• Embed Doc
  • Readcast
  • Collections
  • CommentGo Back
Download
 
A&A 385, 156–165 (2002)DOI: 10.1051/0004-6361:20011624
c
ESO 2002
Astronomy
&
AstrophysicsEvolution of “51Pegasusb-like” planets
T. Guillot
1
and A. P. Showman
2
1
Observatoire de la Cˆote d’Azur, Laboratoire Cassini, CNRS UMR 6529, 06304 Nice Cedex 4, France
2
University of Arizona, Department of Planetary Sciences and Lunar and Planetary Lab, Tucson, AZ 85721 USAe-mail:
showman@lpl.arizona.edu
Rceived 23 March 2001 / Accepted 7 November 2001
Abstract.
About one-quarter of the extrasolar giant planets discovered so far have orbital distances smaller than0.1AU. These “51Pegb-likeplanets can now be directly characterized, as shown by the planet transiting infront the star HD 209458. We review the processes that affect their evolution. We apply our work to the case of HD 209458b, whose radius has been recently measured. We argue that its radius can be reproduced only when thedeep atmosphere is assumed to be unrealistically hot. When using more realistic atmospheric temperatures, anenergy source appears to be missing in order to explain HD 209458b’s large size. The most likely source of energyavailable is not in the planet’s spin or orbit, but in the intense radiation received from the parent star. We showthat the radius of HD 209458b can be reproduced if a small fraction (
1%) of the stellar flux is transformed intokinetic energy in the planetary atmosphere and subsequently converted to thermal energy by dynamical processesat pressures of tens of bars.
Key words.
stars: individual: 51 Peg, HD 209458 – planets and satellites: general – stars: planetary systems
1. Introduction
The detection of planetary-mass companions in small or-bits around solar-type stars has been a major discovery of the past decade. To date, 73 extrasolar giant planets (withmasses
sin
i <
13
J
,
J
being the mass of Jupiterand
i
the inclination of the system) have been detected byradial velocimetry. Fifteen of these (21%) have distancesless than 0.1 AU, and ten (14%) have distances less than0.06 AU (see Marcy et al. 2000 and the discoverers’ webpages). This is for example the case with the first extraso-lar giant planet to have been discovered, 51 Peg b (Mayor& Queloz 1995). These close-in planets form a statisti-cally distinct population: all planets with semi-major axissmaller than 0.06 AU have near-circular orbits while themean eccentricity of the global population is
< e >
0
.
27.This is explained by the circularization by tides raised onthe star by the planet (Marcy et al. 1997). One exceptionto this rule, HD 83443b (
e
= 0
.
079
±
0
.
033), can be at-tributed to the presence of another eccentric planet in thesystem (Mayor et al. 2001). As we shall see, the planetsinside
0.1 AU also have very specific properties due tothe closeness to their star and the intense radiation theyreceive. For this reason, following astronomical conven-tions, we choose to name them after the first object of 
Send offprint requests to
: T. Guillot,e-mail:
guillot@obs-nice.fr
this class to have been discovered: “51Pegb-like” planets,or in short “Pegasi planets”.Such planets provide an unprecendented opportunityto study how intense stellar irradiation affects the evo-lution and atmospheric circulation of a giant planet.Roughly 1% of stars surveyed so far bear Pegasi planets inorbit, suggesting that they are not a rare phenomenon.Their proximity to their stars increases the likelihood thatthey will transit their stars as viewed from Earth, al-lowing a precise determination of their radii. (The prob-ability varies inversely with the planet’s orbital radius,reaching
10% for a planet at 0.05 AU around a solar-type star.) One planet, HD 209458b, has already been ob-served to transit its star every 3.524 days (Charbonneauet al. 2000; Henry et al. 2000). The object’s mass is0
.
69
±
0
.
05
J
, where
J
=1
.
89
×
10
27
kg is the massof Jupiter. Hubble Space Telescope measurements of thetransit (Brown et al. 2001) imply that the planet’s radiusis 96300
±
4000km. An analysis of the lightcurve com-bined with atmospheric models shows that this shouldcorrespond to a radius of 94430km at the 1bar level(Hubbard et al. 2001). This last estimate corresponds to1.349
R
J
, where
R
J
70000kmis a characteristicradius of Jupiter. This large radius, in fair agreement with theoreti-cal predictions (Guillot et al. 1996), shows unambiguouslythat HD 209458b is a gas giant.We expect that the evolution of Pegasiplanets depends more on the stellar irradiation than
 
T. Guillot and A. P. Showman: Evolution of Pegasi planets 157
is the case with Jupiter. HD 209458b and otherPegasi planets differ qualitatively from Jupiter be-cause the globally-averaged stellar flux they absorb is
10
8
ergcm
2
(10
5
Wm
2
), which is
10
4
times greaterthan the predicted intrinsic flux of 
10
4
ergcm
2
. (Incontrast, Jupiter’s absorbed and intrinsic fluxes are thesame within a factor of two.) Several evolution calcu-lations of Pegasi planets have been published (Burrowset al. 2000a; Bodenheimer et al. 2001), but these papersdisagree about whether HD 209458b’s radius can beexplained, and so far there has been no general discussionof how the irradiation affects the evolution. Our aim is tohelp fill this gap.Here, we quantify how atmospheric processes affectthe evolution of Pegasi planets such as HD 209458b. First(Sect. 2), we show that the evolution is sensitive to theassumed atmospheric temperatures. This sensitivity hasnot previously been documented, and quantifying it isimportant because the temperature profiles appropriatefor specific planets remain uncertain (e.g., no atmosphericradiative transfer calculation for HD 209458b yet exists).Our works suggests that the discrepancy between the pre-dictions of Burrows et al. (2000a) and Bodenheimer et al.(2001) can be largely explained by their different assump-tions about atmospheric temperature.Second, the effect of atmospheric dynamics on the evo-lution has to date been neglected. For example, currentmodels assume the day-night temperature difference iszero, despite the fact that substantial day-night temper-ature variations are likely. In Sect. 3.1 we demonstratehow the evolution is modified when a day-night tempera-ture difference is included. Furthermore, the intense stellarirradiation will lead to production of atmospheric kineticenergy, and transport of this energy into the interior couldprovide a substantial energy flux that would counteractthe loss of energy that causes planetary contraction. InSect. 3.2 we investigate this effect.The research has major implications for HD 209458b.Early calculations implied that Pegasi planets contractslowly enough to explain HD 209458b’s large radius(Guillot et al. 1996; Burrows et al. 2000a). But recent cal-culations of irradiated atmospheressuggest that the actualdeep atmosphere is colder than assumed (Goukenleuqueet al. 2000). When such realistic temperatures are adopted(our Sect. 2), the planet contracts too fast and the radiusis
0
.
2–0.3
R
J
too small. Bodenheimer et al. (2001) ar-gued that tidal heating from circularization of the orbitwould slow the contraction, leading to a larger radius, butthis is a transient process that would end
10
8
years afterthe planet’s formation. Instead we argue that kinetic en-ergy produced in the atmosphere is transported into theinterior and dissipated (Sect. 3.2). We show that plau-sible downward energy fluxes can slow or even halt theplanet’s contraction, allowing HD 209458b’s radius to beexplained.In a joint paper (Showman & Guillot 2002, Paper II)we consider the atmospheric dynamics of these planets,with emphasis on how the atmospheres respond to stellarheating and gravitational tidal interactions, and on theobservable consequences.
2. Sensitivity of evolution to atmospherictemperature
The upper boundary condition of evolution models con-sists of a relationship between the effective temperatureand some deeper temperature (say that at 10 bars) towhich the model’s interior temperature profile is attached.Here we show that the evolution is sensitive to the as-sumed relationship (i.e., to the assumed atmospheric tem-perature structure).Before we begin, we provide some definitions. We de-fine the effective temperature of the irradiated planet as4
πR
2
σT 
4eff 
=
L
+
L
int
,
(1)where
R
is the planet’s radius,
eff 
is its effective temper-ature,
σ
is the Stefan-Boltzmann constant,
L
is the partof the stellar luminosity absorbed by the planet and
L
int
isthe intrinsic luminosity of the planet due to its cooling andcontraction (and possibly other processes such as radioac-tivity or thermonuclear reactions in the case of massiveobjects).The temperature corresponding to the intrinsic plan-etary flux, called the “intrinsic” temperature
int
, is de-fined by4
πR
2
σT 
4int
=
L
int
.
(2)Similarly, we define
from the absorbed stellar lumi-nosity
L
.
is the effective temperature towards whichthe planet tends as it cools and
L
int
diminishes (see e.g.Hubbard 1977). It is a function of the Bond albedo (i.e.the ratio of the luminosity directly reflected to the totalluminosity intercepted by the planet).
can be viewedas the effective temperature reached by the planet after ithas lost its internal heat, and is hence sometimes noted
eq
(e.g. Guillot et al. 1996; Saumon et al. 1996).
2.1. Atmospheric boundary conditions 
We consider two evolution models of HD 209458b basedon the parameters listed in Table 1; the two models differonly in their prescription for the atmospheric boundarycondition.Our first evolution sequence, dubbed the “hot” case,uses the standard boundary condition from Guillot et al.(1996) and Burrows et al. (2001a). These papers adoptedan atmospheric structure of an
isolated 
object with theexpected effective temperature, which provides a fair fitto the evolution of Jupiter. The surface boundary condi-tion consists of a relationship between the temperature atthe 10bar pressure level
isolated
, the effective tempera-ture
eff 
and the gravity
g
of an isolated planet/browndwarf derived by several authors (see Marley et al. 1996;Burrows et al. 1997):
(
= 10 bar) =
isolated
(
eff 
,g
)
.
(3)
 
158 T. Guillot and A. P. Showman: Evolution of Pegasi planets
Table 1.
Parameters used in evolution models.Parameter Value References/remarksEvolution model CEPAM Guillot & Morel (1995)Mass
= 0
.
69
Jup
(M
Jup
1
.
89
×
10
30
g)Absorbed stellar heat
= 1400KRadius
R
= 1
.
35
R
J
(
R
J
7
×
10
9
cm)EOS interpolatedSaumon et al. (1995)Helium mass mixing ratio
= 0
.
30Higher than solar in order to mimic a solar abundance of heavy elementsRosseland opacities Alexander & Ferguson (1994) incl. interstellar grainsRotation 0 NeglectedCore mass 0 Not consideredAtmospheric boundary From Marley et al. (1996); Burrows et al. (1997)See Eqs. (3) and (4)
Fig.1.
Surface boundary condition (temperature at the 10 barlevel) that has been used in several published evolution models,and which we dub the “hot” case, as a function of effectivetemperature for three different gravities: log(
g
) = 2
.
5, 3 and3.5. (
g
is in units of cms
2
). (See Saumon et al. 1996 for adiscussion.).
This approximation is exact in the limit when the stellarflux is
entirely 
absorbed at the 10bar level, or in a deepadiabatic (convective) region, as is the case for Jupiter(see Hubbard 1977 for a detailed discussion of the effectof insolation on Jupiter’s evolution). Figure 1 shows thevariation of 
isolated
with effective temperature and grav-ity, for values of interest in the case of Pegasi planets.Unfortunately, the approximation becomes incorrectin the case of strongly irradiated planets because of thegrowth of a thick external radiative zone. Another bound-ary condition has therefore to be sought: either partof the stellar flux is able to penetrate to deeper levels(
0
>
10bar) and lead to a boundary condition definedby
(
0
)
>
isolated
, or most of the stellar flux is absorbedat
0
<
10bar, yielding
(
0
)
<
isolated
. (This is due tothe fact that in the radiative zone d
T/
d
, where
is the flux to be transported.) It will be shown hereafter(see Sect. 3.3) that Eq. (3) is effectively an upper limitto the boundary temperature because continuum opac-ity sources only effectively limit the penetration of thestellar photons. Indeed, more detailed models of the at-mospheres of Pegasi planets have shown that most of thestarlight is absorbed at pressures less than 10 bar, andthat Eq. (3) overestimates the atmospheric temperaturesby as much as 300 to 1000K (Seager & Sasselov 1998,2000; Goukenleuque et al. 2000; Barman et al. 2001).Because these atmospheric models do not presentlyspan the effective temperature and gravity range that isneeded, and more importantly because they assume un-realistic intrinsic temperatures, we chose to construct anarbitrary boundary condition based on the results of theisolated case. For a given
eff 
, the isolated case providesan upper bound to the “surface” temperature and by ex-tension to the temperatures in the planetary interior. Inorder to have an approximate lower bound that agreeswith atmospheric models of irradiated giant planets, weassume (i) a lower value of 
0
= 3bar, and (ii) that thetemperature at that level is given by:
(
= 3 bar) =
isolated
(
eff 
,g
)
1000K
.
(4)Evolution calculations made with this boundary conditionare dubbed the “cold” case.Note that we found a posteriori that the choice of 
0
is almost unconsequential for the evolution calculations.This is because the external radiative region quickly be-comes almost isothermal (see Fig. 4 hereafter). However,the consequences of the cooler temperatures are profound,and as we shall see lead to a much faster evolution.In this context, Bodenheimer et al. (2000, 2001) as-sume that the temperature at optical depth 2
/
3 (cor-responding in their model to a pressure of the or-der
1mbar) is equal to the effective temperature
eff 
,an approximation that leads to an underestimation of the actual atmospheric temperatures. As a consequence,their 1 bar temperatures are of the order of 
1400K, i.e.even lower than what is implied by Eq. (4). This wouldimply an extremely inefficient penetration of the stellarflux in the planetary atmosphere, in disagreement withdetailed models of these atmospheres. We therefore preferto use Eq. (4) as our “cold” boundary condition.
of 00

Leave a Comment

You must be to leave a comment.
Submit
Characters: ...
You must be to leave a comment.
Submit
Characters: ...