0707.0535/m9.tex
1: 
2: \documentclass[aps,float,prd,psfig]{revtex4}
3: %\documentclass[aps,twocolumn,float,prd,psfig]{revtex4}
4: %%%%%%%%%%%%%%%%%%%%%%%%%%%
5: 
6: 
7: \input epsf
8: \usepackage{graphicx}% Include figure files
9: \newcommand{\beq}{\begin{equation}}
10: \newcommand{\beqa}{\begin{eqnarray}}
11: 		  \newcommand{\eeq}{\end{equation}}
12: \newcommand{\eeqa}{\end{eqnarray}}
13: \newcommand{\etal}{{\it et al. }}
14: \newcommand{\siml}{\lesssim}
15: \newcommand{\simg}{\gtrsim}
16: \newcommand{\lsim}{\lesssim}
17: \newcommand{\gsim}{\gtrsim}
18: \newcommand{\psim}{\mbox{\raisebox{-1.0ex}{$~\stackrel{\textstyle \propto}
19: {\textstyle \sim}~$ }}}
20: \newcommand{\vect}[1]{\mbox{\boldmath${#1}$}}
21: \newcommand{\lmk}{\left(}
22: \newcommand{\rmk}{\right)}
23: \newcommand{\lnk}{\left\{ }
24: %\newcommand{\n(T_{obs}/10{\rm yr})^{1/2}n}{\nonumber}
25: \newcommand{\rnk}{\right\} }
26: \newcommand{\lkk}{\left[}
27: \newcommand{\rkk}{\right]}
28: \newcommand{\lla}{\left\langle}
29: \newcommand{\p}{\partial}
30: \newcommand{\rra}{\right\rangle}
31: \newcommand{\so}{M_\odot}
32: \newcommand{\mch}{{\cal M}}
33: \newcommand{\vex}{{\vect x}}
34: \newcommand{\ver}{{\vect r}}
35: \newcommand{\vue}{\hat{\vect e}}
36: \newcommand{\vel}{\vect l}
37: \newcommand{\vez}{\vect z}
38: \newcommand{\ven}{\vect n}
39: \newcommand{\vem}{\vect m}
40: \newcommand{\vep}{{\vect p}}
41: \newcommand{\veu}{{\vect u}}
42: \newcommand{\vev}{{\vect v}}
43: \newcommand{\veq}{{\vect q}}
44: \newcommand{\veo}{{\vect \Omega}}
45: \newcommand{\ved}{{\vect d}}
46: \newcommand{\ve}{{\vect e}}
47: \newcommand{\hf}{{\hat f}}
48: 
49: 
50: \begin{document}
51: %\baselineskip 8mm
52: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
53: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
54: \title{
55: Measuring a Parity Violation Signature in the Early Universe via 
56:    Ground-based Laser Interferometers } 
57: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
58: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
59: %
60: %
61: %
62: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
63: \author{Naoki Seto$^{1,2}$ and Atsushi Taruya$^3$}
64: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
65: \affiliation{$^1$National Astronomical Observatory, 2-21-1
66: Osawa, Mitaka, Tokyo, 181-8588, Japan\\ 
67: $^2$Department of Physics and Astronomy, 4186 Frederick Reines
68: Hall, University of California, Irvine, CA 92697\\
69: $^3$Research Center for the Early Universe, School of Science, The University of Tokyo, Tokyo 113-0033, Japan
70: }
71: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
72: \date{\today}
73: %
74: %
75: %
76: %
77: %
78: %
79: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
80: \begin{abstract}
81:  We show that pairs of widely separated  interferometers 
82:  are advantageous for measuring the Stokes parameter $V$  of a stochastic
83:  background of gravitational waves. This parameter characterizes asymmetry of 
84:  amplitudes of right- and left-handed  waves and generation of the asymmetry 
85:  is closely related to parity violation in the early universe. 
86:  The advantageous pairs include  LIGO (Livingston)-LCGT and AIGO-Virgo
87:  that are relatively insensitive to $\Omega_{\rm\scriptscriptstyle GW}$ 
88: (the simple intensity of 
89:  the background). Using at least three detectors,  information of the 
90:  intensity $\Omega_{\rm\scriptscriptstyle GW}$ and the degree of asymmetry 
91:  $V$ can be separately measured. 
92: \end{abstract}
93: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
94: \pacs{PACS number(s): 95.55.Ym 98.80.Es,95.85.Sz}
95: 
96: \maketitle
97: 
98: 
99: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
100: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
101: \section{Introduction}
102: %\underline{\em 1)  Introduction}
103: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
104: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
105: 
106: 
107: Stochastic background of gravitational waves is one of the most important
108: targets for gravitational wave astronomy.  
109: In the last decade, the detection threshold for the background has been 
110: rapidly improved around $\sim 100$Hz by 
111: continuous upgrades of ground-based  interferometers \cite{Abbott:2005ez}. 
112: This trend will be continued 
113:  with advent of next-generation interferometers currently planned
114: worldwide, such as advanced LIGO \cite{adv} and LCGT
115: \cite{Kuroda:1999vi}. Due to the weakness of 
116: gravitational interaction, our universe is transparent to the
117: background up to very early epoch, and we might uncover interesting 
118: nature of the universe at extremely high-energy scales, 
119: through observational studies of the stochastic background. 
120: To extract the information as much as possible, we need to 
121: characterize the background efficiently in a model independent manner,
122: and  investigation 
123: beyond simple spectral analysis might yield a great discovery. 
124: In this respect, circular polarization degree,  
125: which describes the asymmetry between the amplitudes of right- and 
126: left-handed waves, may be a fundamental characteristic of the 
127: background to probe the early universe. 
128: Because the parity transformation relates these two polarization modes, 
129: the asymmetry in the stochastic gravitational waves 
130: directly reflects a parity violation in the early universe,  for instance,
131: generated through the gravitational  Chern-Simons term 
132:  (e.g., \cite{Alexander:2004us}). 
133: In other words, one can detect a signature 
134: of parity violation by measuring the circular polarization degree of a 
135: gravitational wave background. 
136: Since the observed universe is highly isotropic and homogeneous, we shall 
137: focus on the monopole component of the circular polarization as our primary 
138: target, and report principle aspects for its measurement with a 
139: network of ground-based interferometers (see  \cite{Lue:1998mq,Caprini:2003vc} for CMB polarization
140: and \cite{Seto:2006hf} for space missions). 
141: 
142: 
143: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
144: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
145: \section{Circular polarization}
146: %\underline{\em 2)  Circular polarization}
147: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
148: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
149: 
150: 
151: Let us first describe circular polarization of a gravitational 
152: wave background. We use a plane wave expansion of the background as
153: \cite{Flanagan:1993ix,Allen:1997ad} 
154: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
155: \beq
156: h_{ij}(t,\vex)=\sum_{P=+,\times} \int^{\infty}_{-\infty} df \int_{S^2} d\ven~
157: h_P(f,\ven) e^{2\pi i f (-t+\ven \cdot \vex) } \ve^P_{ij}(\ven).
158: \label{plane}
159: \eeq
160: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
161: Here, the amplitude $h_P$ is the mode coefficient that is stochastic and 
162: random variable. 
163: The bases for transverse-traceless tensor $\ve^P$ $(P=+,\times)$ 
164: are given as 
165: $\ve^+_{}={\hat \ve}_\theta \otimes {\hat \ve}_\theta- {\hat \ve}_\phi
166: \otimes  {\hat \ve}_\phi$ and $\ve^\times_{}={\hat \ve}_\theta \otimes
167: {\hat 
168: \ve}_\phi+{\hat  
169: \ve}_\phi \otimes {\hat \ve}_\theta$ with unit vectors $ {\hat
170: \ve}_\theta$ and $ {\hat \ve}_\phi$. These vectors are 
171: normal to the propagation
172: direction $\ven$, associated with  a right-handed Cartesian 
173: coordinate as usual. As an alternative characterization, 
174: we can use the circular polarization bases 
175: $\ve^R=(\ve^++i\ve^\times)/\sqrt2$ (right-handed mode) and 
176:  $\ve^L=(\ve^+-i\ve^\times)/\sqrt2$ (left-handed mode) for the plane wave
177: expansion (\ref{plane}). The corresponding amplitudes $h_{R}$ and $h_{L}$ 
178: are given by 
179: $h_R=(h_+-i\,h_\times)/\sqrt2$  and $h_L=(h_++i\,h_\times)/\sqrt2$. 
180: The ensemble average of their amplitudes is classified as 
181: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
182: \beq
183: \left( \begin{array}{@{\,}c@{\,}}
184:              \lla h_R(f,\ven) h_R(f',\ven')^* \rra  \\
185:               \lla h_L(f,\ven) h_L(f',\ven')^* \rra \\ 
186:            \end{array} \right)
187: =\frac{\delta_{\ven,\ven'}\delta_{f,f'}}{4\pi}\left( 
188:            \begin{array}{@{\,}c@{\,}}
189:            I(f,\ven)+V(f,\ven)  \\
190:            I(f,\ven)-V(f,\ven)  \\ 
191:            \end{array} \right)  \label{matrix}
192: \eeq
193: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
194: with  the functions $\delta_{Y,Z}$ being delta functions. 
195: %The bracket $\lla \cdots \rra$ stands for an ensemble average. 
196: In the above expression, the real function $V$ characterizes the 
197: asymmetry between the amplitudes of right- and  the left-handed waves, while 
198: the function $I(\ge |V|)$ represents their total amplitude. 
199: Note that the other combinations such as $\lla h_R h_L^* \rra$ and 
200: $\lla h_L h_R^* \rra$ describe the 
201: linear polarization mode and are proportional to $Q\pm i\,U$, 
202: which constitute the well-known Stokes parameter, 
203: together with the $I$- and $V$- modes  
204: (see e.g., \cite{radipro} for electromagnetic counterpart). 
205: In this paper, we do not study the linear polarization $Q\pm i\,U$, since 
206: they do not have an isotropic component. 
207: We will focus on the detectability of the isotropic 
208: components $I(f)$ and $V(f)$ as our primary target. 
209: Using the normalized logarithmic energy density of the background 
210: $\Omega_{\rm \scriptscriptstyle GW}(f)$ \cite{Flanagan:1993ix,Allen:1997ad}, 
211: the two functions $I$ and $V$ are expressed as  
212: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
213: \beq
214: I(f)=\frac{\rho_{\rm c}}{4\pi f^3}\,\,\Omega_{\rm\scriptscriptstyle GW}(f),
215: \quad
216: V(f)=\frac{\rho_{\rm c}}{4\pi f^3}\,\,\Omega_{\rm\scriptscriptstyle GW}(f)\,
217: \Pi(f),
218: \eeq
219: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
220: where $\rho_{\rm c}$ is the critical density of the Universe,  
221: $\rho_{\rm c}=3H_0^2/8\pi$ with $H_0=70h_{70}$\,km/sec/Mpc 
222: being the Hubble parameter. 
223: The ratio $\Pi(f)=V(f)/I(f)$ characterizes the circular polarization degree. 
224: For simplicity, we assume the flat spectra, 
225: $\Omega_{\rm \scriptscriptstyle GW}(f)\propto f^0$ 
226: and $\Pi(f)\propto f^0$ as our fiducial model. Thus, our main interest is the simultaneous 
227: determinations or constraints on the parameters 
228: $\Omega_{\rm\scriptscriptstyle GW}$ and $\Pi$. 
229: 
230: 
231: We next consider how to detect the isotropic components of $I$- and 
232: $V$-modes with laser interferometers. Let us recall that the 
233: output signal $s_a$ of a detector $a$ at the position $\vex_a$ is written as 
234: $
235: s_a(f) =\sum_{P=+,\times}\int_{S^2} d\ven \,h_P(f,\ven)\, F^P_{a}(\ven, f)
236: \,e^{i\,\,2\pi \, f\, \ven\cdot\vex_a}. 
237: $
238: Here, the function  $F_a^P$ is the beam pattern function and it 
239: represents the response of the detector to a polarization mode $\ve^{P}$. 
240: Provided the data streams $s_a$ and $s_b$ taken from two detectors 
241: $a$ and $b$, the detection of stochastic signals can be achieved by 
242: taking a cross-correlation,   
243: $
244:  \lla  s_a(f) s_b(f')^* \rra \equiv C_{ab}(f) \delta_{f,f'}.
245: $
246: Keeping the signals from the isotropic components,  
247: the correlation signal $C_{ab}(f)$ is written as 
248: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
249: \beq
250:  C_{ab}(f)=
251: \gamma_{I,ab}(f)I(f)+\gamma_{V,ab}(f)V(f),
252: \eeq 
253: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
254: where the quantity $\gamma_I$ is the overlap  function
255: given by \cite{Flanagan:1993ix,Allen:1997ad} 
256: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
257: \beq
258: \gamma_{I,ab}(f)=\frac{5}{8\pi}\int_{S^2} d\ven  \lkk  \lnk
259: F_a^+F_{b}^{+*}+
260: F_a^\times F_{b}^{\times*} \rnk e^{i\,y\,\ven\vem} \rkk, \label{gi1}
261: \eeq
262: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
263: with $y\equiv2\pi f\,D/c$. Here, we have expressed 
264: $\vex_a-\vex_b$ as $D \vem$ ($D$:\,\,distance, $\vem$:\,\,unit vector). 
265: Similarly, the function $\gamma_{V,ab}(f)$ is 
266: obtained by replacing the kernel $[\cdots]$ in Eq.~(\ref{gi1}) 
267: with $\lkk i \lnk 
268: F_a^+F_{b}^{\times*}-
269: F_a^\times F_{b}^{+*} \rnk e^{iy \ven\vem}  \rkk $.
270: 
271: 
272: 
273: 
274: 
275: 
276: 
277: 
278: 
279: 
280: 
281: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
282: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
283: \section{Overlap functions for ground based detectors}
284: %\underline{\em 3)  Overlap functions for ground based detectors}
285: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
286: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
287: 
288: 
289: Now, specifically consider the response of an L-shaped 
290: interferometer $a$ on the 
291: Earth. We assume that the detector has two orthogonal arms with equal 
292: arm-length. 
293: \if0%%%%%%%%%%%%%%%%
294: In reality an interferometer might has an opening angle different from
295: $90^\circ$. For example the  angle of GEO600 is $94.3^\circ$. But
296: its response can be effectively regarded  as an interferometer with
297: $90^\circ$ opening angle.
298: \fi%%%%%%%%%%%%%%%%
299: Denoting the unit vectors parallel to the two arms by $\veu$ and 
300: $\vev$, the beam pattern function takes a simple form as 
301: $F_a^P=\ved_a:\ve^P(\ven)$ with 
302: $\ved_a=({\veu} \otimes {\veu}- {\vev} \otimes  {\vev})/2$, 
303: where the colon represents a double contraction.  
304: This expression is always valid as long as the wavelength of the 
305: gravitational waves for our interest is much longer than 
306: the arm-length of the detectors.  
307: \if0%%%%%%%%%%%%%%%%%% 
308: Table \ref{tab:detectors} provides a list of the positions and the 
309: orientations for ongoing (and planned) kilometer-size interferometers 
310: (see e.g., \cite{det}).  We use a spherical coordinate system
311: $(\theta,\phi)$ in which the north pole is located at 
312: $\theta=0^{\circ}$, with $\phi$ representing the longitude. 
313: The angle $\alpha$  characterizes the orientation of the detector between 
314: the local east direction and the bisecting line of the two arms 
315: measured counterclockwise.
316: \fi%%%%%%%%%%%%%%%%%%%%%%%
317: In this paper we study the following five ongoing (and planned)
318: kilometer-size interferometers as concrete examples; AIGO(A), LCGT(C),
319: LIGO-Hanford(H), LIGO-Livingston(L) and Virgo(V) (see {\it e.g.}
320: \cite{det} for their basic information). 
321:  Hereafter, 
322: we mainly use their abbreviations (A,C,H,L,V). 
323: 
324: 
325: \if0
326: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
327: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Table 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
328: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
329: \begin{table}[!tb]
330: %\begin{ruledtabular}
331: \begin{tabular}{lccc}
332: detector & $\theta$ & $\phi$ & $\alpha$ \\
333: \hline
334: \ AIGO (A) & 121.4 & 115.7 & -45.0\\
335: \ LCGT (C) & 53.6 & 137.3 & 70.0 \\
336: \ LIGO\ Hanford (H) & 43.5 & -119.4 & 171.8 \\
337: \ LIGO\ Livingston (L) & 59.4 & -90.8 & 243.0 \\
338: \ Virgo (V) & 46.4 & 10.5 & 116.5 
339: \end{tabular}
340: %\end{ruledtabular}
341: \caption{Positions $(\theta,\phi)$ and orientation angles $\alpha$ of
342:  detectors (in units of degree).   
343: \label{tab:detectors}}
344: \end{table}
345: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
346: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
347: \fi
348: 
349: %We can explicitly calculate correlation $C_{ab}$ of data streams for two
350: %detectors $a$ and $b$. 
351: For the isotropic component of the stochastic background, only the 
352: relative configuration of two detectors is relevant with the correlation 
353: signal $C_{ab}$ and we do not care about the overall rotation. 
354: Hence, the sensitivity of each pair of detectors to the stochastic background 
355:  can be characterized by the three angular parameters 
356: $(\beta,\sigma_1,\sigma_2)$ shown in Fig.~\ref{fig:detector_config}. Here, 
357: $\beta$ is the separation angle between two detectors measured 
358: from the center of the Earth. The angle $\sigma_1$ ($\sigma_2$) is the
359: orientation of the bisector of two arms for detector $a$ ($b$) 
360: measured in counter-clockwise manner relative to the great 
361: circle connecting $a$ and $b$. Their distance is given by
362: $
363: D=2R_E\sin(\beta/2)
364: $
365: ($R_E=6400$km : the radius of the Earth), which 
366: determines a characteristic frequency $f_D\equiv c/(2\pi\,D)$ for the overlap
367: functions. 
368: %We have $f_D=91$Hz, 33Hz and 23Hz
369: %for $\beta=30^\circ$, $90^\circ$ and $180^\circ$ respectively.
370: Following Ref.\cite{Flanagan:1993ix}, we define the  angles 
371: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
372: \beq
373: \Delta\equiv ({\sigma_1+\sigma_2})/2,~~~\delta\equiv ({\sigma_1-\sigma_2})/2.
374: \eeq
375: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
376: The geometrical information about pairs of detectors 
377: among the five interferometers is presented in Table \ref{tab:angles}. 
378: 
379: 
380: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
381: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Figure 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
382: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
383: \begin{figure}
384:   \begin{center}
385: \epsfxsize=4.cm
386: \begin{minipage}{\epsfxsize} \epsffile{ang.eps} \end{minipage}
387:  \end{center}
388:   \caption{  Geometrical configuration of ground-based detectors $a$ and 
389:     $b$ for the cross-correlation analysis. Detector planes are 
390:     tangential to the Earth. Two detectors 
391:  $a$ and $b$ are separated by the angle $\beta$ measured from the center
392:  of the Earth. The angles 
393:  $\sigma_1$ and $\sigma_2$ describe the orientation of bisectors of
394:  interferometers in a counter-clockwise manner relative to the great 
395:  circle joining two sites. \label{fig:detector_config}
396:  }
397: %\label{f2}
398: \end{figure}
399: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
400: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
401: 
402: 
403: 
404: 
405: 
406: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
407: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Table 2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
408: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
409: \begin{table}[!t]
410: %\begin{ruledtabular}
411: \begin{tabular}{l|c|c|c|c|c}
412:   & A &  C & H & L & V \\
413: \hline
414: \ A &* & 70.8$^{\circ}$,~-0.61 & 135.6$^{\circ}$,~-0.82 &157.3$^{\circ}$,~-0.88  
415: &121.4$^{\circ}$,~0.23 \\
416: \hline
417: \ C & -0.58,~0.81 &*  & 72.4$^{\circ}$,~1.00  &99.2$^{\circ}$,~-0.98 & 
418: 86.6$^{\circ}$,~-0.43 \\
419: \hline
420: \ H &-1.00,~-0.007  &-0.21,~0.98  & * &  27.2$^{\circ}$,~-1.00 & 
421: 79.6$^{\circ}$,~-0.43 \\
422: \hline
423: \ L &0.99,~0.15  &0.04,~-1.00  & -0.36,~-0.93   & * &76.8$^{\circ}$,~-0.29 \\
424: \hline
425: \ V &-0.45,~-0.89  & 0.92,~0.38 & -0.76,~-0.65  &0.89,~-0.46 &  *
426: \end{tabular}
427: %\end{ruledtabular}
428: \caption{Upper right $(\beta, \cos(4\delta))$. Lower left 
429: $(\cos (4\Delta),~\sin(4\Delta))$. \label{tab:angles}}
430: \end{table}
431: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
432: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
433: 
434: 
435: In the expression (\ref{gi1}), the angular integral can be performed 
436: analytically with explicit forms of the pattern functions. 
437: A long but straightforward calculation leads to \cite{Flanagan:1993ix}  
438: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
439: \beq
440: \gamma_{I,ab}= \Theta_1(y,\beta)\,\cos(4\delta)+
441: \Theta_2(y,\beta) \,\cos(4\Delta), 
442: \label{gi}
443: \eeq
444: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
445: with
446: $
447:  \Theta_1(y,\beta)=\cos^4\lmk\frac{\beta}2  \rmk \lmk j_0+\frac57
448: j_2+\frac{3}{112} j_4 \rmk ,
449: $
450: and 
451: $
452:  \Theta_2(y,\beta)=\lmk -\frac38 j_0+\frac{45}{56}
453: j_2-\frac{169}{896} j_4 \rmk
454: +\lmk \frac12 j_0-\frac57j_2-\frac{27}{224}j_4  \rmk \cos\beta \nonumber\\
455: + \lmk-\frac18 j_0-\frac5{56}j_2-\frac{3}{896}j_4  \rmk \cos(2\beta).
456: $
457: The function $j_n$ is the $n$-th spherical Bessel function with its argument
458: $
459: y=f/f_D.% ={4\pi f R_E\sin(\beta/2)}/{c}.
460: $
461: %The expression (\ref{gi}) coincides with a corresponding formula  in 
462: 
463: 
464: \cite{Flanagan:1993ix}. 
465: On the other hand, the overlap function for the $V$-mode is
466: given by 
467: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
468: \beq
469: \gamma_{V,ab}=\Theta_3(y,\beta)\,\sin(4\Delta) 
470: \label{gv}
471: \eeq
472: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
473: with
474: $
475: \Theta_3(y,\beta)=-\sin\lmk \frac{\beta}2 \rmk \lkk \lmk-j_1+\frac78
476: j_3  \rmk + \lmk j_1+\frac38 j_3 \rmk\cos\beta   \rkk.
477: $
478: In Fig.~\ref{fig:overlap},  
479: the overlap functions for the two representative pairs are 
480: shown in top (HL) and middle (CL) panels.  
481: 
482: 
483: Here, we give a simple interpretation for the angular 
484: dependence of Eqs.~(\ref{gi}) and (\ref{gv}).
485: The beam pattern functions $F^P_{a}$ and $F^P_{b}$ are given
486: by  linear combinations of $(\,\cos(2\sigma_1),~\sin(2\sigma_1)\,)$ and
487: $(\,\cos(2\sigma_2),~\sin(2\sigma_2)\,)$ respectively, reflecting their 
488: spin-2 like nature. Then, with Eq.~(\ref{gi1}) and addition formulas of
489: trigonometric functions, the overlap functions should be linear
490: combinations of $\cos[2(\sigma_1\pm \sigma_2)]$ and $\sin[2(\sigma_1\pm
491: \sigma_2)]$, namely, $\cos(4\Delta)$, $\cos(4\delta)$, $\sin(4\Delta)$ 
492: and $\sin(4\delta)$. Since the expectation value $C_{ab}(f)$ is a real
493: function for our beam 
494: pattern functions, we have $\lla s_a s_b^*\rra=\lla s_b s_a^*\rra$. This
495: essentially results in replacing the roles of $\sigma_1$ and
496: $\sigma_2$, and the functions $\gamma_I$ and $\gamma_V$ cannot contain
497: terms proportional to $\sin(4\delta)= \sin[2(\sigma_1-\sigma_2)]$.
498: 
499: On the other hand, while the observable $C_{ab}(f)$ and the amplitude $I$ 
500: are invariant under the parity transformation of a coordinate system, 
501: the sign of the parameter $V$ flips, because the transformation 
502: interchanges right-and left-handed waves. Therefore, the function 
503: $\gamma_{V,ab}$ must change its sign while keeping the quantity 
504: $C_{ab}(f)$ invariant. Geometrically,   
505: this corresponds to the re-definition of the azimuthal angles 
506: $\sigma_{1,2}$ in a clockwise direction (or putting $\sigma_1\to
507: -\sigma_1$ and $\sigma_2\to -\sigma_2$). As a result, the function
508: $\gamma_{V,ab}$ should be odd functions of $\delta$ and $\Delta$, and it
509: must be proportional to $\sin (4\Delta)$ as shown in Eq.~(\ref{gv}) (the term
510: proportional to $\sin(4\delta)$ is already prohibited as explained 
511: earlier). With similar arguments,  we find that  the 
512: function $\gamma_I$ is a linear combination of $\cos (4\Delta)$
513: and $\cos (4\delta)$ as in Eq.~(\ref{gi}). 
514: 
515: 
516: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
517: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
518: \subsection{Special cases}
519: %\underline{\em A)  Special cases}
520: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
521: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
522: 
523: 
524: To stress the importance of the geometric configuration,  
525: it is instructive to consider several simple examples for idealistic pair of 
526: detectors. When a pair of detectors is co-located ($\beta=0^{\circ}$ 
527: and $D=0$), the functions $(\Theta_1,\Theta_2)$ defined after Eq.~(\ref{gi}) 
528:  become $(1,0)$ and we have $\gamma_{I,ab}=\cos(4\delta)$. 
529: The identity $\Theta_2=0$ at $\beta=0^{\circ}$ implies that 
530: the function $\gamma_I$ depends very weakly on the parameter 
531: $\Delta$ at a small angle $\beta \sim 0^{\circ}$. 
532: On the other hand, the overlap function $\gamma_{V,ab}$ always 
533: vanishes for a pair of detectors in the same plane $(\beta=0)$. 
534: This is even true with a finite separation $D\ne0$.  
535: This exact cancellation comes from 
536: the geometric symmetry of the beam pattern function with respect 
537: to the detector plane \cite{Seto:2006hf,Kudoh:2005as}. 
538: 
539: 
540: For two detectors at antipodal positions ($\beta=180^\circ$), we have
541: $\Theta_1=0$ and the angle $\delta$ becomes geometrically 
542: meaningless. One can expect that the function 
543: $\gamma_{I,ab}$ is almost proportional to $\cos(4\Delta)$ near $\beta=180^\circ $. 
544: 
545: 
546: 
547: 
548: 
549: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
550: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Figure 2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
551: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
552: \begin{figure}[!tb]
553: \begin{center}
554: \epsfxsize=6.cm
555: \begin{minipage}{\epsfxsize} \epsffile{red.eps} \end{minipage}
556:  \end{center}
557: \caption{Overlap  functions for the un-polarized $I$ mode (dashed
558:  curves), and the circularly polarized $V$-mode (solid curves). The
559:  upper panel shows 
560:  the results for the Hanford-Livingston (HL) pair (the characteristic
561:  frequency $f_D=100$Hz). The middle one
562:  is results for the LCGT-Livingston (CL) pair ($f_D=31$Hz). The
563:  normalized SNRs ${\it S}_{I,V}$ (with the adv LIGO noise spectrum) are 
564: also presented.  The bottom one show the compiled
565:  functions $\Gamma_{I,V}$ (eq.(\ref{co})) made from  both pairs.  
566: \label{fig:overlap} }
567: \end{figure}
568: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
569: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
570: 
571: 
572: 
573: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
574: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
575: \subsection{Broadband SNR}
576: %\underline{\em B)  Broadband SNR}
577: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
578: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
579: 
580: 
581: Now, we turn to focus on a broadband sensitivity to the $I$- and $V$-modes. 
582: In the weak signal limit, the total signal-to-noise ratio (SNR) 
583: for the correlation signal $C_{ab}(f)$ is given by \cite{Flanagan:1993ix}
584: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
585: \beq
586: {\rm SNR}^2=\lmk\frac{3H_0^2}{10\pi^2}  \rmk^2 T_{\rm obs} \lkk  
587: 2\int_0^\infty df
588: \frac{X^2}{f^6 N_a(f)N_b(f)}  \rkk \label{broad}
589: \eeq
590: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
591: with $X=\gamma_{I}\,\Omega_{\rm\scriptscriptstyle GW}+
592: \gamma_{V}\,\Omega_{\rm\scriptscriptstyle GW}\Pi$. We denote 
593: the noise spectra for detectors $a$ and $b$ by $N_a(f)$ and $N_b(f)$, 
594: assuming no noise correlation between them. 
595: In what follows, for simplicity of our analysis, 
596: we further assume that 
597: all the detectors have the same sensitivity comparable to 
598: the noise spectral curves of  advanced LIGO. The analytical fit 
599: from Fig.~1 of Ref.\cite{adv} leads to 
600: $
601: N(f)= 10^{-44} \lmk {f }/{\rm 10 Hz} \rmk^{-4}+10^{-47.25 }
602: \lmk {f }/{\rm 10^2 Hz} \rmk^{-1.7} {\rm Hz^{-1}}$ for $10\,{\rm Hz}\le
603: f \le 240\,{\rm Hz}$,  $N(f)=10^{-46} \lmk {f }/{\rm 10^3 Hz}
604: \rmk^{3}    ~~ {\rm Hz^{-1}}$ for  $240\,{\rm Hz}\le
605: f \le 3,000\,{\rm Hz}$, and otherwise $N(f)=\infty$. Note that 
606: the combination $f^6 N(f)^2$ becomes 
607: minimum around $f\sim 50$Hz with its bandwidth $\Delta f\sim 100$Hz. 
608: For a pair of coincident detectors 
609: (i.e., $\gamma_{I,ab}=1$ and $\gamma_{V,ab}=0$), 
610: the total SNR is evaluated by setting 
611: $X=\Omega_{\rm\scriptscriptstyle GW}$ in Eq.~(\ref{broad}), and we 
612: numerically obtain $
613: {\rm SNR}_0=4.8\,\lmk {T_{\rm obs}}/{3\,\rm yr}  \rmk^{1/2}\lmk
614: {\Omega_{\rm\scriptscriptstyle GW}h_{70}^2}/{10^{-9}}  \rmk.  
615: $
616: 
617: 
618: The total SNR depends strongly on model parameters of the background, 
619: including the polarization degree $\Pi$. 
620: In order to present our numerical results concisely, we first calculate 
621: ${\rm SNR}_{\{I,V\},ab}$ by plugging $X=\gamma_{\{I,V\},ab}$ into 
622: Eq.~(\ref{broad}) and then normalize them as 
623: ${\it S}_{\{I,V\},ab}\equiv {\rm SNR}_{\{I,V\},ab}/{\rm SNR}_0$.
624: The normalized SNRs can be regarded as  rms values of 
625: $\gamma_{\{I,V\},ab}$ with a weight function $[f^6\, N(f)^2]^{-1}$.
626: 
627: 
628: 
629: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
630: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
631: \subsection{Optimal configuration}
632: %\underline{\em C)  Optimal configuration}
633: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
634: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
635: 
636: 
637: Let us discuss optimal configurations of two detectors $(a,b)$ for
638: measuring the $I$- and $V$-modes with the correlation signal $C_{ab}$. 
639: There are two relevant issues here: maximization of the 
640: signals ${\it S}_{I,ab}$ and ${\it S}_{V, ab}$, 
641: and switching off either of them (${\it S}_{I,ab}=0$ or 
642: ${\it S}_{I,ab}=0$) for their decomposition. To deal with the situation
643: comprehensively, we consider how to set the second detector  $b$ relative to
644: the fixed first one $a$ with a given separation angle $\beta$. 
645: In this case, the sensitivity to the $I$- and $V$-modes 
646: is characterized by the remaining adjustable parameters, 
647: $\sigma_1$ and $\sigma_2$. 
648: The former determines the position of the detector $b$, while the latter 
649: specifies its orientation (see Fig.~\ref{fig:detector_config}). 
650: Based on the expressions (\ref{gi}) and (\ref{gv}), 
651: one finds that there are three possibilities for the optimal detector 
652: orientation:  
653: $\cos (4\Delta)=-\cos (4\delta)=\pm 1$ (type I) or $\cos (4\Delta)=\cos 
654: (4\delta)=\pm 1$ (type II) to maximize the normalized 
655: SNR ${\it S}_{I,ab}$ \cite{Flanagan:1993ix}, and   
656: $\cos{(4\Delta)}=\cos{(4\delta)}=0$ (type III) 
657: to erase the contribution from $I$-mode. For type I, the solutions 
658: of the two angles $\sigma_{1,2}$ are $\sigma_1=\sigma_2=45^\circ$ 
659: (mod $90^\circ$) and the detector $b$ must be sited in one 
660: of the two great circles passing through the detector $a$, 
661: parallel to one of the two arms. 
662: For type II, the second detector must reside in two great 
663: circles parallel or perpendicular to the bisecting line of 
664: each detector. Similarly, the type III configuration is 
665: realized by placing the second detector on one of the 
666: four great circles defined for types I and II, with 
667: rotating $45^{\circ}$ relative to the first detector. 
668: 
669: 
670: 
671: Note that the sensitivity to the $V$-mode is automatically switched off 
672: for the type I and II configurations and is conversely maximized for 
673: the type III configuration. This is because the normalized SNR 
674: ${\it S}_{V,ab}$ is proportional to $\sin (4\Delta)$. 
675: While a definite detection of a weak $V$-mode signal requires  
676: a careful removal of the $I$-mode signal from observed 
677: data, it turns out that the geometrical requirement for type III 
678: configuration is severe. 
679: As we see later, however, we can easily control the contribution 
680: from the $I$- (or $V$-)mode by introducing a third detector. 
681: 
682: 
683: In Fig.~\ref{fig:SNR}, we present the normalized SNRs 
684: for the optimal geometries; types I, II and III (short-dashed,
685: long-dashed, and solid curves, respectively). One noticeable point is
686: that  
687: a widely separated  ($\beta\sim 180^\circ$) pair is powerful to search 
688: for the $V$-mode (recall the cancellation $\gamma_V=0$ at $\beta=0$). To
689: reduce the contribution from the $I$-mode,  
690: pairs that are usually disadvantageous to measuring the total intensity 
691: $\Omega_{\rm\scriptscriptstyle GW}$ now play a very  
692: important role. In Fig.~\ref{fig:SNR}, we also show the 
693: normalized SNRs for representative pairs made from the five detectors,  
694: in which several interesting combinations are found. 
695: The HL (with $\cos(4\delta)\sim 1$ and $\sin(4\Delta)\sim 0.93$)
696: realizes nearly maximum values simultaneously for ${\it S}_{I,ab}$ and 
697: ${\it S}_{V,ab}$ at its separation $\beta=27.2^\circ$. 
698: This is because ${\it S}_{I,ab}$ is mainly determined by the angle 
699: $\delta$ at a small  $\beta$, while ${\it S}_{V,ab}$ depends only on 
700: $\Delta$. 
701: %We have $\cos 4\Delta\sim 1$ and $\sin 4\Delta\sim 0.93$ for HL (see
702: %Table 2). 
703: The CL has good sensitivity to the $V$-mode and relatively insensitive to 
704: the $I$-mode with $\sin(4\Delta)\sim1$. 
705: In contrast, AH is almost insensitive to the $V$ mode with 
706: $\sin 4\Delta=-0.007$. In this sense, 
707: LCGT and AIGO detectors are suitably oriented
708: to probe the $I$- and $V$-modes, respectively. 
709: 
710: 
711: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
712: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Figure 3 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
713: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
714: \begin{figure}[!bth]
715: \begin{center}
716: \epsfxsize=6.cm
717: \begin{minipage}{\epsfxsize} \epsffile{nsr.eps} \end{minipage}
718:  \end{center}
719: \caption{ Normalized signal to noise ratios (${\it S}_{I,ab}$ and 
720: ${\it S}_{V,ab}$) with optimal configurations  
721: for the $I$-mode (short dashed curve: type I, long dashed curve: type II)
722:  and for  the $V$-mode (solid curve: type III with setting $\Pi=1$ for
723:  illustrative purpose). We use the noise curve for
724:  the advanced LIGO.   For each detector pair,
725:  ${\it S}_I$ and ${\it S}_V$ are given with a  triangle and a  circle
726:  respectively at its separation $\beta$. There are four other pairs not
727:  shown here; CH
728:  with $(S_I, S_V)=(0.04,0.08)$, LV with (0.08,0.04), HV with (0.07,0.06)
729:  and CV with (0.09,0.04).
730: \label{fig:SNR}} 
731: \end{figure}
732: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
733: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
734: 
735: 
736:  
737: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
738: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
739: \subsection{Separating $I$- and $V$-modes}
740: %\underline{\em D)  Separating $I$- and $V$-modes}
741: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
742: %-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-%-
743: 
744: 
745: As a final mention, we will address the issue of $I$- and $V$-mode separation 
746: by combining several pairs of detectors. 
747: For preliminary investigation, we consider the case that two pairs of 
748: interferometers 
749: $(a,b)$ and $(c,d)$ are available. Detectors $a$ and $c$ can be identical, 
750: but we need at least three independent detectors for the study below. 
751: First note that the correlation signals are given by 
752: $C_{ab}(f)=\gamma_{I,ab}(f)\,I(f)+\gamma_{V,ab}(f)\,V(f)$ and 
753: $C_{cd}(f)=\gamma_{I,cd}(f)\,I(f)+\gamma_{V,cd}(f)\,V(f)$. 
754: From this, one can easily find that the contribution from the 
755: $I$-mode is canceled by taking a 
756: combination $W\equiv\gamma_{I,ab}\,C_{cd}-\gamma_{I,cd}\,C_{ab}=
757: (\gamma_{V,cd}\,\gamma_{I,ab}-\gamma_{V,ab}\,\gamma_{I,cd})\,V(f)$. 
758: The statistical analysis based on the combination $W$ would be a robust 
759: approach for actual $V$-mode search, although a further 
760: refinement may be possible by combining more pairs, 
761: which we will report elsewhere. 
762: 
763: Since the rms amplitude of the detector noise for the combination $W$
764: becomes $N(f)\,(\gamma_{I,ab}^2+\gamma_{I,cd}^2)^{1/2}$, we define the 
765: {\it compiled} overlap function for the $V$ mode by 
766: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
767: \beq
768: \Gamma_{V,ab:cd}\equiv
769: \frac{\gamma_{V,cd}\gamma_{I,ab}-\gamma_{V,ab}\gamma_{I,cd}}{[\gamma_{I,ab}^2 
770: +\gamma_{I,cd}^2]^{1/2}}.\label{co}
771: \eeq
772: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
773: This expression should be used  in Eq.~(\ref{broad}) when evaluating 
774: the broadband SNR for the $V$-mode with the combination $W$. 
775: %This function is proportional to the area of a triangle shaped by two
776: %vectors 
777: %$(\gamma_{I,ab},\gamma_{I,cd})/[\gamma_{I,ab}^2+\gamma_{I,cd}^2]^{1/2}
778: %$ and  $(\gamma_{V,ab},\gamma_{V,cd})$ with  $|\Gamma_{V,ab:cd}|\le 
779: %[\gamma_{V,ab}^2 %+\gamma_{V,cd}^2]^{1/2}$. 
780: %This sounds reasonable since the
781: %right-hand-side represents the simple summation of two independent
782: %sensitivities. 
783: In a similar way, we define the compiled function 
784: $\Gamma_{I,ab:cd}$ for the $I$ mode by 
785: interchanging the subscripts $V$ and $I$ in Eq.~(\ref{co}). 
786: Bottom panel of Fig.~\ref{fig:overlap} shows the compiled overlap functions 
787: $\Gamma_{\{I,V\},ab:cd}$ from two pairs of detectors, CL-HL.  
788: % ???Based on the
789: %geometric interpretation for the compiled functions, we intend to use CL
790: %to mainly extract the $V$ mode and HL for the $I$ mode.??? 
791: With this combination, the normalized SNR becomes $0.11$ for the $V$-mode 
792: and $0.31$ for the $I$-mode. 
793: Using numerical results below eq.(\ref{broad}), the detection limit for
794: the 
795: polarization degree $\Pi$  is given as
796: $\Pi= (T/3{\rm yr})^{-1/2}(SNR_V/5)(\Omega_{\rm\scriptscriptstyle
797: GW}h_{70}^2/10^{-8})^{-1}$ with signal-to-noise ratio $SNR_V$. 
798: These numerical results are almost the same 
799: values as in ${\it S}_V$ for CL and ${\it S}_I$ for HL, and in this
800: sense,  
801: the $I$-, $V$-mode separation can be performed efficiently with 
802: naively expected sensitivities ${\it S}_{\{I,V\},ab}$. 
803: %Note that the sensitivity to 
804: %the $V$-mode by the HL-CL network is about three times worse than 
805: %the $I$-mode probed by HL. 
806: Note  that the other combinations, such as AV-HL, AV-HV and CL-HV, 
807: also provide the normalized value $\sim0.11$ for the $V$-mode, 
808: but AH-AL has only $0.015$.  
809: 
810: 
811: In summary, we  reported principle aspects for measuring a
812: circular polarization degree of a gravitational wave background that is
813: related to parity violation. We find that pairs of ground-based 
814: interferometers that are widely separated and relatively insensitive to the 
815: total intensity $\Omega_{\rm\scriptscriptstyle GW}$ are advantageous 
816: for the measurement. With at 
817: least three detectors, the polarization degree and the intensity 
818: $\Omega_{\rm\scriptscriptstyle GW}$ can be separately detected. 
819: 
820: 
821: We would like to thank N. Kanda and M. Ando for supplying information on LCGT 
822: and comments. This work was supported in part by  a Grant-in-Aid for Scientific Research from the Japan Society for the 
823: Promotion of Science (No.~18740132). 
824: 
825: %\include{ref}
826: \begin{thebibliography}{DUM}
827: 
828: 
829: \if0
830: %\cite{Allen:1996vm}
831: \bibitem{Allen:1996vm}
832:   M.~Maggiore,
833:   %``Gravitational wave experiments and early universe cosmology,''
834:   Phys.\ Rept.\  {\bf 331}, 283 (2000);
835:   %%CITATION = GR-QC 9909001;%%
836: B.~Allen,
837:   %``The stochastic gravity-wave background: Sources and detection,''
838:   arXiv:gr-qc/9604033.
839:   %%CITATION = GR-QC 9604033;%%
840: \fi
841: 
842: %\cite{Abbott:2005ez}
843: \bibitem{Abbott:2005ez}
844:   B.~Abbott { et al.}  
845:   %``Upper limits on a stochastic background of gravitational waves,''
846:   Phys.\ Rev.\ Lett.\  {\bf 95}, 221101 (2005).
847:   %%CITATION = PRLTA,95,221101;%%
848: 
849: 
850: \bibitem{adv}E.  Gustafson et al. 1999,
851: LIGO project document
852: T990080-00-D.
853: 
854: %\cite{Kuroda:1999vi}
855: \bibitem{Kuroda:1999vi}
856:   K.~Kuroda {\it et al.} 
857:   %``Large-scale cryogenic gravitational wave telescope,''
858:   Int.\ J.\ Mod.\ Phys.\  D {\bf 8}, 557 (1999).
859:   %%CITATION = IMPAE,D8,557;%%
860: 
861: 
862: 
863: %\cite{Alexander:2004us}
864: \bibitem{Alexander:2004us}
865:   S.~H.~S.~Alexander et al.
866:   %``Leptogenesis from gravity waves in models of inflation,''
867:   Phys.\ Rev.\ Lett.\  {\bf 96}, 081301 (2006); M.~Satoh, S.~Kanno and
868:  J.~Soda, 
869:   arXiv:0706.3585 [astro-ph].
870:   %%CITATION = ARXIV:0706.3585;%%
871: 
872: 
873:  % [arXiv:hep-th/0403069].
874:   %%CITATION = PRLTA,96,081301;%%
875: 
876: 
877: 
878: 
879: 
880: 
881: 
882: 
883: %\cite{Lue:1998mq}
884: \bibitem{Lue:1998mq}
885:   A.~Lue, L.~M.~Wang and M.~Kamionkowski,
886:   %``Cosmological signature of new parity-violating interactions,''
887:   Phys.\ Rev.\ Lett.\  {\bf 83}, 1506 (1999).
888: 
889: %\cite{Caprini:2003vc}
890: \bibitem{Caprini:2003vc}
891:   C.~Caprini, R.~Durrer and T.~Kahniashvili,
892:   %``The Cosmic Microwave Background and Helical Magnetic Fields: the tensor
893:   %mode,''
894:   Phys.\ Rev.\  D {\bf 69}, 063006 (2004);
895:   %%CITATION = PHRVA,D69,063006;%%
896: S.~Saito, K.~Ichiki and A.~Taruya,
897:   arXiv:0705.3701 [astro-ph].
898: 
899: 
900: 
901: 
902: 
903: %\cite{Seto:2006hf}
904: \bibitem{Seto:2006hf}
905:   N.~Seto,
906:   %``Prospects for direct detection of circular polarization of
907:   %gravitational-wave background,''
908:   Phys.\ Rev.\ Lett.\  {\bf 97}, 151101 (2006); 
909:   %%CITATION = PRLTA,97,151101;%%
910:  Phys.Rev.D {\bf 75}, 0601302 (2007).
911: 
912: 
913: 
914: 
915: 
916: 
917: 
918: 
919: % \cite{Flanagan:1993ix,Allen:1997ad}
920: %\cite{Flanagan:1993ix}
921: \bibitem{Flanagan:1993ix}
922:   E.~E.~Flanagan,
923:   %``The Sensitivity of the laser interferometer gravitational wave observatory
924:   %(LIGO) to a stochastic background, and its dependence on the detector
925:   %orientations,''
926:   Phys.\ Rev.\ D {\bf 48}, 2389 (1993).
927: %  [arXiv:astro-ph/9305029];
928:   %%CITATION = ASTRO-PH 9305029;%%
929: 
930: 
931: %\cite{Allen:1997ad}
932: \bibitem{Allen:1997ad}
933:   B.~Allen and J.~D.~Romano,
934:   %``Detecting a stochastic background of gravitational radiation: Signal
935:   %processing strategies and sensitivities,''
936:   Phys.\ Rev.\  D {\bf 59}, 102001 (1999).
937:   %%CITATION = PHRVA,D59,102001;%%
938: 
939: 
940: 
941: \bibitem{radipro}
942: G. B. Rybicki and A. P. Lightman, {\it Radiative Process in
943:  Astrophysics} (Wiley, New York, 1979). 
944: 
945: \bibitem{det}
946: N. Arnaud et al. Phys.Rev.D {\bf 65}, 042004 (2002).
947: 
948: 
949: %\cite{Kudoh:2005as}
950: \bibitem{Kudoh:2005as}
951: H.~Kudoh and A.~Taruya,
952:   %``Probing anisotropies of gravitational-wave backgrounds with a  space-based
953:   %interferometer: Geometric properties of antenna patterns  and their angular
954:   %power,''
955:   Phys.\ Rev.\  D {\bf 71}, 024025 (2005).
956:   %%CITATION = PHRVA,D71,024025;%%
957: 
958: 
959: 
960: 
961: 
962: 
963: 
964: \end{thebibliography}
965: 
966: 
967: 
968: 
969: \end{document}
970: