1: % new changes marked by %wyq

2: %\documentclass[12pt,preprint]{aastex}

3: %\documentclass[preprint2]{aastex}

4: %\usepackage{natbib}

5: \documentclass{emulateapj}

6: \usepackage{natbib,graphicx,psfig}

7: 

8: %\def\topfraction{.9}

9: %\def\bottomfraction{.1}

10: %\def\textfraction{.1}

11: %\def\floatpagefraction{.5}

12: % if don't want \bf, change to {}

13: \def\bfedit {\bf \large}

14: %\def\bfedit { }

15: \def\refnew#1{(\ref{#1})}

16: \def \micron {\, \mu {\rm m}}

17: \def \ev {\, \rm ev}

18: \def \yrs {\, \rm yrs}

19: \def \cm {\, \rm cm}

20: \def \m {\, \rm m}

21: \def \mm {\, \rm mm}

22: \def \g {, \rm g}

23: \def \s {\, \rm s}

24: \def \K {\, \rm K}

25: \def \erg {\, \rm erg}

26: \def \mearth {\, M_\oplus}

27: \def \be {\begin{equation}}

28: \def \ee {\end{equation}}

29: \def\etal{et al.\,}

30: \def\lesssim{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$<$}}}}

31: \def\gtrsim{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$>$}}}}

32: \newcommand       \bea          {\begin{eqnarray}}

33: \newcommand       \eea          {\end{eqnarray}}

34: \newcommand       \kms          {{\rm\, km/s}}

35: \newcommand       \au           {\rm \, AU}

36: %\newcommand       \apj          {{\it Astrophys. J.}}

37: %\newcommand       \apjl          {{\it Astrophys. J.}}

38: %\newcommand       \aj          {{\it Astron. J.}}

39: %\newcommand       \aap          {{\it Astron. Astrophys.}}

40: %\newcommand       \mnras          {{\it Mon. Not. R. Astron. Soc.}}

41: %\newcommand       \nat          {{\it Nature}}

42: \newcommand{\vdag}{(v)^\dagger}

43: 

44: \begin{document}

45: 

46: \title{Hot Jupiters in binary star systems}

47: 

48: \author{Yanqin Wu\altaffilmark{1}, Norman W. Murray\altaffilmark{2,3} 

49: \& J. Michael Ramsahai\altaffilmark{2}}

50: \altaffiltext{1}{Department of Astronomy \& Astrophysics, 

51:  University of Toronto, Toronto, ON M5S 3H4, Canada}

52: \altaffiltext{2}{Canadian Institute of Theoretical Astrophysics,

53:  University of Toronto, Toronto, ON M5S 3H8, Canada}

54: \altaffiltext{3}{Canada Research Chair in Astrophysics}

55: \email{wu@astro.utoronto.ca; murray@cita.utoronto.ca}

56: 

57: \begin{abstract}

58: 

59: Radial velocity surveys find Jupiter mass planets with semi-major axes

60: $a$ less than 0.1 AU around $\sim 1\%$ of solar-type stars;  %

61: counting planets with $a$ as large as 5 AU, the fraction of stars

62: having planets reaches $\sim 10\%$ \citep{Marcy,Butler}. An  %

63: examination of the distribution of semi-major axes shows that there is

64: a clear excess of planets with orbital periods around 3 or 4 days,

65: corresponding to $a\approx 0.03$ AU, with a sharp cutoff at shorter

66: periods (see Figure 1). It is believed that Jupiter mass planets form

67: at large distances from their parent stars; some fraction then migrate

68: in to produce the short period objects.  We argue that a significant

69: fraction of the `hot Jupiters' ($ a<0.1$AU) may arise in binary star

70: systems in which the orbit of the binary is highly inclined to the

71: orbit of the planet. Mutual torques between the two orbits drive down

72: the minimum separation or periapse $r_p$ between the planet and its

73: host star (the Kozai mechanism). This periapse collapse is halted when

74: tidal friction on the planet circularizes the orbit faster than Kozai

75: torque can excite it. The same friction then circularizes the planet

76: orbit, producing hot Jupiters with the peak of the semimajor axis

77: distribution lying around 3 days. For the observed distributions of

78: binary separation, eccentricity and mass ratio, roughly $ 2.5\%$ of

79: planets with initial semimajor axis $a_p\approx 5\au$ will migrate to

80: within $0.1\au$ of their parent star. Kozai migration could

81: account for $10\%$ or more of the observed hot Jupiters. %

82: \end{abstract}

83: 

84: \keywords{binaries:general;planetary systems;celestial mechanics}

85: 

86: 

87: \section{Introduction to Kozai Migration}

88: \label{sec:intro}

89: 

90: Statistics from radial velocity planet searches \citep{Marcy,Butler}

91: show that

92: %Among the 1330 FGKM dwarfs studied by Marcy et al. (2005a), 

93: the occurrence rate of giant planets within 0.1 AU (``hot-Jupiters'')

94: is $\sim 1\%$; extrapolating to 20 AU the occurrence is $12\%$.  There

95: % 

96: is a clear "pile-up" of planets with orbital periods near 3 days

97: (Fig. \ref{fig:fig1}). Transit observations yield a similar fraction

98: of hot Juptiers \citep{Gould,Fressin}. %

99: %add to the inventory of

100: %hot-Jupiters but their statistics are yet unclear. 

101: What migration mechanisms can produce the observed feature in semi-major %

102: axis distributions represented by hot Jupiters? In this article we

103: focus on the mechanism known as Kozai migration. %

104: 

105: Consider a planet circling a star that is a member of a binary %

106: system. The mutual torques between the binary and planetary orbits

107: transfer angular momentum between the two while leaving the orbital

108: energies nearly unchanged. For mutual inclinations $I\gtrsim 40^\circ$

109: a resonance between the precession rate of the planet's nodal and

110: apsidal lines greatly enhances the effectiveness of this exchange of

111: angular momentum, producing large oscillations in the planet's angular

112: momentum \citep[Kozai cycles,][]{Kozai}.  The planet eccentricity %

113: ($e_p$) and periapse ($r_p\equiv a_p(1-e_p)$) oscillate with a

114: characteristic timescale \citep{HolmanTouma} 

115: %

116: \be %$

117: P_{\rm Kozai} \approx

118: {m_*\over{m_c}} {{P_c^2}\over{P_p}} (1-e_c^3)^{3/2},

119: \label{eq:kozaip}

120: \ee  %$

121: %

122: where $m_*$ and $m_c$ are the masses of the central and companion

123: stars, while $P_c$ and $P_p$ are the periods of the binary and

124: planetary orbits, respectively.  The binary eccentricity is denoted by

125: $e_c$.

126: %\citet{Takeda} investigated the role of binary companion in raising the

127: %planet eccentricity episodically to high values and compared the

128: %resulting eccentricity values against observed distributions.

129: \citet{HolmanTouma} and \citet{Takeda}, among others, have studied the %

130: role of these Kozai cycles in producing the eccentricities observed in

131: known exo-planets.

132: 

133: For sufficiently large $I$, $r_p$ can reach very small values,

134: allowing tidal dissipation to erode the orbit of the planet. 

135: \citet{EKE} were the first to propose that Kozai cycles, in

136: combination with tidal friction, can shrink the orbit of a inner

137: binary in a hierarchical triple system, leading to the formation of

138: contact binaries. \citet[][hereafter WM03]{WuMurray} have studied 

139: Kozai migration in application to exo-planets and found it to be the

140: only plausible explanation for the migration of the planetary object

141: HD80606b.

142: 

143: In the absence of any other modification of the gravitational

144: potential, the minimum $r_p$ may fall below the Roche radius ($r_R$)

145: and the planet may be destroyed.  However, there are a number of

146: competing torques that can limit the amount of angular momentum that

147: the Kozai torque can extract from the orbit of the planet, including

148: general relativistic (GR) corrections to Newtonian gravity, and

149: torques associated with the extended mass distribution of both the

150: primary star and the planet. The latter includes rotationally induced

151: planetary oblateness, the tidal bulge raised by the star on the

152: planet, the misalignment of this bulge produced by friction, and the

153: stellar counterparts of all these. These torques can halt the

154: Kozai-induced collapse in $r_p$ and promote planetary survival.

155: 

156: Which torque becomes competitive with the Kozai torque depends on the

157: system; for systems with very large binary semi-major axis ($a_c$) and

158: therefore very weak Kozai torque, the GR precession can halt the

159: reduction of $r_p$ before tides become important. However, for tighter

160: or more inclined binaries,

161: %systems with $a_p\sim 1$AU, $a_c\sim100$ AU, and

162: %$I\gtrsim40^\circ$, 

163: tidal friction sets the minimum $r_p$. Since the tidal torques depend

164: strongly on $r_p$, binary systems with a wide range of $a_c/a_p$ will

165: be stalled at essentially the same $r_p$, leading to a pile-up of hot

166: Jupiters at $a_p \sim 2 r_p$ when the planet orbits are later

167: circularized.

168: 

169: \section{Numerical Experiments}

170: \label{sec:experiment}

171: 

172: We quantify the effect of Kozai migration by considering an ensemble

173: of binary systems following that in

174: \citet{Takeda}.  These binaries are initially comprised of a solar-mass host star, a

175: jupiter-mass planet ($m_p = M_J$) orbiting at $5$ AU with an

176: eccentricity of $0.05$, and a binary companion of mass $0.23 M_\odot$

177: -- this is the peak of the observed mass ratio distribution in the

178: solar neighbourhood

179: \citep{Mayor}. The distribution in binary separation ($P(a_c)$) is assumed flat in

180: logarithmic $a_c$ ($a_c$ ranging from $20$ to $20,000$ AU). We set

181: $P(e_b) = 2 e_b$, a thermal distribution often adopted in binary

182: population synthesis. This latter choice hardly affects the

183: results. The last ansatz, our most sensitive yet most uncertain

184: assumption, takes $I$ to be isotropically distributed. Based on

185: studies of stellar spin and binary orbits

186: \citep{Hale}, this seems reasonable for $a_c>40$ AU, but may be less

187: appropriate for tighter binaries; polarimetry studies of protostellar

188: disks suggest that the circumstellar disk and the binary plane are

189: correlated for $a_c$ up to a few hundred AU

190: \citep{Jensen,Monin}. However, polarimetry estimates only

191: the projected angle between the two planes, and is strongly plagued by

192: interstellar polarization. The results should be taken with caution at

193: present.

194: 

195: \begin{figure}[t]

196: \centerline{\psfig{figure=fig1.ps,width=1.00\hsize}}

197: \caption[]{The histogram of the planet semi-major axis (logarithmic)

198: distribution. The thick solid curve is the observed radial velocity

199: planet distribution.  The planets detected by the transit technique

200: are added on top (dotted line) assuming that the detection

201: efficiencies are the same between the two techniques. The shaded area

202: shows the simulation result, with the vertical axis read at the

203: right. The peak at $\sim 3$ day orbital period corresponds to planets

204: that are Kozai migrated and later circularized.  The position and

205: width of this peak depends on a number of parameters (see

206: Eq. [\ref{eq:bcrit}]).  In particular, if planet radius is a

207: decreasing function of planet age, the width of the peak shrinks (see

208: Fig. \ref{fig:fig2}).  The narrow peak at $\sim 5$ AU corresponds to

209: planets that are unmigrated, remaining at their initial $a_p$.}

210: \label{fig:fig1}

211: \end{figure}

212: 

213: We produce an ensemble of $100,000$ systems. Out of these we %

214: select systems that can potentially perturb the planet to $r_p\lesssim

215: 0.1$ AU. To reach this distance, a planet starting at semi-major axis

216: $a_p$ (with a small eccentricity) will have to attain $e_{\rm max}

217: \geq 1-0.1/a_p$. Ignoring tidal dissipation,\footnote{Tidal

218: dissipation increases the Kozai integral and slightly raises the

219: minimum requirement on $I$ (WM03).} the Kozai integral (the planet's

220: orbital angular momentum in the normal of the binary plane) $H_K =

221: \sqrt{(1-e_p^2)} \cos I = {\rm constant}$. Taking a minimum $I

222: \approx 40 \deg$ during the Kozai cycles \citep[see,

223: e.g.,][]{HolmanTouma}, this yields a minimum initial inclination

224: required for producing hot Jupiters: $I \gtrsim 81 \deg$. This value

225: is independent of the binary separation or mass.  The fraction of

226: isotropically inclined systems that have such a misalignment is $\sim

227: 15\%$.

228: 

229: We then weed out planets that are likely dynamically unstable

230: according to the following fitting formula 

231: \be

232: {{a_p}\over{a_c}} \geq 0.330-0.417 e_c+0.069 e_c^2.

233: \label{eq:apmax-vert}

234: \ee 

235: % taking <e_c> = 2/3 = 0.677, get a_p/a_c = 0.05, or a_c < 100 AU are gone.

236: This expression is obtained by integrating the orbits of our initial

237: system for $10^4$ binary orbits, taking $I = 90 \deg$.  This

238: non-coplanar stability limit is $15\%$ to $30\%$ more restrictive than

239: the coplanar stability limit found by \citet{HolmanWiegert}. It is

240: used here as a rough proxy for systems that either eject their planets

241: quickly after formation, or are unable to form planets due to the

242: strong tidal influence of the companion.  This proceedure eliminates

243: many systems with $a_c < 100$ AU; we are left with $\sim 10\%$ of the

244: original ensemble that could potentially reach $< 0.1$ AU, if they are

245: not stalled by other torques at larger distances.

246: 

247: These remaining systems are integrated using secular equations

248: obtained by averaging over the orbital motions of both the planet and

249: the binary companion \citep{EKE}. These equations include the effects

250: of Kozai perturbation, tidal dissipation, GR precession, and tidal and

251: rotational bulge precessions.\footnote{In this study, we rely

252: exclusively on these secular equations. The actual dynamics may

253: deviate due to short term noises and mean-motion perturbations and

254: should be studied with N-body integration codes.} We use a Runge-Kutta

255: integrator with an adaptive step size set to keep the integration

256: error below a preset limit. We follow the procedure described in

257: WM03, which also lists values for the various parameters involved. In

258: particular, we choose the initial stellar spin direction to be aligned

259: with the initial orbit normal for the planet.

260: %Initial conditions, let stellar spin pointing initially at planet

261: %orbital direction -- very important, change the minimum distance the 

262: %system can reach, but didn't change the distribution too much...

263: 

264: The integration is stopped after $5$ Gyrs have passed, or when $5$

265: million timesteps are exhausted, or when $r_p<2 R_\odot$. The last

266: condition roughly corresponds to the planet overflowing its Roche %

267: lobe; however, {\it none} of the planet in our simulation reached this

268: state.\footnote{This is due to the strong dependence of the tidal timescale

269: on $r_p$; tidal distortions act as a barrier,

270: maintaining $r_p\gtrsim r_R$.}  % 

271: %%

272: The limit on the number of integration timesteps is usually reached if

273: Kozai oscillations have been effectively halted by rapid tidal or

274: other precessions; in that case the subsequent dynamical evolution of

275: the planet simply reduces $e_p$.  We then use a simplified code,

276: including only the effects of tidal dissipation on the planet orbit

277: and planet/stellar spins, to finish integrating to $5$ Gyrs.

278: 

279: We find that about $2.5\%$ of our ensemble eventually migrate %

280: inward of $0.1$ AU. The distribution of final semi-major axes is

281: concentrated between $0.02$ AU and $0.05$ AU with a peak at $0.03$

282: AU. Our hot Jupiters exhibit a pile-up at $\sim3$ day periods similar

283: to the observed population (Fig. \ref{fig:fig1}). %

284: 

285: Given the same initial $I$, tighter binaries produce a closer-in hot Jupiter

286: in a shorter amount of time.  Many of the hot Jupiters are tidally

287: ensnared on their first close approach to the host star

288: (Fig. \ref{fig:fig3}), with the Kozai period between $10^4$ to $10^8$

289: yrs.  Tidal circularization of these orbits then takes upward of

290: $10^7$ years.

291: % both numbers from tide_hut_followup.f output

292: 

293: The 3-day feature in the computed $a_p$ distribution appears wider

294: than the observed distribution. However, as just noted, closer-in

295: planets are migrated in earlier, so they still have larger radii and

296: larger stalling peraipses. Experimenting with the following time

297: evolution of planet radius,

298: \be

299: R_p = R_J \left[ 1 + \exp\left(- {t\over{\tau_{\rm shrink}}}\right)\right],

300: \ee

301: with $\tau_{\rm shrink}$ taken to be $3\times 10^7 \yrs$, we find that

302: the 3-day bump narrows significantly (Fig. \ref{fig:fig2}).

303: 

304: \begin{figure}[t]

305: \centerline{\psfig{figure=fig2.ps,width=1.00\hsize}}

306: \caption[]{Similar to Fig. \ref{fig:fig1}, except where we have taken the 

307: planet radius to shrink as $R_p = R_J [1 + \exp(-t/3\times 10^7 yrs)]$. 

308: The 3-day feature narrows significantly.}

309: \label{fig:fig2}

310: \end{figure}

311: 

312: 

313: \begin{figure}[t]

314: \centerline{\psfig{figure=fig3.ps,width=1.00\hsize}}

315: \caption[]{Migration history for a system that in the absence of tides

316: would have reached a minimum distance of $0.0004$ AU ($0.1 R_\odot$)

317: and been declared lost; in the presences of tides it

318: % ac=600 AU, cos i = 1.0e-2, ec=0.5

319: reaches a minimum distance of $0.013$ AU and

320: is later circularized at $a_p = 0.026$ AU. The four panels %

321: are: top-left, planet eccentricity as a function of time (in years);

322: top-right, relative inclination between the two orbit normals;

323: bottom-left, planet semi-major axis (solid, in AU) and periapse

324: (dotted); bottom-right, planet total orbital angular momentum ($J$,

325: solid) and its component along the orbit normal of the binary ($J_z$,

326: dotted), both in arbitrary units.  Kozai oscillation (which conserves

327: $J_z$) has proceeded for barely half a cycle before the orbital energy

328: of the planet is significantly dissipated and the planet is removed

329: from the influence of the binary companion.  Tidal dissipation

330: operates afterwards (during which $J$ is conserved). The inclination

331: angle evolves little in this example. }

332: \label{fig:fig3}  

333: \end{figure}

334: 

335: \section{Discussion}

336: \label{sec:discuss}

337: 

338: \subsection{Stalling Radius and the 3-day Pile-up}

339: \label{subsec:stall}

340: 

341: The periapse of a Kozai-migrating planet is stalled at a distance where the

342: eccentricity forcing due to the binary companion is counteracted by

343: the eccentricity damping by tidal dissipation.  Kozai forcing yields

344: \citep{EKE} 

345: \be

346: {1\over e_p} {{de_p}\over{dt}} \approx 5 (1-e_p^2) {{m_c

347: n_c^2}\over{m_p+m_*+m_c}} {1\over{4 n_p \sqrt{1-e_p^2}

348: (1-e_c^2)^{3/2}}},

349: \label{eq:taue}

350: \ee

351: where $n_c = 2\pi/P_c$, $n_p = 2\pi/P_p$. The rate of tidal

352: eccentricity damping depends strongly on the periapse distance. Considering

353: only tides raised on the planet, we obtain\citep{Hut1981}

354: %

355: \be %$

356: {1\over e_p} {{de_p}\over{dt}} \approx - {{27 k_p G m_p}\over{2 R_p^3

357: Q_p n_p}} {1\over q}\left(1+{1\over q}\right)

358: \left({{R_p}\over{a_p}}\right)^8 {1\over{(1-e_p^2)^{13/2}}},

359: \label{eq:edothut}

360: \ee %$

361: %

362: where $k_p$ is the planet's tidal Love number, $Q_p$ its tidal

363: dissipation factor and $R_p$ its radius (see WM03). The mass ratio $q

364: = m_p/m_*$.  Equating the two rates, we obtain the stalling periapse

365: value,

366: %

367: \begin{eqnarray}

368: r_{p, \rm stall} & = & 0.015 AU \left[

369: \left({{m_*}\over{M_\odot}}\right)^{3/2}\,

370: \left({{M_J}\over{m_p}}\right)\,

371: \left({{3 \times 10^5}\over{Q_p}}\right)\,

372: \left({{k_p}\over{0.5}}\right)\, 

373: \right. \nonumber \\

374: & & \hskip-0.3in \left. \times  \left({{R_p}\over{R_J}}\right)^5\,

375: \left({{5 AU}\over{a_p}}\right)\,

376: \left({{0.23 M_\odot}\over{m_c}}\right)\,

377: \left({{a_c}\over{270 AU}}\right)^3\,

378: \right]^{1/6.5},

379: %\hskip0.2in \propto \hskip0.2in

380: %a_c^{3/6.5}\, m_c^{-1/6.5}\, Q_p^{-1/6.5}\, R_p^{5/6.5}.

381: \label{eq:bcrit}

382: \end{eqnarray}

383: %

384: where we have scaled variables by their representative values ($R_J$

385: is the radius of Jupiter). Coincidentally, $r_{p,\rm stall} \sim r_R$,

386: and it depends little on a variety of parameters, including stellar

387: mass, companion mass, planet mass, planet tidal $Q$ factor, and planet

388: initial orbit. This justifies our choices for these parameters in the

389: numerical experiment.

390: %%

391: \footnote{This mechanism works for other types of planets like hot

392: Neptunes or super-earths. Substituting into Eq. \refnew{eq:bcrit}

393: values appropriate for Neptune and Earth,

394: % Neptune, 17 M_E, Q=2e4, k_2 = 0.1, R = 0.3 R_J

395: % Earth, 1 M_E, Q=100, k_2=0.05, R=0.1 R_J

396: we obtain similar values for $r_{p,\rm stall}$.} %

397: %%

398: In our simulation, most binaries that give rise to hot Jupiters have

399: $a_c \in [100,1000]$ AU (Fig. \ref{fig:fig4}) and we have scaled $a_c$

400: here by roughly the median value.  Tighter binaries are relatively

401: unimportant -- planets in many of these systems are dynamically

402: unstable and are excluded from our study. 

403: 

404: In the subsequent tidal circularization, orbital angular momentum is

405: roughly conserved and the final $a_p \sim 2 r_{p, \rm stall}

406: \sim 0.03$ AU. 

407: 

408: 

409: \begin{figure}[t]

410: \centerline{\psfig{figure=fig4.ps,width=1.00\hsize}}

411: \caption[]{Parameters for the binary systems that produce Kozai

412: migration. In the top panel, the final $a_p$ (horizontal

413: axis) is plotted against $a_c$.  Smaller values of the former are in

414: general correlated with closer binaries (eq. \ref{eq:bcrit}), with

415: most hot Jupiters arising from binaries with $a_c \in [100,1000]$ AU.

416: %%

417: The bottom left panel shows the distributions of initial (thin lines)

418: and final (thicker lines) inclinations between the two orbital planes

419: -- the solid curves include systems with $a_p<0.025$ AU, and the

420: dotted curves all systems with final $a_p< 0.1$ AU, (similarly in the

421: right panel). The final inclination angles are much more spread out,

422: as the Kozai cycles convert inclination to eccentricity --

423: \citet{Fabrycky} gives a detailed explanation for the features. The

424: bottom right panel shows the distribution of final angles $\psi$

425: between the stellar spin axis and the planetary orbit normal. Most

426: systems (especially the tightest ones) have $\psi<50^\circ$, although

427: some stars may spin retrograde relative to the planet orbit.

428: %%

429: }

430: \label{fig:fig4}

431: \end{figure}

432: 

433: %The fraction of systems that produce hot Jupiters depends strongly on

434: %the fraction of systems that have high initial inclinations, as well

435: %as on the distributions of binary semi-major axis, but it depends

436: %little on the distributions of binary mass and eccentricity. 

437: 

438: The fraction of stars with Kozai migrated hot Jupiters is given by

439: \be  

440: f_{< 0.1} = f_b\cdot f_p\cdot f_{Kozai},

441: \ee  

442: where $f_b$ is the fraction of stars in binary systems, $f_p$ is the

443: fraction of solar type stars with Jupiter mass planets formed at a few

444: AU, and $f_{\rm Kozai}$ is the fraction of planets in binary star

445: systems that undergo Kozai migration to $a_p<0.1$AU.  Taking $f_b

446: \approx 0.65$ \citep{Mayor},

447: $f_p \gtrsim 0.07$ \citep{Marcy}, and $f_{\rm Kozai} \approx0.025$

448: (this work), we suggest that, at a minimum, $10\%$ of the known hot

449: Jupiters may be due to Kozai migration. The most uncertain number is

450: $f_p$.

451: % the fraction of

452: %solar-type stars that harbor Jupiter mass planets with initial $a_p

453: %\lesssim 30\au$, the latter chosen since that is the largest planetary

454: %semimajor axis we know of (that of Neptune). 

455: The value of $f_p$ we have quoted is the observed fraction in the Keck

456: sample, which is substantially complete up to $a_p\approx 3\au$.

457: Assuming the number of planets per $\au$ is flat up to $a_p=30\au$

458: gives $f_p=0.12$ and $f_{< 0.1}=0.002$. There is some indication that

459: the number of planets per $\au$ is an increasing function of $a_p$.

460: If $f_p=0.5$, more than half the hot Jupiters could be produced by the

461: Kozai mechanism.

462: 

463: \subsection{Predictions of the Kozai Migration Scenario}

464: \label{subsec:predict}

465: 

466: The number of hot Jupiters produced by Kozai migration can be

467: determined by observations in the near future, since Kozai migrated

468: planets must have a number of attributes. First, candidate Kozai hot

469: Jupiters will reside in binary star systems, although the binary mass

470: ratio may well be small; a brown dwarf companion can be dynamically as

471: effective as a solar-type companion (eq. [\ref{eq:bcrit}]). The study

472: by \citet{Mayor} establishes that $\sim 60\%$ of the stars in the

473: solar neighbourhood are actually binary or triple systems.  While

474: radial velocity surveys select against close binaries, studies by

475: \citet{Raghavan} show that at least $23\%$ of radial velocity planet

476: hosts have stellar companions. The discoveries of brown-dwarf

477: companions to the planet bearing stars HD 3651 \citep{Mugrauer} and HD

478: 89744 \citep{Mugrauer2} highlight the possibility that the existence

479: of dim companions will increase the known binary fraction of planet

480: bearing stars significantly. %

481: %%

482: The Kozai scenario predicts that the

483: binary fraction of hot Jupiters will be higher than that of systems

484: with more distant planets.  

485: %%

486: %

487: Binary-induced radial velocity trends induced on the primary by a

488: stellar companion will be of order

489: %

490: \be  %$

491: 5f\left({m_c\over 0.3M_\odot}\right)

492: \left({100AU\over a_c}\right)^2\m/\s/yr,

493: \ee  %$

494: %

495: where $f$ is the sine of the angle between the line of sight and the

496: stellar velocity. This is clearly detectable at the current

497: sensitivity of radial velocity surveys \citep{Wright07}. The companion

498: will also induce an astrometric acceleration of a few

499: micro-arcsecond/yr/yr, detectable by SIM or GAIA.

500: 

501: %star, and potential stellar companions only await being found. This is

502: %evident in the case of HD 3651 B, a dwarf companion found after its

503: %primary star and planet (Mugrauer, Seifahrt et al., 

504: %

505: %Though one worries about the brown-dwarf desert

506: %being real and there are not that many brown-dwarf companions as

507: %naively think. \citet{Desidera} discover little difference between

508: %planets in binary systems and planets outside, however, their binary

509: %mass-ratio does not involve many low-mass companions, and a mass ratio

510: %of $0.23$ seems to be the preferred value. So we may still be missing

511: %a lot of M-dwarf companions.

512: %Current statistics seems to suggest that planets do not avoid binary

513: %systems, at least systems which are wider than $\sim $ (references,

514: %astro-ph)...

515: %the number of binary stars with mass ratio

516: %$b$ increases as $b^{-\alpha}$ with $\alpha\approx xx$ (Duquonnoy and

517: %Mayor). 

518: 

519: Second, Kozai systems have %$I\gtrsim40^\circ$, 

520: $I \in [30^\circ,150^\circ]$, with $I\approx 90^\circ$ not uncommon

521: (Fig. \ref{fig:fig4}). In transiting systems the binary orbit will be

522: in or near the plane of the sky.  This can be tested via both radial

523: velocity and astrometry. %

524: % but,  the velocity is in the plnae..

525: 

526: % long term radial velocity trends and/or

527: %astrometric orbits of a single component, or by measuring relative

528: %radial velocities and/or relative astrometric motions of both

529: %components.

530: 

531: %{\bf The following useul? -- I don't think the rv trend is measurable

532: %-- the period will be 1000 yrs.  Remove? also the transiting system

533: %will likely have binaries on the plane, therefore no radial velocity

534: %possible}  while measurements of radial velocities accurate to

535: %$\lesssim1\kms$ are becoming commonplace. The binary star-planet

536: %system HD 189733ABb is a promising candidate for this type of analysis

537: %(Bakos et al?{\bf check reference}). Our 100 AU binary, The accerleration of this motion can be

538: %measured by GAIA or SIM to determine the binary orientation. {\bf

539: %also, this is not quite exact...}

540: % the reduced mass particle moves 6 milli-arcsec, the host star

541: % 1 milli-arcsec

542: % check reference above... is this a hot Jupiter?

543: 

544: % none of the hot jupiters currently have companions

545: 

546: Third, the angle between the spin axis of the primary star (assumed to

547: be the orbit normal of the planet at formation) and the present-day

548: planet orbit normal will range from 0 to 130 degrees

549: (Fig. \ref{fig:fig4}) with the values between 0 and 50 degrees being

550: preferred.  This angle can be determined if both the spin period of

551: the star as well as its rotational velocity $v\sin i$ can be

552: independently measured. The angle projected onto the plane of the sky,

553: measurable using the Rossiter-McLaughlin effect, will have a similar

554: range.

555: 

556: Fourth, the semimajor axis ratio $a_p/a_p'$ with any second planet

557: will be small. This results from the requirement that the precession

558: rate induced by the second planet not break the Kozai resonance

559: \citep{WuMurray}. Radial velocity measurements can constrain the mass

560: and semimajor axis of any nearby planetary companions to hot Jupiters

561: \citep{Wright07}. A corollary is that the fraction of multiplanet

562: systems having hot Jupiters will be smaller than the fraction of

563: single planet systems with hot Jupiters.

564: 

565: %%{\bfedit \subsection{Two Concerns} %

566: %%\label{subsec:caveat}

567: 

568: %

569: Kozai-migrated planets dissipate many times their own binding energy

570: during tidal circularization.  \citet{OgilvieLin} find that tidally

571: dissipated energy is deposited throughout the bulk of the planet,

572: raising the possibility that the planet will expand

573: catastrophically. In contrast, \citet{WuTide} concludes that energy is

574: deposited exclusively near the photosphere, which would leave the

575: planet intact.

576: %

577: 

578: %

579: The theoretical situation is unclear, but the existence of hot

580: Jupiters suggests an answer. A plot of $e_p$ versus $a_p$ strongly

581: suggests that the low $e_p$'s of the hot Jupiters are the result of

582: tidal circularization, 

583: %wyq

584: {as the observed $e_p$'s follow closely the upper-bound set by the

585: tidal process \citep[see, e.g., Fig. 1 of][]{Wu03}.} If so, most or

586: all hot Jupiters have experienced rapid tidal heating and survived.

587: %

588: %\footnote{Triton, the retrograde moon of Neptune, has also

589: %survived the deposition of 30 times its own binding energy during

590: %$\sim 10^7$ years of tidal circularization.}

591: 

592: Another concern with the Kozai picture is raised by the

593: Rossiter-McLaughlin measurement of stellar obliquity, currently

594: available for 5 transiting planets \citep[see Table 2

595: of][]{Fabrycky}. All are consistent with zero obliquity.

596: %Collecting from literature, results for various transiting planets

597: %are, HD209458b (G0, -4.4+/- 1.4), HD189733 (K1, -1.4+/-1.1, true angle

598: %< 27), HD149026 (G0, 1.3M_sun, 11+/-15), TrEs1 (0.87 M_sun, 0+/-30),

599: %HD147506b (HAT-P-2b, 1.3M_sun, 1.2+/-14). 

600: Taken at face value, this is at variance with the above Kozai

601: prediction.

602: %

603: \footnote{HD147506 \citep{Winn,Loeillet}, a $1.3 M_\odot$

604: star with a massive planet, may have experienced tidal synchronization

605: in its surface-layer that would alter its apparent rotation axis.}

606: 

607: \subsection{Alternatives to Kozai Migration}

608: \label{subsec:alternative}

609: 

610: In Kozai migration, it is important that $r_p$ evolves on a time scale

611: no shorter than the tidal precession time scale; if $r_p$ were to

612: suddenly plunge from above to below the Roche radius, as for example

613: would be the case if two planets suffered a close encounter, the

614: inward scattered planet would not be stalled outside $r_R$. Instead

615: it would suffer rapid mass loss and likely be lost. In that case there

616: will be a cut-off in the distribution of $a_p$ at $2

617: r_R$\citep{FordRasio}, but not a pile-up.

618: 

619: Migration in a gas disk may also produce hot Jupiters.  If the disk

620: extends all the way to the star, one would observe a cut-off at $a_p

621: \sim r_R$; if the disk is truncated, e.g., by stellar magnetic fields

622: \citep{Linetal}, a feature will appear at an orbital period half that

623: of the inner edge of the disk.  However, spin periods and magnetic

624: fields of accreting stars show a substantial dispersion, which would

625: lead to a rather broad distribution in the disk inner radii, and hence

626: a smeared out feature in the distribution of planetary semimajor axis.

627: 

628: We have studied the role of a binary companion in increasing $e_p$ and

629: causing a gradual collapse in $r_p$.  But it is also plausible that

630: soft planet-planet scattering can gradually decrease $r_p$

631: \citep{JuricTremaine,Fordetal}. Moreover, Kozai oscillations can also

632: be excited by a second planet,\footnote{This second planet can be

633: placed on a highly inclined orbit by, e.g., planet-planet scattering.}

634: %wyq

635: in the absence of a binary stellar companion. As long as these or

636: other processes produce gentle eccentricity driving on $10^4$ to

637: $10^8$ year timescales, tidal effects will halt the periapse evolution

638: when $r_p

639: \sim r_R$. Tidal circularization then pushes the planets out to $a_p

640: \sim 2 r_R$ and produces a narrow pile-up of hot Jupiters there.

641: 

642: \acknowledgements 

643: %Most numerical computations in this work were performed on CITA's 

644: %McKinsey cluster. 

645: We acknowledge helpful discussions with Scott Gaudi, Daniel Fabrycky

646: and Andrew Gould, as well as NSERC discovery grants to YW and NM, and

647: an NSERC undergraduate fellowship to MR (summer 2006).

648: 

649: %\bibliographystyle{apj}

650: %\bibliography{mike}

651: 

652: 

653: \begin{thebibliography}{28}

654: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi

655: 

656: \bibitem[{{Butler} {et~al.}(2006){Butler}, {Wright}, {Marcy}, {Fischer},

657:   {Vogt}, {Tinney}, {Jones}, {Carter}, {Johnson}, {McCarthy}, \&

658:   {Penny}}]{Butler}

659: {Butler}, R.~P., {Wright}, J.~T., {Marcy}, G.~W., {Fischer}, D.~A., {Vogt},

660:   S.~S., {Tinney}, C.~G., {Jones}, H.~R.~A., {Carter}, B.~D., {Johnson}, J.~A.,

661:   {McCarthy}, C., \& {Penny}, A.~J. 2006, \apj, 646, 505

662: 

663: \bibitem[{{Chatterjee} {et~al.}(2007){Chatterjee}, {Ford}, \&

664:   {Rasio}}]{Fordetal}

665: {Chatterjee}, S., {Ford}, E.~B., \& {Rasio}, F.~A. 2007, ArXiv Astrophysics

666:   e-prints

667: 

668: \bibitem[{{Duquennoy} \& {Mayor}(1991)}]{Mayor}

669: {Duquennoy}, A. \& {Mayor}, M. 1991, \aap, 248, 485

670: 

671: \bibitem[{{Eggleton} \& {Kiseleva-Eggleton}(2001)}]{EKE}

672: {Eggleton}, P.~P. \& {Kiseleva-Eggleton}, L. 2001, \apj, 562, 1012

673: 

674: \bibitem[{{Fabrycky} \& {Tremaine}(2007)}]{Fabrycky}

675: {Fabrycky}, D. \& {Tremaine}, S. 2007, ArXiv e-prints, 705

676: 

677: \bibitem[{{Ford} \& {Rasio}(2006)}]{FordRasio}

678: {Ford}, E.~B. \& {Rasio}, F.~A. 2006, \apjl, 638, L45

679: 

680: \bibitem[{{Fressin} {et~al.}(2007){Fressin}, {Guillot}, {Morello}, \&

681:   {Pont}}]{Fressin}

682: {Fressin}, F., {Guillot}, T., {Morello}, V., \& {Pont}, F. 2007, ArXiv

683:   e-prints, 704

684: 

685: \bibitem[{{Gould} {et~al.}(2006){Gould}, {Dorsher}, {Gaudi}, \&

686:   {Udalski}}]{Gould}

687: {Gould}, A., {Dorsher}, S., {Gaudi}, B.~S., \& {Udalski}, A. 2006, Acta

688:   Astronomica, 56, 1

689: 

690: \bibitem[{{Hale}(1994)}]{Hale}

691: {Hale}, A. 1994, \aj, 107, 306

692: 

693: \bibitem[{{Holman} {et~al.}(1997){Holman}, {Touma}, \&

694:   {Tremaine}}]{HolmanTouma}

695: {Holman}, M., {Touma}, J., \& {Tremaine}, S. 1997, \nat, 386, 254

696: 

697: \bibitem[{{Holman} \& {Wiegert}(1999)}]{HolmanWiegert}

698: {Holman}, M.~J. \& {Wiegert}, P.~A. 1999, \aj, 117, 621

699: 

700: \bibitem[{{Hut}(1981)}]{Hut1981}

701: {Hut}, P. 1981, \aap, 99, 126

702: 

703: \bibitem[{{Jensen} {et~al.}(2004){Jensen}, {Mathieu}, {Donar}, \&

704:   {Dullighan}}]{Jensen}

705: {Jensen}, E.~L.~N., {Mathieu}, R.~D., {Donar}, A.~X., \& {Dullighan}, A. 2004,

706:   \apj, 600, 789

707: 

708: \bibitem[{{Juric} \& {Tremaine}(2007)}]{JuricTremaine}

709: {Juric}, M. \& {Tremaine}, S. 2007, ArXiv Astrophysics e-prints

710: 

711: \bibitem[{{Kozai}(1962)}]{Kozai}

712: {Kozai}, Y. 1962, \aj, 67, 591

713: 

714: \bibitem[{{Lin} {et~al.}(1996){Lin}, {Bodenheimer}, \& {Richardson}}]{Linetal}

715: {Lin}, D.~N.~C., {Bodenheimer}, P., \& {Richardson}, D.~C. 1996, \nat, 380, 606

716: 

717: \bibitem[{{Loeillet} {et~al.}(2007){Loeillet}, {Shporer}, {Bouchy}, {Pont},

718:   {Mazeh}, {Beuzit}, {Boisse}, {Bonfils}, {Da Silva}, {Delfosse}, {Desort},

719:   {Ecuvillon}, {Forveille}, {Galland}, {Gallenne}, {Hebrard}, {Lagrange},

720:   {Lovis}, {Mayor}, {Moutou}, {Pepe}, {Perrier}, {Queloz}, {Segransan},

721:   {Sivan}, {Santos}, {Tsodikovich}, {Udry}, \& {Vidal-Madjar}}]{Loeillet}

722: {Loeillet}, B., {Shporer}, A., {Bouchy}, F., {Pont}, F., {Mazeh}, T., {Beuzit},

723:   J.~L., {Boisse}, I., {Bonfils}, X., {Da Silva}, R., {Delfosse}, X., {Desort},

724:   M., {Ecuvillon}, A., {Forveille}, T., {Galland}, F., {Gallenne}, A.,

725:   {Hebrard}, G., {Lagrange}, A.~M., {Lovis}, C., {Mayor}, M., {Moutou}, C.,

726:   {Pepe}, F., {Perrier}, C., {Queloz}, D., {Segransan}, D., {Sivan}, J.~P.,

727:   {Santos}, N.~C., {Tsodikovich}, Y., {Udry}, S., \& {Vidal-Madjar}, A. 2007,

728:   ArXiv e-prints, 707

729: 

730: \bibitem[{{Marcy} {et~al.}(2005){Marcy}, {Butler}, {Fischer}, {Vogt}, {Wright},

731:   {Tinney}, \& {Jones}}]{Marcy}

732: {Marcy}, G., {Butler}, R.~P., {Fischer}, D., {Vogt}, S., {Wright}, J.~T.,

733:   {Tinney}, C.~G., \& {Jones}, H.~R.~A. 2005, Progress of Theoretical Physics

734:   Supplement, 158, 24

735: 

736: \bibitem[{{Monin} {et~al.}(2006){Monin}, {M{\'e}nard}, \& {Peretto}}]{Monin}

737: {Monin}, J.-L., {M{\'e}nard}, F., \& {Peretto}, N. 2006, \aap, 446, 201

738: 

739: \bibitem[{{Mugrauer} {et~al.}(2004){Mugrauer}, {Neuh{\"a}user}, {Mazeh},

740:   {Guenther}, \& {Fern{\'a}ndez}}]{Mugrauer2}

741: {Mugrauer}, M., {Neuh{\"a}user}, R., {Mazeh}, T., {Guenther}, E., \&

742:   {Fern{\'a}ndez}, M. 2004, Astronomische Nachrichten, 325, 718

743: 

744: \bibitem[{{Mugrauer} {et~al.}(2006){Mugrauer}, {Seifahrt}, {Neuh{\"a}user}, \&

745:   {Mazeh}}]{Mugrauer}

746: {Mugrauer}, M., {Seifahrt}, A., {Neuh{\"a}user}, R., \& {Mazeh}, T. 2006,

747:   \mnras, 373, L31

748: 

749: \bibitem[{{Ogilvie} \& {Lin}(2004)}]{OgilvieLin}

750: {Ogilvie}, G.~I. \& {Lin}, D.~N.~C. 2004, \apj, 610, 477

751: 

752: \bibitem[{{Raghavan} {et~al.}(2006){Raghavan}, {Henry}, {Mason}, {Subasavage},

753:   {Jao}, {Beaulieu}, \& {Hambly}}]{Raghavan}

754: {Raghavan}, D., {Henry}, T.~J., {Mason}, B.~D., {Subasavage}, J.~P., {Jao},

755:   W.-C., {Beaulieu}, T.~D., \& {Hambly}, N.~C. 2006, \apj, 646, 523

756: 

757: \bibitem[{{Takeda} \& {Rasio}(2005)}]{Takeda}

758: {Takeda}, G. \& {Rasio}, F.~A. 2005, \apj, 627, 1001

759: 

760: \bibitem[{{Winn} {et~al.}(2007){Winn}, {Asher Johnson}, {Peek}, {Marcy},

761:   {Bakos}, {Enya}, {Narita}, {Suto}, {Turner}, \& {Vogt}}]{Winn}

762: {Winn}, J.~N., {Asher Johnson}, J., {Peek}, K.~M.~G., {Marcy}, G.~W., {Bakos},

763:   G.~A., {Enya}, K., {Narita}, N., {Suto}, Y., {Turner}, E.~L., \& {Vogt},

764:   S.~S. 2007, ArXiv e-prints, 707

765: 

766: \bibitem[{{Wright} {et~al.}(2007){Wright}, {Marcy}, {Fischer}, {Butler},

767:   {Vogt}, {Tinney}, {Jones}, {Carter}, {Johnson}, {McCarthy}, \&

768:   {Apps}}]{Wright07}

769: {Wright}, J.~T., {Marcy}, G.~W., {Fischer}, D.~A., {Butler}, R.~P., {Vogt},

770:   S.~S., {Tinney}, C.~G., {Jones}, H.~R.~A., {Carter}, B.~D., {Johnson}, J.~A.,

771:   {McCarthy}, C., \& {Apps}, K. 2007, \apj, 657, 533

772: 

773: \bibitem[Wu(2003)]{Wu03} Wu, Y.\ 2003, Scientific Frontiers 

774: in Research on Extrasolar Planets, 294, 213

775: 

776: \bibitem[{{Wu}(2005)}]{WuTide}

777: {Wu}, Y. 2005, \apj, 635, 688

778: 

779: \bibitem[{{Wu} \& {Murray}(2003)}]{WuMurray}

780: {Wu}, Y. \& {Murray}, N. 2003, \apj, 589, 605

781: 

782: \end{thebibliography}

783: 

784: \end{document}

785: