it can be taken into
64: account if more precise measurements of the primordial signal are
65: required.
66: 
67: \end{abstract}
68: 
69: \keywords{Cosmology: cosmic microwave background ---  Galaxy: kinematics and dynamics}
70: 
71: \section{Introduction and Background}
72: The motion of the ionized gas in our Galaxy leaves its imprint on the CMB 
73: through the Kinetic Sunyaev-Zeldovich Effect \citep{SZ1, Hogan}. 
74: These distortions are spectrally indistinguishable from the CMB. 
75:  Thompson scattering of
76: CMB photons from a line element $ds$ of free electrons with optical
77: depth $d\tau$ along the line of sight determined by the unit vector
78: $\hat{\bf n}$, gives a correction to the CMB temperature:
79: %
80: \begin{equation}
81: dT(\hat{\bf n})=-d\tau \delta T_{CMB}(\hat{\bf n}) - d\tau \frac{\bf v}{c}\cdot \hat{\bf n} T_{CMB}(\hat{\bf n}).
82: \end{equation}
83: %
84: Thus there are two terms, the first is a blurring of the CMB
85: anisotropies ($\delta T_{CMB}(\hat{\bf n})=T_{CMB}(\hat{\bf n})-T_0$ where $T_0=\langle T_{CMB}\rangle$), while the
86: second is a generation of new anisotropies: the kinetic SZ (kSZ) effect.
87: 
88: If $\delta T_{CMB}\ll T_{0}$ (i.e., we are in a coordinate system where the
89: CMB has zero dipole\footnote{Note that this will not be the case in a
90: coordinate system where the CMB has a large dipole, like for example
91: in a coordinate system where the Sun is at rest}) and $v/c > 10^{-5}$, the second term dominates. Thus the integrated effect along the line of sight, for $\tau \ll 1$ is given by:
92: \begin{equation}
93: \label{eqn:kszint}
94: \frac{\delta T}{T}=-\int n_e \sigma_T \frac{v_r}{c} dl \sim \tau(\hat{\bf n})\frac{v_r(\hat{\bf n})}{c}
95: \end{equation}
96: where $v_r \equiv {\bf v}\cdot \hat{\bf n}$,  $n_e$ denotes the electron density and $\sigma_T$ the Thompson
97: cross section.  Given that plausible velocities are of the orders of
98: few $\times 100$ km/s (the largest velocity, $620$ km/s, is that of
99: the infall of the local group to the Great Attractor) only  $\tau > 10^{-3}$ yields a measurable signal. Such high optical depth
100: implies a number density of electrons and baryonic mass that is only
101: consistent with that present in the disk of our own Galaxy.
102: 
103: 
104: In this case, the velocity can be written as a sum of the Sun's
105: velocity as measured from the CMB dipole, $v_{\rm dipole}$, and the gas velocity 
106: with respect to the Sun.
107: Note that in a coordinate system where  the Sun is at rest, the  contribution for $v_{\rm dipole}$ is interpreted as a blurring of anisotropies (first term in the LHS of eq 1), the anisotropies being the  CMB dipole.
108: 
109: 
110: Here, we compute the kSZ signature of the Galaxy on the CMB
111: sky. While this signal is sub-dominant compared to other foreground
112: signals and small compared to the intrinsic CMB anisotropy, it is a
113: component that must be there and that can be modeled as we illustrate
114: below.  We assume a simple model for the velocities of the free
115: electrons in our Galaxy that includes an axisymmetric model for the
116: rotation around the Galactic center with the same speed as the
117: luminous matter, $\mathbf{v}_{rot}$, and a bulk motion,
118: $\mathbf{v}_{bulk}$ of the galactic center with respect to the Hubble flow.  
119: We neglect the higher order corrections to these
120: assumptions  such as blurring of intrinsic CMB anisotropies (first term in eq. 1). 
121: The total velocity of the free electrons with respect to
122: the CMB rest frame is the vector sum of the above velocities,
123: $\mathbf{v}=\mathbf{v}_{rot}+\mathbf{v}_{bulk}$. The bulk velocity,
124: $\mathbf{v}_{bulk}$, can be derived from the observed dipole of the
125: CMB as $\mathbf{v}_{bulk}=\mathbf{v}_{dipole}-\mathbf{v}_\odot$, where
126: $\mathbf{v}_\odot$ is the linear velocity of the rotation of the Sun
127: around the galactic center. Therefore we have
128: $\mathbf{v}=\mathbf{v}_{dipole}+(\mathbf{v}_{rot} - \mathbf{v}_\odot)$. 
129: The radial part of this velocity is the $v_r$
130: that appears in eqn.(\ref{eqn:kszint}) and is given by 
131: \be
132: \label{eqn:vr}
133: v_r=v_{los}+\mathbf{v}_{dipole}\cdot\hat{\mathbf{n}}, 
134: \ee 
135: where $v_{los}=(\mathbf{v}_{rot} - \mathbf{v}_\odot)\cdot\hat{\mathbf{n}}$
136: is the line-of-sight velocity with respect to the Sun.  
137: 
138: %In order to predict the kSZ pattern due to the free electrons in our
139: %Galaxy, we will need a model of the electron distribution in the
140: %Galaxy, $n_e$, and a model for the rotation of the Galaxy to calculate
141: %the $v_{los}$. We present these in sec. 2 and 3 respectively. In
142: %section 4 we present our model for the kSZ galactic signal and we draw
143: %our conclusion in sec. 5.
144: 
145: 
146: 
147: \section{The Model for the Galactic Distribution of Free Electrons}
148: %
149: We use the Cordes and Lazio \citep{Cordes:2002wz} model (hereafter NE2001) 
150:  for the distribution of free electrons in our Galaxy. 
151: \cite{Cordes:2002wz} combine the measurements of the 
152: dispersion measures (DM) of pulsars,
153: temporal broadening of pulses from pulsars with large DM,
154: scintillation bandwidth measurements of low-DM pulsars, angular
155: broadening of Galactic and extragalactic sources and emission
156: measures  to infer the distribution of the free electrons in the Galaxy
157: responsible for pulsar dispersion measures and a spatial model
158: of the warm ionized component of the interstellar gas. 
159: \begin{figure}[h]
160: %  \includegraphics[width=0.8\textwidth, angle=0]
161: \plotone{ne_4paper_label.eps}
162: \caption{\label{fig:ne} Distribution of the free electrons in our
163: Galaxy, $log_{10}(n_e)$, based on NE2001. The $n_e$ is shown at the
164: Galactic plane, $z=0$, and is maximum at the Galactic center. The
165: position of the Sun is denoted by a star, $\star$, and the small
166: overdense region close to the Sun is the Gum Nebula.
167: }
168: \end{figure}
169: 
170: The electron density distribution is given by the sum of two
171: axisymmetric components (the thick disk and the thin disk) and a
172: spiral arm component, combined with terms that describe specific
173: regions in the Galaxy. The local interstellar medium
174: is modeled with multiple components: 
175: (a) four low density regions near the Sun: a local hot
176: bubble centered on the Sun's location, the North Polar Spur, a local
177: superbubble and another low density region, (b) the Galactic center
178: component, (c) the Gum Nebula and  the Vela Supernova Remnant, (d) regions of
179: intense scattering (clumps) and (e) regions of low density (voids).
180: 
181: Figure \ref{fig:ne} shows the Galactic distribution of the free
182: electrons, $n_e$,at the Galactic plane,  based on the NE2001 model. The position of the Sun is denoted by a red star. The high electron density spot on the left side of the Sun is the  Gum nebula. 
183: 
184: \section{Kinematics of the Differential Rotation of the Galaxy}
185: 
186: We consider an axisymmetric model for the rotation of the Milky
187: Way. The orbits in this model are circular and the angular velocity at
188: point $\mathbf{R}$ is $\mathbf{\Omega}(R)$, where $R=|\mathbf{R}|$. We
189: are interested in the line-of-sight velocity, $v_{los}$, the
190: projection of the velocity of each particle relative to the Sun along the vector connecting the Sun to
191: that point, $({\mathbf R} - {\mathbf R}_{\odot})$. Here, ${\mathbf R}_{\odot}$
192: denotes the position of the Sun in the Galactocenteric coordinates.
193: Therefore, 
194: \be
195: v_{los}=\hat{{\mathbf R}}_{los}\cdot ({\mathbf v}_{rot}-{\mathbf v}_{\odot}),
196: \ee 
197: where $\hat{{\mathbf R}}_{los}=({\mathbf R} - {\mathbf R}_{\odot})/|{\mathbf R} - {\mathbf R}_{\odot}|$.
198: Following the standard discussion in  \cite{GalacticAstronomy}, the  line of sight velocity $v_{los}$  can be expressed as a function of longitude, $l$, 
199: on the celestial sphere and $R$
200: \be
201: \label{vlos}
202: v_{los}(l,R) = \left[ \Omega(R) - \Omega(R_{\odot})\right] R_{\odot} \sin{l}.
203: \ee
204: %where $l$ is the longitude of the point $\mathbf{R}$.
205: %\cite{Binney:1996fb}.
206: % This  equation provides the means to determine the line-of-sight velocity 
207: %field of the Galaxy. 
208: 
209: The disk of the Milky Way consists of three parts: the central disk at
210: $R<3$ kpc, the inner disk, $R_0>R>3$ kpc and the outer disk,
211: $R>R_0$. The determination of the $v_{los}$ is different in each of
212: the three parts.
213: 
214: For the inner disk, the observable is the terminal velocity,
215: $v_{los}^t(l_1)$, along each line of sight, $l_1$. The terminal
216: velocity is defined as follows: consider the smallest ring -centered
217: on the galactic center- that intercepts the line of sight. The line of
218: sight is thus tangent to this ring, and the radius of the ring is $R = R_{\odot} \sin{l_1}$. The terminal velocity is the line of sight velocity of the ring at the tangent point. 21-cm line and  CO emission line observations 
219: are used to estimate  $v_{los}$ (e.g., \cite{rougoor}) 
220: %\cite{vanwoerden, rougoor}) 
221: and  the distance to the tangent point is determined by geometry.
222: The circular velocity of the ring can be calculated using $v_{los}^t(l)$ values reported in Chapter 9 of \citet{GalacticAstronomy} and eq.(\ref{vlos}),
223: %
224: \be 
225: v_c(R) =
226: v_{los}^{t} (l_1) + v_c(R_0) \sin{l_1}, \,\,\,\,\,\,\sin{l_1}=R/R_0,
227: \ee 
228: %
229: and therefore 
230: %
231: \be 
232: v_{los}(l,R) = \left[v_c(R) \cdot R_0/R -v_c(R_0)\right] \sin{l}.  
233: \ee
234: %
235: 
236: In the outer disk, distances  are not so easily determined and thus distances and  velocities are obtained either by 
237: observations of Cepheid variables or  by main sequence fitting of a 
238: young cluster and radio-line observations of  associated molecular 
239: gas (see \citet{bb93} and references therein).
240: 
241: The measurement of $v_{los}$ is then given by 
242: \be
243: v_{los}(l,R) =  W(R) \sin{l}, 
244: \ee
245: where $W(R) \equiv R_0[\Omega(R)-\Omega(R_0)]$ is given as a function of $R/R_0$ in \citet{bb93}. Note however that the errors associated to the distance $R$ and the errors assigned to a given measurement of $W(R)$ increase dramatically for $R>1.5R_0$, and therefore very little can be said about $v_{los}$ for radii larger than 2$R_0$, \citep{binneydehnen97}.
246: 
247: \begin{figure}[h]
248:   %\includegraphics[width=0.8\textwidth, angle=0]
249:   \plotone{vlos_4paper_label.eps}
250:   \caption{\label{fig:vlos} The line-of-sight velocity relative to the Sun, $v_{los}$, on the plane of the Galaxy.The position of the Sun is denoted by $\star$.   $v_{los}$ is negative when a point moves towards the Sun.  }
251: \end{figure}
252: 
253:  
254: The velocities in the central disk are more complicated. Here, we
255: calculate them in the same way as the inner disk velocities: we
256: extrapolate the terminal velocities of the central disk from those of
257: the inner disk. This approximation however does not significantly affect our final
258: result because the $v_{los}$ is small in the vicinity of the Galactic
259: center. In addition, the effect of non-circular motions such as
260: axisymmetric expansion, oval distortions of the orbits and random
261: motions, act as a correction to the above model. For
262: simplicity we ignore these corrections and only consider circular
263: motions: as discussed in Chapter 9 of \citet{GalacticAstronomy}, this simplified model works well outside of the central (3$kpc$) disk.
264: %should introduce errors in $v_{los}$ of $\lap$ 40\%.
265: 
266: 
267: \section{The kinetic SZ Pattern}
268: %
269: We now separately compute the contribution to the kSZ signal due to the 
270: rotation of the Galaxy (\ref{sub1}) and the contribution due to the motion of the Galaxy with respect to the CMB rest frame (\ref{sub2}).
271:  We use the free electron density model of NE2001 and the above model
272: for the velocity field of the Galaxy to obtain the Galaxy kSZ pattern.
273: As the velocity field, $v_{los}$ is defined with respect to the Sun, there will be an additional contribution due to ${\bf v}_{dipole}$. 
274: 
275: \subsection{Motion within the Galaxy}
276: \label{sub1}
277: %{\bf Contribution due to $v_{los}$}\\
278: We assume the Sun is located at $(x,y,z) = (0.0,
279: 8.5, 0.0)$ kpc in Galactocenteric coordinates,
280: $R_0=8.5$ kpc, and  $v_{\odot}=220$ km/s. The kSZ effect is calculated using eq.(\ref{eqn:kszint}). 
281: We integrate  eq.(\ref{eqn:kszint}) along the line of sight: 
282: %
283: \bea
284: \label{eqn:kszdiscrete}
285: \Delta T(\hat{n}) &=& -  \frac{\sigma_T}{c}  T_0 \sum_{i=1}^{N} n_e(r_i \hat{n})v_{los}(r_i \hat{n}) \Delta r\\ \nonumber
286: &\simeq&- (4.1 \mu K)
287: \sum_{i=1}^{N} \left(\frac{n_e(r_i \hat{n})}{1 cm^{-3}}\right) \left(\frac{v_{los}(r_i \hat{n})}{220 {\rm km/s}}\right)  \left(\frac{\Delta r}{kpc}\right),
288: \eea
289: %
290: where $r_i=r_{i-1}+\Delta r$ with $r_0=0$.  
291: While the velocities in the outer disk are not known beyond $r\sim 1.5-2
292: R_0$, most of the contribution comes from the central and inner disks, hosting the largest electron densities and peculiar velocities. Thus  we only consider electrons that are within the $r=2R_0$ sphere. 
293: The result is shown in Fig. \ref{fig:ksz}.
294: \begin{figure}[h]
295:   \includegraphics[width=0.3\textwidth, angle=90]{ksz.ps}
296:   \caption{\label{fig:ksz} Map of the predicted kinetic SZ signal
297:   caused by the rotation of the free electrons in the disk of
298:   our Galaxy.}
299: \end{figure}
300:  The kSZ pattern is  antisymmetric: its main
301: features are a cold and a hot spot at
302: the two sides of the Galactic center with $|l|<90^\circ$ and
303: $\delta T_{max}\sim 2\mu K$. These spots come from the regions in the inner
304: disk with the highest $v_{los}$ (see Fig. \ref{fig:vlos}).  There are
305: weaker warm and cold spots at larger longitudes, $|l|>90^\circ$,
306: caused by the rotating electrons in the outer disk. The kSZ signal is
307: zero at $l=0^\circ$ and $l=180^\circ$ where $v_{los}=0$. Both the
308: Galactic center and the Gum Nebula, which have the larger $n_e$, happen
309: to fall in regions with very small $v_{los}$ and therefore do not
310: contribute much to the kSZ  due to the rotation of the Galaxy.
311: 
312: \subsection{Motion of the Galaxy}
313: \label{sub2}
314: %{\bf Contribution due to $v_{dipole}$}\\
315: The second term of 
316: eq.(\ref{eqn:vr}) arises from the bulk motion of the Galaxy with 
317: respect to the CMB rest frame.
318: We consider the dipole velocity to be
319: $v_{dipole}=371km/s$ in the direction of 
320: $(l,b)=(264^\circ,48^\circ)$.
321: \begin{figure}[h]
322:   \includegraphics[width=0.3\textwidth, angle=90]{ksztotal_HammurabiDipole_minusdipole.ps}
323:   \caption{\label{fig:final} Dipole subtracted map of the effect of the free electrons in our Galaxy due to the rotation and the bulk motion of the 
324: Milky Way.}
325: \end{figure}
326: This bulk motion generates a kSZ signal bigger than the one due to the
327: rotation of the Galaxy. The signal can be interpreted as a suppression of the dipole due to the free electrons in the Galaxy, thus it is a  negative dipole modulated by the line of sight integration of the electron distribution $n_e$.
328: The signal of the  Galactic Center and the Gum nebula feature prominently because of their  high density. 
329: 
330: \subsection{The Combined effect}
331: %{\bf The Combined effect}\\ 
332:  We add this to the map of Figure \ref{fig:ksz}
333: to obtain the total kSZ signal generated by the free electron
334: distribution in our Galaxy. The resulting  map {\em after dipole subtraction}
335: is shown in Figure \ref{fig:final}. This pattern is still antisymmetric
336: and mostly lies in the plane of the Galaxy but it 
337: overrides the four-fold pattern of Figure \ref{fig:ksz}.
338: The cold spot centered at $(l,b)=(-90^\circ,0^\circ)$ is
339: due to the overdensity of the free electrons in the Gum Nebula and the
340: warm spot at the center of the Galaxy  reflects the high $n_e$ there.
341: Figure \ref{fig:kszcl} shows the angular
342: power spectrum  of the kSZ signal. Most of the power resides at large scales
343: (low $l$'s), but its contribution to the intrinsic CMB power spectrum
344: remains around the 0.01\% level.
345: \begin{figure}[h]
346:   \includegraphics[width=0.4\textwidth, height=2in, angle=0]{cl_ksz.eps}
347:   \caption{\label{fig:kszcl} Angular power spectrum $C_{l}$ for the  predicted Kinetic SZ signal (red filled circles). Measurements from WMAP (3yr data) are displayed as black triangles.}
348: \end{figure}
349: An interesting feature of the above pattern is the saw-tooth shape
350: of its  angular power spectrum. This is due to the
351: large-scale antisymmetry of the kSZ pattern which suppresses the even
352: multipoles with respect to the odd ones.  The pattern quickly
353: disappears at higher multipoles.
354: 
355: 
356: \section{Discussion and Conclusion}
357: We have computed the kinetic SZ contribution to the CMB anisotropies
358: due to the coherent motion of the free electrons in the disk of our
359: Galaxy. We used the model of \cite{Cordes:2002wz} for the galactic
360: distribution of free electron and kinematics of the differential
361: rotation of our Galaxy to compute the line of sight velocity with
362: respect to the Sun. The bulk motion of our Galaxy causes another effect
363: that is comparable to this effect. We calculate the anisotropic pattern due to 
364: the combination  of these effects and show the result in Fig. \ref{fig:final}.
365: 
366: This kSZ signal  is subdominant compared to the
367: primary CMB and other foregrounds, but the model used to compute it has
368: no free parameters and thus this contribution to the anisotropy can be
369: easily modeled and subtracted out. This effect has been independently 
370: studied by another group \citep{Waelkens} and our results agree. The polarization signal of the Thompson scattering from the free electrons in the local Universe was studied in \cite{Hirata:2005nw}.
371: \begin{figure}[h]
372:   %\includegraphics[width=0.8\textwidth, angle=0]
373:   \plotone{ilcnvlos.ps}
374:   \caption{\label{fig:vlos_map} {\emph{top:}} The ILC map based on the WMAP 3yr data, smoothed with a $420'$ beam.  {\emph{bottom:}} A map of the integral of the line-of-sight velocity relative to the Sun in Galactic coordinates. Both maps happen to have  a similar structure on the large scales.} 
375: \end{figure}
376: In Figure \ref{fig:vlos_map} we show WMAP's  ILC map \cite{Hinshaw:2006ia} smoothed with a $420'$ beam and a map in galactic coordinates of the integral of the line of sight velocity due to the rotation of the Galaxy. Note the  similarity between the maps.  
377: However, any scattering process would flip the sign of the radial velocities.
378: % and
379: % that the dipole correction removes this quadrupolar pattern.
380: %LV:removed the above: a dipole correction does not kill a quardupole, this phrase was misleading
381: We find another similarity in the saw-tooth pattern of the angular
382: power spectrum of the kSZ map (Fig \ref{fig:kszcl}). This pattern has
383: the same shape as the full-sky power spectrum derived from WMAP's ILC
384: map: at low-$l$, even multipoles are suppressed with respect to the
385: odd ones. Although the kSZ signal is too small to explain the
386: suppression of the even multipoles in the WMAP data, it suggests that
387: a mechanism that boosts the above kSZ signal by about two orders of
388: magnitude might explain the saw-tooth pattern of the ILC power
389: spectrum and other peculiarities of the two point correlation of the
390: ILC map (such as the anticorrelation at $\theta=180^\circ$,
391: \cite{Hajian_in_prep,Hajian_ctheta}).  
392: Any element in our Galaxy that does not emit
393: but only scatters CMB photons would be a good candidate for the above
394: process, since it would preserve the thermal CMB spectrum and enhance
395: the kSZ signal. We do not know any physical mechanism with the above properties.The intriguing similarity in figure  \ref{fig:vlos_map} is, therefore more likely a warning about the dangers of  {\it a posteriori} analysis  than the signature of novel physics.
396: \acknowledgements 
397: AH thanks Joseph Taylor and  Andre Waelkens for enlightening discussions.  Some of the results in this paper have been derived using the HEALPix package \citep{Gorski:2004by}. 
398: We acknowledge the use of the Legacy Archive for Microwave Background Data Analysis (LAMBDA). Support for LAMBDA
399: is provided by the NASA Office of Space Science. AH acknowledges 
400: support from NASA grant LTSA03-0000-0090. LV and CHM are supported
401: in part by NASA grant ADP03-0000-009 and ADP04-0000-093. RJ, LV, CHM
402: and DNS are supported in part by NSF PIRE-0507768 grant.  LV, RJ and
403: CHM thank the Physics department of Princeton University for
404: hospitality while part of this work was being carried out. 
405: 
406: \begin{thebibliography}{}
407: 
408: 
409: \bibitem[Sunyaev \& Zeldovich(1980)]{SZ1} Sunyaev, R.~A., 
410: \& Zeldovich, I.~B.\ 1980, \mnras, 190, 413 
411: 
412: %\bibitem[Sunyaev \& Zeldovich(1981)]{SZ2} Sunyaev, R.~A., 
413: %\& Zeldovich, I.~B.\ 1981, Astrophysics and Space Physics Reviews, 1, 1 
414: 
415: 
416: \bibitem[Hogan(1992)]{Hogan} Hogan, C.~J.\ 1992, \apjl, 398, 
417: L77 
418: 
419: \bibitem[Binney \& Dehnen(1997)]{binneydehnen97} Binney, J., \& 
420: Dehnen, W.\ 1997, \mnras, 287, L5  
421: 
422: \bibitem[Binney \& Merrifield(1998)]{GalacticAstronomy} Binney, J., \& 
423: Merrifield, M.\ 1998, Galactic Astronomy, %/ James Binney and Michael 
424: %Merrifield.~ Princeton, NJ : 
425: Princeton University Press, 1998.%~ (Princeton 
426: %series in astrophysics) %QB857 .B522 1998  (\$35.00),  
427: 
428: \bibitem[Brand \& Blitz(1993)]{bb93} Brand, J., \& Blitz, 
429: L.\ 1993, \aap, 275, 67   
430: 
431: \bibitem[Cordes \& Lazio(2002)]{Cordes:2002wz}
432:   Cordes, J. M., Lazio T., J., W.,
433:   %%``NE2001.I. A New Model for the Galactic Distribution of Free Electrons and
434:   %%its Fluctuations,''
435:   arXiv:astro-ph/0207156.
436: 
437: %\cite{Gorski:2004by}
438: \bibitem[{Gorski et al.} (2005)]{Gorski:2004by}
439:   Gorski, K. M., Hivon, E.,  Banday, A. J., Wandelt, B. D., Hansen, F. K., Reinecke, M., Bartelman, M.,
440:   %``HEALPix -- a Framework for High Resolution Discretization, and Fast
441:   %Analysis of Data Distributed on the Sphere,''
442:   Astrophys.\ J.\  {\bf 622}, 759 (2005)
443: 
444: \bibitem[{Hajian} (2007)]{Hajian_ctheta}
445:   Hajian, A.,
446:   %``Analysis of the apparent lack of power in the cosmic microwave background
447:   %anisotropy at large angular scales,''
448:   arXiv:astro-ph/0702723.
449: 
450: \bibitem[{Hajian et al.} (2007)]{Hajian_in_prep}
451:   Hajian et al., in preparation.
452: 
453: %\bibitem[Hawkins \& Wright(1988)]{wright88} Hawkins, I., \& 
454: %Wright, E.~L.\ 1988, \apj, 324, 46 
455: 
456: %\cite{Hirata:2005nw}
457: \bibitem[{Hirata et al.} (2005)]{Hirata:2005nw}
458:   C.~M.~Hirata, A.~Loeb and N.~Afshordi,
459:   %``CMB B-mode polarization from Thomson scattering in the local universe,''
460:   Phys.\ Rev.\  D {\bf 71}, 063531 (2005)
461:   [arXiv:astro-ph/0501167].
462: 
463: %\bibitem[Li(2003)]{needles} Li, A.\ 2003, \apj, 584, 593 
464: 
465: \bibitem[Rougoor \& Oort (1960)]{rougoor} Rougoor, G.W. \& Oort, J.H., 1960, {\it Proc. Nat. Acad.Sci. 46, 1.}
466: 
467: %\cite{Taylor:1993my}
468: \bibitem[Taylor \& Cordes(1993)]{Taylor:1993my}
469:   Taylor J., H. , Cordes, J.~M.~,
470:   %%``Pulsar distances and the galactic distribution of free electrons,''
471:   Astrophys.\ J.\  {\bf 411}, 674 (1993).
472: 
473: 
474: %\cite{Binney:1996fb}
475: %\bibitem{Binney:1996fb}
476:  % J.~Binney and W.~Dehnen,
477:  % %``The outer rotation curve of the Milky Way,''
478:   %arXiv:astro-ph/9612060. 
479: 
480:  
481: %\bibitem[Wright(1982)]{wright82} Wright, E.~L.\ 1982, \apj, 255, 401 
482: 
483: %\cite{Hinshaw:2006ia}
484: \bibitem[Hinshaw et al. (2006)]{Hinshaw:2006ia}
485:   G.~Hinshaw {\it et al.}  [WMAP Collaboration],
486:   %``Three-year Wilkinson Microwave Anisotropy Probe (WMAP) observations:
487:   %Temperature analysis,''
488:   arXiv:astro-ph/0603451.
489: 
490: %\cite{Spergel:2006hy}
491: %\bibitem[Spergel et al. (2006)]{Spergel:2006hy}
492: %  D.~N.~Spergel {\it et al.}  [WMAP Collaboration],
493:   %``Wilkinson Microwave Anisotropy Probe (WMAP) three year results:
494:   %Implications for cosmology,''
495:  % arXiv:astro-ph/0603449.
496: 
497: \bibitem[Waelkens et al.(2007)]{Waelkens}
498: Waelkens et al. (2007) in preparation, (private communication).
499: 
500: \end{thebibliography}
501: 
502: \end{document}
503: