Gravitation
We already discussed gravitation when deriving hydrostatic pressure. But we used an
oversimplification that F = mg. Here we dive into more details that consider the
following:
“True” Gravitational acceleration isn’t constant with position -- it goes like 1/r2
The Earth’s surface is somewhat nonspherical, and the interior doesn’t have a
spherically symmetric density distribution
Earth’s rotation produces an apparent centrifugal force that is factored into our
calculation of gravity (this is what makes gravity different than gravitation).
The latter two bullets cause gravity to not always point towards the center of the
Earth – there may be a small horizontal component.
Since our last discussion was on the kinetic theory of gases, we’ll start gravitation with an
insightful application of Boltzmann’s law. Remember, Boltzmann postulated that
Ei
e kT
pi M Ej
e
j 1
kT
where pi is the probability of finding a molecule in one of its available states i (associated
with one of its degrees of freedom) and Ei is the energy contained by state i. Let’s apply
this to an entire atmospheric column at some fixed temperature T, instead of thinking of
molecules in a box at temperature T. Now, if we consider the gravitational energy of the
molecule to be Eg = mgz, and further postulate that there are an equal number of possible
states in any interval of height dz, then we can rewrite this:
mgz
e kT
dz
p ( z ) dz mgz '
e
0
kT
dz '
ez / H d ( z / H )
where H is the familiar scale height of the atmosphere (H = kT/mg = RT/g)
This is an amazing result that – without invoking the pressure force at all – Boltzmann’s
Law predicts the hypsometric equation for an isothermal atmosphere at temperature T!
Do NOT go so far as to apply the equipartition of energy to potential energy – it doesn’t
work that way. In fact, on a molecule by molecule basis, the average potential energy in
an infinite column of air at fixed temperature is given by
mgz mg zp ( z ) dz mgH kT
0
So while each kinetic degree of freedom stores an average energy of (1/2)kT, the added
degree of freedom for potential energy gets a full kT, when in equilibrium.
What are the implications of this gravitational energy for heat capacity? Remember that
if a molecule has n degrees of freedom for energy, then the heat capacity at fixed volume
per molecule is equal to 1/2nkT. Add in the average gravitational potential energy of a
column of air and you get a total stored energy of 1/2nkT + kT per molecule. So if you
heat up the gas, you not only increase its kinetic energy per degree of freedom, you also
increase the scale height (expand the column to higher altitudes) and hence increase the
average potential energy per molecule. The temperature-dependence of the system’s
gravitational energy has implications for the heat capacity of the system.
The above result is quantitatively equivalent to our result that cp = cv + RT, which we
derived from the ideal gas law and the definition of work (remember, k/m = R*/M = R;
units J/(kg K)) What makes these two concepts equivalent is that the work done when
expanding a parcel in a hydrostatic atmosphere isn’t lost to the column – the expansion is
lifts the overlying air – the energy in the work goes into the gravitational potential energy
of the column.
A closer look at gravitation
Let’s recall Newton’s Law of Gravitation
Gm1m2
F12 r̂12
r122
This basically states that if you have two point masses, the force of the 2nd on the first is
proportional to each of their masses, is inversely proportional to the square of their
distances, and is attractive (i.e. directed towards the other).
One very useful property of the fact that the force is central is that the work done when
going from point A to point B in this force field (using mass 2 as the center of our
reference frame) is independent of the path taken. Since the work done to get from any
one point to any other point depends only on the location of the two points (relative to the
gravitating mass), we define a scalar gravitational potential which is equal to the work
done by a 2nd body as it moves from some infinite distance away (i.e. rA = infinity), per
unit mass of the 1st body.
Gm2
r12
The flip side of this coin is that the force on an object is equal to the down-gradient of
this gravitational potential energy times its mass.
F12 m1
You can convince yourself that the latter two equations reduce to Newton’s law of
gravitation.
The Earth is not a single point mass, of course. So instead, we have to separately
consider the force by each infinitesimal element of mass. Fortunately, forces are vector
quantities that add. So all we have to do is integrate over the force contributed each
element of mass within the earth’s Volume, V. Furthermore, because vector operators are
linear, we can either add the forces due to each sub-mass, or we can add their potentials.
It’s easier to do the latter. We have substituted the identity dm = dv
(r2 )dv(r2 )
E (r1 ) G
V
r2 r1
You don’t really have to worry about solving this equation… The point is that
FE1 m1 E
The gravitational force on a body can be related to the gradient of a single, scalar field,
which we call the Earth’s Gravitational Potential.
To this we will add the effects of Earth’s rotation. Like gravitation, the Centrifugal
pseudo-force is a geometric effect – thus the acceleration is independent of mass.
Fc = mR
where R is the radial vector from Earth’s axis, and is the angular rotation rate
(=2/day). In a rotating coordinate system whose origin is the Earth’s center of mass, we
can derive a potential by integrating the force along some radial path, yielding:
c =-(1/2)R2
= -(1/2)r2cos2
where now r is the distance to Earth’s center and is latitude.
If we were to suppose that Earth was a perfect sphere with a gravitational potential of
-Gme/r, we would have a geopotential of
= -Gme/r - (1/2)r2cos2
Geopotential is the equivalent of gravitational potential, but after including the effects of
centrifugal acceleration. Gravity is the sum of gravitation and centrifugal acceleration.
g = -
Note that there are both radial and longitudinal components to gravity. The former have
to do with what we perceive as gravity – the latter have effects on the shape of constant
geopotential surfaces, such as the oceans and the coarse structure of the Earth itself
gz = -ddz = TBC
Here are some applications we will discuss in lecture:
1) How do we get from F = -Gmem/r2 to F = -mgz?
2) If you plot geopotential as a function of distance from Earth’s surface, it stops
increasing, and starts decreasing with distance. What is the significance of this?
3) What is gravity at the equator compared to that at the poles?
4) What is a constant geopotential surface? How does this affect the shape of the Earth?
5) Certain applications for satellite orbits:
a) What is an orbit?
b) Conservation of energy (kinetic + gravitational potential)
c) relationship between the angular velocity and altitude of an orbit
d) escape velocity for satellites and gases
e) The GRACE mission