1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3: %\documentclass[manuscript]{aastex}
4:
5: %\usepackage{emulateapj5}
6: % \usepackage{aastexug}
7: \usepackage{amsmath}
8: \usepackage{amssymb}
9: \usepackage{latexsym}
10: \usepackage{textcomp}
11: \usepackage{graphicx}
12: \usepackage{subfigure}
13: \usepackage{lscape}
14: \usepackage{longtable}
15: \usepackage{apjfonts}
16:
17: \slugcomment{Accepted to ApJ, July 26, 2008}
18:
19: %\null\voffset=+0.0pc % +0.0pc
20:
21: \begin{document}
22: \title{Theoretical Radii of Extrasolar Giant Planets: \\
23: the Cases of TrES-4, XO-3b, and HAT-P-1b}
24:
25: \shorttitle{THEORETICAL RADII OF EXTRASOLAR GIANT PLANETS}
26: \shortauthors{LIU ET AL.}
27: \author{\sc Xin Liu\altaffilmark{1}, Adam Burrows\altaffilmark{1,2}, and Laurent Ibgui\altaffilmark{2}}
28:
29: \altaffiltext{1}{Department of Astrophysical Sciences, Princeton
30: University, Peyton Hall -- Ivy Lane, Princeton, NJ 08544;
31: xinliu@astro.princeton.edu, burrows@astro.princeton.edu}
32:
33: \altaffiltext{2}{Department of Astronomy and Steward Observatory,
34: The University of Arizona, Tucson, AZ \ 85721; laurent@as.arizona.edu}
35:
36: \begin{abstract}
37: To explain their observed radii, we present theoretical radius-age
38: trajectories for the extrasolar giant planets (EGPs) TrES-4,
39: XO-3b, and HAT-P-1b. We factor in variations in atmospheric
40: opacity, the presence of an inner heavy-element core, and possible
41: heating due to orbital tidal dissipation. A small, yet non-zero,
42: degree of core heating is needed to explain the observed radius of
43: TrES-4, unless its atmospheric opacity is significantly larger
44: than a value equivalent to that at 10$\times$solar metallicity
45: with equilibrium molecular abundances. This heating rate is
46: reasonable, and corresponds for an energy dissipation parameter
47: ($Q_p$) of $\sim$10$^{3.8}$ to an eccentricity of $\sim$0.01,
48: assuming 3$\times$solar atmospheric opacity and a heavy-element
49: core of $M_c = 30$ $M_{\oplus}$. For XO-3b, which has an observed
50: orbital eccentricity of $0.26$, we show that tidal heating needs
51: to be taken into account to explain its observed radius.
52: Furthermore, we reexamine the core mass needed for HAT-P-1b in
53: light of new measurements and find that it now generally follows
54: the correlation between stellar metallicity and core mass
55: suggested recently. Given various core heating rates, theoretical
56: grids and fitting formulae for a giant planet's equilibrium radius
57: and equilibration timescale are provided for planet masses $M_p=$
58: 0.5, 1.0, and 1.5 $M_J$ with $a =$ 0.02-0.06 AU, orbiting a G2V
59: star. When the equilibration timescale is much shorter than that
60: of tidal heating variation, the ``effective age'' of the planet is
61: shortened, resulting in evolutionary trajectories more like those
62: of younger EGPs. Motivated by the work of
63: \citet{jackson08a,jackson08b}, we suggest that this
64: effect could indeed be important in better
65: explaining some observed transit radii.
66: \end{abstract}
67:
68:
69: \keywords{planetary systems --- planets and satellites: general
70: --- stars: individual (HAT-P-1, GSC 02620-00648, XO-3)}
71:
72:
73: %------------------------------------------------------------------------------
74: \section{Introduction}\label{sec:intro}
75:
76: As of the writing of this paper, an astounding 47 transiting
77: extrasolar giant planets (EGPs) have been discovered\footnote{See
78: J. Schneider's Extrasolar Planet Encyclopaedia at
79: http://exoplanet.eu, the Geneva Search Programme at
80: http://exoplanets.eu, and the Carnegie/California compilation at
81: http://exoplanets.org}. For transiting planets, the
82: inclination/planet-mass degeneracy is resolved and the photometric
83: dip in the stellar flux during the transit yields the planet's
84: radius ($R_p$). Theory then attempts to explain the measured radii
85: \citep{guillot96,burrows00,burrows03,burrows04b,burrows07b,bodenheimer01,
86: bodenheimer03,baraffe03,baraffe04,baraffe08,chabrier04,fortney07a,laughlin05}.
87: Importantly, the comparison between theory and measurement must be
88: done for a given stellar type, orbital distance, planet
89: mass, and age. The latter is poorly measured, but crucially
90: important (see, e.g. \S\ref{sec:sum} re HD 209458b).
91:
92: \citet{burrows07b} modeled the theoretical evolution of the radii
93: of the 14 transiting EGPs known at the time. These authors suggest
94: that there are two radius anomalies in the transiting EGP family,
95: of which the smaller-radius anomaly is a result of the presence of
96: dense cores \citep{mizuno80,pollack96,hubickyj04}, whereas the
97: larger-radius anomaly might be explained by the enhanced
98: atmospheric opacities which slow down the heat loss of the core.
99: They also discussed the effects on planet structure of the
100: possible extra heat source in the interior, yet found no obvious
101: correlation between the requisite power and the intercepted
102: stellar power. Note that none of the transiting EGPs modeled by
103: \citet{burrows07b} is known to have a highly eccentric orbit.
104: However, recently several transiting EGPs with significantly
105: non-zero eccentricities ($e \gtrsim 0.15$) have been discovered,
106: including the EGPs XO-3b \citep{johnskrull07}, HAT-P-2b
107: \citep[a.k.a. HD 147506b;][]{bakos07b,loeillet07}, HD 17156b
108: \citep{barbieri07,gillon07d,narita07,irwin08}, and the ``hot
109: Neptune'' GJ 436b \citep{gillon07c,deming07a,demory07}.
110:
111:
112: At least for those systems with highly eccentric orbits, heating
113: due to orbital tidal dissipation
114: \citep{bodenheimer01,bodenheimer03} should be incorporated into
115: theoretical models for the radius-age trajectories. A preliminary
116: exploration of this in a restricted context motivates the present
117: paper. When the theoretical radius evolution calculations are
118: tailored to a system's specific planet mass, age, primary stellar
119: properties, and orbital distance (as they must), the current radii
120: of most of the known transiting EGPs can be explained by the
121: theoretical radius models of \citet{burrows07b}. In many
122: instances, a higher-density core mass provides an even better fit,
123: and extra internal heat sources are not required. However, at
124: least three transiting EGPs seem to be exceptional in some way.
125: The planets TrES-4 and XO-3b (for its large-radius solution based
126: on stellar parameters from spectral synthesis modeling) are cited
127: by their discoverers \citep{mandushev07,johnskrull07} as
128: anomalously large and inconsistent with extant theoretical models.
129: In addition, \citet{burrows07b} found that HAT-P-1b deviated from
130: the core-mass stellar-metallicity relationship followed by many of
131: the transiting EGPs that they modeled \citep[for the core-mass
132: stellar-metallicity correlation, also see][]{guillot06}.
133:
134: Therefore, with this paper, we focus on this small subset of three
135: interesting objects to determine how their radii can indeed be
136: explained with minimal assumptions, that nevertheless can include
137: tidal heating. We find that tidal heating in the core, given
138: measured (XO-3b) and possibly non-zero (TrES-4 and HAT-P-1b)
139: eccentricities, which nevertheless are still consistent with the
140: upper limits, can naturally explain the measured radii. For
141: HAT-P-1b with its new age estimate, we find that a core of
142: reasonable size can now be accommodated. Importantly, when tidal
143: heating needs to be invoked, a canonical tidal dissipation
144: parameter, ``$Q_p$''\citep{goldreich66}\footnote{Note that the
145: $Q_p$ parameter used throughout this paper corresponds to the $Q_p'$
146: in \citet{mardling07}.}, with a value near 10$^{4-6}$, along with
147: the measured eccentricity (or reasonable values consistent with
148: its current bounds), suffices to explain the measurements. Hence,
149: only simple extensions of the default evolution models that
150: incorporate a known process with canonical parameters are
151: required. We postulate that all measured transiting EGP radii can
152: be explained by available theory when proper account is taken of
153: the measured planet-star system parameters, reasonable core masses
154: that follow the relationship between core mass and stellar
155: metallicity \citep{burrows07b,guillot06}, and tidal heating at the
156: expected rate for non-zero, but measured, eccentricities.
157: Exceptions to this might arise if it is determined that tidal
158: dissipation occurs predominantly in the atmosphere, not the
159: convective core, and/or if the eccentricity and semi-major axis
160: history must be factored into the tidal heating history of the
161: planet. The latter effect is intriguing and has been suggested by
162: \citet{jackson08a,jackson08b}.
163:
164:
165: We describe our computational techniques and model assumptions in
166: \S\ref{sec:tech}. In \S\ref{sec:obs}, we review the measurements
167: of these three EGP systems, identify the discrepancies between the
168: observed planetary radii and those predicted by previous
169: theoretical models and present new theoretical radius-age
170: trajectories for them using tailored atmospheric boundary
171: conditions. These new trajectories and theoretical radii include
172: the effects of tidal heating in the convective core for measured
173: or reasonable values of the orbital eccentricities and for a range
174: of values for the tidal parameter, $Q_p$. In \S\ref{sec:heating},
175: we provide theoretical grids and fitting models for the
176: equilibrium planetary radius and the equilibration timescale,
177: given a certain set of planet mass, orbital distance, and tidal
178: heating rate for a G2V primary star. Finally, in \S\ref{sec:sum}
179: we summarize our results for each system, discuss the relevant
180: constraints obtained on their structural properties, and list
181: caveats concerning our model assumptions.
182:
183:
184: %------------------------------------------------------------------------------
185: \section{Computational Techniques and Model Assumptions}\label{sec:tech}
186:
187:
188: A detailed discussion of our computational techniques can be found
189: in \citet{burrows03} and \citet{burrows07b}. Here, we present only
190: a brief summary, along with our model assumptions. We generate
191: realistic atmospheres customized for the three EGPs, their
192: time-averaged orbital separations, and primary stars. The adopted
193: atmospheric boundary conditions incorporate irradiation using the
194: observed stellar luminosity and spectrum, and the measured planet
195: orbital distance. For planets with eccentric orbits, the
196: time-averaged insolation flux is employed in constructing the
197: atmospheric boundary conditions. The theoretical stellar spectra
198: of \citet{kuruz94} are adopted. For the given stellar spectrum and
199: flux (inferred from the luminosity and the planet orbital
200: distance), an $S$-$T_{eff}$-$g$ grid is calculated for the core
201: entropy, $S$, effective temperature, $T_{eff}$, and gravity, $g$,
202: using the discontinuous finite element (DFE) variant of the
203: spectral code TLUSTY \citep{hubeny95}. It is assumed that both the
204: stellar spectrum and flux are constant during the evolution.
205:
206: We employ the Henyey evolutionary code of \citet{burrows97} for
207: the radius-age evolutionary calculations, using the function
208: $T_{eff}(S, g)$ for the interior flux, inverted from the table of
209: $S$, $T_{eff}$, and $g$, referred to above. The helium fraction
210: $Y_{\rm He}$ is assumed to be 0.25. We calculate models with
211: different atmospheric opacities, the effect of which can be
212: conveniently mimicked by using 1$\times$solar, 3$\times$solar, and
213: 10$\times$solar abundance atmospheres. Note that the increase in
214: atmospheric opacity does not need to, and should not be due solely
215: to, increased metallicity. The effects of increased atmospheric
216: opacity and increased envelope heavy-element abundances are
217: decoupled, so that the implied increases in the heavy-element
218: burden of the envelope, if any, will not cancel the expansion
219: effect of enhanced atmospheric opacity \citep{burrows07b}. To
220: model the presence of a heavy-element core, a compressible ball of
221: olivine is placed in the center of the model planet, and pressure
222: continuity between the heavy-element core and the gaseous envelope
223: is ensured throughout the evolution. We adopt the \citet{saumon95}
224: equation of state (EOS) for the H$_2$/He envelope and the ANEOS by
225: \citet{thompson72} for olivine.
226:
227:
228: As noted, to model the atmospheric opacity, we use supersolar
229: metallicities (e.g., $3\times$solar, $10\times$solar) to mimic the
230: expansion effects of enhanced atmospheric opacity
231: \citep{burrows07b}. Possible causes for such enhanced opacities
232: might be supersolar metallicities in the atmosphere,
233: nonequilibrium chemistry, errors in the default opacities, and
234: thick hazes or absorbing clouds. Note that the
235: expansion effects of enhanced atmospheric opacity and the
236: shrinkage effects of increased envelope metallicities conceptually
237: decoupled in our models and that an increase of the envelope
238: heavy-element burden, will not necessarily cancel the expansion
239: effect due to enhanced atmospheric opacity. The effect of a
240: central heavy-element core on the planet radius is to shrink it
241: monotonically with core mass.
242:
243: Given a non-zero orbital eccentricity, the tidal dissipation rate
244: is calculated using the formulation summarized in the Appendix
245: \citep{bodenheimer01,bodenheimer03,gu04}. We assume that the
246: planet is in synchronous rotation and that all the tidal heating
247: is in the convective core. Note that the effects of other core
248: energy dissipation mechanisms on planet structural evolution are
249: also implicitly addressed. As indicated in the Appendix, the tidal
250: heating rate is proportional to $f(e)/Q_p$, where $f(e) = e^2$
251: when $e \ll 1$. Also, the values of $Q_p$ for EGPs with masses
252: $M_p\sim$ $M_J$, although very uncertain, are estimated to be
253: $\sim10^4-10^7$ \citep{adams06,gu03,jackson08a,jackson08b}.
254: Therefore, we calculate typical heating rates using the
255: combination $e^2/Q_p$ for TrES-4 and HAT-P-1b, of which the
256: orbital eccentricities have been estimated to be $\ll 1$, if
257: nonzero at all. Models without external heat sources ($e^2/Q_p =
258: 0$) are also presented.
259:
260:
261:
262:
263: %------------------------------------------------------------------------------
264: \section{Observed Properties and Theoretical Planetary Radii}\label{sec:obs}
265:
266: Table \ref{tab:planet} displays the relevant observed quantities
267: and the corresponding references for the EGPs TrES-4, XO-3b, and
268: HAT-P-1b and their parent stars. These properties include
269: semi-major axis ($a$), orbital period and eccentricity ($e$),
270: stellar mass ($M_{\ast}$), radius ($R_{\ast}$), effective
271: temperature $T_{eff}$, metallicity ([Fe/H]$_{\ast}$), age, and
272: planetary mass ($M_p$) and radius ($R_p$). Also shown are the
273: stellar flux at the planet's substellar point, $F_p$, in units of
274: $10^9$ erg cm$^{-2}$ s$^{-1}$, and the ratio between the possible
275: tidal energy dissipation rate within the planet and the insolation
276: rate $\dot{E}_{{\rm tide}}/\dot{E}_{{\rm insolation}}$ in the unit
277: of $(Q_p/10^5)^{-1}$. $Q_p^{-1} \equiv \frac{1}{2\pi E_0}\oint
278: (-\frac{dE}{dt})dt$ is the specific dissipation function of the
279: planet, where $E_0$ is the maximum energy stored in the tidal
280: distortion and $-\frac{dE}{dt}$ is the rate of dissipation
281: \citep{goldreich66}. The parameters for TrES-4 and HAT-P-1b are
282: drawn from \citet{torres08}. These authors provide a uniform
283: analysis of transit light curves and stellar parameters based on
284: stellar evolution models, and a critical examination of the
285: corresponding errors. Since XO-3b has not been studied by
286: \citet{torres08}, we generate models using parameters
287: reported by \citet{johnskrull07} and \citet{winn08}.
288:
289: Theoretical evolutionary trajectories are presented for TrES-4,
290: XO-3b, and HAT-P-1b under various assumptions about the
291: atmospheric opacity, the presence of a heavy-element core and
292: possible tidal heating. The transit radius effect
293: \citep{burrows07b} is also included in the models.
294:
295:
296: \subsection{TrES-4}
297:
298:
299:
300: TrES-4 is the current record-holder for the lowest EGP density
301: \citep{mandushev07,torres08}. In the discovery work,
302: \citet{mandushev07} carried out spectroscopic observations with
303: the CfA Digital Speedometer \citep{latham92} , radial velocity
304: (RV) measurements with Keck, and transit photometry in the $z$
305: band with KeplerCam at the F. L. Whipple Observatory (FLWO) and in
306: the $B$ band using NASACam on the 0.8-m telescope at the Lowell
307: Observatory. Assuming [Fe/H]$_{\ast}$ = $0.0_{-0.2}^{+0.2}$ dex,
308: Mandushev et al. derived $T_{eff} = 6100_{-150}^{+150}$ K, $M_{\ast}
309: = 1.22_{-0.17}^{+0.17}$ $M_{\odot}$, $R_{\ast} =
310: 1.738_{-0.092}^{+0.092}$ $R_{\odot}$, and an age of
311: $4.7_{-2.0}^{+2.0}$ Gyr for the star, and $M_p =
312: 0.84_{-0.10}^{+0.10}$ $M_J$ and $R_p = 1.674_{-0.094}^{+0.094}$
313: $R_J$ for the planet. The orbital eccentricity was assumed to be
314: exactly zero in the fit. \citet{mandushev07} suggested that its
315: observed radius is too large to be explained by the theoretical
316: EGP models of \citet{burrows07b}, given its estimated mass, age
317: and insolation, even when the effects of higher atmospheric
318: opacities and the transit radius correction are considered.
319:
320:
321: The parameters of TrES-4 listed in Table \ref{tab:planet} are from
322: \citet{torres08}. These authors derived these parameters using the
323: RV measurements and transit photometry from \citet{mandushev07},
324: and the stellar atmospheric properties from Sozzetti {\it et al.}
325: (2008; in preparation). Note that their estimated planetary radius
326: $R_p = 1.751_{-0.062}^{+0.064}$ $R_J$ is $\sim$1-$\sigma$ larger
327: than the value of $R_p = 1.674_{-0.094}^{+0.094}$ $R_J$ of
328: \citet{mandushev07}.
329:
330:
331: \subsubsection{Results for TrES-4}\label{sec:res:tres4}
332:
333: \begin{figure}
334: \centering
335: \includegraphics[width=89mm]{f1a.eps}
336: \includegraphics[width=89mm]{f1b.eps}
337: \caption{Theoretical planet radius $R_p$ ($R_J$) versus age (Gyr)
338: for TrES-4. Also shown on both panels with error bars are the observed radius and
339: age from \citet{torres08}. Various values are assumed concerning
340: the atmospheric opacity, the presence of a heavy-element core, and the core heating due
341: to tidal dissipation. Different colors correspond to different tidal heating rates,
342: which are proportional to $e^2/Q_p$ when $e \ll 1$ (see the Appendix for more details).
343: {\it Top}: This panel demonstrates the effect of enhanced atmospheric opacity under various
344: heating powers. Models assuming 3$\times$solar ($10\times$solar) atmospheric
345: opacities are plotted as solid (dashed) curves, whereas the black, blue, and red curves
346: correspond to $(e/0.01)^2(10^5/Q_p) = 0.0$, 0.1, and 6.0, all without a heavy-element core
347: ($M_c$ = 0). {\it Bottom}: The effect of the presence of a heavy-element core is illustrated, where the
348: dashed (solid) curves denote models with $M_c$ = 30 (0) $M_{\oplus}$. If TrES-4 follows the
349: core-mass stellar-metallicity relation found by \citet{burrows07b}, then it should contain
350: a heavy-element core with $M_c\sim$$20-40$ $M_{\oplus}$, given its stellar metallicity
351: [Fe/H]$_{\ast}$ ($+0.14_{-0.09}^{+0.09}$ dex). Assuming 3$\times$solar atmospheric opacity,
352: the model with $Q_p = 10^{3.8}$, $e \sim$0.01, and a heavy-element core of $M_c = 30$ $M_{\oplus}$
353: ({\it black-dashed} curve) explains the observed radius well.}
354: \label{fig:tres4}
355: \end{figure}
356: %-------------
357:
358:
359:
360: We calculate the radius-age trajectories for TrES-4 using the
361: parameters from \cite{torres08}, taking into account the possible
362: effects of enhanced atmospheric opacities, and the presence of
363: tidal dissipation given a small, yet non-zero, orbital
364: eccentricity. Models with the presence of a heavy-element core are
365: also calculated. The value of $\dot{E}_{{\rm tide}}/\dot{E}_{{\rm insolation}}$
366: quoted for TrES-4 in Table \ref{tab:planet} is calculated assuming
367: $e = 0.01$, but using all the other parameters from
368: \cite{torres08} in which a circular orbit is assumed. More and
369: better transit observations are needed to better constrain its
370: true orbital eccentricity.
371:
372: The first panel of Fig. \ref{fig:tres4} shows the theoretical
373: radii $R_p$ (in units of $R_J$) as a function of age (in units of Gyr) for
374: TrES-4, under various assumptions concerning the atmospheric opacity and the
375: level of tidal dissipation, without a heavy-element core
376: ($M_c = 0$). Models assuming $3\times$ ($10\times$) solar
377: atmospheric opacities are shown as solid (dotted) curves. The
378: black curves show the models without any heat sources, whereas the
379: blue (red) curves depict those with a tidal heating rate assuming
380: $(e/0.01)^2(10^5/Q_p) = 0.1$ ($6.0$). Also shown with error bars
381: are the observed radius and age from \citet{torres08} (Table
382: \ref{tab:planet}).
383:
384: It can be seen from Fig. \ref{fig:tres4} that if the radius and
385: age estimates of \citet{torres08} do not deviate much from their
386: true values (within the uncertainties), our models without any
387: heat sources produce radii which are $\sim$3 $\sigma$ too small.
388: Assuming the \citet{mandushev07} parameters, the theoretical radii
389: are still $\sim$2 $\sigma$ too small. The discrepancy will become
390: smaller for models with even higher atmospheric opacities. So it
391: is concluded that either the atmospheric opacity of TrES-4 is
392: unusually large (much higher than the equivalent of a
393: $10\times$solar metallicity, equilibrium mixture), or there are
394: extra heat sources in the core. The required heating power is very
395: modest; the model with $(e/0.01)^2(10^5/Q_p) = 0.1$ and
396: 10$\times$solar atmospheric opacity produces theoretical radii
397: consistent with the 1-$\sigma$ lower bound of $R_p$ from
398: \citet{torres08}.
399:
400: The models shown in the first panel of Fig. \ref{fig:tres4} do not
401: include any heavy-element cores. If TrES-4 follows the core-mass
402: stellar-metallicity relation studied by \citet{burrows07b}, then
403: there should be a heavy element core with $M_c\sim$$20-40$
404: $M_{\oplus}$, given its stellar metallicity [Fe/H]$_{\ast}$
405: ($+0.14_{-0.09}^{+0.09}$ dex). The presence of such a
406: heavy-element core will shrink the planetary radius, and,
407: therefore, would require a higher tidal heating rate than do
408: models without a core to explain the observed radius. This effect
409: of including a heavy-element core in the models for TrES-4 is
410: shown in the second panel of Fig. \ref{fig:tres4}. Assuming
411: 3$\times$solar atmospheric opacity, the model with $Q_p =
412: 10^{3.8}$, $e \sim$0.01, and a heavy-element core of $M_c = 30$
413: $M_{\oplus}$ ({\it black-dashed}) explains the observed radius
414: well. Within 1 $\sigma$ uncertainties, the model with $Q_p =
415: 10^{4.0}$, $e \sim$0.01, and a heavy-element core of $M_c = 30$
416: $M_{\oplus}$ ({\it red-dashed} curve) can also fit the observed
417: radius.
418:
419: In summary, unless the atmospheric opacity of TrES-4 is unusually
420: large, core heating is required to explain its observed radius.
421: However, the required heating power is modest. A non-core model
422: with $(e/0.01)^2(10^5/Q_p) = 0.1$ produces radii consistent with
423: the 1-$\sigma$ lower boundary of $R_p$ from \citet{torres08},
424: assuming $10 \times$solar atmospheric opacity. The required energy
425: dissipation rates become larger for models with a heavy-element
426: core, but are still reasonable. For instance, the model with $Q_p
427: = 10^{3.8}$, $e \sim$0.01, and a heavy-element core of $M_c = 30$
428: $M_{\oplus}$ produces the observed radius well, assuming
429: 3$\times$solar atmospheric opacity. To better constrain models of
430: TrES-4, definitive measurements of, or stronger limits to, its
431: orbital eccentricity are needed.
432:
433:
434: \subsection{XO-3b}
435:
436: XO-3b has been observed to be supermassive and on an eccentric
437: orbit \citep[$M_p = 13.25_{-0.64}^{+0.64}$ $M_J$, $e =
438: 0.260_{-0.017}^{+0.017}$;][]{johnskrull07}. The discoverers
439: obtained transit light curves with relatively small 0.3-m
440: telescopes, spectroscopic observations using the 2.7-m Harlan J.
441: Smith (HJS) telescope and the 11-m Hobby-Eberly Telescope (HET) ,
442: and RV measurements with the HJS telescope. Based on theoretical
443: spectral models of the HJS data, \citet{johnskrull07} derive
444: $T_{eff} = 6429_{-50}^{+50}$ K, [Fe/H]$_{\ast}$ =
445: $-0.177_{-0.027}^{+0.027}$ dex, and $\log g_{\ast} =
446: 3.95_{-0.062}^{+0.062}$ for the star. Combined with the RV
447: measurements, they arrive at $M_{\ast} = 1.41_{-0.08}^{+0.08}$
448: $M_{\odot}$ and $R_{\ast} = 2.13_{-0.21}^{+0.21}$ $R_{\odot}$ for
449: the star, and $M_p = 13.25_{-0.64}^{+0.64}$ $M_J$ and $R_p =
450: 1.95_{-0.16}^{+0.16}$ $R_J$ for the planet. These authors have
451: commented that XO-3b is observed to be so large that in all cases
452: analyzed by \citet{fortney07a}, their models predict a much
453: smaller radius. However, due to the absence of a precise
454: trigonometric parallax of XO-3, its distance is very uncertain.
455: Assuming a smaller distance, and, hence, a reduced stellar mass and
456: radius than obtained using the isochrone method,
457: \citet{johnskrull07} found a best fit to their transit light
458: curves with $\log g_{\ast} = 4.19$, $M_{\ast} = 1.24 M_{\odot}$,
459: and $R_{\ast} = 1.48$ $R_{\odot}$, with the corresponding
460: estimates for $M_p$ of $12.03_{-0.46}^{+0.46}$ $M_J$ and $R_p$ of
461: $1.25_{-0.15}^{+0.15}$ $R_J$. These light-curve-based results
462: have recently been strengthened by \citet{winn08}
463: using larger aperture telescopes. These authors
464: observed 13 transits photometrically using the 1.2-m
465: telescope at the FLWO, along with 0.4-0.6-m telescopes. Based
466: on these more precise transit light curves, they concluded that
467: $\log g_{\ast} = 4.244$, $M_{\ast} = 1.213 M_{\odot}$, and
468: $R_{\ast} = 1.377$ $R_{\odot}$, with the corresponding estimates
469: for $M_p$ of $11.79_{-0.59}^{+0.59}$ $M_J$ and for $R_p$ of
470: $1.217_{-0.073}^{+0.073}$ $R_J$. Since a trigonometric parallax
471: measurement of XO-3 is still lacking that could distinguish
472: these two different methods, we make models for both the
473: spectroscopically-determined and light-curve-based parameter sets.
474:
475: Both the spectroscopical results of \citet{johnskrull07} and the
476: light-curve-based results of \citet{winn08} are listed in Table
477: \ref{tab:planet}. We note that the $M_p$ and $R_p$ values derived
478: using the two different methods differ significantly from one
479: another ($\sim$10\% in $M_p$ and $\sim$50\% in $R_p$). Therefore,
480: separate models and discussions for these different sets of planetary
481: properties are in order and we calculate theoretical radii for
482: XO-3b for both estimates of the planetary radius and mass.
483:
484:
485: \subsubsection{Results for XO-3b}\label{sec:res:xo3}
486:
487: \begin{figure}
488: \centering
489: \includegraphics[width=89mm]{f2a.eps}
490: \includegraphics[width=89mm]{f2b.eps}
491: \caption{Theoretical planet radius $R_p$ ($R_J$) versus age (Gyr) for
492: XO-3b. In the first panel, a planet mass of $M_p = 13.24$ $M_J$ based on the spectral
493: synthesis model of \citet{johnskrull07} and 3$\times$solar atmospheric opacity are assumed,
494: whereas in the second panel the corresponding values are $M_p = 11.79$ $M_J$ according
495: to the light-curve fitting results of \citet{winn08} and 1$\times$solar atmospheric opacity. Also shown with error
496: bars on both panels are the observed radii and age estimates from \citet{johnskrull07} and \citet{winn08},
497: based on the two different analyses. In both panels, models assuming various tidal
498: heating rates corresponding to
499: $Q_p = \infty$ (no heating), 10$^{7.0}$, 10$^{6.0}$, 10$^{5.7}$, and 10$^{5.5}$ are color coded.
500: In all the models, $e = 0.260$ \citep{johnskrull07} is assumed. Note that the observed radii, based on the
501: two different methods, differ quite a bit from one another ($\sim$50\%). The very large radius based on
502: spectral synthesis modeling can be fit by the model with $Q_p\sim$$10^{5.7}$, whereas the much
503: smaller one inferred from the light-curve fit can be explained by models with $Q_p$ values down
504: to $10^6$, within 1-$\sigma$ errors. See \S\ref{sec:res:xo3} for more discussion.}
505: \label{fig:xo3}
506: \end{figure}
507: %-------------
508:
509:
510:
511: We include the possible heating due to
512: orbital tidal dissipation, assuming reasonable values of $Q_p$.
513: Our results for XO-3b are shown in Fig. \ref{fig:xo3}, where in
514: the first panel a planet mass of $M_p = 13.24$ $M_J$, derived from
515: the spectral synthesis method of \citet{johnskrull07}, and
516: an atmospheric opacity associated with 3$\times$solar metallicity are assumed, whereas in the
517: second panel the corresponding values are $M_p = 11.79$ $M_J$ from
518: the light-curve fitting results of \citet{winn08} and
519: 1$\times$solar atmospheric opacity. In both cases, we assume that
520: a massive heavy-element core is absent. For such massive planets,
521: the effect of reasonable core masses on planetary radii is small.
522:
523: As shown in the first panel of Fig. \ref{fig:xo3}, the very large
524: radius derived by \citet{johnskrull07}, based on spectral
525: synthesis modeling, can be explained by energy dissipation due to
526: tidal heating when the $Q_p$ parameter of XO-3b is
527: $\sim$$10^{5.7}$. For the case of the light-curve-based planetary
528: mass ({second panel} of Fig. \ref{fig:xo3}), the inferred radius
529: is broadly consistent to within 1-$\sigma$ errors with the absence
530: of an internal heat source, or with tidal heating with $Q_p
531: \gtrsim 10^6$. These specific constraints on $Q_p$ assume
532: 3$\times$solar in the former case and 1$\times$solar atmospheric
533: opacity in the latter. For our first model of XO-3b assuming $M_p
534: = 13.24$ $M_J$ and 3$\times$solar atmospheric opacity, it would be
535: surprising to find a planet with $R_p/R_J \lesssim 1.2$ or
536: $\gtrsim 2.2$, since this will require $Q_p$ to be either too large
537: or too small. For the second model assuming $M_p = 11.79$ $M_J$
538: and 1$\times$solar atmospheric opacity, such a region would be
539: $R_p/R_J \lesssim 1.1$ or $\gtrsim 1.6$, for the same reason.
540: Given significant eccentricity, it is important to appropriately
541: account for tidal heating in order to model the planet's
542: structural evolution. We want to emphasize our adopted model
543: assumptions that 1) the planet is in synchronous rotation and that
544: 2) all the tidal heating is in the convective core. Even though the
545: $e^2/Q_p$ degeneracy is broken due to the known value of $e$,
546: detailed radius evolution models could be used to constrain $Q_p$
547: for EGPs, but only if it is determined that tidal dissipation occurs
548: predominantly either in the convective core or in the atmosphere,
549: and if the uncertainties in the core mass and atmospheric opacity are
550: both resolved.
551:
552:
553:
554: \subsection{HAT-P-1b}
555:
556:
557: Using photometry conducted by the Hungarian-made Automated
558: Telescope Network (HATNet) project, \citet{bakos07a} discovered
559: HAT-P-1b transiting one member of the stellar binary ADS 16402.
560: These authors suggested that HAT-P-1b was too large to be
561: explained by theoretical EGP models. Spectral synthesis modeling
562: of the parent star ADS 16402B, based on its Keck spectra, yielded
563: $T_{eff} = 5975_{-45}^{+45}$ K and [Fe/H]$_{\ast}$ =
564: $+0.13_{-0.02}^{+0.02}$ dex. \citet{bakos07a} also fit both
565: stellar members in the binary to evolutionary tracks and based on
566: the Subaru and the Keck spectra derived $M_{\ast} =
567: 1.12_{-0.09}^{+0.09} M_{\odot}$ and $R_{\ast} =
568: 1.15_{-0.07}^{+0.10} R_{\odot}$ for ADS 16402B and a best-fit age
569: of $3.6$ Gyr for the binary. Using the $z$-band transit curves
570: from KeplerCam \citep{holman06}, combined with RV measurements
571: from Subaru and Keck, these authors derive $M_p =
572: 0.53_{-0.04}^{+0.04} M_J$ and $R_p = 1.36_{-0.09}^{+0.11} R_J$,
573: where the errors in the planetary radius include both statistical
574: and systematic errors in both the stellar radius and mass. Note
575: that in the above fits, a circular orbit ($e = 0$) was assumed.
576: However, the $\chi^2$ fitting of the RV data in \citet{bakos07a}
577: favors a small, yet non-zero, eccentricity:
578: $e=0.09_{-0.02}^{+0.02}$. These authors did estimate the heating
579: rate assuming $e = 0.09$ and suggested that if this non-zero
580: orbital eccentricity is confirmed, the observed large $R_p$ could
581: be explained by tidal heating. However, note that a non-zero
582: eccentricity is only suggestive, mainly due to the small number of
583: RV observations (13 velocities, for which the typical S/N is about
584: 150 per pixel). Since RV-based eccentricity estimates are
585: positively biased due to noise \citep[e.g.,][]{shen08}, it is very
586: likely that the true $e$ is smaller than $0.09$. In fact,
587: \citet{johnson08} find a upper limit on $e$ of 0.067, with 99\%
588: confidence, by combining their new and previous RV measurements.
589: Therefore, we assume a smaller value, $e = 0.01$, in our baseline
590: model and see where such an assumption leads. Such a small
591: eccentricity could result from Kozai cycles with tidal friction
592: \citep[e.g.,][]{fabrycky07}, though there is evidence that the
593: spin-orbit misalignment is small \citep{johnson08}.
594:
595:
596: Based on more high-precision transit observations, however,
597: \citet{winn07d} report that HAT-P-1b is less ``bloated'' than
598: originally thought. Their observations include three transits
599: observed in $z$ band with the 1.2-m telescope at the FLWO, three
600: observed through the ``Gunn Z'' filter \citep{pinfield97} using
601: the Nickel 1-m telescope at Lick Observatory, and three observed
602: through the Johnson $I$ filter using the 1-m telescope at the Wise
603: Observatory. \citet{winn07d} derived $R_{\ast}/M_{\ast}^{1/3}$ by
604: fitting the transit light curves, and concluded that $R_{\ast} =
605: 1.115_{-0.043}^{+0.043} R_{\odot}$ and $R_p =
606: 1.203_{-0.051}^{+0.051} R_J$. Note that in their fits the orbital
607: eccentricity was assumed to be zero. These authors suggest that
608: the updated radius can be explained by the structural models of
609: \citet{burrows07b}, unless the planet has a very massive core of
610: heavy elements. Indeed, \citet{burrows07b} calculated radius-age
611: trajectories for HAT-P-1b. They included different core masses and
612: atmospheric opacities in their models, and found that in order to
613: fit the observed radius, HAT-P-1b deviates from the
614: stellar-metallicity versus core-mass sequence otherwise roughly
615: followed by the transiting EGPs included in their paper \citep[see
616: Fig. 9 of][]{burrows07b}. However, the stellar and planetary
617: parameters of the HAT-P-1 system adopted by \citet{burrows07b} are
618: from the discovery work of \citet{bakos07a}. The parameters of the
619: HAT-P-1 system derived by \citet{torres08} are listed in Table
620: \ref{tab:planet}. These authors compile the $z$-band light curves
621: of \citet{winn07d} and the RV data and the atmospheric parameters
622: of \citet{bakos07a}, but with increased uncertainties for
623: $T_{eff}$, [Fe/H]$_{\ast}$, and $\log g_{\ast}$. As we show in
624: \S\ref{sec:res:hatp1}, using these new parameters and the new
625: planet radius, we now find inferred core masses that are roughly
626: consistent with the stellar-metallicity versus core-mass
627: relationship followed by the EGPs studied by \citet{burrows07b}.
628:
629:
630: %-------------
631: \begin{figure}
632: \centering
633: \includegraphics[width=89mm]{f3a.eps}
634: \includegraphics[width=89mm]{f3b.eps}
635: \caption{Theoretical planet radius $R_p$ ($R_J$) versus age (Gyr)
636: for HAT-P-1b. Also shown with error bars are the observational stellar age and planet
637: radius from \citet{torres08}, as listed in Table \ref{tab:planet}. We reexamine the best-fit
638: core masses for HAT-P-1b, under various assumptions concerning the atmospheric opacity and the
639: possible core heating rate due to tidal dissipation. The top (bottom) panel shows models assuming
640: 3$\times$solar (10$\times$solar) atmospheric opacities. In both panels, the black, blue, and
641: red curves denote models assuming $(e/0.01)^2(10^5/Q_p) = $ 0, 0.8, and 8, respectively.
642: Different line styles correspond to models with various heavy-element core masses in units
643: of the Earth mass, $M_{\oplus}$, as labeled on the plot. If there is tidal heating assuming
644: reasonable values of $(e/0.01)^2(10^5/Q_p) \sim$$0.8$, then the core mass required to fit
645: the observed radius is $\sim$30 $M_{\oplus}$ ($\sim$40 $M_{\oplus}$), assuming 3$\times$solar
646: (10$\times$solar) atmospheric opacity, in which case HAT-P-1b does follow the core-mass
647: stellar-metallicity relation found by \citet{burrows07b}. See \S \ref{sec:res:hatp1} for a
648: discussion.}
649: \label{fig:hatp1}
650: \end{figure}
651: %-------------
652:
653:
654: \subsubsection{Results for HAT-P-1b}\label{sec:res:hatp1}
655:
656: Given the new measured radii and stellar age of \citet{torres08},
657: we have reexamined the best-fit core masses for HAT-P-1b. The
658: effects of the possible heating due to tidal dissipation on the
659: planet's structural evolution are considered, assuming a small yet
660: non-zero eccentricity $e = 0.01$. Theoretical evolutionary
661: trajectories of planet radius with age for HAT-P-1b are shown in
662: Fig. \ref{fig:hatp1}, where the first (second) panel displays the
663: results assuming a 3$\times$solar (10$\times$solar) atmospheric
664: opacity. Models with different tidal heating rates proportional to
665: $(e/0.01)^2(10^5/Q_p)$ are color-coded as labeled. For both of the
666: panels, different line styles represent ``no heavy-element
667: cores," or the presence of a heavy-element core with a range of
668: masses in units of Earth masses, $M_{\oplus}$. For clarity, only
669: selected models are presented in the figure. Table \ref{tab:hatp1}
670: lists the best estimates for the core mass under various
671: assumptions.
672:
673: As demonstrated in Fig. \ref{fig:hatp1} and Table \ref{tab:hatp1},
674: there are multiple solutions to explain the observed radius of
675: HAT-P-1b. In all the cases considered, a non-zero heavy-element
676: core mass is needed, which, without any external heat sources, is
677: $\sim$15 $M_{\oplus}$ ($\sim$20 $M_{\oplus}$) assuming
678: 3$\times$solar (10$\times$solar) atmospheric opacity,. This
679: best-fit core mass becomes larger when there is external heating.
680: If HAT-P-1b does follow the approximate
681: core-mass/stellar-metallicity relation found by
682: \citet{burrows07b}, then given its [Fe/H]$_{\ast}$
683: ($+0.13_{-0.08}^{+0.08}$ dex), the core mass would be $\sim$30
684: $M_{\oplus}$ ($\sim$40 $M_{\oplus}$) assuming 3$\times$solar
685: (10$\times$solar) atmospheric opacity. These core masses
686: correspond to the cases with $(e/0.01)^2(10^5/Q_p) = 0.8$.
687:
688: In summary, if there is tidal heating with reasonable values of
689: $Q_p$, the core-mass estimates suggest that HAT-P-1b follows the
690: correlation between stellar metallicity and core mass found by
691: \citet{burrows07b}, or if there is no extra heating, deviates
692: mildly from the correlation sequence. However, a larger core mass,
693: more in keeping with the correlation found by \citet{burrows07b},
694: is more consistent with reasonable values of
695: $(e/0.01)^2(10^5/Q_p)$, as long as $e$ is non-zero and $Q_p$ is
696: not anomalously small.
697:
698:
699:
700: %------------------------------------------------------------------------------
701:
702: \section{Equilibrium Planetary Radii and Equilibration Timescales for Various Heating Rates: a Parameter Study}\label{sec:heating}
703:
704:
705: %------------------------------------------------------------------------------
706:
707:
708: \begin{figure*}
709: \centering
710: \includegraphics[width=166mm]{f4a.eps}
711: \includegraphics[width=166mm]{f4b.eps}
712: \caption{Equilibrium planetary radius $R_{eq}$ ($R_J$) and equilibration timescale
713: (Gyr) assuming various ratios between the core heating
714: power and the insolation power. The equilibration timescale is defined as the
715: time it takes the planet to evolve from 1.25 $R_{eq}$ to 1.05 $R_{eq}$. Models are calculated
716: for planets with masses $M_p =$ 0.5, 1.0, and 1.5 $M_J$, and semi-major axes $a =$ 0.02,
717: 0.03, 0.04, 0.05, and 0.06 AU, orbiting a G2V star. Filled circles represent results
718: calculated from radius-age trajectories, whereas the curves are fits to them given by
719: eq. (\ref{eq:fitting}) and the corresponding parameters in Table \ref{tab:fitting}.
720: See \S \ref{sec:heating} for more information.}
721: \label{fig:equiradiage}
722: \end{figure*}
723: %-------------
724:
725:
726:
727: In this section, we investigate the effects of generic core
728: heating on the planet's equilibrium radius, and the time to reach
729: this equilibrium. We refer to the latter as the equilibration
730: timescale. The heat source discussed in this section could be due
731: to orbital tidal heating, but does not have to be. Figure
732: \ref{fig:equiradiage} shows equilibrium planetary radii ($R_{eq}$)
733: and equilibration timescales ($\tau_{eq}$) with various
734: core-heating powers, quantified by
735: $\dot{E}_{heating}/\dot{E}_{insolation}$. Models are calculated
736: for planets with masses $M_p =$ 0.5, 1.0, and 1.5 $M_J$, with
737: semi-major axes $a =$ 0.02, 0.03, 0.04, 0.05, and 0.06 AU,
738: orbiting a G2V star. Equilibrium is defined as the state after
739: which the planet radius is constant to within a part in $10^{5}$.
740: We define the equilibration timescale as the time it takes the
741: planet to evolve from $1.25$ $R_{eq}$ to 1.05 $R_{eq}$. Note that
742: for those rare models for which the planetary radii still change by more than
743: $10^{-5}$ $R_{eq}$ at the end of calculation (10 Gyr), the
744: equilibrium state is assumed to have been reached at the final age
745: of the evolutionary trajectory.
746:
747: In Fig. \ref{fig:equiradiage}, filled circles represent the values
748: of $R_{eq}$ and $\tau_{eq}$ calculated from our theoretical
749: trajectories, whereas the curves are least-square fits to them. We
750: adopt a fourth-order polynomial in fitting $R_{eq}/R_J$ as a
751: function of log$(\dot{E}_{heating}/\dot{E}_{insolation})$, and a
752: linear model for log$(\tau_{eq}/{\rm Gyr})$ versus
753: log$(\dot{E}_{heating}/\dot{E}_{insolation})$, given by:
754: %
755: \begin{equation}\label{eq:fitting}
756: \begin{array}{rcl}
757: \displaystyle \frac{R_{eq}}{R_J} & = & \displaystyle C_0 + C_1 x + C_2 x^2 + C_3 x^3 + C_4 x^4, \\[3mm]
758: \displaystyle \log \bigg(\frac{\tau_{eq}}{{\rm Gyr}}\bigg) & = & \displaystyle b + k x,
759: \end{array}
760: \end{equation}
761: %
762: where $x \equiv \log(\dot{E}_{heating}/\dot{E}_{insolation})$. The
763: model fits of the parameters are given in Table \ref{tab:fitting}.
764:
765: For an extreme close-in EGP with $M_p = 0.5$ $M_J$ at $a = 0.02$
766: AU orbiting a G2V star, the equilibrium planetary radii range from
767: $\sim$1.3 $R_J$ for little heating
768: ($\dot{E}_{heating}/\dot{E}_{insolation} \lesssim 10^{-6}$) to
769: $\sim$2.5 $R_J$ for strong heating
770: ($\dot{E}_{heating}/\dot{E}_{insolation} \sim 10^{-3}$). The
771: corresponding timescales for the planet to reach these radii are
772: $\sim$3 Gyr for the former and $\sim$3 Myr for the latter. At $a =
773: 0.06$ AU, the equilibrium radii are smaller and the relevant
774: timescales are longer $-$ from $R_{eq} \sim 1.25$ $R_J$ and
775: $\tau_{eq}\sim 2$ Gyr for $\dot{E}_{heating}/\dot{E}_{insolation}
776: \lesssim 10^{-5}$ to $R_{eq} \sim 1.8$ $R_J$ and $\tau_{eq}\sim 3$
777: Myr for $\dot{E}_{heating}/\dot{E}_{insolation} \sim 10^{-2}$. For
778: more massive planets, the equilibrium radii are smaller and the
779: timescales are longer. For an EGP with $M_p = 1.5$ $M_J$ at $a =
780: 0.02$ AU, the values are $R_{eq}\sim 1.8$ $R_J$ and $\tau_{eq}\sim
781: 5$ Myr for $\dot{E}_{heating}/\dot{E}_{insolation} \sim 10^{-2}$.
782: Our theoretical model grids along with the fitting curves provided
783: in eq. (\ref{eq:fitting}) and the parameters listed in Table
784: \ref{tab:fitting} can be used to calculate the equilibrium
785: planetary radius and the typical timescale to reach it, given
786: different combinations of planet mass, orbital distance, and the
787: ratio of core-heating power to insolation power.
788:
789:
790:
791:
792:
793: %------------------------------------------------------------------------------
794: \section{Summary and Discussion}\label{sec:sum}
795:
796: We have calculated theoretical radius-age trajectories for three
797: EGPs: TrES-4, XO-3b, and HAT-P-1b, under various assumptions
798: concerning atmospheric opacity, the presence of an inner
799: heavy-element core, and possible heating due to orbital tidal
800: dissipation. The main model results are the following:
801: \begin{enumerate}
802:
803: \item[1.] Unless the atmospheric opacity of TrES-4 is unusually
804: large (much higher than 10$\times$solar equivalent), core heating
805: is required to explain its observed radius \citep[$R_p =
806: 1.751_{-0.062}^{+0.064}$ $R_J$;][]{torres08}. However, the
807: required heating power is modest. A non-core model with
808: $(e/0.01)^2(10^5/Q_p) = 0.1$ produces radii consistent with the
809: 1-$\sigma$ lower boundary of $R_p$ from \citet{torres08}, assuming
810: 10$\times$solar atmospheric opacity. The required energy
811: dissipation rates become larger for models with a heavy-element
812: core. The model with $Q_p = 10^{3.8}$, $e \sim$0.01, and a
813: heavy-element core of $M_c = 30$ $M_{\oplus}$, reproduces the
814: observed radius well, assuming 3$\times$solar atmospheric opacity.
815: If TrES-4 follows the core-mass stellar-metallicity correlation
816: found by \citet{burrows07b}, then the models with a non-zero
817: heavy-element core mass are favored, considering its stellar
818: metallicity [Fe/H]$_{\ast}$ = $+0.14_{-0.09}^{+0.09}$ dex
819: \citep{torres08}. Ongoing Spitzer photometry of its secondary
820: eclipse will put more stringent constraints on $e$ and can either
821: confirm or rule out these possibilities.
822:
823: \item[2.] For XO-3b, we have shown that orbital tidal heating is a
824: key factor in explaining the planet radius. The very large radius
825: ($R_p = 1.95_{-0.16}^{+0.16}$ $R_J$) derived by
826: \citet{johnskrull07} based on spectral synthesis modeling can be
827: explained by energy dissipation due to tidal heating. In this
828: case, the $Q_p$ parameter of XO-3b is near $\sim$$10^{5.7}$, a not
829: unreasnoable value. On the other hand, the much smaller radius
830: ($1.217_{-0.073}^{+0.073}$ $R_J$) based on light-curve fit by
831: \citet{winn08} is consistent with no core heating sources, or with
832: tidal heating assuming $Q_p \gtrsim 10^6$, within 1-$\sigma$
833: errors. These constraints on $Q_p$ assume 3$\times$solar
834: atmospheric opacity for the former case and 1$\times$solar for the
835: latter, but are only weakly dependent on this.
836:
837:
838: \item[3.] We have reexamined the core mass required for HAT-P-1b
839: using the updated data (importantly, its radius) from
840: \citet{torres08}, and now find it generally follows the
841: correlation between core mass and stellar metallicity found by
842: \citet{burrows07b}. In all the cases considered, a non-zero
843: heavy-element core mass is needed to explain the observed radius
844: \citep[$R_p = 1.242_{-0.053}^{+0.053}$ $R_J$;][]{torres08}. The
845: core mass is $\sim$15 $M_{\oplus}$ ($\sim$20 $M_{\oplus}$)
846: assuming 3$\times$solar (10$\times$solar) atmospheric opacity when
847: there is no external heating. If there is tidal heating
848: corresponding to reasonable values of $(e/0.01)^2(10^5/Q_p)
849: \sim$$0.8$, then the core mass required to fit the observed radius
850: is $\sim$30 $M_{\oplus}$ ($\sim$40 $M_{\oplus}$) assuming
851: 3$\times$solar (10$\times$solar) atmospheric opacity, in which
852: case HAT-P-1b follows the core-mass stellar-metallicity relation
853: found by \citet{burrows07b} and \citet{guillot06}.
854:
855: \end{enumerate}
856:
857:
858: In addition, we have carried out a parameter study of the effects
859: of core heating and provided theoretical grids and fitting
860: formulae for the equilibrium planet radius and equilibration
861: timescale, given various core heating powers for planets with
862: masses $M_p =$ 0.5, 1.0, and 1.5 $M_J$ with $a =$ 0.02-0.06 AU,
863: orbiting a G2V star. The fitting formula for the equilibrium
864: planet radius can be used for a theoretical zeroth-order estimate,
865: without carrying out detailed evolutionary calculations. The
866: equilibration timescale $\tau_{eq}$ characterizes the time it
867: takes the planet to adjust its structure in response to a given
868: degree of core heating.
869:
870: Recently \citet{jackson08a} considered the effect of the
871: co-evolution of the orbital eccentricity and the semi-major axis
872: on the tidal dissipation history. In the past, the semi-major
873: axis had been assumed to be constant when conducting tidal
874: evolution studies
875: \citep[e.g.][]{bodenheimer01,bodenheimer03,gu04}.
876: \citet{jackson08b} calculate the evolutionary histories of the
877: tidal dissipation rate for several EGPs, and find that in most
878: cases the tidal heating rate increases as a planet moves inward
879: and then decreases as the orbit circularizes. The relevant
880: timescale, $\tau_{heating}$, is the time it takes the tidal
881: heating rate to decay by a factor of $e$. If $\tau_{eq} \gg
882: \tau_{heating}$, then it is valid to take a constant effective
883: tidal heating rate in the planet radius-age trajectory
884: calculation. If $\tau_{eq} \sim$$\tau_{heating}$, then in order to
885: account for the effect of a varying tidal heating rate, different
886: values should be used at each time step of the radius-age
887: trajectory calculation. If $\tau_{eq} \ll \tau_{heating}$, then
888: the planet will have enough time to reach an equilibrium radius
889: before the tidal heating rate decays significantly. In this case,
890: the planet's structure evolves in a quasi-equilibrium manner.
891: Theoretical planet radius-age trajectory models will easily be able to
892: account for the effect of varying tidal heating rate by
893: adopting different core heating rates at each time interval of
894: $\tau_{heating}$ during the calculation. In effect, there is a
895: reset of the ``clock'' right after the time of maximum heating $-$
896: the planet becomes most extended on a timescale $\sim$$\tau_{eq}$
897: after the tidal heating rate achieves this maximum. Because of the
898: intense heating and the quick response, the planet loses the
899: memory of its shrinkage history before maximum heating, which
900: is effectively a reset of its ``age.''
901:
902:
903: \citet{burrows07b} calculated theoretical radii for HD 209458b.
904: They found that the measured radius deviated at the
905: $\sim$1.5-$\sigma$ level for the age they assumed, even when
906: employing 10$\times$solar atmospheric opacity, no inner solid
907: core, and no core heating. However, the updated age measurement
908: for HD 209458b by \citet{torres08} of $3.1_{-0.7}^{+0.8}$ Gyr is
909: much smaller than the one adopted by \citet{burrows07b}
910: ($5.5_{-1.5}^{+1.5}$ Gyr), whereas the updated radius,
911: $1.359_{-0.019}^{+0.016}$, is similar to that used by
912: \citet{burrows07b} ($1.32_{-0.03}^{+0.03}$). As a result, the
913: \citet{torres08} radius and age measurement for HD 209458b can be
914: fit by the 10$\times$solar-opacity model of \citet{burrows07b}
915: within $\sim$1-$\sigma$ uncertainties, without the need of any
916: core heating sources. Moreover, if the tidal heating rate of HD
917: 209458b decayed from $\sim$4$\times10^{26}$ erg s$^{-1}$ to
918: $\sim$4$\times10^{24}$ erg s$^{-1}$ during the past 2 Gyr
919: \citep{jackson08b}, then its effective age would be 2 Gyr younger,
920: due to the ``clock reset'' effect. Based on our parameter study
921: in \S\ref{sec:heating}, the equilibration timescale of HD 209458b
922: during maximum heating would have been $\sim$0.05 Gyr. This is
923: small enough compared with the decay timescale of the tidal
924: heating rate ($\sim$1 Gyr from $\sim$4$\times10^{26}$ erg s$^{-1}$
925: to $\sim$1.5$\times10^{26}$ erg s$^{-1}$) for our models to
926: reproduce HD 209458b's current radius. In this case, the
927: \citet{torres08} radius measurement for HD 209458b, along with an
928: ``effective age'' of $\sim$1.1 Gyr, can explain HD 209458b's
929: radius within $\sim$1-$\sigma$ uncertainties, even with the
930: 1$\times$solar-opacity model of \citet{burrows07b} and a core mass
931: of 10-20 $M_{\oplus}$, but without the need for any {current} core
932: heating. The latter comports with the very small limit of
933: $\sim$0.001 on its current orbital eccentricity. A possible caveat to the
934: tidal evolution scenarios described in \citet{jackson08a,jackson08b} is
935: that they can be dramatically changed due to even a small undetected
936: perturbing body \citep[e.g.,][]{mardling07}. Sensitive searches
937: for such companion bodies are needed to further constrain this
938: possibility. Nevertheless, future studies should combine orbital semi-major
939: axis, eccentricity, and planet radius evolution models in a more
940: coupled fashion. It is not only more consistent to consider the
941: co-evolution of $a$, $e$, and $R_p$, but also important to factor
942: in the feedback of the associated tidal heating power on $R_p$ and
943: its radius-age trajectory \citep{jackson08a,jackson08b}. Such a
944: project is in progress.
945:
946:
947:
948: %------------------------------------------------------------------------------
949: \acknowledgments
950:
951: We thank G\'{a}sp\'{a}r Bakos for comments on the eccentricity of
952: HAT-P-1b, and Georgi Mandushev for insights concerning the
953: possible eccentricity range of TrES-4. We also thank Josh Winn for
954: a careful reading of an earlier version of the manuscript, and an
955: anonymous referee for a careful and useful report that improves
956: the paper. This study was supported in part by NASA grants
957: NNG04GL22G, NNX07AG80G, and NNG05GG05G and through the NASA
958: Astrobiology Institute under Cooperative Agreement No.
959: CAN-02-OSS-02 issued through the Office of Space Science.
960:
961:
962:
963: %----------------------------------------------------------------------------
964:
965: %-------------------------------------------------------------
966:
967: \begin{appendix}
968:
969: \section{External Heating due to Tidal Dissipation}
970:
971: The total tidal energy dissipation rate within the planet in its
972: rest frame assuming equilibrium tides with constant lag angle and
973: synchronous rotation is
974: \citep[e.g.][]{goldreich66,bodenheimer01,bodenheimer03,gu04}:
975: %
976: \begin{equation}
977: \begin{array}{rcl}
978: \displaystyle \dot{E}_{{\rm tide}} & = & \frac{GM_{\ast}\mu f(e)}{a\tau_{{\rm
979: circ}}} \\[3mm]
980: & \approx & \displaystyle 1.1 \times 10^{24}\, {\rm erg}\,{\rm s}^{-1}
981: \bigg(\frac{e}{0.01}\bigg)^2\bigg[\frac{f(e)}{e^2}\bigg]\,
982: \bigg(\frac{M_{\ast}}{M_{\odot}}\bigg)\bigg(\frac{M_p}{M_J}\bigg)\bigg(\frac{a}{0.05
983: \, {\rm AU}}\bigg)^{-1}\bigg(\frac{\tau_{{\rm circ}}}{{\rm
984: Gyr}}\bigg)^{-1} \,
985: %&\approx 1.1 \times 10^{25}\,
986: %\bigg(\frac{e}{0.01}\bigg)^2\bigg[\frac{f(e)}{e^2}\bigg]\,\bigg(\frac{Q_p}{10^5}\bigg)^{-1}
987: %\bigg(\frac{M_{\ast}}{M_{\odot}}\bigg)^{5/2}\bigg(\frac{R_p}{R_J}\bigg)^{5}\bigg(\frac{a}{0.05
988: %\, {\rm AU}}\bigg)^{-15/2} \, ,
989: \end{array}
990: \end{equation}
991: %
992: where $\mu\equiv M_{\ast}M_p/(M_{\ast} + M_p)$ is the reduced
993: mass, $f(e) \equiv \frac{2}{7}[h_3(e) - 2h_4(e) + h_5(e)]$ is a
994: function of orbital eccentricity with
995: $h_3(e)=(1+3e^2+3e^4/8)(1-e^2)^{-9/2}$,
996: $h_4(e)=(1+15e^2/2+45e^4/8+5e^6/16)(1-e^2)^{-6}$, and
997: $h_5(e)=(1+31e^2/2+255e^4/8+185e^6/16+25e^8/64)(1-e^2)^{-15/2}$
998: \citep{gu04}. Note that $f(e) \rightarrow e^2$ as $e \rightarrow 0$.
999: $\tau_{{\rm circ}}$ denotes the circularization timescale, which
1000: is
1001: %
1002: \begin{equation}
1003: %\begin{split}
1004: \tau_{{\rm circ}} \approx 0.10 \,{\rm Gyr} \times
1005: \bigg(\frac{Q_p}{10^5}\bigg)\bigg(\frac{M_{\ast}}{M_{\odot}}\bigg)^{-3/2}\bigg(\frac{M_p}{M_J}\bigg)\bigg(\frac{R_p}{R_J}\bigg)^{-5}\bigg(\frac{a}{0.05
1006: \, {\rm AU}}\bigg)^{13/2}\, .
1007: %\end{split}
1008: \end{equation}
1009: %
1010:
1011: A more informative quantity is the ratio of the tidal energy
1012: dissipation rate and the insolation rate, given by
1013: %
1014: {\footnotesize
1015: \begin{equation}
1016: \begin{array}{rcl}
1017: \displaystyle \frac{\dot{E}_{{\rm tide}}}{\dot{E}_{{\rm insolation}}} & = &
1018: \displaystyle \frac{GM_{\ast}\mu f(e)}{\pi F_p R_p^2 \,a\tau_{{\rm circ}}} \\[3mm]
1019: & \approx & \displaystyle 6.9 \times 10^{-5}
1020: \bigg(\frac{e}{0.01}\bigg)^2\bigg[\frac{f(e)}{e^2}\bigg]\,\bigg(\frac{Q_p}{10^5}\bigg)^{-1}
1021: \bigg(\frac{M_{\ast}}{M_{\odot}}\bigg)^{5/2}\bigg(\frac{R_p}{R_J}\bigg)^{3}\bigg(\frac{a}{0.05
1022: \, {\rm AU}}\bigg)^{-15/2}\bigg(\frac{F_p}{10^9 \,{\rm erg}\,{\rm
1023: cm}^{-2}\,{\rm s}^{-1}}\bigg)^{-1}.\\
1024: \end{array}
1025: \end{equation}
1026: }
1027: %
1028:
1029: \end{appendix}
1030:
1031:
1032:
1033:
1034: \bibliographystyle{apj}
1035: \bibliography{apj-jour,epa}
1036:
1037:
1038:
1039:
1040:
1041: \clearpage
1042: %----------------------------------------------------------------------------------------------
1043: \begin{landscape}
1044: \begin{deluxetable}{cccccccccccccc}
1045: \tabletypesize{\scriptsize} \tablewidth{0pc} \tablecaption{
1046: Observational Properties of the Transiting Planet Systems.
1047: \label{tab:planet} } \tablehead{ \colhead{} & \colhead{$a$} &
1048: \colhead{Period} & \colhead{} & \colhead{$M_*$} & \colhead{$R_*$}
1049: & \colhead{$T_{eff}$} & \colhead{[Fe/H]$_*$} & \colhead{Age} &
1050: \colhead{$M_p$} & \colhead{$R_p$} &
1051: \colhead{$F_p$\tablenotemark{a}} & \colhead{$\dot{E}_{{\rm
1052: tide}}/\dot{E}_{{\rm insolation}}$\tablenotemark{b}} &
1053: \colhead{} \\
1054: \colhead{~~~~~System~~~~~} & \colhead{(AU)} & \colhead{(days)} &
1055: \colhead{$e$} & \colhead{($M_{\odot}$)} & \colhead{($R_{\odot}$)}
1056: & \colhead{(K)} & \colhead{(dex)} & \colhead{(Gyr)} &
1057: \colhead{($M_J$)} & \colhead{($R_J$)} & \colhead{($10^9$ erg
1058: cm$^{-2}$ s$^{-1}$)} & \colhead{[$(Q_p/10^5)^{-1}$]} &
1059: \colhead{References} } \startdata
1060: XO-3\dotfill & $0.0476_{-0.0005}^{+0.0005}$ & 3.19154 & $0.260_{-0.017}^{+0.017}$ & $1.41_{-0.08}^{+0.08}$ & $2.13_{-0.21}^{+0.21}$ & $6429_{-50}^{+50}$ & $-0.177_{-0.027}^{+0.027}$ & $2.69_{-0.16}^{+0.14}$ & $13.25_{-0.64}^{+0.64}$ & $1.95_{-0.16}^{+0.16}$ & $4.20_{-0.84}^{+0.84}$ & $5.9\times 10^{-1}$ & \tablenotemark{c} \\ [+1.2ex]
1061: \dotfill & $0.0454_{-0.00082}^{+0.00082}$ & 3.19152 & $0.260_{-0.017}^{+0.017}$ & $1.213_{-0.066}^{+0.066}$ & $1.377_{-0.083}^{+0.083}$ & $6429_{-100}^{+100}$ & $-0.177_{-0.080}^{+0.080}$ & $2.82_{-0.82}^{+0.58}$ & $11.79_{-0.59}^{+0.59}$ & $1.217_{-0.073}^{+0.073}$ & $1.93_{-0.27}^{+0.27}$ & $3.1\times 10^{-1}$ & \tablenotemark{d} \\ [+1.2ex]
1062: TrES-4\dotfill & $0.05092_{-0.00069}^{+0.00072}$ & 3.553945 & $\sim 0.0$ & $1.394_{-0.056}^{+0.060}$ & $1.816_{-0.062}^{+0.065}$ & $6200_{-75}^{+75}$ & $+0.14_{-0.09}^{+0.09}$ & $2.9_{-0.4}^{+0.4}$ & $0.920_{-0.072}^{+0.073}$ & $1.751_{-0.062}^{+0.064}$ & $2.31_{-0.20}^{+0.21}$ & $2.3\times 10^{-4}$ \tablenotemark{e} & \tablenotemark{f} \\ [+1.2ex]
1063: HAT-P-1\dotfill & $0.0553_{-0.0013}^{+0.0012}$ & 4.46543 & $\sim 0.0$ & $1.133_{-0.079}^{+0.075}$ & $1.135_{-0.048}^{+0.048}$ & $5975_{-120}^{+120}$ & $+0.13_{-0.08}^{+0.08}$ & $2.7_{-2.0}^{+2.5}$ & $0.532_{-0.030}^{+0.030}$ & $1.242_{-0.053}^{+0.053}$ & $0.66_{-0.08}^{+0.08}$ & $9.2\times 10^{-5}$ \tablenotemark{e} & \tablenotemark{f} \\ [+1.2ex]
1064: \enddata
1065: \tablenotetext{a}{The stellar flux at the planet's substellar
1066: point.} \tablenotetext{b}{$\dot{E}_{{\rm tide}}$ is the total
1067: tidal energy dissipation rate within the planet in its rest frame
1068: (Gu et al. 2004). $\dot{E}_{{\rm insolation}} \equiv \pi
1069: R_p^2F_p$.} \tablenotetext{c}{\citet{johnskrull07}; Inferred from
1070: spectroscopically derived stellar parameters.}
1071: \tablenotetext{d}{\citet{winn08}; Determined from light-curve
1072: fits.} \tablenotetext{e}{Assuming $e = 0.01$.}
1073: \tablenotetext{f}{\citet{torres08}; Note that they assume $e = 0$
1074: exactly in deriving the parameters, since the radial-velocity data
1075: are consistent with a circular orbit.}
1076: \end{deluxetable}
1077: \clearpage
1078: \end{landscape}
1079: %-------------
1080:
1081: %----------------------------------------------------------------------------------------------
1082:
1083: \begin{deluxetable}{cccccccc}
1084: \tabletypesize{\scriptsize} \tablecaption{Best-estimate Core
1085: Masses for HAT-P-1b under Various Assumptions. \label{tab:hatp1} }
1086: \tablewidth{0pt} \tablehead{ \colhead{Atmospheric Opacity} &
1087: \multicolumn{3}{c}{$3 \times$solar} & \colhead{} &
1088: \multicolumn{3}{c}{$10 \times$solar} \\
1089: \cline{2-4}
1090: \cline{6-8} \\
1091: \colhead{$\bigg(\frac{e}{0.01}\bigg)^2\bigg(\frac{10^5}{Q_p}\bigg)$}
1092: & \colhead{0} & \colhead{0.8} & \colhead{8} & \colhead{} &
1093: \colhead{0} & \colhead{0.8} & \colhead{8} } \startdata
1094: $M_c/M_{\oplus}$\dotfill & 15 & 30 & 60 & & 20 & 40 & 80 \\
1095: \enddata
1096: %%%%%%%
1097: \tablecomments{The tidal dissipation rate is proportional to
1098: $e^2/Q_p$ when $e \ll 1$.}
1099: \end{deluxetable}
1100:
1101:
1102: %----------------------------------------------------------------------------------------------
1103: \begin{deluxetable}{ccccccccc}
1104: \tabletypesize{\scriptsize} \tablewidth{0pc} \tablecaption{
1105: Fitting Parameters for the Equilibrium Planetary Radius and
1106: Equilibration Timescale for EGPs Undergoing Core Heating.
1107: \label{tab:fitting} } \tablehead{ \colhead{$M_p$} & \colhead{$a$}
1108: & \colhead{} & \colhead{} & \colhead{} & \colhead{} & \colhead{} &
1109: \colhead{} &
1110: \colhead{} \\
1111: \colhead{($M_J$)} & \colhead{(AU)} &
1112: \colhead{$C_0$\tablenotemark{a}} &
1113: \colhead{$C_1$\tablenotemark{a}} &
1114: \colhead{$C_2$\tablenotemark{a}} &
1115: \colhead{$C_3$\tablenotemark{a}} &
1116: \colhead{$C_4$\tablenotemark{a}} & \colhead{$b$\tablenotemark{b}}
1117: & \colhead{$k$\tablenotemark{b}} } \startdata
1118: $0.5$ & $0.02$ & $15.4$ & $9.69$ & $2.63$ & $0.328$ & $0.0156$ & $-4.78$ & $-0.845$ \\ [+1.2ex]
1119: \dotfill & $0.03$ & $7.74$ & $4.43$ & $1.22$ & $0.157$ & $0.00785$ & $-4.26$ & $-0.794$ \\ [+1.2ex]
1120: \dotfill & $0.04$ & $6.42$ & $4.00$ & $1.25$ & $0.182$ & $0.0101$ & $-3.95$ & $-0.762$ \\ [+1.2ex]
1121: \dotfill & $0.05$ & $4.61$ & $2.66$ & $0.866$ & $0.135$ & $0.00809$ & $-3.77$ & $-0.746$ \\ [+1.2ex]
1122: \dotfill & $0.06$ & $3.60$ & $1.79$ & $0.573$ & $0.0905$ & $0.00560$ & $-3.71$ & $-0.758$ \\ [+1.2ex]
1123: \tableline
1124: $1.0$ & $0.02$ & $6.42$ & $4.51$ & $1.58$ & $0.255$ & $0.0155$ & $-4.10$ & $-0.783$ \\ [+1.2ex]
1125: \dotfill & $0.03$ & $4.44$ & $3.02$ & $1.16$ & $0.208$ & $0.0139$ & $-3.70$ & $-0.742$ \\ [+1.2ex]
1126: \dotfill & $0.04$ & $3.42$ & $2.19$ & $0.894$ & $0.170$ & $0.0122$ & $-3.42$ & $-0.701$ \\ [+1.2ex]
1127: \dotfill & $0.05$ & $2.72$ & $1.54$ & $0.669$ & $0.137$ & $0.0106$ & $-3.27$ & $-0.686$ \\ [+1.2ex]
1128: \dotfill & $0.06$ & $2.33$ & $1.15$ & $0.512$ & $0.110$ & $0.00897$ & $-3.17$ & $-0.677$ \\ [+1.2ex]
1129: \tableline
1130: $1.5$ & $0.02$ & $4.20$ & $2.54$ & $0.888$ & $0.145$ & $0.00899$ & $-3.59$ & $-0.702$ \\ [+1.2ex]
1131: \dotfill & $0.03$ & $3.15$ & $1.76$ & $0.663$ & $0.118$ & $0.00796$ & $-3.21$ & $-0.654$ \\ [+1.2ex]
1132: \dotfill & $0.04$ & $2.57$ & $1.34$ & $0.543$ & $0.105$ & $0.00762$ & $-3.01$ & $-0.630$ \\ [+1.2ex]
1133: \dotfill & $0.05$ & $2.13$ & $0.924$ & $0.391$ & $0.0808$ & $0.00634$ & $-2.90$ & $-0.622$ \\ [+1.2ex]
1134: \dotfill & $0.06$ & $1.91$ & $0.693$ & $0.292$ & $0.0610$ & $0.00486$ & $-2.82$ & $-0.612$ \\ [+1.2ex]
1135: \enddata
1136: \tablenotetext{a}{$R_{eq}/R_J = C_0 + C_1 x + C_2 x^2 + C_3 x^3 +
1137: C_4 x^4$, where $x \equiv
1138: \log(\dot{E}_{heating}/\dot{E}_{insolation})$.}
1139: \tablenotetext{b}{$\log(\tau_{eq}/{\rm Gyr}) = b + k x$.}
1140: \end{deluxetable}
1141: %-------------
1142:
1143:
1144:
1145:
1146:
1147:
1148:
1149:
1150:
1151:
1152:
1153:
1154:
1155:
1156:
1157:
1158: %-------------
1159: %-------------
1160:
1161:
1162: \end{document}
1163: