1: \documentclass[aps,float,prl,psfig]{revtex4}
2: %\documentclass[aps,twocolumn,float,prl,psfig]{revtex4}
3: \input epsf
4: \usepackage{graphicx}% Include figure files
5: \newcommand{\beq}{\begin{equation}}
6: \newcommand{\beqa}{\begin{eqnarray}}
7: \newcommand{\eeq}{\end{equation}}
8: \newcommand{\eeqa}{\end{eqnarray}}
9: \newcommand{\etal}{{\it et al. }}
10: \newcommand{\siml}{\lesssim}
11: \newcommand{\simg}{\gtrsim}
12: \newcommand{\lsim}{\lesssim}
13: \newcommand{\gsim}{\gtrsim}
14: \newcommand{\psim}{\mbox{\raisebox{-1.0ex}{$~\stackrel{\textstyle \propto}
15: {\textstyle \sim}~$ }}}
16: \newcommand{\vect}[1]{\mbox{\boldmath${#1}$}}
17: \newcommand{\lmk}{\left(}
18: \newcommand{\rmk}{\right)}
19: \newcommand{\lnk}{\left\{ }
20: \newcommand{\nn}{\nonumber}
21: \newcommand{\rnk}{\right\} }
22: \newcommand{\lkk}{\left[}
23: \newcommand{\rkk}{\right]}
24: \newcommand{\lla}{\left\langle}
25: \newcommand{\p}{\partial}
26: \newcommand{\rra}{\right\rangle}
27: \newcommand{\so}{M_\odot}
28: \newcommand{\mch}{{\cal M}}
29: \newcommand{\vex}{{\vect x}}
30: \newcommand{\ver}{{\vect r}}
31: \newcommand{\vue}{\hat{\vect e}}
32: \newcommand{\vel}{\vect l}
33: \newcommand{\vez}{\vect z}
34: \newcommand{\ven}{\vect n}
35: \newcommand{\vem}{\vect m}
36: \newcommand{\vep}{{\vect p}}
37: \newcommand{\veq}{{\vect q}}
38: \newcommand{\veo}{{\vect \Omega}}
39: \newcommand{\ve}{{\vect e}}
40:
41: \begin{document}
42: \title{
43: Prospects for direct detection of circular polarization of
44: gravitational-wave background
45: }
46: \author{Naoki Seto
47: }
48: \affiliation{Department of Physics and Astronomy, 4186 Frederick Reines
49: Hall, University of California, Irvine, CA 92697
50: }
51:
52: \begin{abstract}
53:
54: We discussed prospects for directly detecting circular
55: polarization signal of gravitational wave background. We found it is
56: generally difficult to probe the monopole mode of the signal due to
57: broad directivity of gravitational wave detectors.
58: But the dipole $(l=1)$ and octupole $(l=3)$ modes of the signal can be
59: measured in a simple manner by combining outputs of two unaligned
60: detectors, and we can dig them deeply under confusion and detector
61: noises. Around $f\sim 0.1$mHz LISA will provide ideal data
62: streams to detect these anisotropic components whose magnitudes are as
63: small as $\sim1$ percent of the detector noise level in
64: terms of the non-dimensional energy
65: density $\Omega_{GW}(f)$.
66: \end{abstract}
67: \pacs{PACS number(s): 95.55.Ym 98.80.Es, 95.85.Sz }
68: \maketitle
69:
70: %\baselineskip 8mm
71: \section{introduction}
72: As gravitational interaction is very weak, significant efforts have
73: been made
74: to detect gravitational waves. But, on the other hand, we will be able
75: to get rich
76: information of the universe by observing gravitational waves that
77: directly propagate
78: to us with almost no absorption. Various astrophysical and cosmological
79: models predict existence of stochastic gravitational wave background,
80: and it is an interesting target for gravitational wave astronomy
81: \cite{Allen:1996vm}. For
82: its observational prospects, we need to understand how we characterize the background
83: and what aspects we can
84: uncover with current and future observational facilities.
85:
86: One of such aspects is circular polarization that describes whether the
87: background has asymmetry with respect to magnitudes of right- and
88: left-handed waves. Circular polarization of gravitational wave
89: background might be generated by helical
90: turbulent motions (see {\it e.g.} \cite{Kahniashvili:2005qi} and
91: references therein). Inflation
92: scenario predicts gravitational wave background
93: from quantum fluctuations during acceleration phase in the early
94: universe, but asymmetry of left and right-handed waves can be produced
95: with the gravitational Chern-Simon term that might be derived from
96: string theory and might be related to
97: creation of baryon number (see {\it e.g.} \cite{Lyth:2005jf} and
98: references therein). Primordial gravitational wave background, including
99: information of its
100: circular polarization \cite{Lue:1998mq}, can be indirectly
101: studied with CMB measurement \cite{Seljak:1996gy} at very low frequency
102: regime $f\sim
103: 10^{-17}$Hz that is largely different from the regime directly
104: accessible with
105: gravitational wave detectors studied in this Letter. Gravitational wave
106: background from
107: Galactic binaries can be polarized, if orientation of their angular
108: momentum have coherent distribution, such as, correlation with Galactic
109: structure. While observational samples of local Galactic binaries
110: do not
111: favor such correlation \cite{inc}, this will also be an interesting
112: target that can be directly studied with LISA.
113:
114: It is well known that observation of gravitational wave is
115: intrinsically sensitive to its polarization state \cite{Thorne_K:1987}.
116: This is because we measure spatial expansion and
117: contraction due to the wave, and
118: polarization determines the direction of the oscillation perpendicular
119: to its propagation direction. Another
120: important nature of the observation is that we have to simultaneously
121: deal with waves basically coming from all the directions, in contrast
122: to observations with typical electromagnetic wave telescopes that have
123: sharp directivity. Therefore, polarization information and directional
124: information couple strongly in observational analysis of gravitational
125: wave background. In general, orientation of a gravitational wave
126: detector
127: changes with time, and induced modulation of data stream is useful to
128: probe the polarization and directional information. For ground based detectors, such as, LIGO, this is
129: due to the daily rotation of the Earth \cite{Allen:1996gp}. For a space mission like LISA,
130: this
131: change is determined by its orbital choice \cite{lisa,Cornish:2001qi}. In addition, we
132: can also
133: expect that
134: several independent data streams of gravitational waves will be taken at
135: the same time
136: \cite{Cutler:1997ta,Prince:2002hp}. In this
137: Letter, in view of these observational characters, we study how well we
138: can
139: extract information of circular
140: polarization of the background
141: in a model independent manner about its origin.
142: \section{formulation}
143:
144:
145: The standard plane wave expansion of metric perturbation by
146: gravitational waves is given as
147: \beq
148: h_{ab}(t,\vex)=\sum_{P=+,\times} \int^{\infty}_{-\infty} df \int_{S^2} d\ven
149: h_P(f,\ven) e^{2\pi i f (t-\ven \cdot \vex) } \ve^P_{ab}(\ven),
150: \eeq
151: where $S^2$ is the unit sphere for the angular integral, the unite vector
152: $\ven=(\sin\theta\cos\phi, \sin\theta\sin\phi,\cos\theta)$ is
153: propagation direction, and $\ve^+_{ab}$ and $\ve^\times_{ab}$ are the basis for the transverse-traceless tensor. We fix them as
154: $\ve^+_{ab}={\hat \ve}_\theta \otimes {\hat \ve}_\theta- {\hat \ve}_\phi
155: \otimes {\hat \ve}_\phi$ and $\ve^\times_{ab}={\hat \ve}_\theta \otimes
156: {\hat
157: \ve}_\phi+{\hat
158: \ve}_\phi \otimes {\hat \ve}_\theta$ where ${\hat \ve}_\theta$ and ${\hat
159: \ve}_\phi$ are two unit vectors with a fixed spherical coordinate
160: system.
161: As the metric perturbation $h_{ab}(t,\vex)$ is real, we have a relation
162: for complex conjugate;
163: $h_P(-f,\ven)=h_P(f,\ven)^*$.
164: When we replace the direction $\ven\to -\ven$, the matrices have
165: correspondences $ \ve^+_{ab} \to \ve^+_{ab}$ (even parity) and
166: $\ve^\times_{ab}\to
167: -\ve^\times_{ab}$ (odd parity).
168:
169: To begin with, we study gravitational wave modes at a
170: frequency $f$. Here,
171: we omit explicit frequency
172: dependence for notational simplicity, unless we need to keep it. The
173: covariance matrix
174: $\left(
175: \begin{array}{@{\,}cc@{\,}}
176: \lla h_+(\ven) h_+^* (\ven') \rra & \lla h_+ (\ven)
177: h_\times^* (\ven') \rra \\
178: \lla h_+^*(\ven) h_\times(\ven')\rra & \lla h_\times(\ven)
179: h_\times^*(\ven')\rra \\
180: \end{array} \right)
181: $
182: for two polarization modes $h_+(\ven)$ and $h_\times(\ven)$
183: is decomposed as
184: \beq
185: \frac{\delta_{drc}(\ven-\ven')}{4\pi}\left(
186: \begin{array}{@{\,}cc@{\,}}
187: I+Q & U-iV \\
188: U+iV & I-Q \\
189: \end{array} \right),
190: \eeq
191: where the symbol $\lla \dots\rra$ represents to take an ensemble average
192: for superposition of stationary incoherent waves, and
193: $\delta_{drc}(\cdot)$ is the delta
194: function on the unit sphere $S^2$.
195: \if0
196: The Pauli
197: matrices are %basis for $2\times2$ matrices and
198: written as
199: $
200: \sigma_0=\left(
201: \begin{array}{@{\,}cc@{\,}}
202: 1 & 0 \\
203: 0 & 1 \\
204: \end{array} \right),~
205: \sigma_1=\left(
206: \begin{array}{@{\,}cc@{\,}}
207: 1 & 0 \\
208: 0 & -1 \\
209: \end{array} \right),~
210: \sigma_2=\left(
211: \begin{array}{@{\,}cc@{\,}}
212: 0 & 1 \\
213: 1 & 0 \\
214: \end{array} \right)
215: $ and $ \sigma_3=\left(
216: \begin{array}{@{\,}cc@{\,}}
217: 0 & -i \\
218: i & 0 \\
219: \end{array} \right).
220: $
221: \fi
222: We have defined the following Stokes parameters \cite{radipro} analog to
223: electromagnetic
224: waves
225: as
226: $
227: I(\ven)= \lla |h_+|^2+|h_\times|^2\rra /2, ~~Q(\ven)=\lla |h_+|^2-|h_\times|^2\rra
228: /2, ~~U(\ven)= \lla h_+ h_\times^*+ h_+^* h_\times \rra /2, $ and
229: $V(\ven)=i \lla h_+
230: h_\times^*- h_+^* h_\times \rra /2.
231: $
232: The parameter $I(\ven)$ represents the total intensity of the wave, while $Q(\ven)$
233: and $U(\ven)$ is related to linear polarization. The parameter $V(\ven)$ is related
234: to circular
235: polarization, and its sign shows whether right- or left-handed
236: waves dominate. When we rotate the basis vectors ${\hat \ve}_\theta
237: $ and $
238: {\hat \ve}_\phi$ around the axis $\ven$ by angle $\psi$, the parameters
239: $I$
240: and $V$ are invariant (spin 0). But the parameters $Q$
241: and $U$ are transformed as
242: $
243: (Q\pm iU)'(\ven)=e^{\mp 4i \psi}(Q\pm iU)(\ven),
244: $
245: and the combinations $(Q\pm iU)(\ven)$ have spin $\pm 4$
246: \cite{radipro}.
247: %Note that the
248: %combinations $(Q\pm iU)(\ven)$ of the stokes parameters for
249: %electromagnetic waves have spin $\pm
250: %2$ \cite{Seljak:1996gy,radipro}.
251:
252: Next we discuss response of a two-arm interferometer $J$ with
253: $90^\circ$ vertex angle and equal arm-length $L$. We put $\vel_1$ and
254: $\vel_2$ as unit vectors for the directions of the arms. The beam pattern
255: functions $F_J^P=(\vel_1\cdot \ve^P\cdot \vel_1-\vel_2\cdot \ve^P\cdot \vel_2)/2$ $(P=+,\times)$
256: represent relative sensitivities to two linearly polarized gravitational
257: waves
258: ($P=+,\times$) with various directions $\ven$
259: \cite{Thorne_K:1987}. Note that the function $F_J^+$ has even
260: parity
261: and $F_J^\times$ has odd parity with respect to the
262: direction $\ven$ (see {\it e.g.} \cite{Kudoh:2004he}).
263:
264: In this Letter we mainly deal with low frequency regime with
265: $f/f_*\ll 1$. Here, using the arm-length $L$, we have defined a characteristic frequency
266: $f_*=1/(2\pi
267: L)$ that corresponds to 10mHz for LISA and 1Hz for Big Bang Observer
268: (BBO) \cite{bbo,Seto:2005qy}.
269: LISA is formed by three spacecrafts that nearly keep a regular triangle
270: configuration \cite{lisa}. From its six one-way data streams, we can make
271: Time-Delay-Interferometer (TDI) variables that cancel laser frequency
272: noises. We can select three TDI variables $A$, $E$ and $T$
273: whose detector noises are not correlated and can be regarded as
274: independent \cite{Prince:2002hp}. At the low frequency regime the
275: responses of the $A$ and $E$ modes can be effectively regarded as those of
276: two-arm interferometers whose configuration are shown in figure 1
277: \cite{Cutler:1997ta}. In
278: this figure we
279: put the whole system on $XY$-plane ($\theta=\pi/2$). The beam pattern
280: functions of the $A$
281: mode are explicitly given as
282: \beq
283: F_A^+(\theta,\phi)=\frac12 (1+\cos^2\theta) \cos2\phi,~~~
284: F_A^\times(\theta,\phi)=-\cos\theta \sin2\phi,\label{f2}
285: \eeq
286: while those for the $E$ mode are given by replacing $\phi\to\phi-\pi/4$
287: in the
288: above expressions.
289: The beam pattern functions for the $T$ mode are quite different from the
290: $A$
291: and
292: $E$ modes, and given as $F_T^P=\sum_{i=1}^3 (\vem_i\cdot \ve^P\cdot
293: \vem_i)(\vem_i\cdot \ven)$ $(P=+,\times)$ where directions of three unit
294: vectors
295: $\vem_i$ are shown in figure 1 \cite{Kudoh:2004he,Tinto:2001ii}. At low
296: frequency regime
297: sensitivity of the $T$ mode to gravitational waves is $\sim (f/f_*)^{-1}$ times worse than those
298: for the $A$ and $E$ modes \cite{Prince:2002hp,Tinto:2001ii}. Therefore we put this mode aside for a while.
299:
300: We can express responses of the $A$ and $E$ modes to gravitational
301: waves from a single direction $\ven$ as $(J=A,E)$
302: \beq
303: r_J(\ven) = \lmk F_J^+ h_+ (\ven)+ F_J^\times h_\times
304: (\ven) \rmk \lkk
305: 1+i(f/f_*)D_J(\ven)+O(f^2/f_*^2)\rkk, \label{ha}
306: \eeq
307: where the second factor $\lkk\dots \rkk$ is corrections caused by
308: the
309: finiteness
310: of the arm-length, and two real functions $D_A(\theta,\phi)$ and
311: $D_E(\theta,\phi)$ depend on the propagation directions $\ven$ of waves
312: (see {\it e.g.} \cite{Kudoh:2004he}).
313: In eq.(\ref{ha}) we neglected
314: overall factors that depend only on frequency $f$ and are irrelevant for
315: our
316: study. These formal expressions for the perturbative expansion with
317: respect to the ratio $(f/f_*)$ can be generally used with relevant beam
318: pattern functions, including the
319: case for the $T$-mode or Fabri-Perot detectors as LIGO.
320:
321: %\end{document}%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
322:
323:
324: The information $V$ cannot be produced from the response $r_A$ or $r_E$
325: alone.
326: To get it
327: we need independent linear combination of $h_+$ and
328: $h_\times$.
329: For example the term $ \lla r_A(\theta,\phi) r_A(\theta,\phi)^* \rra$
330: can
331: be
332: written only with $I,Q$ and $U$. We now take
333: the low frequency limit,
334: and keep only the leading order terms of the expansions. The cross term
335: $\lla r_A(\ven)
336: r_E(\ven')^*\rra$ is given as
337: \begin{widetext}
338: \beqa
339: \lla r_A(\ven) r_E(\ven')^*\rra
340: &=&\frac{\delta_{drc}(\ven-\ven')}{4\pi}{\Bigg\{}\frac12 \lkk\lmk
341: \frac{1+\cos^2\theta}2
342: \rmk^2 -\cos^2\theta
343: \rkk \sin4\phi
344: I(\theta,\phi)
345: + \frac12\lkk\lmk \frac{1+\cos^2\theta}2
346: \rmk^2 +\cos^2\theta
347: \rkk \sin4\phi
348: Q(\theta,\phi)\nonumber\\
349: & & + \lkk\lmk \frac{1+\cos^2\theta}2
350: \rmk \cos\theta
351: \rkk \cos4\phi
352: U(\theta,\phi)
353: -i \lkk\lmk \frac{1+\cos^2\theta}2
354: \rmk \cos\theta
355: \rkk
356: V(\theta,\phi) {\Bigg\}}.\label{corr}
357: \eeqa
358: \end{widetext}
359: %The corresponding expression without polarization can be seen in
360: %\cite{Kudoh:2004he}.
361: With the following combination
362: \beq
363: % -i \lkk h_A(\theta,\phi)h_E(\theta,\phi)^*-h_A(\theta,\phi)^*
364: %h_E(\theta,\phi) \rkk =
365: {\rm Im}\lkk \lla r_A(\ven) r_E(\ven')^* \rra \rkk= - \frac{\delta_{drc}(\ven-\ven')}{4\pi} \lkk\lmk
366: \frac{1+\cos^2\theta}2
367: \rmk \cos\theta
368: \rkk V(\theta,\phi)\label{defs}
369: \eeq
370: we can extract the circular parameter $V$ alone.
371: %Note that a similar
372: %combination from any two-arm interferometers will depends only
373: %on $V$.
374: As shown in figure 1, the effective detector $E$ is obtained by
375: rotating the detector $A$ around $Z$-axis by $\pi/4$
376: \cite{Cutler:1997ta}. If this angle is
377: $\delta$,
378: there appears a factor $\sin 2\delta$ in the final expression in
379: eq.(\ref{defs}). Therefore, in some sense, LISA will provide an optimal set $(A,E)$ to
380: study the parameter $V$.
381: In contrast, sensitivity of correlation analysis to the monopole
382: intensity $I_{00}$ is
383: proportional to $\cos 2\delta$, and LISA cannot probe it with the
384: method, as is well known.
385:
386:
387:
388: We now discuss responses of detectors to gravitational waves
389: from all directions
390: $\ven$.
391: Firstly, what we can get observationally form the $A$ and $E$ modes
392: are the following integrals
393: $R_A=\int r_A(\ven)d\ven$ and $R_A=\int r_A(\ven)d\ven$.
394: Considering the spin of quantities $I$, $Q\pm iU$ and $V$ we can
395: expand them in terms of the spin weighted spherical harmonics ${}_s Y_{lm}$
396: as
397: $
398: I(\theta,\phi)=\sum I_{lm}Y_{lm}(\theta,\phi),~~~(Q\pm
399: iU)(\theta,\phi)=\sum K_{\pm ,lm}\cdot{}_{\pm 4}Y_{lm}(\theta,\phi),~~~V(\theta,\phi)=\sum V_{lm}Y_{lm}(\theta,\phi),
400: $
401: where the coefficients $K_{\pm ,lm}$ are defined for $l\ge 4$.
402: With these harmonic expansions and eq.(\ref{defs}), the combination
403: ${\cal C}={\rm Im} \lla R_A R_E^*\rra$ is evaluated as
404: $
405: {\cal C}=-\frac85\sqrt{\frac{\pi}3}V_{10}-\frac25\sqrt{\frac{\pi}7}V_{30}.
406: $
407: When we rotate two interferometers $A$ and $E$ with Euler angles
408: $(\alpha,\beta,\gamma)$, the signal ${\cal C}$ becomes
409: \beq
410: {\cal C}(\alpha,\beta,\gamma)=-\frac{16\pi}{15}V_{1m}Y_{1m}(\alpha,\beta)-\frac{4\pi}{35}V_{3m}Y_{3m}(\alpha,\beta). \label{defsa}
411: \eeq
412: This result is obtained by relating the coefficients $V_{lm}$ in
413: original coordinate with those in rotated coordinate
414: \cite{Allen:1996gp}.
415: LISA moves around the Sun with changing the orientation of its detector
416: plane that is
417: inclined to the
418: ecliptic plane by 60 degree \cite{lisa}. To describe these motions in
419: the ecliptic
420: coordinate, the Euler
421: angles are given as
422: $\alpha=2\pi (t/1{\rm yr})+\alpha_0$, $\beta=\pi/3$ and $\gamma=-2\pi
423: (t/1{\rm yr})+\gamma_0$ with time $t$ and constants $\alpha_0$ and
424: $\gamma_0$ \cite{Cutler:1997ta}. This parameterization is also valid for
425: BBO \cite{bbo}. The
426: parameter $\gamma$ determines the
427: so-called cartwheel motion, but the combination ${\cal
428: C}(\alpha,\beta,\gamma)$
429: does not depend on it. From eq.(\ref{defsa}) we can understand that
430: the observed signal ${\cal C}(\alpha,\beta,\gamma)$ with LISA is
431: decomposed
432: into modulation patterns with frequencies $f=0,
433: \pm1/3,\pm1/2$ and $\pm1{\rm yr^{-1}}$ from $m=0,\pm1,\pm2$ and $\pm 3$ modes respectively.
434:
435:
436: \begin{figure}
437: \begin{center}
438: \epsfxsize=8.cm
439: \begin{minipage}{\epsfxsize} \epsffile{1.eps} \end{minipage}
440: \end{center}
441: \caption{ The effective two-arm interferometers corresponding to the
442: TDI modes $A$ and $E$. Three spacecrafts of LISA are shown with
443: circles that are put on the $XY$-plane ($\theta =\pi/2$). The beam
444: pattern functions for $T$ mode are given with three unit
445: vectors $\vem_i$ as $F_T^P=\sum_{i=1}^3 (\vem_i\cdot \ve^P\cdot
446: \vem_i)(\vem_i\cdot \ven)$ $(P=+,\times)$.}
447: %\label{f2}
448: \end{figure}
449:
450: So far we have used the low frequency approximation with $f/f_*\ll 1$.
451: When we increase the frequency $f$, the correction terms
452: $i(f/f_*)D_A(\theta,\phi)$ and $i(f/f_*)D_E(\theta,\phi)$
453: in eq.(\ref{ha}) change the phases of
454: $r_A(\theta,\phi)$ and
455: $r_E(\theta,\phi)$ as a function of propagation directions $\ven$. With
456: these
457: correction
458: terms, the combination $R_A R_E^*-R_A^* R_E$ has contributions of
459: $I$, $Q$ and $U$ modes, and we cannot extract the circular polarization
460: $V$ in a clean manner. Note also that the first order term $O(f/f_*)$ for the
461: combination $R_A R_E^*+R_A^* R_E$ depends on the parameter $V$ due to the corrections. We can
462: restate the situation as follows; Roughly
463: speaking,
464: the circular polarization is measured by correlating two
465: data with phase difference $\pi/2$ \cite{radipro}. This is related to
466: the fact
467: $V(\ven)\propto {\rm Re}\lla r_A(\theta,\phi)r_E(\theta,\phi )^* e^{i \pi/2}\rra$.
468: But the finiteness of the arm-length modulates the phase as a function of
469: direction, and we cannot keep the phase difference $\pi/2$ simultaneously
470: for all the
471: directions. When we
472: consider the spatial separation ${\vect d}$ of two interferometers, same
473: kind of arguments hold by perturbatively expanding the phase
474: difference
475: $\exp[2\pi i f {\vect d}\cdot \ven]$ with a expansion parameter $(f|{\vect d}|)$ in addition to that with $(f/f_*)$ for the effects of arm-length. Therefore, in our
476: analysis, the requirement for the low frequency
477: regime
478: is not for simplicity of calculation, but is a crucial condition to
479: extract the
480: circular polarization alone with gravitational wave detectors that have
481: broad directional response and are remarkably
482: different from electromagnetic wave telescopes.
483:
484:
485: As shown in eq.(\ref{defs}), we cannot measure the monopole moment
486: $V_{00}$
487: of circular polarization using the signal $\cal C$ made from the $A$ and
488: $E$
489: modes due to parity reason at low frequency limit. More specifically,
490: products, such as
491: $F_A^+ F_E^\times$,
492: have odd parity. This simple results hold for signal $\cal C$
493: with any two two-arm interferometers at low frequency limit. It does not matter whether their
494: vertex angles are not $\pi/2$, whether two arm-lengths of each detector
495: are not
496: equal, or whether two interferometers are on unparallel planes
497: ({\it e.g.} for LIGO and VIRGO combination).
498: The signals $\cal C$
499: with ($A,T$) or ($E,T$) modes depend on the coefficients $V_{lm}$
500: with even
501: $l$ at their leading order. But, in the case of LISA, the combinations
502: do not have monopole
503: moment $V_{00}$. This is because of the apparent symmetry
504: of these data streams \cite{Prince:2002hp,Kudoh:2004he}. Furthermore we
505: cannot get the moment $V_{00}$
506: even using their higher order terms with
507: $O((f/f_*)^n)$ $(n\ge 1)$, as long as LISA is symmetric at each vertex.
508: A future mission might use multiple LISA-type sets with data streams
509: $\{A,E,T\}, \{A',E',T'\},...$ \cite{bbo}. We confirmed that even if detector
510: planes for $A$ and $T'$ modes are not parallel, the monopole $V_{00}$ can
511: not be captured by the signal $\cal C$ with their combination at their
512: leading order.
513: \if0%%%%%%%%%%%%%%%%%%%%%%%%%%%
514: If we have two detectors $A$ and $E$ separate along $Z$-axis from
515: the
516: configuration given in figure 1, the combination $R_A R_{E}^*+R_A^*
517: R_{E}$ have monopole mode $V_{00}$ at order $O((f/f_*)^0 (f|{\vect
518: d}|)^1)$, as expected from eq.(\ref{corr}).
519: Note also that this combination does not have potentially strong
520: contribution $I_{00}$ at its leading order $O((f/f_*)^0 (f|{\vect
521: d}|)^0)$.
522: \fi%%%%%%%%%%%%%%%%%%%%%%%%%%%%
523:
524: As we discussed so far, it is not straightforward to capture the monopole
525: $V_{00}$ in a simple manner. But, in principle, we can manage this. For
526: example, we add a detector $E_2$ that is given by
527: moving the original detectors $E$ in figure 1 by distance $d$
528: toward $+
529: Z$-direction, and then take the signal $R_{A}R_{E_2}^*$.
530: % There are no terms of order $O((fd)^0
531: %(f/f_*)^n)$ $(n\ge 0)$
532: % due to cancellation at $d=0$. But with its leading
533: At order $O((fd)^1 (f/f_*)^0)$, we can probe the monopole
534: $V_{00}$ \cite{prep}.
535: This kind of
536: arrangement for detector configuration might be important to study
537: anisotropy and polarization of gravitational wave background in the
538: long run.
539:
540: \if0
541: In contrast to difficulties for observing the monopole $V_{00}$ of the
542: circular polarization, the monopole $I_{00}$ of the intensity appears in
543: the simple
544: amplitudes $\lla R_A R_A^*\rra$ or $\lla R_E R_E^*\rra$. Therefore we
545: can imagine the following
546: interesting situation. For a strong and isotropic background with 100\%
547: circular polarization $(|V_{00}|=I_{00}=I_t \equiv
548: \sqrt {\sum_{lm}
549: |I_{lm}|^2})$ we might easily detect the
550: background without identifying its extreme
551: polarization nature.
552: \fi
553:
554: \section{ observation with LISA}
555:
556: Next we discuss how well we can analyze the information $V_{lm}$ of
557: circular polarization of stochastic gravitational wave background with
558: LISA.
559: The observed inclinations of local Galactic binaries are known to be
560: consistent with having random distribution with no correlation to global
561: Galactic
562: structure \cite{inc}. This suggests that the confusion background by
563: galactic
564: binaries is not polarized. In future, LISA itself will provide basic
565: parameters for thousands of Galactic binaries at $f\ge 3$mHz, including
566: information
567: of their
568: orientations \cite{lisa}. This can be used to further constrain the
569: allowed
570: polarization degree of the Galactic confusion noise.
571: While this information will help us to discriminate the origin of
572: detected polarization signal with LISA, our analysis below does not
573: depend on this outlook.
574: %Here, based on thisoutlook, we assume that the Galactic confusion noise
575: %is not polarized and study a scenario to search circularly polarized
576: %cosmological background with LISA.
577: Our target here is a combination $p\Omega_{GW,b}(f)$ where
578: $\Omega_{GW,b}(f)$
579: is the
580: normalized energy density of gravitational wave background (see
581: \cite{Allen:1996vm} for its definition), regardless
582: of its origin. The parameter
583: $p=(\sqrt{\sum_m |V_{1m}|^2+\sum_m |V_{3m}|^2})/I_t$ is its circular
584: polarization power in $l=1$ and 3 multipoles ($I_t \equiv
585: \sqrt {\sum_{lm}
586: |I_{lm}|^2}$ : total intensity). For simplicity, we do not
587: discuss technical aspects in relation to the time modulation of the
588: signal ${\cal C}$
589: due to the motion of LISA that was
590: explained earlier around eq.(\ref{defsa}). We can easily make an
591: appropriate extension to deal with it \cite{Cornish:2001qi}.
592: % (see {\it e.g.}
593: %\cite{Kudoh:2004he,Cornish:2001qi} for a similar situation).
594:
595: As we will see below, the signal $\cal C$ is a powerful probe to study
596: the target $p\Omega_{GW,b}(f)$ in a frequency regime
597: where the amplitude
598: $p\Omega_{GW,b}(f)$ itself may be dominated by the confusion background noise or detector noise.
599: We take a summation of the signal $\cal C$ for Fourier modes around a
600: frequency $f$
601: in a bandwidth $\Delta f\sim f$. As the frequency resolution is inverse
602: of the observational time $T_{obs}$ and the total number of Fourier
603: modes in
604: the band is $(T_{obs} \Delta f)$, the expectation value for the
605: summation becomes ${\cal C} (T_{obs} \Delta f)\sim p I_t (T_{obs}\Delta f)$.
606: Here we used the fact that the detector noises for the $A$ and $E$ modes are not correlated.
607: In contrast the
608: fluctuation $N$
609: for this summation is
610: given as
611: $
612: N=\sum_f \lmk n_A(f)n_E(f)^*-n_A(f)^*n_E(f) \rmk,
613: $
614: where the total noises $n_A(f)$ and $n_E(f)$ include both detector noise and
615: circularly unpolarized potion of the confusion noise.
616: The root-mean-square value of $N$ is written as
617: $
618: \lla |N|^2 \rra^{1/2}\sim S_n(f)(T_{obs }\Delta f)^{1/2}
619: $
620: with $S_n(f)$ being the total noise spectrum for the $A$ and $E$ modes.
621: Thus the signal to noise ratio for the measurement is $SNR\sim p I_T
622: (T_{obs }\Delta f)^{1/2}/S_n(f)$. We can rewrite this expression with using
623: the normalized energy density $\Omega_{GW}$ and obtain
624: $
625: SNR\sim \lmk \frac{p \Omega_{GW,b} }{\Omega_{GW,n}} \rmk (T_{obs}\Delta
626: f)^{1/2} ,
627: $
628: where the magnitude ${\Omega_{GW,n}}$ corresponds to the total noise
629: level around frequency $f$.
630: When the detector noise is dominated by the background noise, we have
631: $\Omega_{GW,n}\sim \Omega_{GW,b}$. This will be the case for LISA
632: around $0.3 {\rm mHz} \siml f \siml 2$mHz.
633: As a concrete example, we put $f\sim 0.1$mHz with the expected noise level
634: ${\Omega_{GW,n}}\sim
635: 10^{-10}$ \cite{lisa}, and take bandwidth
636: $\Delta f \sim f$. Then we have
637: $
638: SNR\sim \lmk \frac{p \Omega_{GW,b} }{10^{-12}} \rmk (T/3{\rm
639: yr})^{1/2}.
640: $
641: The improvement factor $(T_{obs}\Delta
642: f)^{-1/2}$ for the detectable level of the target $p \Omega_{GW,b}$ is caused by the same reason as standard
643: correlation technique
644: for detecting stochastic background
645: \cite{Flanagan:1993ix}. Interestingly, this factor
646: $(T_{obs}\Delta f)^{-1/2}\sim 0.01$ around $f\sim 0.1$mHz is almost same
647: as the
648: maximum
649: level of relative contamination $(f/f_*)\sim 0.01$ for LISA by other
650: modes ($I$,
651: $Q$,
652: $U$) due to finiteness of arm-length.
653: A Japanese future project DECIGO \cite{Seto:2001qf} is planed to use
654: Fabri-Perot type
655: design with its characteristic sensitivity $f_*= 50$Hz \cite{Kawamura},
656: while its best
657: sensitivity is around $\sim 0.3$Hz similar to BBO whose characteristic
658: frequency is $f_*=1$Hz.
659: Therefore, DECIGO is expected to be less affected by other modes and has
660: potential to
661: reach the level $p \Omega_{GW,b} \sim 10^{-16}$ with one year observation.
662:
663: If circular polarization parameter $V({\theta,\phi})$ of cosmological
664: background is highly isotropic in the
665: CMB-rest frame, its dipole pattern will be induced by our peculiar
666: motion \cite{Kudoh:2005as}. This might be an interesting observational
667: target in future. In this case
668: we can predict
669: the time modulation pattern ${\cal
670: C}(\alpha,\beta,\gamma)$ for rotating detectors. A consistency
671: check using this modulation will
672: be powerful approach to discriminate its cosmological origin. Actually a
673: similar
674: method can be
675: used for LISA to analyze circular polarization of Galactic
676: confusion noise.
677:
678:
679: The author would like to thank A. Taruya for comments, and
680: A. Cooray for various supports.
681: This research was funded by McCue Fund at the Center for Cosmology,
682: University of California, Irvine.
683:
684: %\include{ref}
685:
686: %%%%%%%%% references %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
687: \begin{thebibliography}{DUM}
688:
689:
690: %\cite{Allen:1996vm}
691: \bibitem{Allen:1996vm}
692: M.~Maggiore,
693: %``Gravitational wave experiments and early universe cosmology,''
694: Phys.\ Rept.\ {\bf 331}, 283 (2000).
695: %%CITATION = GR-QC 9909001;%%
696:
697:
698: %\cite{Kahniashvili:2005qi}
699: \bibitem{Kahniashvili:2005qi}
700: T.~Kahniashvili, G.~Gogoberidze and B.~Ratra,
701: %``Polarized cosmological gravitational waves from primordial helical
702: %turbulence,''
703: Phys.\ Rev.\ Lett.\ {\bf 95}, 151301 (2005).
704: % [arXiv:astro-ph/0505628].
705: %%CITATION = ASTRO-PH 0505628;%%
706:
707:
708: %\cite{Lyth:2005jf}
709: \bibitem{Lyth:2005jf}
710: S.~H.~S.~Alexander, M.~E.~Peskin and M.~M.~Sheikh-Jabbari,
711: %``Leptogenesis from gravity waves in models of inflation,''
712: Phys.\ Rev.\ Lett.\ {\bf 96}, 081301 (2006).
713: %%CITATION = HEP-TH 0403069;%%
714: % D.~H.~Lyth, C.~Quimbay and Y.~Rodriguez,
715: %``Leptogenesis and tensor polarisation from a gravitational Chern-Simons
716: %term,''
717: % JHEP {\bf 0503}, 016 (2005).
718: % [arXiv:hep-th/0501153].
719: %%CITATION = HEP-TH 0501153;%%
720:
721:
722:
723: %\cite{Lue:1998mq}
724: \bibitem{Lue:1998mq}
725: A.~Lue, L.~M.~Wang and M.~Kamionkowski,
726: %``Cosmological signature of new parity-violating interactions,''
727: Phys.\ Rev.\ Lett.\ {\bf 83}, 1506 (1999).
728: % [arXiv:astro-ph/9812088].
729: %%CITATION = ASTRO-PH 9812088;%%
730:
731:
732: %\cite{Seljak:1996gy}
733: \bibitem{Seljak:1996gy}
734: U.~Seljak and M.~Zaldarriaga,
735: %``Signature of gravity waves in polarization of the microwave background,''
736: Phys.\ Rev.\ Lett.\ {\bf 78}, 2054 (1997);
737: % [arXiv:astro-ph/9609169];
738: %%CITATION = ASTRO-PH 9609169;%%
739: M.~Kamionkowski, A.~Kosowsky and A.~Stebbins,
740: %``A probe of primordial gravity waves and vorticity,''
741: Phys.\ Rev.\ Lett.\ {\bf 78}, 2058 (1997).
742: % [arXiv:astro-ph/9609132].
743: %%CITATION = ASTRO-PH 9609132;%%
744:
745:
746:
747:
748: \bibitem{inc}S.~S. Huang and C. Wade, Astrophys.J, {\bf 143},
749: 146 (1966).
750:
751:
752:
753:
754: \bibitem{Thorne_K:1987}
755: K.~S. Thorne, in {\em Three hundred years of gravitation}, (Cambridge University Press, Cambridge, 1987),
756: pp.\ 330--458.
757:
758:
759:
760: %\cite{Allen:1996gp}
761: \bibitem{Allen:1996gp}
762: B.~Allen and A.~C.~Ottewill,
763: %``Detection of anisotropies in the gravitational-wave stochastic
764: %background,''
765: Phys.\ Rev.\ D {\bf 56}, 545 (1997).
766: % [arXiv:gr-qc/9607068].
767: %%CITATION = GR-QC 9607068;%%
768:
769:
770:
771:
772: %\cite{lisa}
773: \bibitem{lisa}
774: P.~L.~Bender { et al},
775: {\it LISA Pre-Phase A Report,} Second edition, July 1998.
776:
777:
778: %\if0
779: %\cite{Cornish:2001qi}
780: \bibitem{Cornish:2001qi}
781: N.~J.~Cornish and S.~L.~Larson,
782: %``Space missions to detect the cosmic gravitational-wave background,''
783: Class.\ Quant.\ Grav.\ {\bf 18}, 3473 (2001);
784: % [arXiv:gr-qc/0103075];
785: %%CITATION = GR-QC 0103075;%%
786: C.~Ungarelli and A.~Vecchio,
787: %``Studying the anisotropy of the gravitational wave stochastic background
788: %with LISA,''
789: Phys.\ Rev.\ D {\bf 64}, 121501 (2001);
790: % [arXiv:astro-ph/0106538];
791: %%CITATION = ASTRO-PH 0106538;%%
792: N.~Seto,
793: Phys.\ Rev.\ D {\bf 69}, 123005 (2004);
794: % [arXiv:gr-qc/0403014];
795: %%CITATION = GR-QC 0403014;%%
796: A.~Taruya and H.~Kudoh,
797: %``Probing anisotropies of gravitational-wave backgrounds with a space-based
798: %interferometer. II: Perturbative reconstruction of a low-frequency skymap,''
799: Phys.\ Rev.\ D {\bf 72}, 104015 (2005).
800: % [arXiv:gr-qc/0507114].
801: %%CITATION = GR-QC 0507114;%%
802: %\fi
803:
804: %\cite{Cutler:1997ta}
805: \bibitem{Cutler:1997ta}
806: C.~Cutler,
807: %``Angular resolution of the LISA gravitational wave detector,''
808: Phys.\ Rev.\ D {\bf 57}, 7089 (1998).
809: % [arXiv:gr-qc/9703068].
810: %%CITATION = GR-QC 9703068;%%
811:
812:
813:
814: %\cite{Prince:2002hp}
815: \bibitem{Prince:2002hp}
816: T .~A.~Prince, et al,
817: %``The LISA optimal sensitivity,''
818: Phys.\ Rev.\ D {\bf 66}, 122002 (2002).
819: % [arXiv:gr-qc/0209039].
820: %%CITATION = GR-QC 0209039;%%
821:
822:
823:
824: \bibitem{radipro}
825: G. B. Rybicki and A. P. Lightman, {\it Radiative Process in Astrophysics} (Wiley, New York, 1960).
826:
827: %\cite{Kudoh:2004he}
828: \bibitem{Kudoh:2004he}
829: H.~Kudoh and A.~Taruya,
830: %``Probing anisotropies of gravitational-wave backgrounds with a space-based
831: %interferometer: Geometric properties of antenna patterns and their angular
832: %power,''
833: Phys.\ Rev.\ D {\bf 71}, 024025 (2005).
834: % [arXiv:gr-qc/0411017].
835: %%CITATION = GR-QC 0411017;%%
836:
837:
838:
839:
840: %\cite{bbo}
841: \bibitem{bbo}
842: E. S. Phinney et al. The Big Bang Observer, NASA Mission Concept Study
843: (2003).
844:
845:
846: %\cite{Seto:2005qy}
847: \bibitem{Seto:2005qy}
848: N.~Seto,
849: %``Correlation analysis of stochastic gravitational wave background around
850: %0.1-Hz - 1-Hz,''
851: Phys.\ Rev.\ D {\bf 73}, 063001 (2006);
852: % [arXiv:gr-qc/0510067];
853: %%CITATION = GR-QC 0510067;%%
854: V.~Corbin and N.~J.~Cornish,
855: %``Detecting the cosmic gravitational wave background with the big bang
856: %observer,''
857: Class.\ Quant.\ Grav.\ {\bf 23}, 2435 (2006).
858: % [arXiv:gr-qc/0512039].
859: %%CITATION = GR-QC 0512039;%%
860:
861:
862:
863:
864:
865: %\cite{Tinto:2001ii}
866: \bibitem{Tinto:2001ii}
867: M.~Tinto, J.~W.~Armstrong and F.~B.~Estabrook,
868: %``Discriminating a gravitational wave background from instrumental noise in
869: %the LISA detector,''
870: Phys.\ Rev.\ D {\bf 63}, 021101 (2001).
871: %%CITATION = PHRVA,D63,021101;%%
872:
873: \if0
874: %\cite{Seto:2002uj}
875: \bibitem{Seto:2002uj}
876: N.~Seto,
877: %from MBH binaries,''
878: Phys.\ Rev.\ D {\bf 66}, 122001 (2002);
879: %%CITATION = GR-QC 0210028;%%
880: N.~J.~Cornish and L.~J.~Rubbo,
881: %``The LISA response function,''
882: Phys.\ Rev.\ D {\bf 67}, 022001 (2003).
883: \fi
884:
885: \bibitem{prep}
886: N.~Seto, in preperation
887:
888:
889:
890: %\cite{Flanagan:1993ix}
891: \bibitem{Flanagan:1993ix}
892: E.~E.~Flanagan,
893: %``The Sensitivity of the laser interferometer gravitational wave observatory
894: %(LIGO) to a stochastic background, and its dependence on the detector
895: %orientations,''
896: Phys.\ Rev.\ D {\bf 48}, 2389 (1993).
897: % [arXiv:astro-ph/9305029];
898: %%CITATION = ASTRO-PH 9305029;%%
899:
900:
901:
902: %\cite{Seto:2001qf}
903: \bibitem{Seto:2001qf}
904: N.~Seto, S.~Kawamura and T.~Nakamura,
905: %``Possibility of direct measurement of the acceleration of the universe
906: %using 0.1-Hz band laser interferometer gravitational wave antenna in
907: %space,''
908: Phys.\ Rev.\ Lett.\ {\bf 87}, 221103 (2001).
909: % [arXiv:astro-ph/0108011].
910: %%CITATION = ASTRO-PH 0108011;%%
911:
912:
913: %\cite{Kawamura}
914: \bibitem{Kawamura}S. Kawamura, et al,\
915: Class. Quant. Grav. {\bf 23}, 125 (2006).
916:
917: %\cite{Kudoh:2005as}
918: \bibitem{Kudoh:2005as}
919: H.~Kudoh, A.~Taruya, T.~Hiramatsu and Y.~Himemoto,
920: %``Detecting a gravitational-wave background with next-generation space
921: %interferometers,''
922: Phys.\ Rev.\ D {\bf 73}, 064006 (2006).
923: % [arXiv:gr-qc/0511145].
924: %%CITATION = GR-QC 0511145;%%
925:
926:
927:
928:
929: \end{thebibliography}
930:
931:
932:
933:
934:
935: \end{document}
936:
937:
938:
939: