1: \documentclass[prl,twocolumn,showpacs]{revtex4}

2: 

3: \usepackage{graphicx}

4: \usepackage{dcolumn}

5: \usepackage{amsmath}

6: 

7: 

8: \begin{document}

9: 

10: 

11: \title{Dynamic Resonance of Light in Fabry-Perot Cavities} 

12: 

13: \author{M. Rakhmanov} \author{R.L. Savage, Jr.}

14: \altaffiliation{LIGO Hanford Observatory, P.O. Box 1970, M/S S9-02, 

15: Richland, WA 99352}

16: \author{D.H. Reitze} \author{D.B. Tanner}

17: \affiliation{Physics Department, P.O. Box 118440, 

18: University of Florida, Gainesville, FL 32611}

19: 

20: 

21: \date{January, 2001}

22: 

23: \begin{abstract}

24: The dynamics of light in Fabry-Perot cavities with varying length 

25: and input laser frequency are analyzed and the exact condition for

26: resonance is derived. This dynamic resonance depends on the light

27: transit time in the cavity and the Doppler effect due to the mirror

28: motions. The response of the cavity to length variations is very

29: different from its response to laser frequency variations. If the

30: frequency of these variations is equal to multiples of the cavity free

31: spectral range, the response to length is maximized while the response

32: to the laser frequency is zero. Implications of these results for the

33: detection of gravitational waves using kilometer-scale Fabry-Perot

34: cavities are discussed.

35: \end{abstract}

36: 

37: 

38: 

39: 

40: \pacs{07.60.Ly, 07.60.-j, 42.60.Da, 42.60.-v, 04.80.Nn, 95.55.Ym}

41: 

42: \maketitle

43: 

44: 

45: Fabry-Perot cavities, optical resonators, are commonly utilized for

46: high-precision frequency and distance measurements

47: \cite{Vaughan:1989}. Currently, kilometer-scale Fabry-Perot cavities

48: with suspended mirrors are being employed in efforts to detect cosmic

49: gravitational waves \cite{Abramovici:1992, Bradaschia:1990}. This

50: application has stimulated renewed interest in cavities with moving

51: mirrors \cite{Camp:1995, Mizuno:1997, Chickarmane:1998, Pai:2000} and

52: motivated efforts to model the dynamics of such cavities on the

53: computer \cite{Redding:1997, Sigg:1997, Bhawal:1998, Beausoleil:1999,

54: Yamamoto:2000}. Recently, several studies addressed the process of

55: lock acquisition in which the cavity mirrors move through the

56: resonance positions \cite{Camp:1995, Lawrence:1999,

57: Rakhmanov:doppler}. In this process, the Doppler effect due to the

58: mirror motions impedes constructive interference of light in the

59: cavity giving rise to complex field dynamics. In contrast, Fabry-Perot

60: cavities held in the state of resonance have usually been  treated as

61: essentially static. In this letter, we show that resonant cavities

62: also have complex field dynamics and we derive the condition for

63: dynamic resonance. Our analysis is developed for the very long

64: Fabry-Perot cavities of gravitational wave detectors, but the results

65: are general and apply to any cavities, especially when the frequencies

66: of interest are close to the cavity free spectral range.

67: 

68: 

69: We consider a Fabry-Perot cavity with a laser field incident from

70: one side as shown in Fig.~\ref{fpcav}. Variations in the cavity length 

71: are due to the mirror displacements $x_a(t)$ and $x_b(t)$ which are 

72: measured with respect to reference planes $a$ and $b$. The nominal 

73: light transit time and the free spectral range (FSR) of the cavity 

74: are defined by 

75: %

76: \begin{equation}

77: %

78:    T = L/c, \qquad \omega_{\mathrm{fsr}} = \pi/T.

79: %

80: \end{equation}

81: %

82: The field incident upon the cavity and the field circulating in the

83: cavity are described by plane waves with nominal frequency $\omega$

84: and wavenumber $k$ ($k=\omega/c$). Variations in the laser frequency 

85: are denoted by $\delta \omega(t)$. We assume that the mirror

86: displacements are much less than the nominal cavity length and that 

87: the deviations of the laser frequency are much less than the nominal

88: frequency.

89: 

90: 

91: \begin{figure}

92:    \centering\includegraphics[width=0.46\textwidth]{fig1}

93:    \caption{Mirror positions and fields in a Fabry-Perot cavity.}

94:    \label{fpcav}

95: \end{figure}

96: 

97: 

98: At any given place the electric field ${\mathcal{E}}$ in 

99: the cavity oscillates at a very high frequency: 

100: ${\mathcal{E}}(t) \propto \exp(i\omega t)$. For simplicity, we 

101: suppress the fast-oscillating factor and define the slowly-varying 

102: field as $E(t) = {\mathcal{E}}(t) \exp(-i\omega t)$. To properly 

103: account for the phases of the propagating fields, their complex 

104: amplitudes are defined at fixed locations, reference planes $a_1$ 

105: and $a_2$, as shown in Fig.~\ref{fpcav}. (The small offset $\epsilon$

106: is introduced for convenience and can be set to zero at the end of

107: calculations.)

108: 

109: 

110: 

111: 

112: The equations for fields in the cavity can be obtained by tracing

113: a wavefront during its complete round-trip in the cavity (starting

114: from the reference plane $a_2$). The propagation delays $\tau_1$ and

115: $\tau_2$ depend on the mirror positions and are given by

116: %

117: \begin{eqnarray}

118: %

119:    c \; \tau_1 & = & L - \epsilon + x_b(t - \tau_1) ,\label{t1} \\

120: %

121:    c \; \tau_2 & = & \epsilon - x_a(t - \tau_2) .\label{t2}

122: %

123: \end{eqnarray}

124: %

125: Then the fields in the cavity satisfy the equations:

126: %

127: \begin{eqnarray}

128: %

129:    E'(t) & = & - r_b E (t - 2 \tau_1) e^{-2i \omega \tau_1} ,

130:                \label{Epr} \\

131: %

132:    E(t)  & = & - r_a E'(t - 2 \tau_2) e^{-2i \omega \tau_2} +

133:                \nonumber \\

134: %

135:          &   & t_a E_{\mathrm{in}}(t - 2 \epsilon/c) ,

136:                \label{Ecav}

137: %

138: \end{eqnarray}

139: %

140: where $r_a$ and $r_b$ are the mirror reflectivities, and $t_a$ is the

141: transmissivity of the front mirror.

142: 

143: 

144: Because the field amplitudes $E$ and $E'$ do not change significantly 

145: over times of order $x_{a,b}/c$, the small variations in these 

146: amplitudes during the changes in propagation times due to mirror 

147: displacements can be neglected. Furthermore, the reference planes $a$ 

148: and $b$ can be chosen so that the nominal length of the Fabry-Perot 

149: cavity becomes an integer multiple of the laser wavelength, making 

150: $\exp(-2ikL)=1$. Finally, the offset $\epsilon$ can be set to zero, 

151: and Eqs.~(\ref{Epr})-(\ref{Ecav}) can be combined yielding one 

152: equation for the cavity field

153: %

154: \begin{equation}\label{Eiter}

155: %

156:    E(t) = t_a E_{\mathrm{in}}(t) + 

157:           r_a r_b E(t - 2T) \exp[-2ik \delta L(t)] .

158: %

159: \end{equation}

160: %

161: Here $\delta L(t)$ is the variation in the cavity length ``seen'' 

162: by the light circulating in the cavity,

163: %

164: \begin{equation}\label{deltaL2}

165: %

166:    \delta L(t) = x_b(t - T) - x_a(t) .

167: %

168: \end{equation}

169: %

170: Note that the time delay appears in the coordinate of the end mirror, 

171: but not the front mirror. This is simply a consequence of our

172: placement of the laser source; the light that enters the cavity 

173: reflects from the end mirror first and then the front mirror. 

174: For $\delta L=0$, Laplace transformation of both sides of

175: Eq.~(\ref{Eiter}) yields the basic cavity response function

176: %

177: \begin{equation}\label{cavTf}

178: %

179:    H(s) \equiv \frac{\tilde{E}(s)}{\tilde{E}_{\mathrm{in}}(s)} = 

180:       \frac{t_a}{1 - r_a r_b e^{-2sT}},

181: %

182: \end{equation}

183: %

184: where tildes denote Laplace transforms.

185: 

186: 

187: The static solution of Eq.~(\ref{Eiter}) is found by considering 

188: a cavity with fixed length ($\delta L = {\mathrm{const}}$) 

189: and an input laser field with fixed amplitude and frequency

190: ($A, \delta \omega = {\mathrm{const}}$). In this 

191: case the input laser field and the cavity field are given by

192: %

193: \begin{eqnarray}

194: %

195:    E_{\mathrm{in}}(t) & = & A   \; e^{i \delta \omega t}, \\

196:    E(t)               & = & E_0 \; e^{i \delta \omega t},

197: %

198: \end{eqnarray}

199: %

200: where $E_0$ is the amplitude of the cavity field,

201: %

202: \begin{equation}\label{Estat}

203: %

204:    E_0 = \frac{t_a A}{1 - r_a r_b \exp[-2i

205:       (T \; \delta \omega + k \; \delta L)]} .

206: %

207: \end{equation}

208: %

209: The cavity field is maximized when the length and the laser frequency

210: are adjusted so that

211: %

212: \begin{equation}\label{statRes}

213: %

214:    \frac{\delta \omega}{\omega} = - \frac{\delta L}{L} .

215: %

216: \end{equation}

217: %

218: This is the well-known static resonance condition.

219: The maximum amplitude of the cavity field is given by

220: %

221: \begin{equation}\label{barE}

222: %

223:    \bar{E} = \frac{t_a A}{1 - r_a r_b} .

224: %

225: \end{equation}

226: %

227: 

228: 

229: Light can also resonate in a Fabry-Perot cavity when its length and

230: the laser frequency are changing. For a fixed amplitude and variable 

231: phase, the input laser field can be written as

232: %

233: \begin{equation}

234: %

235:     E_{\mathrm{in}}(t) = A \; e^{i \phi(t)} ,

236: %

237: \end{equation}

238: %

239: where $\phi(t)$ is the phase due to frequency variations,

240: %

241: \begin{equation}

242: %

243:    \phi(t) = \int_0^t \delta \omega(t') dt' .

244: %

245: \end{equation}

246: %

247: Then the steady-state solution of Eq.~(\ref{Eiter}) is 

248: %

249: \begin{equation}

250: %

251:     E(t) = \bar{E} \; e^{i \phi(t)} ,

252: %

253: \end{equation}

254: %

255: where the amplitude $\bar{E}$ is given by Eq.~(\ref{barE}) and the 

256: phase satisfies the condition

257: %

258: \begin{equation}\label{phaseT}

259: %

260:    \phi(t) - \phi(t - 2T) = - 2 k \; \delta L(t) .

261: %

262: \end{equation}

263: %

264: Thus resonance occurs when the phase of the input laser field is 

265: corrected to compensate for the changes in the cavity length due to

266: the mirror motions. The associated laser frequency correction is equal 

267: to the Doppler shift caused by reflection from the moving mirrors

268: %

269: \begin{equation}

270: %

271:    \delta \omega(t) - \delta \omega(t - 2T) = 

272:       - 2 \frac{v(t)}{c} \omega , 

273: %

274: \end{equation}

275: %

276: where $v(t)$ is the relative mirror velocity ($v = d\delta L/dt$).

277: The equivalent formula in the Laplace domain is

278: %

279: \begin{equation}\label{dynResF}

280: %

281:    C(s) \frac{\delta \tilde{\omega}(s)}{\omega} = - 

282:       \frac{\delta \tilde{L}(s)}{L} ,

283: %

284: \end{equation}

285: %

286: where $C(s)$ is the normalized frequency-to-length transfer

287: function which is given by

288: %

289: \begin{equation}

290: %

291:    C(s) = \frac{1 - e^{-2sT}}{2 s T} .

292: %

293: \end{equation}

294: %

295: Eq.~(\ref{dynResF}) is the condition for dynamic resonance. It must be

296: satisfied in order for light to resonate in the cavity when the

297: cavity length and the laser frequency are changing.

298: 

299: 

300: 

301: The transfer function $C(s)$ has zeros at multiples of the cavity free 

302: spectral range,

303: %

304: \begin{equation}\label{zeros}

305: %

306:    z_n = i \omega_{\mathrm{fsr}} n ,

307: %

308: \end{equation}

309: %

310: where $n$ is integer, and therefore can be written as the infinite

311: product,

312: %

313: \begin{equation}

314: %

315:    C(s) = e^{-s T} \prod \limits_{n=1}^{\infty} \left( 

316:       1 - \frac{s^2}{z_n^2} \right) ,

317: %

318: \end{equation}

319: %

320: which is useful for control system design\footnote{This formula is 

321: derived using the infinite-product representation for sine:

322: $\sin x = x \;\prod_{n=1}^{\infty}\left(1 - x^2/\pi^2 n^2\right)$.}.

323: 

324: 

325: To maintain resonance, changes in the cavity length must be

326: compensated by changes in the laser frequency according to 

327: Eq.~(\ref{dynResF}). If the frequency of such changes is much 

328: less than the cavity free spectral range, $C(s) \approx 1$ and 

329: Eq.~(\ref{dynResF}) reduces to the quasi-static approximation,

330: %

331: \begin{equation}\label{qstatResF}

332: %

333:    \frac{\delta \tilde{\omega}(s)}{\omega} \approx - 

334:       \frac{\delta \tilde{L}(s)}{L} .

335: %

336: \end{equation}

337: %

338: At frequencies above the cavity free spectral range, 

339: $C(s) \propto 1/s$

340: and increasingly larger laser frequency changes are required to

341: compensate for cavity length variations. Moreover, at multiples of

342: the FSR, $C(s)=0$ and no frequency-to-length compensation is possible.

343: 

344: 

345: In practice, Fabry-Perot cavities always deviate from resonance, and 

346: a negative-feedback control system is employed to reduce the

347: deviations. For small deviations from resonance, the cavity field can

348: be described as

349: %

350: \begin{equation}

351: %

352:    E(t) = [\bar{E} + \delta E(t)] e^{i\phi(t)},

353: %

354: \end{equation}

355: %

356: where $\bar{E}$ is the maximum field given by Eq.(\ref{barE}), and 

357: $\delta E$ is a small perturbation ($|\delta E| \ll |\bar{E}|$). 

358: Substituting this equation into Eq.~(\ref{Eiter}), we see that the 

359: perturbation evolves in time according to

360: %

361: \begin{eqnarray}

362: %

363:    & & \delta E(t) - r_a r_b \delta E(t - 2T) = \nonumber \\

364: %

365:    & & \qquad - i r_a r_b \bar{E} \left[ \phi(t) - \phi(t - 2T) + 

366:        2 k \; \delta L(t) \right] .\label{dE(t)}

367: %

368: \end{eqnarray}

369: %

370: This equation is easily solved in the Laplace domain, yielding

371: %

372: \begin{equation}\label{dE(s)}

373: %

374:    \delta \tilde{E}(s) = - i r_a r_b \bar{E} \; 

375:       \frac{ \left(1 - e^{-2sT} \right) \tilde{\phi}(s) +

376:       2 k \; \delta \tilde{L}(s)}{1 - r_a r_b e^{-2sT}} .

377: %

378: \end{equation}

379: %

380: 

381: \begin{figure}

382:    \centering\includegraphics[width=0.46\textwidth]{fig2}

383:    \caption{Bode plot of $H_L(i\Omega)$ for the LIGO 4-km

384:    Fabry-Perot cavities. The peaks occur at multiples of the 

385:    FSR ($37.5$ kHz) and their half-width ($91$ Hz) is equal to the 

386:    inverse of the cavity storage time.}

387:    \label{tfLength}

388: \end{figure}

389: 

390: 

391: Deviations of the cavity field from its maximum value can be measured

392: by the Pound-Drever-Hall (PDH) error signal which is widely utilized

393: for feedback control of Fabry-Perot cavities \cite{Drever:1983}. The

394: PDH signal is obtained by coherent detection of phase-modulated light

395: reflected by the cavity. With the appropriate choice of the

396: demodulation phase, the PDH signal is proportional to the imaginary

397: part of the cavity field (Eq.~(\ref{dE(s)})) and therefore can be 

398: written as

399: %

400: \begin{equation}\label{PDH}

401: %

402:    \delta \tilde{V}(s) \propto H(s) \left[

403:       \frac{\delta \tilde{L}(s)}{L} + C(s) \;

404:       \frac{\delta \tilde{\omega}(s)}{\omega} \right] ,

405: %

406: \end{equation}

407: %

408: where $H(s)$ is given by Eq.~(\ref{cavTf}). In the presence of length

409: and frequency variations, feedback control will drive the error signal 

410: toward the null point, $\delta \tilde{V} = 0$, thus maintaining

411: dynamic resonance according to Eq.~(\ref{dynResF}).

412:                         

413: 

414: 

415: The response of the PDH signal to either length or laser frequency 

416: deviations can be found from Eq.~(\ref{PDH}). The normalized

417: length-to-signal transfer function is given by

418: %

419: \begin{equation}

420: %

421:    H_L(s) = \frac{H(s)}{H(0)} = 

422:       \frac{1 - r_a r_b}{1 - r_a r_b e^{-2 s T}}.

423: %

424: \end{equation}

425: %

426: A Bode plot (magnitude and phase) of $H_L$ is shown in 

427: Fig.~\ref{tfLength} for the LIGO \cite{Abramovici:1992} 

428: Fabry-Perot cavities with $L=4$ km, $r_a=0.985$, and $r_b=1$.

429: The magnitude of the transfer function,

430: %

431: \begin{equation}\label{airy}

432: %

433:    |H_L(i\Omega)| = \frac{1}{\sqrt{1 + F \sin^2 \Omega T}},

434: %

435: \end{equation}

436: %

437: is the square-root of the well-known Airy function with the 

438: coefficient of finesse $F = 4 r_a r_b/(1 - r_a r_b)^2$. 

439: (In optics, the Airy function describes the intensity profile 

440: of a Fabry-Perot cavity \cite{Born:1980}.) 

441:  

442: 

443: \begin{figure}

444:    \centering\includegraphics[width=0.46\textwidth]{fig3}

445:    \caption{Bode plot of $H_{\omega}(i\Omega)$. The sharp features 

446:    are due to the zero-pole pairs at multiples of the FSR.}

447:    \label{tfFreq}

448: \end{figure}

449: 

450: 

451: 

452: The transfer function $H_L$ has an infinite number of poles:

453: %

454: \begin{equation}\label{poles}

455: %

456:    p_n = - \frac{1}{\tau} + i \omega_{\mathrm{fsr}} n ,

457: %

458: \end{equation}

459: %

460: where $n$ is integer, and $\tau$ is the storage time of the cavity,

461: %

462: \begin{equation}

463: %

464:    \tau = \frac{2T}{|\ln (r_a r_b)|}.

465: %

466: \end{equation}

467: %

468: Therefore, $H_L$ can be written as the infinite product,

469: %

470: \begin{equation}

471: %

472:    H_L(s) = e^{s T} \prod \limits_{n=-\infty}^{\infty}

473:       \frac{p_n}{p_n - s},

474: %

475: \end{equation}

476: %

477: which can be truncated to a finite number of terms for use in 

478: control system design. 

479: 

480: 

481: 

482: The response of a Fabry-Perot cavity to laser frequency variations 

483: is very different from its response to length variations. The

484: normalized frequency-to-signal transfer function is given by 

485: %

486: \begin{equation}

487: %

488:    H_{\omega}(s) = C(s) H_L(s),

489: %

490: \end{equation}

491: %

492: or, more explicitly as

493: %

494: \begin{equation}

495: %

496:    H_{\omega}(s) = \left( \frac{1 - e^{-2 s T}}{2 s T} \right) \;

497:       \left( \frac{1 - r_a r_b}{1 - r_a r_b e^{-2 s T}} \right).

498: %

499: \end{equation}

500: %

501: A Bode plot of $H_{\omega}$, calculated for the same parameters as 

502: for $H_L$, is shown in Fig.~\ref{tfFreq}. The transfer function 

503: $H_{\omega}$ has zeros given by Eq.~(\ref{zeros}) with $n \neq 0$,

504: and poles given by Eq.~(\ref{poles}). The poles and zeros come in

505: pairs except for the lowest order pole, $p_0$, which does not have 

506: a matching zero. Therefore, $H_{\omega}$ can be written as the

507: infinite product,

508: %

509: \begin{equation}

510: %

511:    H_{\omega}(s) = \frac{p_0}{p_0 - s}

512:       {\prod\limits_{n = -\infty}^{\infty}}'

513:       \left( \frac{1 - s/z_n}{1 - s/p_n} \right) ,

514: %

515: \end{equation}

516: %

517: where the prime indicates that $n=0$ term is omitted from the

518: product.

519: 

520: 

521: The zeros in the transfer function indicate that the cavity does not 

522: respond ($\delta E = 0$) to laser frequency deviations if these 

523: deviations occur at multiples of the cavity FSR. In this case, 

524: the amplitude of the circulating field is constant while the 

525: phase of the circulating field is changing with the phase of the 

526: input laser field.  

527: 

528: 

529: In summary, we have shown that resonance can be maintained in a

530: Fabry-Perot cavity even when the cavity length and laser frequency

531: are changing. In this dynamic resonance state, changes in the laser

532: frequency and changes in  the cavity length play very different roles

533: (Eq.~(\ref{dynResF})) in contrast to the quasi-static resonance state

534: where they appear equally (Eq.~(\ref{qstatResF})). Maintenance of

535: dynamic resonance requires that the  frequency-to-length transfer

536: function, $C(s)$, be taken into account when compensating for length

537: variations by frequency changes and vice versa. Compensation for

538: length variations by frequency changes becomes increasingly more

539: difficult at frequencies above the FSR, and impossible at multiples

540: of the FSR.

541: 

542: 

543: As can be seen in Fig.~\ref{tfFreq}, the response of the PDH error 

544: signal to laser frequency variations decreases as $1/\Omega$ for

545: $\Omega\gg\tau^{-1}$ and becomes strongly suppressed at frequencies 

546: equal to multiples of the cavity FSR. In contrast, the response  

547: to length variations is a periodic function of frequency as shown in 

548: Fig.~\ref{tfLength}. For $\Omega\gg\tau^{-1}$, it also decreases as

549: $1/\Omega$ but only to the level of $(1+F)^{-\frac{1}{2}}$ and then 

550: returns to its maximum value. Thus, at multiples of the FSR, the 

551: sensitivity to length variations is maximum while the sensitivity to 

552: frequency variations is minimum.

553: 

554: 

555: Both these features  suggest searches for gravitational waves at

556: frequencies near multiples of the FSR. However, because gravitational

557: waves interact with the light as well as the mirrors, the response of

558: an optimally-oriented interferometer is equivalent to $H_{\omega}(s)$

559: and not to $H_L(s)$ \cite{Mizuno:1997}. Thus, an optimally-oriented

560: interferometer does not respond to gravitational wave at multiples of

561: the FSR. However, for other orientations gravitational waves can be

562: detected with enhanced sensitivity at multiples of the cavity FSR

563: \cite{Schilling:1997}.

564: 

565: 

566: We thank Robert Coldwell, Guido Mueller and David Shoemaker for 

567: comments on the paper. This research was supported by the National 

568: Science Foundation under grants PHY-9210038 and PHY-0070854.

569: 

570: 

571: 

572: \begin{thebibliography}{10}

573: 

574: \bibitem{Vaughan:1989}

575: J.M. Vaughan. {\it The {Fabry-Perot} interferometer: history, theory, 

576: practice, and applications}. (Adam Hilger, Bristol, England, 1989).

577: 

578: \bibitem{Abramovici:1992}

579: A.~Abramovici {\it et~al.}, Science {\bf 256}, 325 (1992).

580: 

581: \bibitem{Bradaschia:1990}

582: C.~Bradaschia {\it et~al.}, Nucl. Instrum. Methods Phys. 

583: Res. A {\bf 289}, 518 (1990).

584: 

585: \bibitem{Camp:1995}

586: J.~Camp, L.~Sievers, R.~Bork, and J.~Heefner, Opt. Lett.

587: {\bf 20}, 2463 (1995).

588: 

589: \bibitem{Mizuno:1997}

590: J.~Mizuno, A.~R\"{u}diger, R.~Schilling, W.~Winkler, and K.~Danzmann,

591: Opt. Commun. {\bf 138}, 383 (1997).

592: 

593: \bibitem{Chickarmane:1998}

594: V.~Chickarmane, S.V. Dhurandhar, R.~Barillet, P. Hello, and J.-Y. Vinet,

595: Appl. Opt. {\bf 37}, 3236 (1998).

596: 

597: \bibitem{Pai:2000}

598: A.~Pai, S.V. Dhurandhar, P.~Hello, and J.-Y. Vinet,

599: Eur. Phys. J. D {\bf 8}, 333 (2000).

600: 

601: \bibitem{Redding:1997}

602: D.~Redding, M.~Regehr, and L.~Sievers,

603: {\it Dynamic models of Fabry-Perot interferometers.}

604: LIGO Technical Report T970227, Calif. Inst. Tech., 1997.

605: 

606: \bibitem{Sigg:1997}

607: D.~Sigg {\it et~al.},

608: {\it Frequency response of the LIGO interferometer.}

609: LIGO Technical Report T970084, Mass. Inst. Tech., 1997.

610: 

611: \bibitem{Bhawal:1998}

612: B.~Bhawal, J. Opt. Soc. Am. A {\bf 15}, 120 (1998).

613: 

614: \bibitem{Beausoleil:1999}

615: R.G. Beausoleil and D. Sigg, J. Opt. Soc. Am. A {\bf 16}, 2990 (1999).

616: 

617: \bibitem{Yamamoto:2000}

618: H.~Yamamoto {\it et~al.}, in {\em Gravitational Wave Detection II.}

619: Proceedings of the 2nd TAMA International Workshop, Tokyo, Japan,

620: edited by S.~Kawamura and N.~Mio (Universal Academy Press, Tokyo, 2000)

621: 

622: \bibitem{Lawrence:1999}

623: M.J. Lawrence, B.~Willke, M.E. Husman, E.K. Gustafson, and R.L. Byer,

624: J. Opt. Soc. Am. B {\bf 16}, 523 (1999).

625: 

626: \bibitem{Rakhmanov:doppler}

627: M.~Rakhmanov, Appl. Opt. {\bf 40}, 1942 (2001).

628: 

629: \bibitem{Drever:1983}

630: R.W.P. Drever, J.L. Hall, F.V. Kowalski, J.~Hough, G.M. Ford, 

631: A.J. Munley, and H.~Ward, Appl. Phys. B {\bf 31}, 97 (1983).

632: 

633: \bibitem{Born:1980}

634: M.~Born and E.~Wolf. {\it Principles of Optics} 

635: (Pergamon Press, Oxford, 1980), 6th edition.

636: 

637: \bibitem{Schilling:1997} R. Schilling, Class. Quant. Grav. {\bf 14},

638: 1513 (1997).

639: 

640: \end{thebibliography}

641: 

642: \end{document}

643: