0804.2259/ms.tex
1: \documentclass[preprint,12pt]{aastex}
2: %\usepackage{latexsym,graphicx,natbib}
3: %\usepackage{apjfonts,emulateapj5,psfig}
4: %\usepackage{apjfonts,emulateapj5}
5: 
6: \shortauthors{Winn et al.~2007}
7: \shorttitle{Prograde Orbit of TrES-2b}
8: 
9: \begin{document}
10: 
11: % ------------------------------------------------------------------------
12: % New commands
13: %
14: \def\ltsima{$\; \buildrel < \over \sim \;$}
15: \def\lsim{\lower.5ex\hbox{\ltsima}}
16: \def\gtsima{$\; \buildrel > \over \sim \;$}
17: \def\gsim{\lower.5ex\hbox{\gtsima}}
18:                                                                                           
19: % -------------------------------------------------------------------------
20: %
21: 
22: \bibliographystyle{apj}
23: 
24: \title{
25: The Prograde Orbit of Exoplanet TrES-2b\altaffilmark{1}
26: }
27: 
28: \author{
29: Joshua N.\ Winn\altaffilmark{2},
30: John Asher Johnson\altaffilmark{3,4},
31: Norio Narita\altaffilmark{5},
32: Yasushi Suto\altaffilmark{5},\\
33: Edwin L.\ Turner\altaffilmark{6},
34: Debra A.\ Fischer,\altaffilmark{7},
35: R.\ Paul Butler\altaffilmark{8},
36: Steven S.\ Vogt\altaffilmark{9},\\
37: Francis T.\ O'Donovan\altaffilmark{10},
38: B.\ Scott Gaudi\altaffilmark{11}
39: }
40: 
41: \altaffiltext{1}{Data presented herein were obtained at the W.~M.~Keck
42:   Observatory, which is operated as a scientific partnership among the
43:   California Institute of Technology, the University of California,
44:   and the National Aeronautics and Space Administration, and was made
45:   possible by the generous financial support of the W.~M.~Keck
46:   Foundation.}
47: 
48: \altaffiltext{2}{Department of Physics, and Kavli Institute for
49:   Astrophysics and Space Research, Massachusetts Institute of
50:   Technology, Cambridge, MA 02139}
51: 
52: \altaffiltext{3}{Department of Astronomy, University of California,
53:   Mail Code 3411, Berkeley, CA 94720}
54: 
55: \altaffiltext{4}{Present address: Institute for Astronomy, University
56:   of Hawaii, 2226 Woodlawn Drive, Honolulu, HI 96822}
57: 
58: \altaffiltext{5}{Department of Physics, The University of Tokyo, Tokyo
59:   113-0033, Japan}
60: 
61: \altaffiltext{6}{Princeton University Observatory, Peyton Hall,
62:   Princeton, NJ 08544}
63: 
64: \altaffiltext{7}{Department of Physics and Astronomy, San Francisco
65:   State University, San Francisco, CA 94132}
66: 
67: \altaffiltext{8}{Department of Terrestrial Magnetism, Carnegie
68:   Institution of Washington, 5241 Broad Branch Road NW, Washington
69:   D.C.\ 20015-1305}
70: 
71: \altaffiltext{9}{UCO/Lick Observatory, University of California at
72:   Santa Cruz, Santa Cruz, CA 95064}
73: 
74: \altaffiltext{10}{California Institute of Technology, 1200 E.\
75:   California Blvd., Pasadena, CA 91125}
76: 
77: \altaffiltext{11}{Department of Astronomy, Ohio State University, 140
78:   W.~18th Ave., Columbus, OH 43210}
79: 
80: \begin{abstract}
81:   
82:   We monitored the Doppler shift of the G0V star TrES-2 throughout a
83:   transit of its giant planet. The anomalous Doppler shift due to
84:   stellar rotation (the Rossiter-McLaughlin effect) is discernible in
85:   the data, with a signal-to-noise ratio of 2.9, even though the star
86:   is a slow rotator. By modeling this effect we find that the planet's
87:   trajectory across the face of the star is tilted by $-9 \pm 12$~deg
88:   relative to the projected stellar equator. With 98\% confidence, the
89:   orbit is prograde.
90: 
91: \end{abstract}
92: 
93: \keywords{planetary systems --- planetary systems: formation ---
94:   stars:~individual (TrES-2, GSC 03549-02811) --- stars:~rotation}
95: 
96: \section{Introduction}
97: 
98: A small fraction of Sun-like stars have giant planets with orbital
99: periods smaller than about 10 days (Marcy et al.~2005, Udry \& Santos
100: 2007). The existence of these planets was a surprise, because it was
101: expected that giant planets would only be found beyond the ``snow
102: line,'' with orbital distances greater than a few astronomical
103: units. Other surprises have come from detailed studies of individual
104: objects. Some are found on highly eccentric orbits (Johnson et
105: al.~2006, Bakos et al.~2007, Maness et al.~2007, Johns-Krull et
106: al.~2008). Some have mean densities that are quite small (Knutson et
107: al.~2007, Mandushev et al.~2007) or large (Sato et al.~2005, Torres et
108: al.~2007) in comparison with Jupiter.
109: 
110: However, in at least one sense, the close-in giant planets have
111: fulfilled prior expectations: they orbit their host stars in the
112: prograde direction, relative to the sense of the stellar
113: rotation. This is true, at least, of the 6 systems for which
114: measurements of spin-orbit alignment have been reported (Queloz et
115: al.~2000; Wolf et al.~2007; Narita et al.~2007a,b; Loeillet et
116: al.~2007; Winn et al.~2005, 2006, 2007a). In all of these cases but
117: one, the sky projections of the orbital axis and the stellar rotation
118: axis are observed to be fairly well-aligned, with measurement
119: precisions ranging from about 1.5 to 30 deg. The exception is
120: HD~17156, for which the angle between those axes was found to be
121: $62\pm 25$~deg (Narita et al.~2007b). In all of these cases, the
122: measurement technique relies upon the Rossiter-McLaughlin (RM) effect,
123: the anomalous Doppler shift that occurs during transits due to stellar
124: rotation (see, e.g., Queloz et al.~2000, Ohta et al.~2005, Gim\'enez
125: 2006, Gaudi \& Winn 2007, Winn 2007).
126: 
127: A close alignment between the orbital and rotational axes seems
128: natural because this pattern prevails in the Solar system, and because
129: the angular momenta of the parent star and the planetary orbits
130: presumably derive from the same protostellar disk. However, some
131: theories of planetary migration---proposed to explain how giant
132: planets attain short-period orbits---predict occasionally large
133: misalignments (Chatterjee et al.~2007, Fabrycky \& Tremaine 2007, Wu
134: et al.~2007, Nagasawa et al.~2008). These theories, as well as the
135: general history of surprises in this field, provide motivation to
136: continue measuring exoplanetary spin-orbit alignment.
137: 
138: In this paper, we present a measurement of the RM effect for the
139: transiting exoplanetary system TrES-2. This system was discovered by
140: O'Donovan et al.~(2006). It consists of a planet with a mass of
141: $1.2$~$M_{\rm Jup}$ and radius $1.2$~$R_{\rm Jup}$ orbiting a G0V star
142: with a period of 2.5~d (O'Donovan et al.~2006, Holman et al.~2007,
143: Sozzetti et al.~2007). It did not stand out as a promising RM target
144: because the star is relatively faint ($V=11.4$) and is a slow rotator
145: ($v\sin i_\star = 2.0\pm 1.5$~km~s$^{-1}$; O'Donovan et al.~2006). On
146: the other hand, the transit occurs at a high impact parameter across
147: the stellar disk ($b=0.8540\pm 0.0062$; Holman et al.~2007), a
148: favorable circumstance for this type of measurement (Gaudi \& Winn
149: 2007). Furthermore, in our continuing effort to measure the spin-orbit
150: angles for a statistically meaningful number of systems, we do not
151: want to ignore stars with small sky-projected rotation rates. This is
152: because a small value of $v\sin i_\star$ might be caused by a small
153: value of $\sin i_\star$, i.e., there might be a large spin-orbit
154: misalignment. For these reasons, we pursued TrES-2. We describe the
155: new data in \S~2, the model that we used to interpret the data in
156: \S~3, and the results in \S~4.
157: 
158: \section{Observations and Data Reduction}
159: 
160: We observed a transit of TrES-2 on UT~2007~April~26 with the Keck~I
161: 10m telescope and the High Resolution Echelle Spectrometer (HIRES;
162: Vogt et al.~1994). We set up the instrument in the same manner that
163: has been used consistently for the California-Carnegie planet search
164: (Butler et al.~1996, 2006). In particular we employed the red
165: cross-disperser and used the I$_2$ absorption cell to calibrate the
166: instrumental response and the wavelength scale. The slit width was
167: $0\farcs85$ and the typical exposure time was 3-4~min, giving a
168: resolution of about 70,000 and a signal-to-noise ratio (SNR) of
169: approximately 200~pixel$^{-1}$. We observed the star for 4~hr
170: bracketing the predicted transit midpoint and obtained a total of 56
171: spectra, of which 30 were taken during the transit.
172: 
173: We also obtained two iodine-free spectra, with a higher SNR and higher
174: resolution. We used the sum of these spectra as a template for the
175: Doppler analysis, which was performed with the algorithm of Butler et
176: al.~(1996). We estimated the measurement error in the Doppler shift
177: derived from a given spectrum based on the scatter among the solutions
178: for individual 2~\AA~sections of the spectrum. The typical error was
179: 6~m~s$^{-1}$. The data are given in Table~1 and plotted in Figs.~1 and
180: 2. Also shown in those figures are data obtained previously by
181: O'Donovan et al.~(2006), consisting of 11 velocities measured with
182: Keck/HIRES using a different setup\footnote{Table~3 of O'Donovan et
183:   al.~(2006) gives incorrect values for the heliocentric Julian dates
184:   of the velocity measurements. The corrected dates were provided to
185:   us by D.~Charbonneau (private communication, 2007).}, as well as the
186: photometric data of Holman et al.~(2007).
187: 
188: %\clearpage
189: \begin{figure}[p]
190: \epsscale{1.0}
191: \plotone{f1.eps}
192: \caption{
193: Radial velocity measurements of TrES-2, from this work and from
194: O'Donovan et al.~(2006), as a function of orbital phase.
195: The best-fitting values of the systemic velocity have been
196: subtracted. The solid line is the best-fitting model.
197: \label{fig:1}}
198: \end{figure}
199: 
200: \begin{figure}[p]
201: \epsscale{0.75}
202: \plotone{f2.eps}
203: \caption{ 
204: {\it Top.} The $z$-band photometry of Holman et al.~(2007), averaged
205: into 1.5~min bins. The solid line is the best-fitting model. {\it
206:   Middle.} A close-up of the radial velocity data shown in Fig.~1,
207: centered on the midtransit time.  {\it Bottom.} Same, but the orbital
208: velocity has been subtracted and the post-midtransit data ($t > 0$)
209: have been inverted about the origin ($t\rightarrow -t$ and $\Delta v
210: \rightarrow -\Delta v$), highlighting the Rossiter-McLaughlin anomaly.
211: Filled symbols denote data from before midtransit, and open symbols
212: denote data from after midtransit.
213: \label{fig:2}}
214: \end{figure}
215: %\clearpage
216: 
217: \section{The Model}
218: 
219: To determine the projected spin-orbit angle and its uncertainty, we
220: simultaneously fitted a parametric model to the radial-velocity data
221: as well as the photometric data of Holman et al.~(2007). We included
222: the photometric data as a convenient way to account for the
223: uncertainties in the photometric parameters and their covariances with
224: the spin-orbit parameters, although in practice the photometric
225: uncertainties were irrelevant for this system.
226: 
227: The model is based on a circular orbit of a star and planet. The
228: photometric transit model was identical to the model used by Holman et
229: al.~(2007). To calculate the anomalous Doppler shift as a function of
230: the positions of the planet and star, we used the technique of Winn et
231: al.~(2005): we simulated in-transit spectra, and determined the
232: Doppler shifts using the same algorithm used on the actual data. The
233: simulations rely on a template spectrum (described below) that is
234: meant to mimic the emergent spectrum from a small portion of the
235: photosphere. At a given moment of the transit, we denote by $\epsilon$
236: the fractional loss of stellar flux, and we denote by $v_p$ the
237: line-of-sight velocity of the occulted portion of the stellar disk. To
238: represent the occulted portion of the stellar spectrum, we scaled the
239: template spectrum in flux by $\epsilon$ and shifted it in velocity by
240: $v_p$. We subtracted the scaled and shifted spectrum from a
241: rotationally-broadened template spectrum and then ``measured'' the
242: anomalous Doppler shift $\Delta v$. This was repeated for a grid of
243: $\{\epsilon, v_p\}$, and a polynomial function was fitted to the
244: resulting grid. We used this polynomial to calculate the anomalous
245: Doppler shift $\Delta v$ as a function of $\epsilon$ and $v_p$, which
246: are themselves functions of time. Differential rotation was ignored,
247: as its effects are expected to be negligible (Gaudi \& Winn 2007).
248: 
249: The template spectrum should be similar to that of TrES-2 but with
250: slightly narrower lines because of the lack of rotational
251: broadening. We experimented with two different empirical templates
252: based on observations of similar stars,\footnote{The two stars were
253:   HD~38858 ($T_{\rm eff} = 5726$~K, $\log g = 4.51\pm 0.08$,
254:   [Fe/H]~=~$-0.23\pm 0.04$, $v\sin i_\star = 0.3\pm 0.5$~km~s$^{-1}$)
255:   and HD~66428 ($T_{\rm eff} = 5752$~K, $\log g = 4.49\pm 0.08$,
256:   [Fe/H]~=~$+0.31\pm 0.04$, $v\sin i_\star = 0.0\pm
257:   0.5$~km~s$^{-1}$). The stellar parameters are from the SPOCS catalog
258:   (Valenti \& Fischer 2005).} finding that both templates gave results
259: consistent with the function $\Delta v = -\epsilon~v_p$. This function
260: is consistent with the analytic expressions of Ohta et al.~(2006) and
261: Gimenez~(2006), even though those analytic expressions do not attempt
262: to account for the spectral deconvolution. It is simpler than the
263: quadratic or cubic functions that we have derived for other systems
264: (Winn et al.~2005, 2006, 2007a). We do not know the reason for the
265: difference but it is possibly related to the much slower projected
266: rotation speed of TrES-2.
267: 
268: The fitting statistic was
269: \begin{eqnarray}
270: \chi^2 & = &
271: \sum_{j=1}^{1033}
272: \left[
273: \frac{f_j({\mathrm{obs}}) - f_j({\mathrm{calc}})}{\sigma_{f,j}}
274: \right]^2
275: +
276: \sum_{j=1}^{67}
277: \left[
278: \frac{v_j({\mathrm{obs}}) - v_j({\mathrm{calc}})}{\sigma_{v,j}}
279: \right]^2
280: ,
281: \end{eqnarray}
282: where $f_j$(obs) and $\sigma_{f,j}$ are the flux measurements and
283: uncertainties of Holman et al.~(2007), and $v_j$(obs) and
284: $\sigma_{v,j}$ are the radial-velocity measurements and uncertainties
285: from our new data and from O'Donovan et al.~(2006). The two model
286: parameters relating to the RM effect are the line-of-sight stellar
287: rotation velocity ($v \sin i_\star$), and the angle between the
288: projected stellar spin axis and orbit normal ($\lambda$). The
289: projected spin-orbit angle $\lambda$ ranges from $-180\arcdeg$ to
290: $+180\arcdeg$, and is measured counterclockwise on the sky from the
291: projected stellar rotational angular-momentum vector to the projected
292: orbital angular-momentum vector (see Ohta et al.~2007 or Gaudi \& Winn
293: 2007 for a diagram). If we define stellar ``north'' by the sky
294: projection of the stellar angular-momentum vector, then when
295: $\lambda=0\arcdeg$ the axes are aligned and the planet moves directly
296: ``eastward'' across the face of the star, for $0\arcdeg < \lambda <
297: 90\arcdeg$ the planet moves ``northeast,'' and so forth.
298: 
299: The other model parameters were the planetary mass ($M_p$); the
300: stellar and planetary radii ($R_\star$ and $R_p$); the orbital
301: inclination ($i$); the mid-transit time ($T_c$); and an additive
302: constant velocity for each of the two different velocity data sets
303: ($\gamma_1$ and $\gamma_2$). We allowed our velocities to have a
304: different additive constant from the velocities of O'Donovan et
305: al.~(2006) in order to account for systematic differences in the
306: spectrograph setup and reduction procedures. We fixed the orbital
307: period to be $2.47063$~days (Holman et al.~2007). We used a Markov
308: Chain Monte Carlo algorithm to solve for the model parameters and
309: their confidence limits, with uniform priors on all parameters. This
310: algorithm and our implementation of it are described in detail
311: elsewhere (see, e.g., Winn et al.~2007b). The minimum $\chi^2$ is
312: 1127.6, with 1091 degrees of freedom, giving $\chi^2/N_{\rm dof} =
313: 1.034$ and indicating an acceptable fit.
314: 
315: \section{Results}
316: 
317: The RM effect is certainly not obvious in Fig.~1, which shows the
318: entire spectroscopic orbit. It is not even very obvious in the middle
319: panel of Fig.~2, which focuses on the velocity data around the time of
320: transit. However, our analysis shows that the RM effect was indeed
321: detected. As mentioned above, for the best-fitting model, $\chi^2_{\rm
322:   min} = 1127.6$. If the parameter $v\sin i_\star$ is set equal to
323: zero, thereby neglecting the RM effect, then $\chi^2_{\rm min} =
324: 1135.8$, with the increase of $\Delta\chi^2 = 8.2$ arising from the
325: velocity data during the transit. We conclude that the RM effect was
326: detected with a signal-to-noise ratio (SNR) of approximately
327: $\sqrt{8.2} = 2.9$. Gaudi \& Winn (2007) have given analytic formulas
328: for the signal-to-noise ratio of RM observations as a function of the
329: system and telescope parameters, under the assumption of Gaussian
330: velocity errors. Using their Eqn.~(26) for this case, the forecasted
331: SNR is 2.9, in agreement with the actual SNR.
332: 
333: One might wonder how much this result was influenced by the inclusion
334: of the photometric data. To check on this, we tried setting aside the
335: photometric data and fitting only the 67 radial-velocity data
336: points. We fixed the photometric parameters ($M_p$, $R_p$, $R_\star$,
337: $i$, $T_c$, and $P$) at the values determined previously. In this case
338: we found $\chi^2_{\rm min} = 63.7$. If $v\sin i_\star$ is set equal to
339: zero, then $\chi^2_{\rm min} = 71.9$, giving $\Delta\chi^2 = 8.2$,
340: just as in the full model fit. This confirms that the lowered $\chi^2$
341: is an effect of a better fit to the transit velocities, and that the
342: uncertainties in the photometric parameters are negligible in this
343: instance.
344: 
345: The best-fitting model parameters are also consistent with good
346: alignment of the spin and the orbit. Specifically, we find $\lambda =
347: -9\pm 12$~deg, and $v\sin i_\star = 1.0\pm 0.6$~km~s$^{-1}$, where the
348: quoted values are the medians of the {\it a posteriori}\,
349: distributions returned by the MCMC algorithm, and the error bars
350: represent 68\% confidence limits. Table~2 gives these results, along
351: with some other revelant system parameters of TrES-2, for
352: convenience. Visually, the RM effect is more apparent in the bottom
353: panel of Fig.~2, in which the orbital velocity has been subtracted
354: from the data, and the sampling rate has been effectively doubled by
355: inverting the data through the origin ($t \rightarrow -t$ and $\Delta
356: v \rightarrow -\Delta v$). This works because for $\lambda\approx 0$,
357: the RM waveform is antisymmetric about the origin.
358: 
359: Figure~3 shows the {\it a posteriori}\, probability distribution for
360: $\lambda$ and the joint distribution of $\lambda$ and $v\sin
361: i_\star$. The distribution for $\lambda$ resembles a slightly
362: asymmetric Gaussian function to which is added a low-level uniform
363: probability distribution.  Although only the region from $-90\arcdeg$
364: to $+90\arcdeg$ is shown in Fig.~3, this low-level uniform
365: distribution extends all the way from $-180\arcdeg$ to
366: $+180\arcdeg$. The uniform background corresponds to the very lowest
367: allowed values of $v\sin i_\star$. This makes sense because when the
368: rotation rate is zero, the Rossiter anomaly vanishes and $\lambda$ is
369: irrelevant. Values of $\lambda$ between $-90\arcdeg$ and $+90\arcdeg$
370: correspond to prograde orbits, for which the stellar and orbital
371: angular momenta are in the same half-plane. The integrated probability
372: between $-90\arcdeg$ and $+90\arcdeg$ is 98\%. We conclude that the
373: TrES-2 orbit is prograde with 98\% confidence. As an illustration of
374: the constraints provided by our analysis, Fig.~4 shows a drawing of
375: the face of the star and the orbit of the transiting planet.
376: 
377: Our result for $v\sin i_\star$ is in agreement with the value reported
378: by O'Donovan et al.~(2006), $2.0\pm 1.5$~km~s$^{-1}$, which was based
379: on an analysis of the line broadening in an out-of-transit spectrum.
380: This finding is also supported by an analysis of our own
381: out-of-transit, iodine-free spectra, using the {\it Spectroscopy Made
382:   Easy}\, (SME) software package of Valenti \& Piskunov~(1996). The
383: automated analysis gave a formal result of $v\sin i_\star = 0.5\pm
384: 0.5$~km~s$^{-1}$, although the true uncertainty may be larger, since
385: with a disk-integrated spectrum of such a slow rotator it is difficult
386: to disentangle the effects of rotation, macroturbulence,
387: microturbulence, and the instrumental profile. In particular, the SME
388: code assumes ``typical'' values for the turbulent broadening
389: mechanisms that are of the same magnitude as the rotation speed of
390: TrES-2 (see \S~4.2-4.4 of Valenti \& Fischer 2005).\footnote{We
391:   investigated the consequences of accepting the SME result at face
392:   value, by imposing a Gaussian prior constraint on the $v\sin
393:   i_\star$ parameter with mean 0.5~km~s$^{-1}$ and standard error
394:   0.5~km~s$^{-1}$. In that case, the MCMC analysis gave
395:   68\%-confidence ranges of $-31$ to $1$~deg for $\lambda$ and $0.3$
396:   to $1.1$~km~s$^{-1}$ for $v\sin i_\star$, and showed that the orbit
397:   is prograde with 95\% confidence. The constraint on $\lambda$ was
398:   weakened because the SME result favors slower rotation rates, for
399:   which the sensitivity of the RM waveform to $\lambda$ is reduced.}
400: 
401: %\clearpage
402: \begin{figure}[p]
403: \epsscale{0.75}
404: \plotone{f3.eps}
405: \caption{
406: {\it Top.}---The probability distribution for $\lambda$, the angle
407: between the sky projections of the orbital axis and the stellar
408: rotation axis. {\it Bottom.}---The joint probability distribution
409: of $\lambda$ and $v\sin i_\star$. The solid dot shows the best-fitting
410: values. The contours represent 68\% and 95\% confidence limits.
411: \label{fig:3}}
412: \end{figure}
413: 
414: \begin{figure}[p]
415: \epsscale{0.75}
416: \plotone{f4.eps}
417: \caption{
418: Scale drawing of the TrES-2 system.  The relative radii of the bodies
419: and the impact parameter of the transit are taken from our
420: best-fitting model.  The ``north pole'' of the star is drawn with an
421: arrow, and the curved arc shows the 68\%-confidence region for its
422: orientation.  The angle $\lambda$ is measured clockwise from the
423: projected orbit normal vector (vertical dashed line) to the projected
424: stellar north pole (tilted dashed line). The best-fitting value of
425: $\lambda$ is negative.
426: \label{fig:4}}
427: \end{figure}
428: %\clearpage
429: 
430: \section{Summary and Discussion}
431: 
432: We have monitored the apparent Doppler shift of TrES-2 throughout a
433: transit of its giant planet and we have detected the
434: Rossiter-McLaughlin effect. Using the available photometric and
435: spectroscopic data, we have found good evidence that the orbit is
436: prograde, as are the other 6 systems that have been measured (with the
437: possible exception of HD~17156), and as are the planets in the Solar
438: system. In this sense, our results for TrES-2 are not surprising.
439: However, as mentioned in \S~1, some theories of planet migration do
440: predict occasionally large misalignments. For example, Nagasawa et
441: al.~(2008) investigated a scenario in which a planet is scattered into
442: an eccentric, inclined orbit with a small periastron distance (as
443: envisioned earlier by Rasio \& Ford 1996 and Marzari \&
444: Weidenschilling 2002), and subsequently a more distant planet forces
445: Kozai oscillations in the inner planet's eccentricity and
446: inclination. If the periastron distance is small enough during the
447: high-eccentricity phases, the orbit may circularize at a small orbital
448: distance with a substantial inclination. Nagasawa et al.~(2008) found
449: that this migration mechanism produces a very broad range of final
450: inclinations, including a significant fraction of retrograde
451: orbits. Of course, prograde orbits are also permitted in this
452: scenario, and our finding of a prograde orbit for TrES-2 cannot be
453: taken as evidence against this mechanism. We raise the issue only to
454: show that a prograde orbit was not a foregone conclusion.
455: 
456: Furthermore, we have shown it is possible to glean this information
457: and measure the projected spin-orbit angle to within $12\arcdeg$, even
458: for an 11th magnitude star with a slow projected rotation rate. A
459: potentially important application of the RM effect is the detection of
460: planets that are too small to be readily detected using other types of
461: ground-based data. For example, in many cases of terrestrial planets
462: detected by the {\it Corot}\, or {\it Kepler}\, satellites, it will be
463: easier to observe the RM effect than to observe the star's orbital
464: Doppler shift (and thereby measure the planet's mass). The theory
465: underlying this idea has been discussed by Welsh et al.~(2004) and
466: Gaudi \& Winn~(2007).
467: 
468: The present work serves to illustrate this point with actual data. If
469: TrES-2 had a rotation rate of 5~km~s$^{-1}$ instead of 1~km~s$^{-1}$,
470: but all other stellar and orbital parameters were the same, then the
471: quantity and quality of data presented in this paper would permit a
472: $\sim$3$\sigma$ detection of a planet with a radius $\sim$$\sqrt{5}$
473: times smaller than TrES-2, or $\sim$6 Earth radii. If the transit were
474: equatorial instead of grazing (the best configuration for detecting
475: the effect, although not for assessing spin-orbit alignment), the
476: duration of the transit would be longer by a factor of $\sim$2 and the
477: amplitude of the RM effect would be larger by a factor of $\sim$2,
478: leading to another factor-of-2 improvement in the detectable planet
479: radius ($\sim$3~$R_\oplus$). Such a planet would produce a photometric
480: transit depth of only $8\times 10^{-4}$, which is smaller than the
481: transit depth of any known transiting planet.
482: 
483: \acknowledgments We thank G.~Marcy for advice and encouragement, and
484: D.~Charbonneau for helpful conversations. We are grateful for support
485: from the NASA Keck PI Data Analysis Fund (JPL 1326712). We recognize
486: and acknowledge the very significant cultural role and reverence that
487: the summit of Mauna Kea has always had within the indigenous Hawaiian
488: community. We are most fortunate to have the opportunity to conduct
489: observations from this mountain. Access to the Keck telescopes for
490: this project was through the Telescope System Instrumentation Program,
491: and was supported by AURA through the National Science Foundation
492: under AURA Cooperative Agreement AST 0132798 as amended.
493: 
494: \begin{thebibliography}{}
495: 
496: \bibitem[Bakos et al.(2007)]{2007ApJ...670..826B} Bakos, G.~{\'A}., et
497:   al.\ 2007, \apj, 670, 826
498: 
499: \bibitem[Butler et al.(1996)]{1996PASP..108..500B} Butler, R.~P.,
500:   Marcy, G.~W., Williams, E., McCarthy, C., Dosanjh, P., \& Vogt,
501:   S.~S.\ 1996, \pasp, 108, 500
502: 
503: \bibitem[Butler et al.(2006)]{2006ApJ...646..505B} Butler, R.~P., et
504:   al.\ 2006, \apj, 646, 505
505: 
506: \bibitem[Chatterjee et al.(2007)]{2007astro.ph..3166C} Chatterjee, S.,
507:   Ford, E.~B., \& Rasio, F.~A.\ 2007, ArXiv Astrophysics e-prints,
508:   arXiv:astro-ph/0703166
509: 
510: \bibitem[Fabrycky \& Tremaine(2007)]{2007ApJ...669.1298F} Fabrycky,
511:   D., \& Tremaine, S.\ 2007, \apj, 669, 1298
512: 
513: \bibitem[Gaudi \& Winn(2007)]{gw07} Gaudi, B.~S.\ \& Winn, J.~N.\
514:   2007, ApJ, 655, 550
515: 
516: \bibitem[Gim\'enez(2006)]{gim06} Gim\'enez, A.\ 2006, ApJ, 650, 408
517: 
518: \bibitem[Holman et al.(2007)]{2007ApJ...664.1185H} Holman, M.~J., et
519:   al.\ 2007, \apj, 664, 1185
520: 
521: \bibitem[Johns-Krull et al.(2007)]{2007arXiv0712.4283J} Johns-Krull,
522:   C.~M., et al.\ 2007, ArXiv e-prints, 712, arXiv:0712.4283
523: 
524: \bibitem[Johnson et al.(2006)]{2006ApJ...652.1724J} Johnson, J.~A.,
525:   Marcy, G.~W., Fischer, D.~A., Henry, G.~W., Wright, J.~T., Isaacson,
526:   H., \& McCarthy, C.\ 2006, \apj, 652, 1724
527: 
528: \bibitem[Knutson et al.(2007)]{2007ApJ...655..564K} Knutson, H.~A.,
529:   Charbonneau, D., Noyes, R.~W., Brown, T.~M., \& Gilliland, R.~L.\
530:   2007, \apj, 655, 564
531: 
532: \bibitem[Loeillet et al.(2007)]{2007arXiv0707.0679L} Loeillet, B., et
533: al.\ 2007, ArXiv e-prints, 707, arXiv:0707.0679
534: 
535: \bibitem[Mandushev et al.(2007)]{2007ApJ...667L.195M} Mandushev, G.,
536:   et al.\ 2007, \apjl, 667, L195
537: 
538: \bibitem[Maness et al.(2007)]{2007PASP..119...90M} Maness, H.~L.,
539:   Marcy, G.~W., Ford, E.~B., Hauschildt, P.~H., Shreve, A.~T., Basri,
540:   G.~B., Butler, R.~P., \& Vogt, S.~S.\ 2007, \pasp, 119, 90
541: 
542: \bibitem[Marcy et al.(2005)]{2005PThPS.158...24M} Marcy, G., Butler,
543:   R.~P., Fischer, D., Vogt, S., Wright, J.~T., Tinney, C.~G., \&
544:   Jones, H.~R.~A.\ 2005, Progress of Theoretical Physics Supplement,
545:   158, 24
546: 
547: \bibitem[Marzari \& Weidenschilling(2002)]{2002Icar..156..570M}
548:   Marzari, F., \& Weidenschilling, S.~J.\ 2002, Icarus, 156, 570
549: 
550: \bibitem[Nagasawa et al.(2008)]{2008arXiv0801.1368N} Nagasawa, M.,
551:   Ida, S., \& Bessho, T.\ 2008, ArXiv e-prints, 801, arXiv:0801.1368
552: 
553: \bibitem[Narita et al.(2007a)]{2007astro.ph..2707N} Narita, N., et
554:   al.\ 2007a, PASJ, 59, 763
555: 
556: \bibitem[Narita et al.(2007b)]{2007arXiv0712.2569N} Narita, N., Sato,
557:   B., Ohshima, O., \& Winn, J.~N.\ 2007b, ArXiv e-prints, 712,
558:   arXiv:0712.2569
559: 
560: \bibitem[O'Donovan et al.(2006)]{2006ApJ...651L..61O} O'Donovan,
561:   F.~T., et al.\ 2006, \apjl, 651, L61
562: 
563: \bibitem[Ohta et al.(2005)]{2005ApJ...622.1118O} Ohta, Y., Taruya, A.,
564:   \& Suto, Y.\ 2005, \apj, 622, 1118
565: 
566: \bibitem[Queloz et al.(2000)]{2000A&A...359L..13Q} Queloz, D.,
567:   Eggenberger, A., Mayor, M., Perrier, C., Beuzit, J.~L., Naef, D.,
568:   Sivan, J.~P., \& Udry, S.\ 2000, \aap, 359, L13
569: 
570: \bibitem[Rasio \& Ford(1996)]{rasio96} Rasio, F.~A., \& Ford, E.~B.\
571:   1996, Science, 274, 954
572: 
573: \bibitem[Sato et al.(2005)]{2005ApJ...633..465S} Sato, B., et al.\
574:   2005, \apj, 633, 465
575: 
576: \bibitem[Sozzetti et al.(2007)]{2007ApJ...664.1190S} Sozzetti, A.,
577:   Torres, G., Charbonneau, D., Latham, D.~W., Holman, M.~J., Winn,
578:   J.~N., Laird, J.~B., \& O'Donovan, F.~T.\ 2007, \apj, 664, 1190
579: 
580: \bibitem[Torres et al.(2007)]{2007ApJ...666L.121T} Torres, G., et al.\
581:   2007, \apjl, 666, L121
582: 
583: \bibitem[Udry \& Santos(2007)]{2007ARA&A..45..397U} Udry, S., \&
584:   Santos, N.~C.\ 2007, \araa, 45, 397
585: 
586: \bibitem[Valenti \& Piskunov(1996)]{1996A&AS..118..595V} Valenti,
587:   J.~A., \& Piskunov, N.\ 1996, \aaps, 118, 595
588: 
589: \bibitem[Valenti \& Fischer(2005)]{2005ApJS..159..141V} Valenti,
590:   J.~A., \& Fischer, D.~A.\ 2005, \apjs, 159, 141
591: 
592: \bibitem[Vogt et al.(1994)]{1994SPIE.2198..362V} Vogt, S.~S., et al.\
593:   1994, \procspie, 2198, 362
594: 
595: \bibitem[Welsh et al.(2004)]{2004AAS...20513516W} Welsh, W.~F., Orosz,
596:   J.~A., \& Wittenmyer, R.~A.\ 2004, Bulletin of the American
597:   Astronomical Society, 36, 1567
598: 
599: \bibitem[Winn(2007)]{2007ASPC..366..170W} Winn, J.~N.\ 2007, in
600:   Transiting Extrasolar Planets Workshop, eds.~C.~Afonso, D.~Weldrake,
601:   \& Th.~Henning, 366, 170 (ASP: San Francisco)
602: 
603: \bibitem[Winn et al.(2005)]{2005ApJ...631.1215W} Winn, J.~N., et al.\
604:   2005, \apj, 631, 1215 % hd 209458
605: 
606: \bibitem[Winn et al.(2006)]{2006ApJ...653L..69W} Winn, J.~N., et al.\
607:   2006, \apjl, 653, L69 % hd 189833
608: 
609: \bibitem[Winn et al.(2007a)]{2007ApJ...665L.167W} Winn, J.~N., et al.\
610:   2007a, \apjl, 665, L167 % hat-2
611: 
612: \bibitem[Winn et al.(2007b)]{2007AJ....133...11W} Winn, J.~N., Holman,
613:   M.~J., \& Fuentes, C.~I.\ 2007b, \aj, 133, 11
614: 
615: \bibitem[Wolf et al.(2007)]{2007ApJ...667..549W} Wolf, A.~S.,
616:   Laughlin, G., Henry, G.~W., Fischer, D.~A., Marcy, G., Butler, P.,
617:   \& Vogt, S.\ 2007, \apj, 667, 549
618: 
619: \bibitem[Wu et al.(2007)]{2007ApJ...670..820W} Wu, Y., Murray, N.~W.,
620:   \& Ramsahai, J.~M.\ 2007, \apj, 670, 820
621: 
622: \end{thebibliography}
623: 
624: %\clearpage
625: \begin{deluxetable}{lcccc}
626: \tabletypesize{\tiny}
627: \tablecaption{Radial Velocities of TrES-2\label{tbl:rv}}
628: \tablewidth{0pt}
629: 
630: \tablehead{
631: \colhead{HJD} & \colhead{Radial Velocity [m~s$^{-1}$]} & \colhead{Measurement Uncertainty [m~s$^{-1}$]}
632: }
633: 
634: \startdata
635:  $  2454216.96599$ & $          42.59$ & $           5.86$ \\
636:  $  2454216.96998$ & $          38.52$ & $           5.76$ \\
637:  $  2454216.97930$ & $          37.06$ & $           5.43$ \\
638:  $  2454216.98973$ & $          26.69$ & $           5.76$ \\
639:  $  2454216.99368$ & $          27.35$ & $           5.82$ \\
640:  $  2454216.99769$ & $          21.46$ & $           5.87$ \\
641:  $  2454217.00168$ & $          22.74$ & $           5.69$ \\
642:  $  2454217.00564$ & $          17.82$ & $           5.75$ \\
643:  $  2454217.00876$ & $          31.44$ & $           6.16$ \\
644:  $  2454217.01102$ & $          16.32$ & $           6.06$ \\
645:  $  2454217.01327$ & $          12.64$ & $           6.24$ \\
646:  $  2454217.01552$ & $           6.15$ & $           6.05$ \\
647:  $  2454217.01779$ & $          17.40$ & $           6.13$ \\
648:  $  2454217.02003$ & $          19.25$ & $           6.35$ \\
649:  $  2454217.02229$ & $           5.79$ & $           6.24$ \\
650:  $  2454217.02453$ & $          14.41$ & $           6.35$ \\
651:  $  2454217.02681$ & $          23.50$ & $           6.26$ \\
652:  $  2454217.02905$ & $           6.40$ & $           6.19$ \\
653:  $  2454217.03131$ & $          35.58$ & $           6.27$ \\
654:  $  2454217.03356$ & $           6.90$ & $           5.92$ \\
655:  $  2454217.03580$ & $          14.89$ & $           6.07$ \\
656:  $  2454217.03803$ & $          16.89$ & $           6.09$ \\
657:  $  2454217.04040$ & $           4.04$ & $           6.25$ \\
658:  $  2454217.04266$ & $           7.37$ & $           6.17$ \\
659:  $  2454217.04492$ & $           3.26$ & $           5.91$ \\
660:  $  2454217.04715$ & $           0.45$ & $           6.44$ \\
661:  $  2454217.04940$ & $           3.51$ & $           6.27$ \\
662:  $  2454217.05165$ & $          -0.38$ & $           6.31$ \\
663:  $  2454217.05403$ & $           2.49$ & $           6.37$ \\
664:  $  2454217.05648$ & $           7.42$ & $           6.13$ \\
665:  $  2454217.05907$ & $          -8.45$ & $           6.16$ \\
666:  $  2454217.06167$ & $          -4.76$ & $           6.17$ \\
667:  $  2454217.06425$ & $          -7.03$ & $           6.06$ \\
668:  $  2454217.06684$ & $         -11.21$ & $           6.06$ \\
669:  $  2454217.06957$ & $          -8.69$ & $           6.32$ \\
670:  $  2454217.07214$ & $          -6.58$ & $           6.15$ \\
671:  $  2454217.07473$ & $         -17.63$ & $           6.14$ \\
672:  $  2454217.07730$ & $         -19.40$ & $           6.14$ \\
673:  $  2454217.07991$ & $         -25.50$ & $           6.18$ \\
674:  $  2454217.08250$ & $         -16.69$ & $           6.14$ \\
675:  $  2454217.08513$ & $         -11.51$ & $           6.33$ \\
676:  $  2454217.08767$ & $         -16.22$ & $           6.21$ \\
677:  $  2454217.09031$ & $         -28.52$ & $           6.52$ \\
678:  $  2454217.09287$ & $         -18.49$ & $           6.62$ \\
679:  $  2454217.09545$ & $          -9.36$ & $           6.50$ \\
680:  $  2454217.09806$ & $         -25.95$ & $           6.43$ \\
681:  $  2454217.10065$ & $         -25.17$ & $           6.26$ \\
682:  $  2454217.10335$ & $         -15.71$ & $           6.36$ \\
683:  $  2454217.10608$ & $         -15.20$ & $           6.34$ \\
684:  $  2454217.10868$ & $         -21.00$ & $           6.26$ \\
685:  $  2454217.11124$ & $         -30.07$ & $           6.31$ \\
686:  $  2454217.11386$ & $         -22.90$ & $           6.24$ \\
687:  $  2454217.11709$ & $         -35.76$ & $           5.99$ \\
688:  $  2454217.12110$ & $         -33.87$ & $           5.88$ \\
689:  $  2454217.12509$ & $         -29.70$ & $           5.79$ \\
690:  $  2454217.12911$ & $         -26.81$ & $           5.88$
691: \enddata 
692: 
693: \tablecomments{Column 1 gives the Heliocentric Julian Date at the
694:   photon-weighted midexposure time, i.e., weighted by the photon count
695:   rate recorded by the HIRES exposure meter.}
696: 
697: \end{deluxetable}
698: 
699: \begin{deluxetable}{lccc}
700: \tabletypesize{\normalsize}
701: \tablecaption{System Parameters of TrES-2\label{tbl:params}}
702: \tablewidth{0pt}
703: 
704: \tablehead{
705: \colhead{Parameter} & \colhead{Value} & \colhead{68\% Confidence Limits} & \colhead{References}
706: }
707: \startdata 
708: $P$~[d]                & $2.470621$ & $\pm 0.000017$ & 1  \\
709: $T_c$ [HJD]            & $2453957.63479$ & $\pm 0.00038$ & 1  \\
710: $(R_p/R_\star)^2$          & $0.0157$ & $\pm 0.0003$             & 1 \\
711: $b \equiv a\cos i/R_\star$ & $0.8540$ & $\pm 0.0062$             & 1 \\
712: $M_\star~[M_\odot]$        & $0.980$  & $\pm 0.062$          & 2 \\
713: $R_\star~[R_\odot]$        & $1.000$  & $+0.036$, $-0.033$ & 2 \\
714: $M_p~[M_{\rm Jup}]$        & $1.198$  & $\pm 0.053$          & 2 \\
715: $R_p~[R_{\rm Jup}]$        & $1.220$  & $+0.045$, $-0.042$ & 2 \\
716: $v\sin i_\star$~[km~s$^{-1}$] & $1.0$    & $\pm 0.6$                & This work \\
717: $\lambda$~[deg]            & $-9$   & $\pm 12$   & This work
718: \enddata
719: \tablecomments{(1) Holman et al.~(2007); (2) Sozzetti et al.~(2007).}
720: 
721: \end{deluxetable}
722: 
723: \end{document}
724: