1: \documentclass[preprint,12pt]{aastex}
2: %\usepackage{latexsym,graphicx,natbib}
3: %\usepackage{apjfonts,emulateapj5,psfig}
4:
5: % ------------------------------------------------------------------------
6: % New commands
7: %
8: \def\ltsima{$\; \buildrel < \over \sim \;$}
9: \def\lsim{\lower.5ex\hbox{\ltsima}}
10: \def\gtsima{$\; \buildrel > \over \sim \;$}
11: \def\gsim{\lower.5ex\hbox{\gtsima}}
12: \def\rs{$R_\star = 1.003\pm 0.027$~$R_\odot$}
13: \def\rp{$R_p = 1.222\pm 0.038$~$R_{\rm Jup}$}
14: \def\ms{$M_\star = 0.980\pm 0.062$~$M_\odot$}
15: \def\mp{$M_p = 1.198\pm 0.053$~$M_\odot$}
16:
17: \def\kms{\ifmmode{\rm km\thinspace s^{-1}}\else km\thinspace s$^{-1}$\fi}
18: %\def\ms{\ifmmode{\rm m\thinspace s^{-1}}\else m\thinspace s$^{-1}$\fi}
19: \def\dur{$\Theta_{\rm I}$}
20: \def\pdur{$\Theta'_{\rm I}$}
21:
22: % -------------------------------------------------------------------------
23: %
24:
25: \shortauthors{Holman et al.\ 2006}
26: \shorttitle{Transit Photometry of TrES-2}
27:
28: \begin{document}
29:
30: \bibliographystyle{apj}
31:
32: \title{
33: The Transit Light Curve (TLC) Project.\\
34: VI.~Three Transits of the Exoplanet TrES-2
35: }
36:
37: \author{
38: Matthew J.\ Holman\altaffilmark{1},
39: Joshua N.\ Winn\altaffilmark{2},
40: David W.\ Latham\altaffilmark{1},\\
41: Francis T.\ O'Donovan\altaffilmark{3},
42: David Charbonneau\altaffilmark{1,6},
43: Guillermo~Torres\altaffilmark{1},
44: Alessandro~Sozzetti\altaffilmark{1,4},
45: Jose Fernandez\altaffilmark{1},
46: Mark E.\ Everett\altaffilmark{5}
47: }
48:
49: \altaffiltext{1}{Harvard-Smithsonian Center for Astrophysics, 60
50: Garden Street, Cambridge, MA 02138; mholman@cfa.harvard.edu}
51: \altaffiltext{2}{Department of Physics, and Kavli Institute for
52: Astrophysics and Space Research, Massachusetts Institute of
53: Technology, Cambridge, MA 02139}
54: \altaffiltext{3}{California Institute of Technology, 1200 East
55: California Blvd., Pasadena, CA 91125}
56: \altaffiltext{4}{INAF - Osservatorio Astronomico di Torino, 10025 Pino
57: Torinese, Italy}
58: \altaffiltext{5}{Planetary Science Institute, 1700 East Fort Lowell,
59: Tucson, AZ 85719}
60: \altaffiltext{6}{Alfred P.\ Sloan Research Fellow.}
61:
62: \begin{abstract}
63:
64: Of the nearby transiting exoplanets that are amenable to detailed
65: study, TrES-2 is both the most massive and has the largest impact
66: parameter.
67: We present $z$-band photometry of three transits of TrES-2. We
68: improve upon the estimates of the planetary, stellar, and orbital
69: parameters, in conjunction with the spectroscopic analysis of the
70: host star by Sozzetti and co-workers. We find the planetary radius
71: to be \rp~ and the stellar radius to be \rs. The quoted
72: uncertainties include the systematic error due to the uncertainty in
73: the stellar mass (\ms). The timings of the transits have an accuracy
74: of 25~s and are consistent with a uniform period, thus providing a
75: baseline for future observations with the NASA {\it Kepler}
76: satellite, whose field of view will include TrES-2.
77:
78: \end{abstract}
79:
80: \keywords{planetary systems --- stars:~individual (GSC~03549-02811) ---
81: techniques: photometric}
82:
83: \section{Introduction}
84:
85: Careful follow-up observations of nearby transiting planet systems
86: have revolutionized our understanding of a whole new kind of planet:
87: hot Jupiters. They have been used to reveal absorption by atmospheric
88: atomic sodium ~\citep{Charbonneau.2002} and the presence of an
89: extended hydrogen exosphere~\citep{Vidal-Madjar.2003} in HD~209458b,
90: as well as to detect the thermal infrared emission from TrES-1,
91: HD~209458b, and
92: HD~189733b~\citep{Charbonneau.2005,Deming.2005a,Deming.2006}. They
93: have been used to investigate the spin-orbit alignment of
94: HD~209458b~\citep{Queloz.2000, Winn.2005} and
95: HD~189733b~\citep{Winn.2006}. Most recently, spectra of the infrared
96: planetary emission of HD~189733b~\citep{Grillmair.2007} and
97: HD~209458b~\citep{Richardson.2007}, obtained with the {\it Spitzer}
98: Space Telescope, have been used to constrain models of the atmospheric
99: content of those planets.
100:
101: Through these observations we are steadily improving our understanding
102: of the interior and atmospheric structure of hot Jupiters. Future
103: measurements, such as reflected-light observations or the detection of
104: other atmospheric constituents through transmission spectroscopy, will
105: continue to advance our knowledge of these planets. One goal of the
106: Transit Light Curve (TLC) Project is to support these efforts by
107: refining the estimates of the planetary, stellar, and orbital
108: parameters, through high-accuracy, high-cadence photometry of
109: exoplanetary transits. We also seek to measure or bound any variations
110: in the transit times and light-curve shapes that would be caused by
111: the influence of additional bodies in the
112: system~\citep{Miralda-Escude.2002, Agol.2005,Holman.2005}. Along the
113: way, we are exploring different techniques for photometry and
114: parameter determination. Previous papers in this series have reported
115: results for the exoplanets XO-1b~\citep{Holman.2006},
116: OGLE-TR-111b~\citep{Winn.2007a}, TrES-1~\citep{Winn.2007b},
117: OGLE-TR-10b~\citep{Holman.2007}, and HD~189733b~\citep{Winn.2007c}.
118:
119: The present paper is concerned with TrES-2, the second transiting hot
120: Jupiter discovered by the Trans-atlantic Exoplanet Survey
121: \citep{O'Donovan.2006b}. The planet orbits a nearby G0~V star
122: (GSC~03549-02811) and transits every $\sim$2.5~days. Although each of
123: the fourteen known transiting exoplanets has its own story to tell,
124: (see \citealt{Charbonneau.2007} for a review), the TrES-2 system has
125: at least three distinguishing characteristics.
126:
127: First, TrES-2 is the first transiting extrasolar planet discovered in
128: the field of view of the NASA {\it Kepler}
129: mission~\citep{Borucki.2003,Basri.2005}. {\it Kepler} will observe
130: nearly six hundred transits of TrES-2 during the nominal 4~yr lifetime
131: of the mission. This opportunity prompts us to improve the
132: determinations of the orbital parameters of TrES-2 for comparison to
133: the future estimates from {\it Kepler}.
134:
135: Second, TrES-2 has the highest impact parameter of any known
136: nearby transiting extrasolar planet. This makes the duration of the transit
137: (as well as the duration of ingress and egress) more sensitive to
138: changes in impact parameter. This, in turn, makes TrES-2 an excellent
139: target for the detection of long-term changes in transit
140: characteristics induced by orbital
141: precession~\citep{Miralda-Escude.2002}.
142:
143: Third, the mass of TrES-2 is the largest of the known nearby
144: transiting extrasolar planets. Furthermore, the radius of TrES-2
145: appears somewhat larger than predicted by simple structural models of
146: irradiate hot Jupiters, as also appears to be the case for HAT-P-1b,
147: WASP-1b, and HD~209458b (although see Burrows et al.~2007 for a
148: contrary view).
149:
150: In what follows we present TLC results for TrES-2. In \S~2 we describe
151: the observations and the data reduction procedures. In \S~3 we
152: describe the model and techniques we used to estimate the physical and
153: orbital parameters of the TrES-2 system, and in \S~4 we summarize our
154: results.
155:
156: \section{The Observations and Data Reduction}
157:
158: We observed four transits of TrES-2. According to the ephemeris
159: provided by \citet{O'Donovan.2006b},
160: \begin{equation}
161: T_c(E) = 2,453,957.6358~\mathrm{[HJD]} + E\times (2.47063~\mathrm{days}),
162: \label{eq:ephemeris}
163: \end{equation}
164: these transits correspond to epochs 13, 15, 32, and 34 on
165: UT~2006~Sept~11, Sept~16, Oct~28, and Nov~2, respectively.
166: Observations of a fifth transit, epoch 17, were scheduled but were
167: not executed due to poor weather.
168:
169: We observed these transits with KeplerCam on the 1.2m (48~inch)
170: telescope of the Fred L.\ Whipple Observatory (FLWO) on Mt.~Hopkins,
171: Arizona. This camera (P.I.\ D.~Latham) was built for a photometric
172: survey of the target field of the {\it Kepler}\/ satellite
173: mission~\citep{Borucki.2003}. It has a single
174: $4\mathrm{K}\times4\mathrm{K}$ Fairchild 486 CCD with a $23\farcm 1
175: \times 23\farcm 1$ field of view. We used $2 \times 2$ binning, for
176: which the readout and reset time is 11.5~s and the typical read noise
177: is 7~$e^{-}$ per binned pixel. The response of each amplifier
178: deviates from linearity by less that 0.5\% over the range of counts
179: from the faintest to brightest comparison star. We observed through
180: the SDSS $z$ filter, the reddest available band, in order to minimize
181: the effect of color-dependent atmospheric extinction on the relative
182: photometry, and to miminize the effect of limb-darkening on the
183: transit light curve.
184:
185: The full-width at half-maximum (FWHM) of a stellar image was typically
186: $\sim$3 binned pixels ($2\arcsec$) on Sept~11, Sept~16, and Nov~2; the
187: FWHM ranged from $\sim$3 to $\sim$8~pixels on Oct~28. We used
188: automatic guiding to maintain the locations of TrES-2 and its
189: comparison stars to within a few pixels over the course of each night.
190: We repeatedly took 30~s exposures for 3.5--5~hr bracketing the
191: predicted transit midpoint. The conditions on UT~2006~Sept~11 were
192: clear during the time of the observations, and the images were taken
193: through airmasses ranging from 1.05 to 1.90. The conditions on
194: UT~2006~Sept~16 were also clear, and the airmass ranged from 1.05 to
195: 1.40. There were clouds passing overhead during the observations on
196: UT~2006~Oct~28, and the airmass ranged from 1.05 to 2.50. The
197: observing conditions were significantly worse during and after egress;
198: the result was essentially observations of only a partial transit.
199: Consequently, the data from Oct~28 were not included in the analysis
200: below. There were very thin clouds during the observations on
201: UT~2006~Nov~2, and the airmass ranged from 1.15 to 1.95.
202:
203: The images were calibrated using standard IRAF\footnote{IRAF is
204: distributed by the National Optical Astronomy Observatories, which
205: are operated by the Association of Universities for Research in
206: Astronomy, Inc., under cooperative agreement with the National
207: Science Foundation.} procedures for the overscan correction,
208: trimming, bias subtraction, and flat-field division. We did not
209: attempt to correct the fringing that was apparent with the $z$ filter.
210: The fringing had a small amplitude and little effect on the final
211: photometry, given the accuracy of the automatic guiding. We then
212: performed aperture photometry of TrES-2 and 20 nearby comparison
213: stars, using an aperture radius of 8.0 pixels ($4\farcs 3$) for each
214: night. We subtracted the underlying contribution from the sky, after
215: estimating its brightness within an annulus ranging from 30 to 35
216: pixels in radius, centered on each star. We divided the flux of
217: TrES-2 by the total flux of the comparison stars.
218:
219: To estimate the uncertainties in our photometry, we computed the
220: quadrature sum of the errors due to Poisson noise of the stars (both
221: TrES-2 and the comparison stars), Poisson noise of the sky background,
222: readout noise, and scintillation noise (as estimated according to the
223: empirical formulas of~\citealt{Young.1967} and
224: \citealt{Dravins.1998}). The dominant term is the Poisson noise from
225: TrES-2. The final time series is plotted in Fig.~1 and is available in
226: electronic form in Table~1. (In that table, the quoted errors have
227: been rescaled such that $\chi^2/N_{\rm dof} = 1$ for the best-fitting
228: model, as explained in the next section.)
229:
230: \begin{figure}[p]
231: \epsscale{0.85}
232: \plotone{fig1.eps}
233: \caption{
234: Photometry of TrES-2 in the $z$ band, using
235: the FLWO 1.2m telescope and Keplercam. These data were used to
236: estimate the planetary, stellar, and orbital parameters (see \S~3).
237: The bottom panel is a composite light curve created from the three data
238: sets, after time-shifting and averaging into 2~min bins. The
239: residuals (observed$-$calculated) are plotted beneath the data.
240: \label{fig:1}}
241: \end{figure}
242:
243: \section{Determination of System Parameters}
244:
245: Our methodology for determining the system parameters has been
246: described in previous TLC
247: papers~\citep{Holman.2006,Winn.2007a,Winn.2007b,Holman.2007,Winn.2007c},
248: and is summarized here. We assume a circular orbit of a planet (mass
249: $M_p$, radius $R_p$) and a star ($M_\star$, $R_\star$), with period
250: $P$ and inclination $i$ relative to the sky plane. We allow each
251: transit to have an independent value of $T_c$ (the transit midpoint)
252: rather than forcing them to be separated by exact multiples of the
253: orbital period. Thus, the only effect of $P$ on the model is to
254: determine the semimajor axis $a$ for a given value of ($M_\star +
255: M_p$). We fixed $P=2.47063$~days \citep{O'Donovan.2006b}; the
256: uncertainty of $0.00001$~days was negligible for this purpose.
257:
258: To calculate the relative flux as a function of the projected
259: separation of the planet and the star, we employed the analytic
260: formulas of Mandel \& Agol~(2002), using a quadratic limb darkening
261: law,
262: \begin{equation}
263: \frac{I_\mu}{I_1} = 1 - u_1(1-\mu) - u_2(1-\mu)^2,
264: \end{equation}
265: where $I$ is the intensity and $\mu$ is the cosine of the angle
266: between the line of sight and the normal to the stellar surface. We
267: chose the values $u_1=0.22$, $u_2=0.32$, based on the tabulated values
268: of \citet{Claret.2004} and the estimates by \citet{Sozzetti.2007} of
269: the stellar effective temperature, surface gravity, and metallicity.
270: We accounted for the color-dependent {\it residual} airmass effects
271: with a parameter $k$ specific to each transit, such that the observed
272: flux is equal to the intrinsic (zero airmass) flux times $\exp(-kz)$,
273: where $z$ is the airmass. The best-fitting values of $k$ were 0.0021,
274: 0.0086, and -0.0005, for Sept~11, Sept~16, and Nov~2, respectively. We
275: also fitted for the out-of-transit flux $f_{\rm oot}$.
276:
277: The light curves cannot be used to determine both the stellar mass and
278: radius; there is a fitting degeneracy $R_\star \propto
279: M_\star^{1/3}$. Our usual approach is to assume a value for $M_\star$
280: (based on external analyses of the stellar spectrum) and then
281: determine $R_\star$ by fitting the light curves. This case was
282: slightly different because we worked in conjunction with
283: \citet{Sozzetti.2007}, who sought to {\it improve}\, the estimates of
284: the stellar parameters based on the results of the light-curve fit. We
285: worked iteratively, as described below in more detail; for our final
286: analysis, we fixed $M_\star=0.98~M_\odot$.
287:
288: Our fitting statistic was \begin{equation}
289: \chi^2 =
290: \sum_{j=1}^{N_f}
291: \left[
292: \frac{f_j({\mathrm{obs}}) - f_j({\mathrm{calc}})}{\sigma_j}
293: \right]^2
294: ,
295: \label{eq:chi2}
296: \end{equation}
297: where $f_j$(obs) is the flux observed at time $j$, $\sigma_j$ controls
298: the relative weights of the data points, and $f_j$(calc) is the
299: calculated value. It is important for $\sigma_j$ to include
300: measurement errors and also any unmodeled systematic effects, and in
301: particular to account for time-correlated noise, which effectively
302: reduces the number of independent measurements. Our approach was as
303: follows. First, we rescaled the instrumental uncertainties such that
304: $\chi^2/N_{\rm dof} = 1$ for the best-fitting model. Table~1 lists the
305: resulting uncertainties. Second, we followed the procedure of
306: \citet{Gillon.2006} to decompose the observed noise into ``white
307: noise'' (that which averages down as $1/\sqrt{N}$, where $N$ is the
308: number of data points) and ``red noise'' (that which does not average
309: down over some specified time interval). Specifically, we calculated
310: the standard deviation of the residuals ($\sigma$) and the standard
311: deviation of the time-averaged residuals ($\sigma_N$). The averaging
312: time was 1~hr (a timescale comparable to the transit event),
313: corresponding to a number $N$ of data points that depended upon the
314: cadence of observations. Then we solved for the white noise
315: $\sigma_w$ and red noise $\sigma_r$ from the system of equations
316: \begin{eqnarray}
317: \sigma_1^2 & = & \sigma_w^2 + \sigma_r^2, \\
318: \sigma_N^2 & = & \frac{\sigma_w^2}{N} + \sigma_r^2.
319: \end{eqnarray}
320: Finally, to account approximately for the effective reduction in the
321: number of independent data points, we rescaled the $\sigma_j$ in
322: Eq.~(\ref{eq:chi2}) by the factor $\sigma_r/(\sigma_w/\sqrt{N})$. In
323: this case, the Sept~11 and Nov~2 transits did not show evidence for
324: red noise according to this criterion, but for the Sept~16 transit the
325: red-noise rescaling factor was 1.14. For that transit, we find
326: $sigma_r = 0.00016$ and $\sigma_w = 0.0014$. To be conservative, we a
327: pplied this same factor 1.14 to the data from all 3 transits.
328:
329: In short there were 12 model parameters: $\{R_\star, R_p, i\}$, as
330: well as $\{T_c, k, f_{\rm oot}\}$ for each of 3 transits. We
331: determined the {\it a posteriori} probability distributions for these
332: parameters using the same Markov Chain Monte Carlo algorithm described
333: in previous TLC papers. We took the median value of each probability
334: distribution to be the best estimate of each parameter, and the
335: standard deviation to be the 1~$\sigma$ uncertainty. In addition to
336: this statistical error, for the special cases of $R_\star$ and $R_p$
337: there is an additional error arising from the uncertainty in
338: $M_\star$, which we add to the statistical error in quadrature.
339:
340: Our choice of $M_\star$ merits further discussion since it is based on
341: a novel iterative procedure conducted in tandem with
342: \citet{Sozzetti.2007}. The underlying idea is that when fitting a
343: light curve, the results for $R_\star$ and $R_p$ depend on the choice
344: of $M_\star$, while the result for $R_\star/a$ is {\it independent}\,
345: of $M_\star$ because both $R_\star$ and $a$ vary as $M_\star^{1/3}$
346: for a fixed value of the orbital period. (There is, however, a minor
347: dependence of $R_\star/a$ on the choice of limb darkening function,
348: which is in turn informed by the estimates of the stellar parameters.)
349: Meanwhile, as \citet{Sozzetti.2007} have shown, $R_\star/a$ is useful
350: for estimating $M_\star$, since it can be directly related to the
351: stellar mean density through Kepler's Third Law (see also
352: \citet{Seager.2003}):
353: \begin{equation}
354: \frac{a}{R_\star} =
355: \left(\frac{G P^2}{4\pi^2}\right)^{1/3}
356: \left(\frac{M_\star + M_p}{R_\star^3}\right)^{1/3}.
357: \end{equation}
358: This makes $R_\star/a$ a useful proxy for $\log g$ for the purpose of
359: comparing the observed stellar properties with theoretical isochrones.
360: The advantage of $R_\star/a$ is that in typical cases it is more
361: precisely determined than the spectroscopic value of $\log g$.
362:
363: We iterated as follows. First, we fitted the light curves using the
364: choices $M_\star=1.08~M_\odot$, $u_1=0.18$, $u_2=0.34$, based on the
365: previous estimates of the relevant stellar parameters by
366: \citet{O'Donovan.2006b}. Next, we passed our results for $R_\star/a$
367: to \citet{Sozzetti.2007}, who used it to refine the estimate of
368: $M_\star$. (We refer the reader to \citet{Sozzetti.2007} for details
369: on how this refinement was achieved.) In return,
370: \citet{Sozzetti.2007} provided us with a new estimate of $M_\star$,
371: along with a new estimate of the stellar surface gravity (which
372: affects the choice of limb darkening law). We refitted the light
373: curves using the updated values of the stellar mass and the slightly
374: adjusted limb darkening law. Then we passed our new result for
375: $R_\star/a$ back to \citet{Sozzetti.2007}, who used it to refine the
376: estimate of $M_\star$ and $\log g$, and so forth. This process
377: converged after a few iterations, leading to the final choices for
378: $M_\star$, $u_1$, and $u_2$ noted above.
379:
380: While it is possible for the value of the stellar radius that
381: minimizes $\chi^2$ to be inconsistent with the theoretical mass-radius
382: relation, in this case we have effectively required consistency with
383: the theoretical mass-radius relation by iterating with
384: \citet{Sozzetti.2007}.
385:
386: \section{Results}
387:
388: The final results are given in Table~\ref{tbl:params}. In addition to
389: the results for the basic model parameters, we have also included in
390: this table a number of interesting derived quantities, such as
391: $a/R_\star$ (which is related to the stellar mean density, as
392: described above) and the calculated durations of the transit and the
393: partial transit phases. The most interesting parameters are the radius
394: of the star, the radius of the planet, the orbital inclination, and
395: the mid-transit times, which we discuss in turn.
396:
397: We find the stellar radius to be \rs, where the quoted error includes
398: both the statistical error ($0.017$) and the systematic error due to
399: the uncertainty in the stellar mass ($0.021$). This estimate agrees
400: with all of the star's observed broadband colors and spectral
401: properties as determined by \citet{Sozzetti.2007}, as it must, given
402: that our analyses were coupled as described in the previous section.
403:
404: We find the planetary radius to be \rp, where (again) the quoted error
405: includes both the statistical error ($0.028$) and the systematic
406: error due to the uncertainty in the stellar mass ($0.026$). The
407: difference between our value, and the value $R_p =
408: 1.220^{+0.045}_{-0.042}$~$R_{\rm Jup}$ presented by
409: \citet{Sozzetti.2007}, is slight indeed, although our figure has a
410: somewhat smaller error bar. The reason why there is any difference at
411: all is subtle. \citet{Sozzetti.2007} determined $R_p$ by taking our
412: result for $(R_p/R_\star)$ and the associated uncertainty, and
413: multiplying by their estimate for $R_\star$ (which in turn was based
414: on matching the observed values of $T_{\rm eff}$, $a/R_\star$, and
415: metallicity to theoretical isochrones). In contrast, we determined
416: $R_p$ and $R_\star$ simultaneously by fitting a parameterized model to
417: the light curves, as described above, and then accounting for the
418: uncertainty in the stellar mass. Our analysis takes into account the
419: correlations between all of the parameters, while that of
420: \citet{Sozzetti.2007} assumes $(R_p/R_\star)$ is independent of
421: $a/R_\star$. In this case, our procedure has yielded somewhat more
422: precise results for $R_p$ and $R_\star$.
423:
424: For an eclipsing single-lined spectroscopic binary the surface gravity
425: of the secondary ($GM_p/R_p^2$, in this case) can be determined nearly
426: independently of any assumptions regarding the properties of the
427: primary\citep{Southworth.2004,Winn.2007a,Beatty.2007,Sozzetti.2007,Southworth.2007}.
428: This result holds because the fitting degeneracy for the
429: radial-velocity data is $M_p \propto M_\star^{2/3}$, and the fitting
430: degeneracy for the photometric data is $R_p \propto M_\star^{1/3}$,
431: and in the ratio $M_p/R_p^2$ the stellar mass cancels out. There
432: remains only a weak dependency of $R_p$ on the choice of limb
433: darkening law, which is based on knowledge of the host star. In this
434: case, the result is $GM_p/R_p^2 = 1976\pm 91$~cm~s$^{-2}$, or $\log
435: g_p = 3.299\pm 0.020$.
436:
437: We confirm the finding by \citet{O'Donovan.2006b} that the transit
438: chord occurs at an unusually large impact parameter, $b \equiv a\cos
439: i/R_\star = 0.8540 \pm 0.0062$. This is of interest because the error
440: in the impact parameter is much smaller when the impact parameter is
441: high than when the transit is near-equatorial (all other things being
442: equal). This facilitates the detection of small changes in the impact
443: parameter due to orbital precession, which can be caused by additional
444: bodies in the system or by the stellar quadrupole field
445: \citep{Miralda-Escude.2002}. A large impact parameter is also
446: advantageous for interpreting the Rossiter-McLaughlin effect, as long
447: as an accurate external measurement of the projected rotation speed of
448: the star ($v\sin i$) is available \citep{Gaudi.2007}.
449:
450: Accurate timing of exoplanetary transits is a promising method to
451: identify additional planets or moons (see, e.g., Holman \& Murray~2005
452: and Agol et al.~2005), and in this case, transit timing takes on
453: special importance because TrES-2 is in the field of view of the {\it
454: Kepler} mission~\citep{Borucki.2003,Basri.2005}. We have tested
455: whether or not our 3 measured transit times and the single transit
456: time reported by \citet{O'Donovan.2006b} are consistent with a uniform
457: period, by fitting a linear function of epoch number to the observed
458: times. The residuals to this linear fit are shown in Fig.~2, and are
459: consistent with zero within the measurement errors. Thus there is not
460: yet any indication of timing anomalies. Based on our fit, we have
461: refined the ephemeris. The new value of $T_c$ is
462: $2,453,957.63479(38)$~[HJD] and the new value of the orbital period is
463: $2.470621(17)$~days, where the numbers in parentheses are the
464: 1~$\sigma$ uncertainties in the last 2 digits of each figure.
465:
466: \begin{figure}[p]
467: \epsscale{1.0}
468: \plotone{fig2.eps}
469: \caption{
470: The timing residuals (observed~$-$~calculated) for 4 observed
471: transits, according to the ephemeris of Eq.~(\ref{eq:ephemeris}). The
472: first point corresponds to the $T_c$ reported by
473: \citet{O'Donovan.2006b}, and the other three points correspond to the
474: three transits reported in this paper. The points lie on a horizontal
475: line, and therefore the data are consistent with a constant period.
476: \label{fig:tc}}
477: \end{figure}
478:
479: We conducted two tests to check the robustness of our results. First,
480: we fitted each of the three transits separately, and examined the
481: scatter in the results. For each of the parameters $\{R_\star, R_p,
482: i\}$, the 3 different results were all within the 1~$\sigma$
483: uncertainty of the result when fitting all the transits together. Thus
484: the results of the 3 transits agree well with one another. Second, we
485: examined the sensitivity of the results to the limb darkening
486: function, finding also that the results are robust. For example, the
487: effect on $R_p$ of allowing the quadratic limb-darkening coefficients
488: to be free parameters (rather than fixing them at the values tabulated
489: by \citet{Claret.2004}) is an increase of 1.0\%. If we use a linear
490: law instead of a quadratic law, $R_p$ is increased by 0.6\%, and if we
491: use the four-parameter ``nonlinear'' law of \citet{Claret.2004} (with
492: coefficients fixed at ATLAS-based values) then $R_p$ is decreased by
493: 0.5\%. None of these changes are very significant compared to the
494: 2.3\% statistical error.
495:
496: \section{Summary}
497:
498: Through observations and analysis of three transits, we have improved
499: upon the estimates of the orbital and physical parameters of TrES-2.
500: Our results are consistent with the estimates of the stellar and
501: planetary radii by \citet{O'Donovan.2006b}, but have smaller
502: uncertainties. We also show that the available transit times are
503: consistent with a uniform period. In our analysis of the photometry
504: we have made use of an improved estimate of the stellar mass from
505: \citet{Sozzetti.2007}. This estimate was obtained by iteratively
506: combining values of $a/R_\star$ determined from the light curves with
507: values of effective temperature and metallicity determined from
508: stellar spectra. This novel technique can be applied to all
509: transiting systems for which high quality stellar spectra and high
510: precision light curves are available. Our observations and analysis
511: help lay the ground work for interpreting the $\sim 600$ transits of
512: TrES-2 that will be observed by {\it Kepler}.
513:
514: \acknowledgments We thank E.\ Falco for accommodating our observing
515: schedule changes. KeplerCam was developed with partial support from
516: the Kepler Mission under NASA Cooperative Agreement NCC2-1390 (P.I.\
517: D.~Latham), and the KeplerCam observations described in this paper
518: were partly supported by grants from the Kepler Mission to SAO and
519: PSI\@. MJH acknowledges support for this work NASA Origins grant
520: NG06GH69G. Work by F.T.O'D. and D.C.\ was supported by NASA under
521: grant NNG05GJ29G, issued through the Origins of Solar Systems Program.
522: AS gratefully acknowledges the Kepler Mission for partial support under
523: NASA Cooperative Agreement NCC 2-1390. GT acknowledges partial support
524: for this work from NASA Origins grant NNG04LG89G.
525:
526:
527: \begin{thebibliography}{30}
528: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
529:
530: \bibitem[{{Agol} {et~al.}(2005){Agol}, {Steffen}, {Sari}, \&
531: {Clarkson}}]{Agol.2005}
532: {Agol}, E., {Steffen}, J., {Sari}, R., \& {Clarkson}, W. 2005, \mnras, 359, 567
533:
534: \bibitem[{{Basri} {et~al.}(2005){Basri}, {Borucki}, \& {Koch}}]{Basri.2005}
535: {Basri}, G., {Borucki}, W.~J., \& {Koch}, D. 2005, New Astronomy Review, 49,
536: 478
537:
538: \bibitem[{{Beatty} {et al.}(2007)}]{Beatty.2007}
539: {Beatty}, T.~G. {et al.} 2007, \apj, in press [arXiv:0704.0059]
540:
541: \bibitem[{{Borucki} {et~al.}(2003){Borucki}, {Koch}, {Basri}, {Caldwell},
542: {Caldwell}, {Cochran}, {Devore}, {Dunham}, {Geary}, {Gilliland}, {Gould},
543: {Jenkins}, {Kondo}, {Latham}, \& {Lissauer}}]{Borucki.2003}
544: {Borucki}, W.~J., et al.
545: %{Borucki}, W.~J., {Koch}, D.~G., {Basri}, G.~B., {Caldwell}, D.~A., {Caldwell},
546: % J.~F., {Cochran}, W.~D., {Devore}, E., {Dunham}, E.~W., {Geary}, J.~C.,
547: % {Gilliland}, R.~L., {Gould}, A., {Jenkins}, J.~M., {Kondo}, Y., {Latham},
548: % D.~W., \& {Lissauer}, J.~J.
549: 2003, in ASP Conf. Ser. 294: Scientific Frontiers
550: in Research on Extrasolar Planets, ed. D.~{Deming} \& S.~{Seager}, 427--440
551:
552: \bibitem[{{Charbonneau} {et~al.}(2005){Charbonneau}, {Allen}, {Megeath},
553: {Torres}, {Alonso}, {Brown}, {Gilliland}, {Latham}, {Mandushev}, {O'Donovan},
554: \& {Sozzetti}}]{Charbonneau.2005}
555: {Charbonneau}, D., et al.
556: %{Charbonneau}, D., {Allen}, L.~E., {Megeath}, S.~T., {Torres}, G., {Alonso},
557: % R., {Brown}, T.~M., {Gilliland}, R.~L., {Latham}, D.~W., {Mandushev}, G.,
558: % {O'Donovan}, F.~T., \& {Sozzetti}, A.
559: 2005, \apj, 626, 523
560:
561: \bibitem[{{Charbonneau} {et~al.}(2007){Charbonneau}, {Brown}, {Burrows}, \&
562: {Laughlin}}]{Charbonneau.2007}
563: {Charbonneau}, D., {Brown}, T.~M., {Burrows}, A., \& {Laughlin}, G. 2007, in
564: Protostars and Planets V, ed. B.~{Reipurth}, D.~{Jewitt}, \& K.~{Keil},
565: 701--716
566:
567: \bibitem[{{Charbonneau} {et~al.}(2002){Charbonneau}, {Brown}, {Noyes}, \&
568: {Gilliland}}]{Charbonneau.2002}
569: {Charbonneau}, D., {Brown}, T.~M., {Noyes}, R.~W., \& {Gilliland}, R.~L. 2002,
570: \apj, 568, 377
571:
572: \bibitem[{{Claret}(2004)}]{Claret.2004}
573: {Claret}, A. 2004, \aap, 428, 1001
574:
575: %\bibitem[{{Cody} \& {Sasselov}(2002)}]{Cody.2002}
576: %{Cody}, A.~M. \& {Sasselov}, D.~D. 2002, \apj, 569, 451
577:
578: \bibitem[{{Deming} {et~al.}(2006){Deming}, {Harrington}, {Seager}, \&
579: {Richardson}}]{Deming.2006}
580: {Deming}, D., {Harrington}, J., {Seager}, S., \& {Richardson}, L.~J. 2006,
581: \apj, 644, 560
582:
583: \bibitem[{{Deming} {et~al.}(2005){Deming}, {Seager}, {Richardson}, \&
584: {Harrington}}]{Deming.2005a}
585: {Deming}, D., {Seager}, S., {Richardson}, L.~J., \& {Harrington}, J. 2005,
586: \nat, 434, 740
587:
588: \bibitem[{{Dravins} {et~al.}(1998){Dravins}, {Lindegren}, {Mezey}, \&
589: {Young}}]{Dravins.1998}
590: {Dravins}, D., {Lindegren}, L., {Mezey}, E., \& {Young}, A.~T. 1998, \pasp,
591: 110, 610
592:
593: \bibitem[{{Gaudi} \& {Winn}(2007)}]{Gaudi.2007}
594: {Gaudi}, B.~S. \& {Winn}, J.~N. 2007, \apj, 655, 550
595:
596: \bibitem[{{Gillon} {et~al.}(2006){Gillon}, {Pont}, {Moutou}, {Bouchy},
597: {Courbin}, {Sohy}, \& {Magain}}]{Gillon.2006}
598: {Gillon}, M., {Pont}, F., {Moutou}, C., {Bouchy}, F., {Courbin}, F., {Sohy},
599: S., \& {Magain}, P. 2006, \aap, 459, 249
600:
601: \bibitem[{{Grillmair} {et~al.}(2007){Grillmair}, {Charbonneau}, {Burrows},
602: {Armus}, {Stauffer}, {Meadows}, {Van Cleve}, \& {Levine}}]{Grillmair.2007}
603: {Grillmair}, C.~J., {Charbonneau}, D., {Burrows}, A., {Armus}, L., {Stauffer},
604: J., {Meadows}, V., {Van Cleve}, J., \& {Levine}, D. 2007, \apjl, 658, L115
605:
606: \bibitem[{{Holman} \& {Murray}(2005)}]{Holman.2005}
607: {Holman}, M.~J. \& {Murray}, N.~W. 2005, Science, 307, 1288
608:
609: \bibitem[{{Holman} {et~al.}(2007){Holman}, {Winn}, {Fuentes}, {Hartman},
610: {Stanek}, {Torres}, {Sasselov}, {Gaudi}, {Jones}, \&
611: {Fraser}}]{Holman.2007}
612: {Holman}, M.~J., et al.
613: %{Holman}, M.~J., {Winn}, J.~N., {Fuentes}, C.~I., {Hartman}, J.~D., {Stanek},
614: % K.~Z., {Torres}, G., {Sasselov}, D.~D., {Gaudi}, B.~S., {Jones}, R.~L., \&
615: % {Fraser}, W.
616: 2007, \apj, 655, 1103
617:
618: \bibitem[{{Holman} {et~al.}(2006){Holman}, {Winn}, {Latham}, {O'Donovan},
619: {Charbonneau}, {Bakos}, {Esquerdo}, {Hergenrother}, {Everett}, \&
620: {P{\'a}l}}]{Holman.2006}
621: {Holman}, M.~J., et al.
622: %{Holman}, M.~J., {Winn}, J.~N., {Latham}, D.~W., {O'Donovan}, F.~T.,
623: % {Charbonneau}, D., {Bakos}, G.~A., {Esquerdo}, G.~A., {Hergenrother}, C.,
624: % {Everett}, M.~E., \& {P{\'a}l}, A.
625: 2006, \apj, 652, 1715
626:
627: \bibitem[{{Miralda-Escud{\' e}}(2002)}]{Miralda-Escude.2002}
628: {Miralda-Escud{\' e}}, J. 2002, \apj, 564, 1019
629:
630: \bibitem[{{O'Donovan} {et~al.}(2006){O'Donovan}, {Charbonneau}, {Mandushev},
631: {Dunham}, {Latham}, {Torres}, {Sozzetti}, {Brown}, {Trauger}, {Belmonte},
632: {Rabus}, {Almenara}, {Alonso}, {Deeg}, {Esquerdo}, {Falco}, {Hillenbrand},
633: {Roussanova}, {Stefanik}, \& {Winn}}]{O'Donovan.2006b}
634: {O'Donovan}, F.~T., et al.
635: %{O'Donovan}, F.~T., {Charbonneau}, D., {Mandushev}, G., {Dunham}, E.~W.,
636: % {Latham}, D.~W., {Torres}, G., {Sozzetti}, A., {Brown}, T.~M., {Trauger},
637: % J.~T., {Belmonte}, J.~A., {Rabus}, M., {Almenara}, J.~M., {Alonso}, R.,
638: % {Deeg}, H.~J., {Esquerdo}, G.~A., {Falco}, E.~E., {Hillenbrand}, L.~A.,
639: % {Roussanova}, A., {Stefanik}, R.~P., \& {Winn}, J.~N.
640: 2006, \apjl, 651, L61
641:
642: \bibitem[{{Queloz} {et~al.}(2000){Queloz}, {Eggenberger}, {Mayor}, {Perrier},
643: {Beuzit}, {Naef}, {Sivan}, \& {Udry}}]{Queloz.2000}
644: {Queloz}, D., {Eggenberger}, A., {Mayor}, M., {Perrier}, C., {Beuzit}, J.~L.,
645: {Naef}, D., {Sivan}, J.~P., \& {Udry}, S. 2000, \aap, 359, L13
646:
647: \bibitem[{{Richardson} {et~al.}(2007){Richardson}, {Deming}, {Horning},
648: {Seager}, \& {Harrington}}]{Richardson.2007}
649: {Richardson}, L.~J., {Deming}, D., {Horning}, K., {Seager}, S., \&
650: {Harrington}, J. 2007, \nat, 445, 892
651:
652: \bibitem[{{Seager} \& {Mall{\'e}n-Ornelas}(2003)}]{Seager.2003}
653: {Seager}, S. \& {Mall{\'e}n-Ornelas}, G. 2003, \apj, 585, 1038
654:
655: \bibitem[{{Southworth} {et~al.}(2004){Southworth}, {Zucker}, {Maxted},
656: \& Smalley}]{Southworth.2004}
657: {Southworth}, J., {Zucker}, S., {Maxted}, P.~F.~L., \& {Smalley}, B.\
658: 2004, \mnras, 355, 986
659:
660: \bibitem[{Southworth} {et~al.}(2007){Southworth}, {Wheatley}, \&
661: {Sams}] {Southworth.2007}
662: {Southworth}, J., {Wheatley}, P., {Sams}, G.\ 2007, \mnras, in press
663: [arXiv:0704.1570]
664:
665: \bibitem[{{Sozzetti} {et~al.}(2007){Sozzetti}, {Torres}, {Latham}, {Holman},
666: {Winn}, J.B., \& {O'Donovan}~F.T.{Charbonneau}}]{Sozzetti.2007}
667: {Sozzetti}, A., {Torres}, G., {Charbonneau}, D., {Latham}, D.~W., {Holman}, M.~J., {Winn}, J.~N.,
668: Laird, J.B., \& {O'Donovan},~F.T. 2007, \apj, in press
669:
670: \bibitem[{{Vidal-Madjar} {et~al.}(2003){Vidal-Madjar}, {Lecavelier des Etangs},
671: {D{\'e}sert}, {Ballester}, {Ferlet}, {H{\'e}brard}, \&
672: {Mayor}}]{Vidal-Madjar.2003}
673: {Vidal-Madjar}, A., {Lecavelier des Etangs}, A., {D{\'e}sert}, J.-M.,
674: {Ballester}, G.~E., {Ferlet}, R., {H{\'e}brard}, G., \& {Mayor}, M. 2003,
675: \nat, 422, 143
676:
677: \bibitem[{{Winn} {et~al.}(2007{\natexlab{a}}){Winn}, {Holman}, \&
678: {Fuentes}}]{Winn.2007a}
679: {Winn}, J.~N., {Holman}, M.~J., \& {Fuentes}, C.~I. 2007{\natexlab{a}}, \aj,
680: 133, 11
681:
682: \bibitem[{{Winn} {et~al.}(2007{\natexlab{c}}){Winn}, {Holman}, {Henry},
683: {Roussanova}, {Enya}, {Yoshii}, {Shporer}, {Mazeh}, {Johnson}, {Narita}, \&
684: {Suto}}]{Winn.2007c}
685: {Winn}, J.~N., et al.
686: %{Winn}, J.~N., {Holman}, M.~J., {Henry}, G.~W., {Roussanova}, A., {Enya}, K.,
687: % {Yoshii}, Y., {Shporer}, A., {Mazeh}, T., {Johnson}, J.~A., {Narita}, N., \&
688: % {Suto}, Y.
689: 2007{\natexlab{c}}, \aj, 133, 1828
690:
691: \bibitem[{{Winn} {et~al.}(2007{\natexlab{b}}){Winn}, {Holman}, \&
692: {Roussanova}}]{Winn.2007b}
693: {Winn}, J.~N., {Holman}, M.~J., \& {Roussanova}, A. 2007{\natexlab{b}}, \apj,
694: 657, 1098
695:
696: \bibitem[{{Winn} {et~al.}(2006){Winn}, {Johnson}, {Marcy},
697: {Butler}, {Vogt}, {Henry}, {Roussanova}, {Holman}, {Enya}, {Narita}, {Suto},
698: \& {Turner}}]{Winn.2006}
699: {Winn}, J.~N., et al.
700: %{Winn}, J.~N., {Johnson}, J.~A., {Marcy}, G.~W., {Butler}, R.~P., {Vogt},
701: % S.~S., {Henry}, G.~W., {Roussanova}, A., {Holman}, M.~J., {Enya}, K.,
702: % {Narita}, N., {Suto}, Y., \& {Turner}, E.~L.
703: 2006, \apjl, 653, L69
704:
705: \bibitem[{{Winn} {et~al.}(2005){Winn}, {Noyes}, {Holman}, {Charbonneau},
706: {Ohta}, {Taruya}, {Suto}, {Narita}, {Turner}, {Johnson}, {Marcy}, {Butler},
707: \& {Vogt}}]{Winn.2005}
708: {Winn}, J.~N., et al.
709: %{Winn}, J.~N., {Noyes}, R.~W., {Holman}, M.~J., {Charbonneau}, D., {Ohta}, Y.,
710: % {Taruya}, A., {Suto}, Y., {Narita}, N., {Turner}, E.~L., {Johnson}, J.~A.,
711: % {Marcy}, G.~W., {Butler}, R.~P., \& {Vogt}, S.~S.
712: 2005, \apj, 631, 1215
713:
714:
715:
716: \bibitem[{{Young}(1967)}]{Young.1967}
717: {Young}, A.~T. 1967, \aj, 72, 747
718:
719: \end{thebibliography}
720:
721: %\bibliography{ms}
722:
723: \begin{deluxetable}{lcc}
724: \tabletypesize{\normalsize}
725: \tablecaption{Relative Photometry of TrES-2\label{tbl:photometry}}
726: \tablewidth{0pt}
727:
728: \tablehead{
729: \colhead{HJD} & \colhead{Relative flux} & \colhead{Uncertainty} \\
730: }
731:
732: \startdata
733: %\input tab1.tex
734: 2453989.63669 & 0.9992 & 0.0013 \\
735: 2453989.63943 & 1.0022 & 0.0013 \\
736: 2453989.64013 & 0.9994 & 0.0013 \\
737: 2453989.64058 & 0.9971 & 0.0013 \\
738: 2453989.64105 & 1.0002 & 0.0013 \\
739: 2453989.64150 & 0.9994 & 0.0013
740: \enddata
741:
742: \tablecomments{The time stamps represent the Heliocentric Julian Date
743: at the time of mid-exposure. The uncertainty estimates are based on
744: the procedures described in \S~2. We intend for this Table to appear
745: in entirety in the electronic version of the journal. A portion is
746: shown here to illustrate its format. The data are also available in
747: digital form from the authors upon request.}
748:
749: \end{deluxetable}
750:
751: \begin{deluxetable}{lcc}
752:
753: \tabletypesize{\small}
754: \tablecaption{System Parameters of TrES-2\label{tbl:params}}
755: %\tablewidth{0pt}
756: \tablewidth{4.5in}
757:
758: \tablehead{
759: \colhead{Parameter} & \colhead{Value} & \colhead{Uncertainty}
760: }
761:
762: \startdata
763: $(R_\star/R_\odot)(M_\star/0.98~M_\odot)^{-1/3}$& $ 1.003$ & $ 0.017 $\\
764: $(R_p/R_{\rm Jup})(M_\star/0.98~M_\odot)^{-1/3}$& $ 1.222$ & $ 0.028 $\\
765: $R_\star/R_\odot$& $ 1.003$ & $ 0.027 $\\
766: $R_p/R_{\rm Jup}$& $ 1.222$ & $ 0.038 $\\
767: $R_p / R_\star$& $ 0.1253$ & $ 0.0010 $\\
768: $(R_p/a)^2$& $ 0.000270$ & $ 0.000012 $\\
769: $M_p/M_{\rm Jup}$\tablenotemark{1}& $ 1.198$ & $ 0.053 $\\
770: $T_{\rm eff}$~[K]\tablenotemark{1}& $ 5850 $ & $ 50 $\\
771: $a/R_\star$& $ 7.63$ & $ 0.12 $\\
772: $i$~[deg]& $ 83.57$ & $ 0.14 $\\
773: $b$& $ 0.8540$ & $ 0.0062 $\\
774:
775: $t_{\rm IV} - t_{\rm I}$~[hr]& $ 1.840$ & $ 0.020 $\\
776: $t_{\rm II} - t_{\rm I}$~[hr]& $ 0.683$ & $ 0.045 $\\
777: $T_c(13)$~[HJD]& $ 2453989.75286$ & $ 0.00029 $\\
778: $T_c(15)$~[HJD]& $ 2453994.69393$ & $ 0.00031 $\\
779: $T_c(34)$~[HJD]& $ 2454041.63579$ & $ 0.00030 $
780: \enddata
781:
782: \tablenotetext{1}{Adopted from Sozzetti et al. (2007).}
783:
784: \tablecomments{The system parameters and their associated
785: uncertainties for TrES-2 are listed. $t_{\rm I}$, $t_{\rm II}$, and
786: $t_{\rm IV}$, correspond to the times of the first, second, and fourth
787: points when the projected limb of the planet contacts that of the
788: star.}
789:
790: \end{deluxetable}
791:
792: \end{document}
793: