ICARUS30, 305--310 (1977)

The Jovian Surface Condition and Cooling Rate

W. B. HUBBARD
Department of Planetary Sciences, Lunar and Planetary Laboratory,
University of Arizona, Tucson, Arizona 85721

Received April 6, 1976; revised June 2, 1976

A theory which is almost fully analytic is used to investigate Jupiter's cooling rate. We find
that a simple model of contraction with adiabatic interior structure gives a total cooling time
to the present which is in good agreement with the age of the solar system. The interplay
between the surface condition and the cooling rate is exhibited and discussed. The current rate
of change of the effective temperature is calculated to be -1°K/0.145 X 10a yr. Discrepancies
with fully numerical investigations of the Jovian age and cooling rate are noted.

I. INTRODUCTION about 10%. The present paper reports the

H u b b a r d (1968) showed t h a t the in- result of independent calculations of Jovian
trinsic thermal radiation from Jupiter cooling using the thermodynamics of M C I V
could be attributed, in order of magnitude, together with an analytic representation
to Kelvin contraction-cooling of a de- of the same surface condition used by
generate hydrogen fluid body. Since t h a t GPGO. Because the calculations are almost
paper, a n u m b e r of investigations have entirely analytic rather than numerical,
a t t e m p t e d to s t u d y the thermal history of any interested worker can reproduce the
Jupiter quantitatively. The most elaborate results and observe the impact of varying
calculations available to date are those of different parameters on the cooling time
Graboske et al. (1975; henceforth, GPGO). scale. These calculations are published here
G P G O found t h a t a solar-composition because the shortfall in the cooling time
object of Jovian mass would evolve from of Jupiter found by G P G O is not con-
an initially gaseous and luminous state to firmed. The discrepancies between the time
a state resembling the present Jupiter in scales of the present investigation and
a time period of 2.6 X 109 yr. The dis- those of G P G O are produced almost en-
crepancy between this cooling time and tirely during the late, degenerate phase of
the expected age of Jupiter, roughly cooling; we have not carried out inde-
4.5 X 10 ~ yr, was a t t r i b u t e d to possible pendent calculations for the early, non-
inaccuracies in constituent physics, in- degenerate phase. Our results indicate t h a t
cluding t h e r m o d y n a m i c equations and
the simplest possible model of Jovian
departures from an adiabatic t e m p e r a t u r e
cooling, namely, fully adiabatic structure
distribution. The thermodynamics of liquid
metallic hydrogen has been intensively determined b y the atmospheric surface
studied recently (DeWitt and Hubbard, condition with no energy sources such as
1976; henceforth, MCIV), and the thermal unmixing due to immiscibility, is ~d)le to
properties are now known well enough to give a cooling time compatible with the
preclude time-scale uncertainties exceeding expected age of Jupiter.

305
Copyright O 1977 by Academic Press, Inc.
All rights of reproduction in any form reserved. I S S N 0019-1035
306 W. B. ItUBBARD

IL APPllOXIMATIONS AND C O N S T I T U E N T equ:tles the Jovian luminosity (neglecting

P I I YSICS solar input for the moment) lo the loss
The initial assumptions are (a) cooling of interior heat :
at, constant radius, and (b) fully adiabatic
L = 4rR2oT, 4
structure determined by the surface con-
dition. Assumption (a) will be examined
.... + (4)
further below. We also assume (e) ttmt
dl p" d}l '
the relevant t h e r m o d y n a m i c parameters
are those for the liquid metallic hydrogen where L is the luminosity, R is the planet's
core, which comprises 80% of the mass radius, ~ is the S t e f a n - B o l t z m a n n con-
of Jupiter at present. Finally, (d) it. is stant, and E is the internal energy per
assumed t h a t the composition of the gram, and the integration is carried out
metallic core is approximately solar', with over the mass of the planet.
a mass fraction 0.77 of hydrogen. Using the thermodynamic identity
F r o m H u b b a r d (1973), the interior tem-
dE = CvdT- (l':p'2)dp- Cv(OT;'Op),dp,
perature distribution can I)e represented
to w i t h i n a~ c,//o by (4) takes the form
T = 1 2 0 T ~ ( p / 5 ) r, (1) 4rrR2~Te 4
where T is the temperature in degrees
Kelvin, o is the total mass density, and = _f<l,,,Cv(¢r 7' zo) ,
"r--- (n)
T~ is the t e m p e r a t u r e at 1 bar of pressure.
The calculations of M C I V indicate th:~t where Cw is tim heat capacity per gram.
the Grfineisen parameter, y, can be taken F r o m (3),
equal to 0.64 with satisfactory accuracy
over the range of conditions of interest .dT
... = T( l d In g d In T,,
here. Calculations for a molecular hydro- dt + 1.243 ----
6 dt (It
gen fluid (Slat tery and Hubbard, 1976)
indicate that, (1) is also accurate to within d In o~
a few per cent for 0 > 0 . 2 g/'em:L
The interior adiabat (1) is then coupled
to the planet's luminosity by means of a The g dependence of the interior tempera-
grid of model atmospheres. For this pur- ture distribution is so weak tlmt it can
pose, we use an analytic representation of be safely neglected. Dropping the term
the surface condition used by GI)(IO. For d In g/dt, then, (5) becomes
a surfaee gravity g ~ 10a era/see'-' and an
dt = -o~T~-a7aTdT,,, (7)
effective temperature T, ~ 10 e °K, G P G O ' s
atmospheric adiabats can be represented by where

p = 0.263g°'~'T,,-:~'TaT a, (2) OL = (4rR2z)-l --/"d ~ Cv (22.4p °'~4)

J

where P is the pressure in bars. Equa- (s)

tion (2) is valid for P ~ 1 bar'. Combining 22.4 f r.,.drCvpi .~4 ,
....
(1) and (2), the interior temperature is 1?'2~
coupled to the atmospheric parameters:
and r is the radial distance from the center
T = 66.Sg-II~'T,:'24ap'~, (3)
of the planet in centimeters. The surface
wdid for p >~ 0.2. gravity (! is assumed constant and equal
We now write an expression which to 2690 em/sec'-', corresponding to a non-
 JOVIAN COOLING 307

rotating model of the present Jupiter. rotating model. The value of a incre,tses
Similarly, we take R = 6.86 X 109 cm. by about 5% as the heavy material in
It is clear that the constant a is the the envelope is transferred to the core.
crucial parameter for the cooling rate In this calculation, we took the lower value
during the late stage of Jovian evolution. corresponding to a chemically homogeneous
To evaluate a, we need the heat capacity model, with the result
per gram, Cv, and an interior model, p(r).
a = 2.79 X 10~a cgs units. (10)
From the results of MCIV, for a hydrogen
mass fraction 0.77, and for the tempera- The time required for the planet to cool
ture range T ~ 10000 to 40000°K, the from To = ~ to the currently observed
heat capacity is, to within 5%, about value is then obtained by integrating (7).
2.0kB per heavy particle; thus For T, = 134°K at present, this time is
4.4 X 109 yr. For To = 127°K, the time
Cv = 1.66ku/mH, (9)
is 5.1 X 109 yr.
where ki~ is Boltzmann's constant and mu
is the mass of a hydrogen atom. This III. VARIOUS CORRECTIONS TO THE TIME
result for the heat capacity is essentially SCALE
independent of the hydrogen density for The calculations presented above give
densities of concern here. In the molecular the basic time scale for Jovian evolution.
hydrogen phase, the heat capacity per Here we consider a number of different
gram (Slattery and Hubbard, 1976) is corrections which can affect the time scale,
lower than the value given by (9) by but by no more than 20 to 30%.
about 50%. However, this region of the
planet contributes only about 5% of the (a) Insolation Correction
integral for a, Eq. (8), so that this ap-
proximation is quite acceptable. A further The observed heat flux from Jupiter is
approximation is the assumption that produced partly from the loss of heat
Jupiter is of solar composition. Detailed from the interior and partly from the
static models constrained by the gravity conversion of sunlight to infrared in the
coefficients and high-precision pressure- Jovian atmosphere. We assume that this
density relations indicate that the actual conversion occurs well within the con-
Jovian interior composition may comprise vective portion of the atmosphere (a fur-
of the order of 10% of elements other ther discussion of this point is given in
than hydrogen and helium by mass (Hub- Section IV). Thus the surface condition
bard and Slattery, 1976). Since the heat for the interior adiabats is given, as before,
capacity per gram is roughly inversely in terms of To, corresponding to the total
proportional to the mean molecular weight, flux through the Jovian photosphere. The
expression (9) may need to be reduced by analysis then proceeds as before, except
as much as 10%, depending upon the that (4) is now replaced by
precise composition of the interior model. 47rR2~r(T~4 - T o 4)
The time scale for Jovian cooling will
dearly be reduced by the same amount. P dp
We evaluated integral (8) for two plau- = - f d m ( ~ q - - - - f f ) , p= (11)
sible Jovian mass distributions which sat-
isfy the constraints imposed by gravity where T+ is the same parameter used by
field measurements (Hubbard and Slat- GPGO to represent the equivalent black-
tery, 1976), for a nonrotating model with body temperature of the eonverted solar
the same pressure-density relation as the radiation. Differential (7) is then re-
308 W. B. ItUBBARD

placed b y radius ('an be, represented in the form

a ~ ao[1 -- 0.S(T~ - 134)/134], (14)
dt = --aT~ -a'vav
X E1 - (r®/TO']='dV,. (12) where ~o is the value of ~ at present,
given b y (10). When this expression is
This equation can be integrated b y ex- substituted in (12) and then integrated,
panding the q u a n t i t y in brackets in powers we find a total correc.tion to the time
of To/T+, with the result required to evolve from To = 200°K to
T,. = 134°K of - 0 . 4 × 10" yr, while ( 1 3 )
t = (a/2.757)T.-2'757[-1 + 0 . 4 1 ( T e / T ~ ) ' gives 3.8 X 10" yr. T h u s we find a net
+ 0.26(Te/T,) ~ + . . . ] . (13) evolution t i m e for this phase of 3.4 X 10 °
yr, c o m p a r e d with ( I P ( i ( l ' s v.due of
T h u s the effect of insolation is to increase 2.0 X 10 '~ yr.
the cooling time. This is not a m a j o r For T+ > 200°K, m o s t of our assump-
effect. For T o = 105°K and 7,, = 134°K tions are vitiated. T h e formal result from
( A u m a n n et al., 1969), the q u a n t i t y in (13) for the time required to cool fl'om
brackets is 1.20. For T o = 108 ° K and T~ = ~ t o 7'+. = 200°K is 1.4 X 109 yr,
T~ = 127°K, it is 1.31. while G P G O finds the corresponding time
to be 0.6 X 1@ y r from a full stellar
(b) Corrections for Contraction evolution calculation. We will simply ac-
E q u a t i o n (12) cannot be validly inte- cept this latter result, which leads to a
grated b a c k to T~.-- ~ because the as- further correction of --0.8 X 10" yr.
s u m p t i o n of cooling at constant radius will
b r e a k down at some point. Accordingly, (c) Summary of Corrections
we integrate (12) only over a r'mge of 7',, The simple theory of Section I I gave :~
for which the assumptions and representa- total time to cool to T~ = 134°K of
tions of constituent relations are expected 4.4 X 10" yr. To th;s we add the correc-
to be accurate. For T, = 200°i(, internal tions + 0 . S N 10 :~ yr for insolation, - 0 . 4
t e m p e r a t u r e s are increased by a factor of X 10" yr for contraction after T~ = 200°K,
a b o u t 1.85 over the present epoch. F r o m - 0 . S X 10" yr for contraction before T~
the calculations of M C I V , this causes an = 200°K, for a total age of 4.0 X 10 '~ yr.
increase of a b o u t 20~c in internal pressures If we contimm the evolution to T.
for a given density. Assuming t h a t J u p i t e r = 127°K, the simple theory gives an age
can be represented a p p r o x i m a t e l y as a of 5.1 N 10'~ yr. The insolation correction
homologously contracting polytrope <>f increases to + 1 . 5 X 10 ~ yr, while the other
index unity (Hubl>ard, 1975), we find t h a t two corrections remain the same, leading
the corresponding radius change is a b o u t to a total age of 5.4 X 10 '~'yr.
1 0 ~ above the present, epoch, in good
agreement with G P G O . I)ensities are re- IV. EFFECT t)F POTATION: LUCY'S LAW
duced by at)out 30~)~. Although these
changes are appreciable and act to reduce T h e foregoing disc, ussion should m a k e
the time scale, their cumulative effect on clear the overriding i m p o r t a n c e of the sur-
the cooling time is not large, priinarily face condition in determining the cooling
because cooling-time increments increase time of a J o v i a n - t y p e planet. We h a v e
rapidly for each degree K e h i n (h,<'rease incorporated the effect of insolation on the
in T¢. surface condition in the same way as
Over the range 130°K <~ 7'~ < 200°K, (IPG(). The Jovian photosphere lies at
the comt)ined effect of varying density and levels of about 0.g bar, and adiabatic
 JOVIAN COOLING 309

conditions prevail at depths greater than V. CONCLUSIONS

about 0.5 bar (Trafton, •967). It is as- The results of this investigation indicate
sumed that solar energy is deposited at that a simple Jovian contraction model
considerably deeper levels, where it is has a time scale which is in agreement
effectively redistributed by convection with the age of the solar system. It does
without directly modifying the surface not appear to be necessary to invoke com-
condition for the adiabats. It is clear that plications such as gravitational unmixing
a consequence of this assumption is that of hydrogen from helium, or a possible
the effective temperature of Jupiter should discontinuity in specific entropy across the
be uniform from equator to pole. 1 If one metallic-molecular hydrogen interface (Ste-
takes into account the variation in the venson and Salpeter, 1976). This may
surface gravity of Jupiter due to its rota- prove to be a useful, if indirect, clue to
tion, one is led to an interesting predic- the nature of the metallic-molecular
tion. If convection is sufficiently vigorous transition.
to maintain the interior very close to an The time scale for Jovian cooling in-
adiabatic temperature distribution, then creases rapidly with decreasing Te. The
the same surface condition must apply at present calculations indicate that the value
equator and pole in order to maintain the of Te for Jupiter which would correspond
same adiabat T ( P ) in both regions of the to a total cooling age equal to the age of
planet. From (2), it is seen that this the solar system should lie somewhere
requires between the Earth-based measurement of
Tog-°.~3~ = const, (15) To = 134°K (Aumann et al., 1969) and
the spacecraft measurement of To --- 127°K
which is equivalent to a relation proposed (Orton, 1975). At present there is no
by Lucy (1967) for the effect of gravity strong reason, from a theoretical point of
on convective stellar envelopes. Equation view, for preferring either measurement.
(15) predicts that the polar effective tern- It should be noted that the uncertainty
perature of Jupiter should be about 2°K in the theoretical calculation presented
higher than the equatorial effective here could easily be 10% because of im-
temperature. precision in constituent relations and ne-
At a subsequent stage, it will be de- glect of effects of rotation.
sirable to incorporate the effects of rota- Our moderate discrepancy with GPGO
tion fully into a Jovian evolution calcula- remains unexplained. After all corrections
tion. For the moment, however, these are made, the time scale for our model
corrections can be ignored since they are to cool to an effective temperature of
at the level of current imprecision in the 134°K is 1.4 X 10° yr greater than the
equations of state and surface conditions. equivalent model of GPGO. At T~
Hubbard (1970) showed that the increase = 134°K, the cooling rate, using (12)
in rotational kinetic energy of Jupiter due with T o = 105°K, is
to present contraction is less than 5% of
(dTo/dt)~o~134
the rate of loss of heat from the interior.
The biggest correction d u e to rotation --- - 1 ° K / 0 . 1 4 5 X 109 yr. (16)
m a y be due to the variation of To over We conclude with experimental recom-
the surface, which could change the cooling mendations. First, it is clearly desirable
time by as much as 0.2 X 109 yr. to measure, by means of an entry probe,
1This effect was first noted and explained by the conversion of sunlight to thermal
Ingemoll (1976). The discussion presented here is radiation in the Jovian atmosphere, and
essentially equivalent. to verify in situ the surface condition used
310 W. B. HUBBARD

HUR~AaD, W. B. (1968). Therm.d structure of

for e v o l u t i o n a r y c a l c u l a t i o n s . Second, it is
Jupiter. Astrophys. J. 152, 745-754.
i m p o r t a n t to a t t e m p t to m e a s u r e t h e v a r i -
Ht;~t),XRD, W. B. 11970). Structure of Jupiter:
a t i o n i n effective t e m p e r a t u r e f r o m pole to
Chemical composition contraction, and rotation.
e q u a t o r w i t h e n o u g h precision to v e r i f y
Asb'ophys. J. 162, 687--697.
L u c y ' s law (this e x p e r i m e n t m:~y p r o v e
Ilt:m~Aau, W. B. (1973). Observational con~tl'aint
m o r e feasible for S a t u r n , where t h e v a r i - on the structure of hydrogen planets. Astrophys.
a t i o n w o u l d be larger). W h e n d a t a of this J. 182, L35 L38.
tltim~iRl), W. B. (1975). Gravitational field ()f a
q u a l i t y are a v a i l a b l e , t h e p r e s e n t J o v i a n
rotating planet with a polytropic index of unity.
i n t e r i o r s t r u c t u r e s h o u l d be t i g h t l y con-
Soviet Astron.--AJ 18, 621-624.
strained by evolutionary considerations. t]UBBARD, W. g., AND SLATTERY, W. L. (1976).
Interior sIrue(ul'e (if Jupiter: Theory of gravity
ACKNOWLIi]DGMENTS somlding. In Jupiter (A. M. J. Gehrels, Ed.).
Univ. of Arizona Press, Tqeson.
This work was supported by Joint lleseareh IXGEItS()LI~, A. P. (1976). Pioneer l0 and I I ol)-
Interchange NCA2-OI1040-602 with Ames Research serw~ti,ms anti the dynamics (if Jupiter's atnm-
Center, and by NASA Grant NSG-7045. sphere. Icarus 29, 245-253.
Lt:cv, l,. B. (1967). Gravity-darkening for ~tars
ILEFEII ENCES wilil ('()nve(.tive enveh)t)es. Z, f. Aslrophys. 65,
89 92.
AI:MANN, 1t. ti., (~ILLESPIE, C..X[., gin, AND LOW, ()awON, G. S. (1975). The thermal strucim'e of
F. J. 11969). The internal powers and effective Jupiter. ]. lmpli(:~tions of Pioneer 10 infrared
temperatures of Jupiter and S~tm'n. Aslrophys. radi(ime(er data. Icarus 26, 12:)-141.
J. 157, L69-1,72. SLX'rTEI~Y, W. 12., AND tttJmaAaD, W. B. (1976).
])EWITT) H. E., ANI) }{UB|I.\I/I), W. B. (1976). Therm(>dynaini(..sof a solar mixture of molecular
Statistical mechanics of light elements at high hydrogen and helium at, high pressure. Icarus
pressure. IV. A mi)del free energy fl)r the me- 29, 187 192.
tallic phase. Astrophys. J. 205, 295-301. STEVENSON) ]). J.) AND SALPETN|¢, ]~. ]4]. (1976).
(In~l~()sl~E, H. C., Ja., ])OLLACK,J. B., (IItOSSMAN, Interior models of Jnpiter. In Jupiler (A. M. J.
A. S., AND OLNESS, R. J. (1975). The structure (lehrels, Fd.). Univ. of Arizon~ Press, Tucson.
and evolution of Jupiter: The fluid i'omraction Ttt.XFTt)N, L. (1967). Model atmospheres of the
siage. Astrophys. J. 199, 265- 281. maj(ir planets. Astrophys. J. 147, 765-781.