Atmosphere, Ocean and Climate Dynamics

Answers to Chapter 2

1. At present the emission temperature of the Earth is 255K, and its albedo
is 30%. How would the emission temperature change if

(a) the albedo were reduced to 10% (and all else were held fixed);
(b) the infra-red opacity of the atmosphere were doubled, but albedo
remains fixed at 30%.

The emission temperature is defined as

· ¸ 14
(1 − αp )S
Te = , (1)
4σ

where αp is the planetary albedo, S the solar flux, and σ the Stefan-
Boltzmann constant.

(a) If albedo were reduced from αp = 30% to α0p = 10%, the emission
temperature would change from Te (at present) to Te0 , where
· ¸ 1 · ¸ 14
Te0 1 − α0p 4 0.9
= = = 1.0648 ,
Te 1 − αp 0.7

so the new emission temperature would be 255×1.0648 = 271.5K.

(b) Emission temperature–the temperature at which the Earth emits
to space–would not change at all if atmospheric IR opacity were
doubled but albedo remained fixed. Emission temperature–unlike
surface temperature–depends only on how much of the solar en-
ergy flux is absorbed by the Earth and, by (1), depends only on
αp , S, and σ.

2. Suppose that the Earth is, after all, flat. Specifically, consider it to be
a thin circular disk (of radius 6370km), orbiting the Sun at the same
distance as the Earth; the planetary albedo is 30%. The vector normal
to one face of this disk always points directly toward the Sun, and the
disk is made of perfectly conducting material, so both faces of the disk

1
 are at the same temperature. Calculate the emission temperature of this
disk, and compare with the result we obtained for a spherical Earth.
Incoming solar flux S0 = 1367Wm−2 ; planetary albedo αp = 0.3. Area
of disk intercepting solar flux= πa2 . So,

Net solar input = S0 πa2 (1 − αp ) .

Disk has temperature on both faces, so area emitting thermal radiation

is 2πa2 . Disk emits σTe4 per unit area, so

Net thermal emission = 2πa2 σTe4 .

Balancing input and emission,

(1 − αp )S0 πa2 = 2πa2 σTe4 ,

i.e.,
· ¸ 14
(1 − αp )S0
Te = = 303.1K .
2σ
1
The expression for Te is a factor 2 4 larger than we found for a spherical
Earth–the disk has the same cross-section as the sphere (and so inter-
cepts the same amount of solar radiation) but one-half of the surface
area, so must increase Te4 by a factor of 2 to compensate.

3. Consider the thermal balance of Jupiter. You will need the following
information about Jupiter: mean planetary radius = 69500km; mean
radius of orbit around the Sun = 5.19A.U. (where 1A.U. is the mean
radius of the Earth’s orbit); planetary albedo= 0.51.

(a) Assuming a balance between incoming and outgoing radiation, cal-

culate the emission temperature for Jupiter.

Solar flux at earth orbit S0 = 1367Wm−2 , so solar flux at Jupiter’s

orbit is
µ ¶2
mean radius of earth’s orbit 1367
SJ = S0 = = 50.75Wm−2 .
mean radius of Jupiter’s orbit (5.19)2

2
Given a Jupiter albedo αJ = 0.51,

net solar input = SJ (1 − αJ ) πa2J = 3.77 × 1017 W.

Assuming blackbody radiation at temperature TJ ,

net thermal emission = 4πa2J σTJ4 .

Assuming these balance gives

· ¸ 14
(1 − αJ )SJ
TJ = = 102.3K .
4σ

b. In fact, Jupiter has an internal heat source resulting from con-

tinued planetary contraction. Using the conventional definition of
emission temperature Te ,

σTe4 = (outgoing flux of planetary radiation per unit surface area)

the measured emission temperature of Jupiter is 130K. Calculate

the magnitude of Jupiter’s internal heat source.

Observations show actual emission temperature is TJactual = 130K, i.e.

Therefore, if the net internal heat source is H,

H = (net thermal emission) − (net solar input)

h ¡ ¢4 i
= πa2J 4σ TJactual − SJ (1 − αJ )
= 6.06 × 1017 W.

c. It is believed that the source of Q on Jupiter is the release of

gravitational potential energy by a slow contraction of the planet.
On the simplest assumption that Jupiter is of uniform density and
remains so as it contracts, calculate the annual change in its radius
ajup required to produce your value of Q. (Only one half of the
released gravitational energy is convertible to heat, the remainder
appearing as the additional kinetic energy required to preserve the
angular momentum of the planet.)

3
 [A uniform sphere of mass M and radius a has a gravitational
2
potential energy of − 35 G Ma where G is the gravitational constant
= 6.7 × 10−11 kg−1 m3 s−2 . The mass of Jupiter is 2 × 1027 kg and
its radius is ajup = 7.1 × 107 m.]
Expressing what we are told in mathematics we have:
µ ¶
1∂ 3 M2
− G = 4πa2 Q = H
2 ∂t 5 a

∂
¡ 1¢ 1 ∂
and so, noting that ∂t
−a = a2 ∂t
a and rearranging we find

∂ 40π a2
a= H.
∂t 3 GM 2
Inserting numbers we have
2
∂ 40π (7.1 × 107 m)
a = × 6.06 × 1017 J s−1
∂t 3 6.7 × 10−11 kg−1 m3 s−2 × (2 × 1027 kg)2
= 4. 8 × 10−10 m s−1 = 1. 5 × 10−2 m per year!

4. Consider the “two-slab” greenhouse model illustrated in the figure below

in which the atmosphere is represented by two perfectly absorbing layers
of temperature Ta and Tb .
Determine Ta , Tb , and the surface temperature Ts in terms of the emis-
sion temperature Te .

4
Each layer (and the surface) radiates as a blackbody, and is totally ab-
sorbing. Therefore fluxes (per unit area) are

A ↑ = A ↓= σTa4 ;
B ↑ = B ↓= σTb4 ;
S ↑ = σTs4 .

Net input per unit area from space to the Earth-atmosphere system is
(1 − αp ) S0 /4. Net output per unit area is σTa4 . Therefore
S0
σTa4 = (1 − αp ) , or
4
· ¸1
S0 4
Ta = (1 − αp ) = Te .
4σ
Consider balance of layer A. Only inputs and outputs are IR; net input
is B ↑; net output is A ↑ +A ↓. So

B ↑= σTb4 = A ↑ +A ↓= 2σTa4 = 2σTe4 ,

so
1
Tb = 2 4 Te .
Consider balance of layer B. Net input is A ↓ +S ↑; net output is
B ↑ +B ↓. Therefore

S ↑= B ↑ +B ↓ −A ↓ ,

i.e.,

σTs4 = 4σTa4 − σTe4

= 3σTe4 .

So
1
Ts = 3 4 Te .
QED. Can also (as a check) consider surface balance. Net input there
is (1 − αp ) S0 /4 + B ↓; net output is S ↑. With the above results,

net input = (1 − αp ) S0 /4 + B ↓= σTe4 + σTb4

= 3σTe4 ;

5
 net output = S ↑= σTs4
= 3σTe4 .

These are balanced, as they should be.

5. Consider an atmosphere that is completely transparent to shortwave

(solar) radiation, but very opaque to infrared (IR) terrestrial radiation.
Specifically, assume that it can be represented by N slabs of atmosphere,
each of which is completely absorbing of IR, as depicted in the following
schematic figure (not all layers are shown).

Assume blackbody radiation. Then each atmospheric layer radiates

both up and down with a flux σTn4 per unit area. The upward flux
from the surface is σTs4 . The net solar flux per unit area, which is
absorbed only at the surface, is 14 S0 (1 − αp ) = σTe4 , where Te is the
equilibrium temperature.

(a) By considering the radiative equilibrium of the surface, show that

the surface must be warmer than the lowest atmospheric layer.
Because IR radiation reaching the surface comes only from the
N th layer, the surface heat budget, in equilibrium, is

net solar input + IR from layer N = net IR loss from surface

6
 i.e.,

σTe4 + σTN4 = σTs4 , or (6.1)

Ts4 = TN4 + Te4 .

Therefore Ts > TN –the surface is necessarily warmer than the

lowest atmospheric layer. This is a simple consequence of the fact
that the surface is heated by solar radiation as well as downwelling
IR from the lowest layer. (If no solar radiation reaches the surface,
Ts = TN .)
(b) By considering the radiative equilibrium of the nth layer, show
that, in equilibrium,

2Tn4 = Tn+1
4 4
+ Tn−1 ,

where Tn is the temperature of the nth layer, for n > 1. Hence

argue that the equilibrium surface temperature is
1
Ts = (N + 1) 4 Te ,

where Te is the planetary emission temperature.

Consider layer n > 1. This layer loses heat by radiating IR both
up and down, so its net rate of heat loss per unit area is 2σTn4 . It
4
receives IR from the layer above (σTn−1 per unit area) and below
4
(σTn+1 per unit area). balancing input and output, therefore,

2Tn4 = Tn−1
4 4
+ Tn+1 . (6.2)

Note that this result is valid for all n > 1, including n = N, in

which case layer n + 1 is the surface.
Now consider the net radiation between the Earth and space,
which must be zero in equilibrium. Net input per unit area from
the Sun is 14 S0 (1 − αp ) = σTe4 ; the output to space per unit area,
from the top of the atmosphere, is σT14 . Balancing these,

T1 = Te , (6.3)

Thus defining T1 , the temperature of the top layer. If we consider

the balance of the top layer itself, its loss of heat per unit area

7
 (via IR up and down) is 2σT14 , while the gain is from layer 2 only,
and is σT24 per unit area. Equating these,
T24 = 2T14 = 2Te4 . (6.4)
Now, (6.2) gives
4
Tn+1 − Tn4 = Tn4 − Tn−1
4
(6.5)
so the difference in T 4 between adjacent layers is the same. Given
4
(6.4) at the top layer, it follows from (6.5) that Tn+1 − Tn4 = Te4
for all n. From (6.3), it then follows that
Tn4 = nTe4 .
With n = N + 1 at the surface, then, the surface temperature is
1
Ts = (N + 1) 4 Te .

6. Determine the emission temperature of the planet Venus. You may

assume the following: the mean radius of Venus’ orbit is 0.72 times
that of the Earth’s orbit. Given the solar flux decreases like the square
of the distance from the sun and given that the planetary albedo of
Venus = 0.77, determine the emission temperature of Venus.
According to the inverse square law,
µ ¶−2
solar flux at Venus radius of Venus orbit
=
solar flux at Earth radius of Earth orbit
whence
1367
solar flux at Venus = SV = = 2637Wm−2 .
0.722
Therefore, in equilibrium, the emission temperature of Venus is
· ¸1
SV (1 − αV ) 4
TV = ,
4σ
where αV = 0.77 is the planetary albedo of Venus and σ the Stefan-
Boltzmann constant. Therefore
· ¸ 14
2637 × 0.23
TV = = 227K.
4 × 5.67 × 10−8

8
 The observed mean surface temperature of the planet Venus is about
750K. This is much greater than the emission temperature of Venus
calculated in Q.3. How many layers of the N −layer model considered
in Q5 would be required to achieve this degree of warming? Comment.
The observation that the surface temperature is about 750K suggests
an extremely efficient greenhouse effect on Venus. (In terms of the N-
layer model, we would need (750/227)4 − 1 ∼ = 100 layers to achieve this
degree of warming.) Indeed Venus has a much thicker atmosphere than
the Earth (surface pressure on Venus ' 90 times that on the Earth),
which consists mostly of CO2 , so it is extremely opaque to IR.

7. Climate feedback due to Stefan-Boltzmann.

(a) Show that the globally-averaged incident solar flux at the ground
is 14 (1 − αp )S0 .
The solar radiation reaching the surface over the globe is (see
Fig.2.4) (1 − αp )S0 πa2 . To obtain the global average, we divide by
the surface area 4πa2 to yield a globally-averaged incident solar
flux at the ground of 14 (1 − αp )S0 .
(b) If the outgoing longwave radiation from the earth’s surface were
governed by the Stefan-Boltzmann law, then we showed in Eq.(2.15)
that for every 1W m−2 increase in the forcing of the surface energy
balance, the surface temperature will increase by about a quarter
of a degree. Use your answer to (a) to estimate by how much
one would have to increase the solar constant to achieve a 1 ◦ C
increase in surface temperature? You may assume that the albedo
of earth is 0.3 and does not change.
To achieve a 1 ◦ C rise in surface temperature we require a 4W m−2
increase in the forcing of the surface. Thus δS0 (1−α
4
p)
= 4W m−2
16
implying that δS0 = 1−0.3 = 22W m−2 . This is 1. 6% of the solar
constant, a significant increase.