Part III - Lectures 27 - 41

Lecture 29: The Snow Line and Formation of the Jovian Planets

The Snow Line

Models of the solar nebula may be calculated with the help of the thermodynamics discussed in Lectures 20 and 22. The classical calculations, developed along the ideas of Harold Urey, assumed that there was a unique temperature and pressure at each distance from the sun. Given these conditions, it is relatively straightforward to calculate the chemical and mineralogical composition of solids that are in equilibrium with the residual gas.

For realistic models, the gas is largely H2 throughout the nebula, with the usual admixture of 10% helium by number. We can set up a model starting with the SAD composition, and fix the temperature and pressure so that we get a density for the solid species at 1 AU from the sun of about 4 gm/cm3. This will match the uncompressed density of the earth. The models then show that water will condense out as ice a little beyond the asteroid belt. This is the location of the snow line.

In the inner solar system, the solids that have formed are essentially rock and metal. It is often thought that these solids are amorphous, ill or loosely formed. An essential point is that we do not really know how quickly these solids will collect to form bodies that can accrete. Collisions may be destructive as well as constructive, and we do not yet know how to fix the ratio between them.

Beyond the snow line, the solids that will come together are intrinsically softer, and perhaps they will have an easier time sticking together.

The Need For Speed

We are quite sure that the terrestrial planets have lost their complement of H and He from the SAD. Generally, strong winds from the early sun are postulated to have blown this excess gas away. The sun is then said to be in its T Tauri phase; the young stars named after the prototype T Tauri are known to have strong winds. However, if this gas blows past the outer solar system, it would also remove that gas, and consequently the Jovian planets could not form as we now observe them.

There are two possibilities. One is that the wind blows mostly out of the plane of the solar system. Perhaps there would have been enough wind to scour out the inner regions, but not where the Jovian planets are. This possibility is not usually considered. The more likely explanation is that the Jovian planets formed rapidly, before the energetic wind blew the nebula away.

Rapid Core Formation

It is generally believed that Jupiter and Saturn have substantial cores that are more like rock and ice in composition than the SAD. These rocky-ice cores are estimated to range from 5 to 15 times the mass of the earth. It is assumed these cores accrete, by much the same mechanism as the metallic-rocky terrestrial planets, that is, by constructive collisions.

There is a big problem here. The terrestrial planets can form on a relatively leisurely time scale, of the order of 107 or 108 years, after the residual gas of the solar nebula has been blown away. The giant planets must accrete much more material, perhaps 5 to 15 earth masses, before the gas has been expelled. Current calculations find a way to do this, but by setting adjustable parameters to favorable values.

We don't really know the distribution of sizes of accumulating planetesimals. Nor do we know the relative rates of constructive to destructive collisions--the fraction of the time when two planitesimals collide that a bigger body results rather than a smaller one. Within this blissful ignorance, we are free to set the parameters so we get the answers we need. We know the Jovian planets did form, and we have an idea of the time scale required for them to form. So we can take the existence of the planets as a way of fixing these unknown parameters. Some day, we may have a better way to do this.

Models of the Giant Planets

A model of a planet is really a table of numerical values of thermodynamic variables as a function of radius. For our purposes, any physical system that has reached an equilibrium, or maximum entropy state may be described by two thermodynamic variables, plus the chemical composition. These two variables might be the temperature and the pressure, or the density and the pressure. The latter two variables turn out to be the most useful in practical calculations. We shall not worry about additional variables that may be necessary in a refined calculation.

The pressure at the surface of a giant planet is much lower than that at its center. It is a good approximation to assume that the surface pressure is zero, and increases toward the interior.

Consider an area 1 cm2 at the planet's surface, and 1 cm in thickness--a cube, one cm on a side. If we know the density of the material, we know the mass of the cube. It is simply the density times the volume, 1 cm3, or just . We assume this density has some small value, characteristic of a value at the center of the cube. Since we know the mass of the planet, M, we can compute the force on the cube using Newton's law of gravitation. Thus

         GM 
F = ---------
          2
         R

Now this force, divided by the area at the base of the cube gives the pressure, P, since P = force/area. This gives us the first step in the creation of a model for the planet. We next need to have a way of estimating how much the density in the next cube down will increase as a result of the increase in pressure. This will enable us to take the next step in the construction of the model planet.

The new pressure thus requires a new density. Relations among thermodynamic variables, such as pressure and density are called equations of state. The simplest equation of state is that for an ideal gas, which we may write in several ways, e.g.


PV = nRT, or P = RT/mu,

where "mu" is the molecular weight, and R the gas constant, 8.314 x 107erg/mole/deg.

It is a good approximation to assume the equation of state is a perfect gas when the pressure is very small, and at that point, the molecular weight should be close to 2.0, for molecular hydrogen.

The model is constructed by requiring that at every level in the planet, the pressure is sufficient to support, or hold up the layers above it. When one gets to the center of the planet, the densities that have been calculated must be such that the known mass of the planet is obtained by adding all of the mass in the spherical shells from the center out to the surface.

Deeper in the planet, more complicated equations of state are required. What is the proper equation of state? This is the most important question for anyone who wants to make a model of a planet. Strictly speaking, we must know the physical conditions--the model, before we can give a valid answer. It would seem that we are in a hopeless situation, but in practice it is possible to make some progress. We proceed in a series of approximations.

First, we assume a very crude equation of state, and with its help, we get a first approximation to a model. The approximate model gives us an idea of the physical conditions, the thermodynamic variables for various points within the model. We can then formulate a better equation of state, and repeat the procedure.

If you have assumed the correct equation of state, then the densities at various radii will be such that all spherical shells will add up to give you the correct mass of the planet when you get out to the correct radius.

You can get the mass and radius to agree even if you have the wrong equation of state! This is the problem with science. You are never 100% sure you've got it right. What you can know for sure is that if your model doesn't give you the correct mass for the known radius, you've got something wrong.

Figure 29 - 1: Pure Hydrogen Models of the Jovian Planets

Figure 29-1 shows the results of constructing models for giant planets under the assumption that they are pure hydrogen. The curve shows that for small masses, the radii increase with the mass. This is what one would expect for familiar materials.

Above masses of about 1031 gm, an increase of the mass results in a decrease of the radius. This happens because of the nature of the equation of state at higher pressures than are relevant for the planets. What happens is that the electrons are squeezed closely together, and quantum effects become relevant. We need not pursue this here. The phenomena are relevant for white dwarf stars.

Modern models of Jupiter have a dense, fluid molecular shell below the outer atmosphere. This material never liquifies in the sense of exhibiting a phase change, where the volume decreases virtually discontinuously with pressure. Indeed, for molecular hydrogen, there is a ``critical temperature'' above which this liquifaction is impossible. This critical temperature, 33K, is lower than any temperature of models of Jupiter, so that liquifaction in the usual sense of the word never occurs. Nevertheless, with increasing pressure, the gas density increases to values typically found in liquids, and it is common for workers in the field to use the term liquid to describe it. Strictly, it would be better to describe this material as a ``fluid,'' a term applicable to either a gas or a liquid. However, this refinement is often eschewed by workers in the field as well as writers of textbooks. At the relevant pressures, the hydrogen assumed metallic properties--it becomes a good conductor of heat and electricity.

You can see from Figure 29-1 that the points for Jupiter and Saturn come close to the theoretical curve for pure hydrogen, but those for Uranus and Neptune do not. You can see clearly that for the radii of the two outer planets, the masses are higher than those predicted for pure hydrogen models. This is the basis for the assumption that the Jovian planets have rocky cores. In the case of Jupiter and Saturn, the need for these cores is less demanding, but such cores are now assumed to be relevant for them also.

Envelopes or Mantles of the Jovian Planets

Stars and planets get divided into three zones, for good reasons. In the case of the earth, the three zones are core, mantle, and crust. The core is metal, the mantle is rock, and the crust is highly differentiated rock. The three zones for stars are called core, envelope, and atmosphere. Energy from the sun's core is transported outward by photons, so it is said to have a radiative core. Energy from the core is transported to the atmosphere mostly by convection, so the sun has a convective envelope.

The distinction between core, mantle, and crust seems appropriate for the terrestrial planets. Workers who model the Jovian planets tend to borrow the term used by the stellar workers, and call the thick, intermediate shell an envelope rather than a mantle. Both terms are in use.

In the envelopes or mantles of Jupiter and Saturn, the molecular hydrogen is squeezed first to a liquid, and finally to a metallic state. The corresponding zones of Uranus and Neptune are thought to have a larger component of rock and ice in their mantles than jupiter and Saturn.

Figure 29-1 shows that pure hydrogen models come close to the correct mass-radius relations for Jupiter and Saturn. Recall that the SAD is some 90% hydrogen, so these planets are probably mostly SAD on top of the rocky cores. These SAD envelopes probably formed by (rapid) gravitational collapse onto the cores. This is a mechanism used long ago to account for the formation of the stars from diffuse clouds of gas. If enough mass is present in a given volume, it will begin to collapse under its own self gravitation, that is, the outer layers will be pulled in by the inner regions. Since gravity is an inverse square force, the smaller the mass becomes, the greater the force pulling the outer layers toward the center.

Gravitational collapse will stop if the inner regions, the cores, can become hot enough or compact enough that high pressures stop the collapse. In stars, high temperatures stop the collapse. In the giant planets, it is stopped by phase changes, from gas to liquid or by the solid (rock or icy) cores.

Modern calculations have refined the picture of gravitational collapse, but the description remains qualitatively valid.

The outer Jovian planets, Uranus and Neptune, are smaller than the inner ones. It is not unreasonable to attribute this to the fact that they are at the edge of the solar nebula, and there was simply less material available.

Sources of Internal Heat

In the next chapter, we will discuss in some detail the question of the heat balance in both the Jovian and terrestrial planets. Here, we will take up the question of internal sources of heat.

When the earliest space probes, the Voyagers, flew past jupiter and Saturn, they found both planets radiated more heat than they received from the sun. This was quite different from the situation that obtained for the terrestrial planets. What could the internal heat sources be?

In the case of the earth, it is possible to get a measurement of the heat flux coming from the interior. It is about 4 x 1013 Joules/sec, or 4 x 1013 watts. We can account for this flux by assuming that radioactive elements are decaying within the earth at known rates. There are other possibilities as well, so as far as the earth is concerned, we have an embarrassment of riches.

Radioactive decay cannot possibly supply the heat fluxes from the Jovian planets. It is generally assumed that the planets are either still shrinking, or that helium is sinking, causing the density at the centers of the planets to increase. Either of these possibilities could be considered primordial heat in the sense that it is derived from the same gravitational energy that was released as the planet formed.

Summary

The snow line is the zone beyond the asteroid belt where water would condense as ice from the solar nebula. The Jovian planets are thought to have begun as rock and/or icy cores in this region. These planets must have formed rapidly, before the early sun, in its T Tauri phase, expelled the excess gas of the solar nebula. Possibly the ices could coagulate more quickly than the metal and rocky materials of the inner solar system.

Jupiter and Saturn can be approximately fit by models consisting of pure hydrogen. They are mostly SAD in composition, with rock and icy cores. The intermediate zones, called envelopes (or sometimes mantles), probably formed by gravitational collapse. Rock-icy cores play a larger role in models of Uranus and Neptune. Their envelopes also contain a larger rock-ice admixture along with the SAD than their more massive congeners.

All four Jovian planets put out more heat than they receive from the sun. Radioactivity can not account for the excess heat. It probably derives from gravitation, but could also have a chemical source.

Lecture 30 - Temperatures and Heat Flow in the Solar System

Heat Flow and Radiation: Key Concepts

We have discussed the three modes of transport of energy, conduction, convection, and radiation. Heat is a manifestation of the energy of matter and radiation. In this context, radiation means photons, or light particles. Photons have energy and momentum, but no rest mass. There is no such thing as a photon with zero velocity. This, then, distinguishes radiation from matter, which may be in the form of electrons, protons, or neutrons. These particles may be in motion, or at rest.

The energy in any volume in space is a measure of the ability of the matter and radiation within it to do work. The second Law of Thermodynamics puts restrictions on the amount of work that can be done with that energy. We shall only make indirect use of those principles in this lecture.

We have seen that the energy of a molecule of a perfect gas is (3/2)kT. In general, ``kT'' is a reasonable measure of the kinetic energy of motion of an atom or molecule in a medium of temperature T. The constant, k, is called Boltzmann's constant, after the nineteenth century physicist, Ludwig Boltzmann.

Note that the gas constant, R (PV = nRT), is equal to Avogadro's constant (Na) times k. It is easy to see that if there are N particles in a volume, V, that N/Na = n, the number of moles. Thus the energy per unit volume, is (3/2)NkT/V = (3/2)nRT/V = (3/2) P. Thus we see that the pressure of an ideal gas is 2/3 of the energy density.

These relations for an ideal gas were well known by the end of the nineteenth century. Physicists also knew that radiation in equilibrium could be described by simple laws that depended only on the temperature, and not on the matter interacting with the radiation. It is essential for the radiation to be in equilibrium for this to be rigorously true. We can be sure that radiation and matter are in equilibrium if they are in an enclosure, and there has been time for all temperature inequalities to even out.

The english word for radiation that is present in equilibrium conditions is blackbody radiation. This word arose from experiments with radiation and soot-covered bodies, and is not especially descriptive. The corresponding german word literally means ``cavity radiation,'' and is a little more helpful. Radiation within a cavity is, by definition, blackbody radiation, provided the material within the cavity is in equilibrium.

One often sees the following definition: a black body is an object that absorbs all of the radiation (of all frequencies) that falls on it. Now if a body doesn't absorb radiation of some frequencies, it must reflect it. Indeed this property of matter can be very characteristic of its composition, and is the basis for reflective spectroscopy used in remote sensing (see Lectures 13 and 27).

A soot-covered ball absorbs most of the light that falls on it, and it emits a spectrum that is approximately that of an ideal black body.

The spectrum of an ideal black body follows a formula that contains mathematical and physical constants and one thermodynamic variable, the temperature. Some blackbody curves are shown in Figure 30-1. They are also called Planck curves, after the physicist Max Planck, who derived the mathematical formula that describes them. The relation is sometimes called the Planck Formula, or Planck's Law.

Figure 30-1: Black Body Curves for Several Temperatures

Both axes in Figure 30-1 are logarithmic. The smallest wavelength shown is 0.1 microns, or 1000 Angstroms. Note that the maximum of the curves shift to the left as the temperatures increase. Indeed, the wavelength where the Planck curves reach a maximum obey a simple relation:

maxT = 0.29

The wavelength of maximum intensity, max, is in centimeters, in this formula, and the temperature in degrees Kelvin.

The total amount of energy that leaves unit area of a black body is directly related to the fourth power of the temperature (in Kelvin):

E = T4.

The constant is called Stefan's constant, or sometimes the Stefan-Boltzmann constant. In cgs units, = 5.670 x 10-5 erg cm2sec-1deg4. In SI units = 5.670 x 10-8 watts meter2sec-1deg4. Remember that a watt is a joule per second, or 107 erg/sec.

Non-black Surfaces

When radiation falls on any "non-(ideal)black surface," some fraction of the energy enters the surface, and the complementary fraction is reflected. The energy that enters the surface may be completely absorbed, or if the surface belongs to a thin layer, the energy that is not absorbed will be "transmitted." Most of the visible light that enters glass is transmitted. If the surface upon which light falls is thick, and opaque, we may safely assume that all light that is not reflected is absorbed. This is the situation with light received from the sun by planets, satellites, and asteroids.

The fraction of the light that falls on a planet that is reflected is called its albedo. There are several ways to define albedo, but we shall use it to mean the fraction of the total radiation that is reflected from a planet. Thus, if the planet receives an amount of energy E, from the sun, and the planet's albedo is A, then (1-A)E is absorbed by the planet, and AE is reflected into space.

Non-black surfaces emit radiation approximately according to Stefan's law, but with a correction factor that is peculiar to each substance, and to each temperature. The correction factor, , is called the thermal emissivity. For soot, is very close to unity, but for other substances, it can vary from values near unity to values under 0.1.

Predicted Surface Temperatures

Consider the planet Jupiter. We can easily calculate the power that is intercepted by a disk with its radius (RJ) at its distance from the sun. The sphere surrounding the sun has an area of 4 r2, while the disk of Jupiter represents an area RJ2. Here, r is the distance of Jupiter from the sun, and RJ is Jupiter's radius. So the fraction of the sun's power that is intercepted by this disk is RJ2/(4 r2)= 2.1 x 10-9.

The power of the radiation intercepted by Jupiter's disk comes if we multiply the above fraction by the sun's power output, 3.84 x 1026 Watts. The total power caught by the disk is 8.08 x 1017 Watts. This is not the full story because some fraction of this light must be reflected--otherwise we wouldn't see Jupiter. The fraction that is reflected is the albedo, and the figure we use for it is 0.70. This means 0.30 of the light falling on Jupiter is absorbed. Doing the multiplication, we get 2.4 x 1017 Watts.

We can use these concepts to estimate the temperatures at the surfaces of all of the planets. The calculations are summarized in the table below. What we do is equate the energy received from the sun to the energy that would be radiated by a black body. In the formula below, we assume the emissivity of the planet is unity. Since it is more common to find planetary diameters, D, in tables, we use R = D/2.

                        2
               \pi (D/2)                  2           4
(1-A) L(sun)x ------------- = 4\pi   (D/2)  \sigma x T        (1)
                       2
                4 \pi r

If we solve this equation for T, after calculating the energies received by the planets, we get the following predicted temperatures T = T(p), which are compared to measured temperatures T(m). In the following table, we give the distances r in astronomical units and the planetary diameters D in kilometers, as is done in the text. But for use in (1) all distances must have the same units. We must use meters, if we take sigma as 5.67 x 10-8 Watts per meter squared per deg4.

Table 30-1 Predicted Surface Temperatures

      Planet    Albedo    r       D      T(p)      T(m)
     Mercury    0.056   0.387    4878    441       440
     Venus      0.72    0.723   12102    238       250 (clouds)
     Earth      0.39    1.00    12756    246       280
     Mars       0.16    1.5237   6786    216       230
     Jupiter    0.70    5.203  142984     90       125
     Saturn     0.75    9.539  120536     64        95

The values T(p) in Table 30-1 are in reasonable agreement with the measurements, though not perfect. Planetologists often call this temperature, T(p), the effective temperature. This means nothing more than that a possible emissivity, that should be on the right side of Equation (1), has been set to unity.

All told, this technique does a reasonable job of predicting the surface temperatures of planets. We got fooled by the temperature at the base of the clouds on venus, but that is another story. We were fooled again, when the first probes, the Voyagers, flew by Jupiter and Saturn. These planets were putting out more heat than they received from the sun.

In order to see how this situation might arise, let us examine the sources of internal heat of planets.

Heat Sources of Planetary Interiors

- i -

The planets all have both internal and external heat sources. The most important external heat source now is the sun, but in the past meteoroid bombardment supplied a good deal of heat. Internal heat sources derive from radioactive decay, chemical energy, and gravitation.

The amount of energy released in a radioactive decay can be measured in the laboratory. The measurements are for specific, radioactive elements. In order to apply these measurements to the earth's heat supply, we must know the relative fraction of the radioactive elements in the earth.

The radioactive elements that are important heat suppliers are uranium, thorium, and potassium. Uranium and potassium have two naturally occurring isotopes. Both uranium isotopes are radioactive, though at different rates. Of the two potassium isotopes, only 40K is radioactive. To calculate the energy released from these two elements, it is necessary to know the abundances of the isotopes individually. Most calculations assume "natural" isotopic abundance ratios, values that can be measured in most substances. This is not necessary for thorium, where there is only one naturally occurring isotope, 232Th.

With the isotopic mixture of uranium and potassium, we can calculate the rate of energy generation, per gram, say of these elements. Note that Rb-87, which provides a very useful means of dating rocks is not an important heat source. It decays too slowly, and there is not enough of it.

The following table gives energy output rates for the important sources.

Table 30-2: Radioactive Energy Release Rates


       0.97 x 10^{-7} Watts per gram of natural uranium
       2.7  x 10^{-8}  "     "    "  "  thorium-232
      36.   x 10^{-13} "     "    "  "  natural potassium

These figures as well as the ones following are from a 1988 book by John Verhoogen, ENERGETICS OF THE EARTH, that is often cited by geophysicists. The abundances are in parts per million (ppm) by weight. The first figures come from assuming the composition of the bulk earth can be estimated from the composition of certain meteorites. For potassium, the abundance is 170 ppm. To get the number of grams of potassium in the earth, we must multiply the 170 by the mass of the earth in grams 5.97 x 10^{27}, and then divide by a million. So the estimate is 1.015 x 10^{24} grams of potassium. To get the energy generated by radioactive decay of potassium, we now multiply by 36 x 10{-13}. If we do similar calculations for uranium and thorium, we obtain the following results.

Table 30-3: Power Output from Radioactivities
 
   Source               Abund(ppm)      Power (Watts)
                                        (whole earth)
 
   Potassium            170             3.7 x 10^{12}
   Uranium               18 x 10^{-3}   1   x 10^{13}
   Thorium               65 x 10^{-3}   1   x 10^{13}
                                        -------------
   Total    
2.4 x 10^{13}

- ii -

The average heat coming from the earth's interior is estimated to be about 4 x 1013 Joules/sec, or 4 x 1013 watts. We can see that the radioactivities just about account for this output. Perhaps, because the abundances of these radioactive sources are themselves uncertain by as much as a factor of two, we may say the heat source is explained.

It is interesting, though to consider another source of heat, core formation. Surely this was important at some point in the history of the earth. In order to calculate this, we find the difference in the potential energy of a sphere with a mean density of 5.5, with all the mass distributed uniformly through it. This calculation is done with a little calculus. Here, we just write down the answer:


               2
             M
PE = -0.6 G ----                                         (2)

R

Here, M is the mass of the earth, and R its radius. The factor 0.6 varies if the sphere isn't uniform in density, but it is nearly always close to unity (within 1/3 and 3).

We say the potential energy is negative, because we choose the zero point of the energy to be when the masses are all separated to infinity. We must put energy into the earth masses to effect that separation, so it's potential energy is negative.

G is the constant of gravitation. Notice that this is very nearly the potential energy of two masses M separated by a distance R. Now, let the mass be uniformly distributed in two layers, one with a density of about 4.3 (mantle) and another with a density of about 12.0 (core). Using the current dimensions for the core and the mantle, we find a mean density of 5.53, satisfactory for the earth.

If the earth were uniform in density, its potential energy would be -2.24 x 1032 Joules. With a core, its potential energy would be some 10% more negative. Consequently, we take the energy of core formation to be about 2.24 x 1031 Joules. Now suppose that energy were released over the 4.5 billion year lifetime of the earth. The rate of release would be:


        2.2 x 10^{31} Joules
  ------------------------------------    = 1.5 x 10^{14} Watts (3)
  (4.5 x 10^9 years)(3.1 x 10^7 sec/year)

This is more by nearly a factor of 4 than the current power output of the earth! But we assumed that the core formation energy would be put out uniformly over the history of the earth. That cannot be true. Probably much of the power was dissipated early in the earth's history. It does seem possible that a small amount might contribute to the current output.

If we took the measured heat flux 4.0 x 1013 watts literally, and also the result 2.4 x 1013 watts from Table 29-3, we would need a little more energy.

Some additional energy may come from chemical changes. One may estimate the energy released per gram of iron, as it solidifies in the core. Other phase changes take place in the mantle. If we add all of these possible sources of chemical energy, we get only 1 to 10 per cent of that released by core formation. So we shall not consider these further for the terrestrial planets.


All of the quantitative concepts here basically use one formula: Amount = rate times time. We inverted this formula to get rate = amount divided by time. Can you point out which derivation used the formula in which way? The familiar formula distance = velocity times time is a special case of the more general formula, where "amount" is replaced by distance. Given any two of these quantities, you should be able to get the third.

Jupiter and the Jovian Planets

Let us now turn to a discussion of the heat flow from Jupiter. Could this heat come from radioactive decay, as it does from the interior of the earth? We may test this hypothesis in the following way.

The power (energy per second) supplied by the dominant three radioactivities was given in Table 29-2. To see if we can get the Jovian heat output from them, we need only find out the masses of these three elements in Jupiter. This means, we need to know Jupiter's composition. Let us assume Jupiter has the same composition as the sun, what we have called the "solar mix" or the SAD --the standard abundance distribution. This isn't strictly true, but it is a reasonable first guess. If you click on the NEWS button on my HomePage, you can get the plot and a table of SAD abundances. In the astronomical convention, hydrogen is set at 1012, so its log is 12.0.

To estimate the mass of thorium in Jupiter, for example, we need the fraction of the mass of the SAD that is thorium, and then we need to multiply that by the mass of Jupiter. Since the SAD is given as numbers of atoms, we also need to consider the atomic weights. The average (atomic or) molecular weight for the SAD as a whole is 1.298. This is nearly the same as that of a mix that is 1 part helium and 10 hydrogen (about 1.27). The total number of atoms in the SAD is 1.098 x 1012. The mass of the atoms in the SAD is therefore 1.425 x 1012 grams.

To get the fraction that is thorium we need to divide this number into the mass of thorium in the SAD. The number of thorium atoms is 1.2, and the molecular weight of thorium is 232.04. So the fraction of the SAD mass that is thorium is 278.4/1.425 x 1012 = 1.95 x 10-10. We have to multiply this fraction by the mass of Jupiter, 1.9 x 1030 grams = 3.70 x 1020 grams of thorium.

The final step in estimating the energy from thorium is to multiply by the figure given in Table 30-2 for the power from thorium per gram: 2.7 x 10-8. The result is about 1.0 x 1013 Watts.

We can proceed in exactly the same way for uranium and potassium. It turns out that potassium gives the major contribution, and the total for the three radioactivities is 5.5 x 1013 Watts.

The amount of heat emerging from Jupiter has been measured to be

                               18
Jovian power output = approx 10    watts           (4)

This is far more than could be supplied by the radioactivities--by more than 4 powers of ten. Consequently, there is no possibility that radioactivities could supply the heat that is currently coming from Jupiter. A similar situation holds for Saturn, Neptune, and possibly Uranus.

The amount of heat generated by radioactive decay within Jupiter is not far from the 4 x 1013 Watts coming from the earth's interior. Actually, this is not surprising. Although the mass of Jupiter is 318 times the mass of the earth, most of the additional matter is hydrogen and helium. Indeed only about 2% or 0.02 of Jupiter's mass is not hydrogen and helium. A good fraction of that 2% is in volatiles, carbon, nitrogen, and oxygen, lost to the earth. If we remove these elements from the SAD, the resulting mass is only about (1/300)th of the original. This is about the ratio of the terrestrial to Jovian mass.

Plausible Sources of Jovian Heat

No one has suggested that any substantial amount of heat is generated by chemical reactions in the Jovian planets. We are thus left with gravitation as the only plausible force capable of generating energy (force x distance).

The excess Jovian heat might be generated by gravitation in one of two ways. First, Jupiter might be in the process of forming a core with a density much greater than the bulk density of the planet. We have seen how this mechanism could work in the case of the earth.

Let us look at what might happen if something analogous to core formation on the earth happened on Jupiter. Consider the potential energy of two different density distributions. One is constant with planetary radius, and the other goes from some central value linearly to zero at the radius of the planet. Both calculations require calculus, so we'll just state the answers. For the constant density case, the potential energy is -0.6GM2/R, where M is the mass of the planet, and R its radius. In the second case, the potential energy is -0.7GM2/R; in the second case, the matter is a little deeper in the well. Let us take -0.1GM2/R as an estimate of how much energy might be released by density rearrangement. Neither of these simple density distributions is realistic, but their difference is probably a realistic (if rough) estimate of the energy that might be released by density readjustments in the real planet.

If we put in the figures for Jupiter, we find, using G = 6.67 x 10-11 Joule-meter / kg2. SI units. Note a Joule-meter = newton-meter2. Jupiter's mass must be in kilograms, and its radius in meters to use this value of G.

 

  Estimate of energy from density distribution readjustment

         2                   -11           54
       GM          (6.67 x 10   )(3.69 x 10   )           35
 = 0.1 ----  = 0.1 ---------------------------  = 3.4 x 10   Joules (5)
        R                        8
                        0.71 x 10

This is the energy we have to work with. To turn it into a power, we have to divide by a time. If we use the age of the solar system, 4.5 billion years, as we did for the earth, we find a power (energy per unit time):

            35
    3.4 x 10                       18
--------------------     = 2.4 x 10   (joules/sec) or watts    (6)
        9           7
(4.5 x 10 )(3.1 x 10 )

This is fortuitously close to the heat output from Jupiter, which we have seen (Equation 4) is about 1018 watts.

It was originally suggested that the mass readjustment in Jupiter might be due to the heavier element helium sinking relative to hydrogen. But recent measurements of Jupiter's He/H ratio at the surface, from the Galileo probe indicated a nearly normal surface helium. By "normal" here, we mean the same as in the sun. Earlier estimates had led people to think that Jupiter's surface helium was below normal, and this could be explained if the helium had sunk relative to the hydrogen. This sinking would provide energy, as we have shown. Everything seemed to fit, and then an observation ruins the beautiful theory!! This isn't the first time this has happened in science.

The final chapter hasn't yet been written. Measurements of the H/He ratio in Jupiter's atmosphere do not necessarily apply to the bulk of the planet. Indeed, the parts of Jupiter's interior that are most important for the gravitational energy are deep within it. So we should not just throw out our beautiful theory. We just have to be cautious.

Most of the Jovian planets radiate more energy than they get from the sun. Here are some figures from a recent textbook:




                          Jupiter  Saturn   Uranus   Neptune
Energy from planet  
-----------------          1.7      1.8      1.1       2.6
Energy from Sun

Observations of Saturn show that it has a helium deficiency. These are not as complete as those of Jupiter, but the sinking of helium is perhaps still viable for that planet. Don't bet a lot on it. Let us withhold judgement, and explore an even simpler scheme to get energy. Again, we take Jupiter for specificity.

Suppose Jupiter were shrinking, without necessarily changing the density distribution within it. How much would it have to shrink to provide the power, some 1.0 x 1018 Watts put out by the planet? We can answer this with a calculation. The potential energy depends on the mass and radius of the planet, along with a constant factor, that depends on the distribution of density. Usually, this factor, which we shall call (\gamma) is near unity. So we don't need its exact value for a rough estimate.

Let us call the change in potential energy \Delta PE (PE). What is this change if the radius of the planet changes by \Delta R (R). This could be an exercise in calculus. But we can just take the difference of two expressions of the form of Equation (2). In the first, the radius is R - minus a small increment, \Delta R. The second term is the original potential energy.


                           2                       2
                         GM                      GM
\Delta PE = -\gamma  --------------   +  \gamma ---------    (7)
                     (R - \Delta R)               R

The first term on the right of (1) is bigger in absolute value, that is, more negative, than the second, because we are dividing by a smaller number. The numerators are the same, and we assume the \gamma's are the same too--similar density distributions.

We can get a power from (1) by dividing the energy by a time interval. Evaluation of the right-hand side is an exercise in the manipulation of small quantities. We assume \Delta R << R, and use 1/(1-x) is about equal to 1+x if x is small. I'll let you work out the details, as for this course, we are only interested in the result. We want to divide the difference in potential energy by a relevant time increment, which we shall call \Delta t. Then


                        2
\Delta PE             GM      \Delta R
---------   = \gamma -------  --------                (2)
\Delta t                 2    \Delta t
                       R
 

The (\Delta R)/(\Delta t) is a distance increment over a time increment, that is, a velocity. What we want to know is, how large must that velocity be for the power output (\Delta PE)/(\Delta t) to be 1.0 x 1018 Watts. We know all of the quantities in (2) on the right hand side except for the (\Delta R)/(\Delta t). So we equate them to the current power output of Jupiter, and solve for the unknown. With \gamma = 1, we find that (\Delta R)/(\Delta t)= 2.1 x 10-9 centimeters per sec!! is all that the rate of shrinkage needs to be to supply the current power output. That rate would not be detected by any measurements currently available.

In the roughly 500 years since Jupiter has been under telescopic observation, it would have shrunk by 2.1 x 10-9 (cm/sec) x 500 (yr) x 3.1 x 107 (sec/yr) = 33 cm. That is only about 2 parts in 108 of the radius of the planet. The radius of the planet cannot be uniquely given with that accuracy, since the atmosphere is gaseous.

We conclude Jupiter could derive its power either from simple shrinkage or a density redistribution. The density redistribution is now thought to be more relevant for Saturn.

Summary

Energy from the sun in the form of radiation falls on the disks of the planets. Because they are not perfect black bodies, some fraction of this energy is not absorbed. The fraction is called the albedo. We can account roughly for the surface temperatures of the planets by equating the solar influx times (1 - albedo) with the output from a black body with a temperature called the ``effective temperature'' of the planet. The Jovian planets put out more energy than they receive from the sun.

The energy flux from the earth's interior can be accounted for by the decay of radioactive elements uranium, thorium, and potassium. These sources are not sufficient for the Jovian planets. Gravitational energy does seem to be an adequate source for them. It could arise either from a redistribution of the planetary masses or a simple shrinking of the planets' radii.

Lecture 31 - Cosmic Magnetic Fields

Theory

In Lecture 12, we discussed a set of four equations developed in the nineteenth century, and now known as Maxwell's equations, after their originator, James Clerke Maxwell. These equations described scalar quantities, electrical and magnetic ``charges'' (see below), and vector quantities, the electrical and magnetic fields. Recall that a scalar quantity, like temperature, can be described by a single number, but a vector requires three, since it has both a magnitude and a direction.

One of the intriguing aspects of Maxwell's equations, is that isolated magnetic ``charge'' does not exist. By contrast, isolated electrical charges are familiar. On a microscopic scale, we may have isolated positive or negative charges, protons or electrons. On larger scales, a net positive or negative charge may reside on some macroscopic object, which can be isolated in space. No similar analogues are known to exist for magnetic charges.

Magnetic fields, on the other hand, emanate from poles of a magnet. These poles are called N and S, or + and -, the latter, rather like electrical charges. The big difference is that no one has ever located an isolated magnetic pole. All known electrical and magnetic phenomena seem to be adequately described without one.

How could you tell if there were such a thing? Here's one way. Suppose you had an isolated bit of electrical charge, which for purposes of argument, we'll say was positive. Consider a sphere of some fixed radius that surrounds the charge. We'll suppose the sphere is much bigger than the charge itself. If we took another positive charge and put it anywhere on this sphere, it would be pushed away from the center of the sphere.

Now let's try the same trick with a magnet. The only magnets we know have two ends, they are dipolar. But we can imagine a very long magnet, and surround its N pole with a sphere. The sphere can be big, but the magnet's length must be greater than the sphere, since we are trying to isolate one pole.

Finally, take (in your imagination) another long magnet, and poke the N end of it all around the sphere that surrounds the first. What every experiment finds is this: While the N pole of the second magnet will be pushed away from the center of the sphere at some points, there will be others for which it is pulled toward the center.

Lines of force give the direction a charge (pole) would be pushed by the field of another charge (pole). If there were an isolated charge, one could surround it by a sphere, and all of the lines of force would leave (or enter) the sphere. This never happens for a magnetic pole. In fact, one of the Maxwell equations states that the net magnetic lines that go into and out of any closed volume will always balance. This also happens for electrical lines of force, provided there is no charge within the volume.

Some exotic theories of physics have postulated the existence of isolated magnetic charge, called magnetic monopoles. These monopoles play a role in some theories of the early big bang. As of this writing, there is no evidence that monopoles manifest themselves in the world in which we live.

- ii -

Magnetic fields can be generated by electrical currents. Even the fields that exist in natural magnets can be thought of as due to tiny currents within atoms. They arise due to what is called the electron spin, but we won't go further into this topic.

A long wire conducting an electrical current will be surrounded by rings of magnetic field lines. This can be tested with a small compass needle. If you turn the current off, the magnetic field will collapse.

There is energy in a magnetic field. If you have a current flowing in a wire loop, there must be a magnetic field about the wire. Then if you break the current with a switch, and the field disappears, what happens to that energy? How does it manifest itself? You can do a simple experiment to show that if you just barely open the switch, a spark will jump across it. In other words, something will make the current try to continue to flow. The energy to do this, it turns out, comes from the collapsing magnetic field.

There is also energy in an electrical field. We have seen in Lecture 12 that photons, particles of light, are manifestations of both electrical and magnetic fields. The energy of photons may be thought of as coming from their associated electrical and magnetic fields.

Electric fields by definition exert forces on charges. Charges in motion give rise to electrical currents. These currents can discharge the fields, so they no longer exist. Suppose we painted one wall of a room with positively charged paint, and painted the opposite side with an equal amount of negative paint. This would create an electric field in the room, since a positive charge in the middle of the room would be pushed away from the positive paint, and pulled toward the negative paint. However, if we filled the room with mercury (which conducts electricity), the field would quickly discharge. All of the electrons in the negatively charged paint would flow through the mercury and neutralize the positive wall.

Since there is no such thing as a magnetic charge, there is no magnetic phenomena that quite corresponds to the example of the painted room. There is no magnetic analogue of an electrical current. In cosmic phenomena, there is often no simple route to the destruction of a magnetic field.

We have seen, in the case of a simple current loop, that magnetic fields are certainly destructible, but the simple route of discharge (equalization) is not open to them. Electrical discharge can take place through a conducting medium-- one capable of carrying current. It turns out that most of the cosmos--plasmas--is rather highly conducting, and this means that strong electrical fields rarely occur. The absence of magnetic currents has the consequence that there are many manifestations throughout the cosmos of magnetic fields and associated phenomena. In this lecture, we shall explore some of them.

The Earth's Magnetic Field

We saw in Lecture 13, in connection with the discussion of the mass spectrograph, that a charge moving in a magnetic field would experience a force. The force is perpendicular to both the direction of motion of the charge, and the field. In a medium with a high electrical conductivity, this force has the effect of binding the matter to the field, or vice-versa. Depending on the strength of the field, the matter will either be bound to the field, or the field will be transported by the matter.

It is common to describe this situation by saying the magnetic field is frozen into the medium, although this is strictly true only for material motion perpendicular to the field direction. Flow along the field direction is not inhibited. If the energy of motion of the medium is larger than the energy in the field, motions of the medium are capable of moving the field in such a way that it becomes either stronger or weaker.

One useful measure of the strength of a magnetic or electric field is the number of lines of force through a unit area. The field can be doubled by doubling the lines of force through some area. If the field is frozen in a medium, it may happen that the fluid motions are such that the magnetic field is twisted round on itself so that the lines reinforce one another. This is illustrated in Figure 31-1 for a ``loop'' of magnetic field.

(see drawing from lecture)

You can do a simple experiment yourself with a rubber band. First mark arrows along it to indicate a direction for the field. next, form a figure eight, and twist the band back on itself. You will see that the arrows will again line up, but where there was once one strand of the band, there are now two. Just as the strength of the band has doubled, a magnetic field will double in strength if it is pushed in this way by the motions of a conducting fluid medium.

Of course, you can also simply take the two ends of the rubber band and stretch them, until the arrows going one way are very close to the arrows going the other. In this case, there would be no net flux, and the field would essentially have destroyed itself.

We believe that motions in the liquid core of the earth act in the former of these two ways, to build and maintain the earth's field. The overall mechanism of building a field is called the dynamo mechanism--the same name that is used to describe a generator of electrical current.

We do not know in detail how the earth's dynamo works, nor are we sure why the field is built up rather than weakened, as in one version of our experiment. One possible explanation postulates that it is natural for energy to distribute itself among its possible manifestations. This means for mechanical energy, that kinetic and potential energy would tend toward equality. For systems obeying Newton's laws, this is generally the case.

If the same principle applies to the magnetic energy, and a conducting fluid, with the field frozen in. Then if there were originally strong fluid motions, and weak magnetic fields, the motions would be such as to strengthen the fields. When the energy in the field and in the fluid motions were comparable, the field strengths would cease to grow.

Planetary Magnetic Fields

The strength of the earth's magnetic field varies somewhat, with position on the surface. Roughly, it is one to ten thousand times less than a strong bar magnet, ample for the operation of a compass. There are two relevant ways to give the relative strengths of planetary magnetism. Table 31-1 gives magnetic field strengths at the surfaces of the planets relative to the earth's surface field. Also given is a measure of the strength of the magnets that would simulate the planetary fields if they were placed within the planets. (This ``strength'' is called the magnetic moment.) The two columns of the table differ by the cube of the radii of the planets.

Table 30-1 Planetary Magnetic Fields


            All Field Measures Relative to Earth

          Mean Surface    Magnetic Strength   Angle of pole to Axis
              Field        (dipole moment)    of Rotation (degrees)

mercury        0.011        6 E-4*                ?
venus         <0.001       <8 E-4
earth          1.0          1 E0                  12
mars          <0.002       <3 E-4
Jupiter       13.8          2 E+4                -10#
Saturn         0.6          5 E+2                  0
Uranus         0.7          4 E+1                 59
Neptune        0.4          2 E+1                -47

* The number following `E' is the power of ten that
should be applied.  Thus 6 E-4 means six times ten
to the power minus four.
#This indicates the NS direction of the dipole is opposite
that of the earth.

Roughly speaking, the fields of the equivalent magnets for the planets are shaped about like the fields of a bar magnet. However, important modifications from this shape occur as a result of pressure from the solar wind, which squeezes the field on one side of the planet, and causes a kind of tail on the other.

Fast, charged particles from solar storms or flares travel outward through the interplanetary medium. Some of these particles will interact with planetary magnetic fields. The nature of this interaction is basically the same as that discussed in Lecture 13. The charges experience a force that is perpendicular both to the field, and to their velocity. This makes them spiral around the field lines.

The volume of space where a planet's magnetic field significantly influences the motion of charged particles is called the magnetosphere. This is a simplified concept, since there is no sharp boundary between regions of space where the trajectory of a charge is influenced and one where it is not. Generally speaking, particles which enter a planetary magnetosphere move along the lines of force. These lines of force typically end on the planetary surfaces near the magnetic poles.

We see the effects of these charged particles when they collide with atoms in the earth's atmosphere. The atoms are excited by these collisions--they absorb energy. This excess energy is radiated away in the form of photons, which may be seen from the ground as aurorae. The most common auroral lines are red or green, and are due to transitions in the neutral oxygen atom.

The magnetic fields of the Jovian planets give rise to aurora that have been seen on all four of them. Important work on these phenomena have been done at the U of M by John Clarke and his colleagues at the Department of Atmospheric and Oceanic Sciences. They have obtained impressive images using the Hubble Space Telescope. (Type "aurorae" into a search engine.)

One of the more spectacular interactions takes place between Jupiter, and the inner Galilean satellite Io. As the satellite moves through the ionized medium between it and Jupiter, electrical currents are generated that follow the magnetic field lines between planet and satellite. The effects of these currents are sometimes manifested as aurorae on Jupiter. Currents were actually measured by the 1979 space probe Voyager I.

Magnetic Signature of Rocky Materials

Only a few minerals are strongly magnetic. Run a magnet through beach sand, and you will quickly find that most of the black grains are strongly magnetic. These grains are mostly magnetite, Fe3O4 (cf. Lecture 23). But you will also find that other grains are attracted to the magnet.

The inherent magnetism of a solid like a rock or mineral depends (interestingly enough) on its history. If the material contains atoms, like iron or nickel that have partially filled inner electron shells, then each atom can act like a little magnet. The magnetism arises ultimately from the spins of the electrons. A grain of material will act like a magnet if the individual atomic spins have some net alignment.

In the case of materials like magnetite, the atoms will align naturally, forming a natural magnet. Other materials can exhibit what is called induced magnetism by placing them near a strong magnet.

Materials will lose their magnetism if they are heated to a high enough temperature. The critical temperature is called the Curie temperature, or Curie point. In general, the Curie temperature is lower than the melting point, so that any igneous material may safely be assumed to have passed through a state of no magnetization (liquid) to one where magnetization is possible.

What will be the direction of the field of magnetic igneous materials? Generally speaking, those materials will preserve the field direction of the earth at the time the temperature dropped below the Curie temperature. This provides the geologist with an interesting way of determining the history of planetary magnetic fields.

We can read the history of the earth's field near the mid atlantic ridge, where ocean floor is constantly being created, and pushed east and west. We discovered in the 1960's that the magnetic pattern was striped, with the average polarity of the rocks of the ocean floor alternating with distance from the ridge. This observation actually cinched the theory of plate tectonics, as well as providing a detailed record of the history of the earth's polar reversals.

For reasons that are unclear, the earth's magnetic pole reverses direction on a time scale of the order of half a million years--short. We do not know what causes these reversals, but the evidence for them is very clear, and can be read in the general geological column (cf. Grand Canyon Layers) as well as in sea floor spreading.

We have yet to exploit the magnetism of planetary materials beyond a few cases. Lunar rocks show evidence of a slight magnetism, whose origin is puzzling, since the Moon now has no general field. Did it have on in the past, or were these rocks somehow magnetized by the solar wind?

Cosmic Magnetic Fields

We have discussed the magnetic field of the sun (Lecture 6). The origin of this field is not entirely certain. The original field may have come from the interstellar cloud from which the sun formed. We now think this field is strengthened and maintained by a dynamo mechanism that is driven by the differential rotation of the sun-- low latitudes rotate more rapidly than the higher latitudes.

It is now known that magnetic fields exist throughout interstellar space. Indeed, our Galaxy is threaded by such fields. Even though typical interstellar fields are quite weak, they play an important role in many cosmic phenomena.

The Mystery of Cosmic Magnetic Fields

The famous equations of Maxwell tell us that magnetic fields may be generated by electrical currents. At the present time, we do not know the nature of the currents that generated most of the cosmic magnetic fields. We have discussed a simple dynamo mechanism that will build up a magnetic field that is threaded through a conducting medium. It was necessary for us to postulate that the medium moved in the proper way to strengthen the field. We could give only a rather general, plausibility argument why this might happen.

The term fossil field is often used to describe a field whose origin is uncertain. The notion of a fossil field is often used when one is willing to accept the existence of a field due to unknown causes. Given the presence of a field, it is possible to work out some of its consequences. For example, given the general magnetic field of the galaxy, we may investigate the effect on star formation in giant molecular clouds.

In principle, the answers to many of our questions lie in Maxwell's equations. Unfortunately, it is not possible to give a general solution to these equations--the kind of solution that is available for the two-body problem, or the falling body problem. We can only make models, and use the equations to try to follow their subsequent behavior. Computers are only now getting the necessary speed and memory to attack such problems.

Summary

Maxwell's equations tell us that magnetic fields may be generated by electrical currents. The earth's magnetism is manifested in a large variety of ways from compass deflections to aurorae to the magnetic stripes of the mid-Atlantic ridge. We think the earth's field is maintained by a dynamo mechanism. We understand in principle how it might work, but we do not know the detailed mechanism. Magnetic fields are frozen in highly conducting media, and those motions are capable of generating or strengthening the fields. The Jovian planets all have magnetic fields, and have aurorae. These are caused by the interaction of particles from the sun with the local magnetospheres.

There are many manifestations of a general interstellar magnetic field in our Galaxy. It acts to constrain the cosmic rays, and provides support through its magnetic pressure to giant molecular clouds. Some stars, and especially white dwarfs and pulsars have very strong magnetic fields.

Lecture 32 - Jupiter and Its Satellites

Jupiter is surely the dominant planet of the solar system. It's mass is nearly three times that of its nearest rival, Saturn, and more than 300 times that of the earth. If we take Neptune as marking the edge of the planetary system, then Jupiter is only 20% of the way out to it. However, in terms of the mass in the solar nebula, Jupiter could have been closer to the halfway point. By this, we mean that the mass in the original solar nebula within 5.2 AU could have been half of the total mass of the nebula. This is because we think the density of the nebular gas would have been higher closer to the sun.

We do not know the conditions in the early solar system, of course. We make models, and follow their evolution with computers. Some of these models show considerable migration of the planetesimals from the inner to the outer solar system. Perhaps Jupiter was once much closer to the sun than it now is. We shall see in Lecture 37 that some extrasolar planetary systems have giant planets closer to the central stars than we ever thought would be possible.

We know a good deal about today's Jupiter. We discussed the Hubble Space Telescope's observations of the Jovian aurorae in Lecture 30. The HST and space probes, especially Galileo, have provided detailed pictures of the complex patterns of flow that have developed in Jupiter's atmosphere. The simplest aspects of this flow have been known for some time. Energy from Jupiter's interior is transported by convective currents. Jupiter is rotating quite rapidly, once about every 10 hours, and this rotation distorts the convection cells into a system of belts and zones.

Belts and Zones

The Galileo probe confirmed that the atmosphere is undergoing convection, the transport of hot gas from below to cooler layers above it. In Jupiter, this gives rise to the bright zone, dark belt structure. Idealized structure is indicated below. We'll only show the southern hemisphere. Nomenclature for the north is the same, only with the word South replaced by North. Assume the surface rotates from left to right --->

Figure 32 - 1: Jupiter's Zones and Belts


--------->-------->----------->-----------------x
Equatorial zone                                  x
-------->--------->----------->---------------- x
South equatorial belt                           x
--------<---------<-----------<----------------x
South tropical zone  (Great Red Spot)          x
-------->--------->----------->---------------x
South temperate belt                         x
--------<---------<-----------<-------------x
South temperate zone                       x
-------->--------->----------->----------x
South South temperate belt              x
--------<---------<-----------<------- x
South South temperate zone           x
                                   x
                                x
                              x
                            x
South Polar              x
Region                x
                  x
           x
x
 

The bright zones represent hot material rising, and the dark belts, material sinking. Let us see how this gives rise to a flow pattern. The concept we need for this is just that the matter near the equator is rotating about Jupiter's axis faster than that at points between the equator and the poles. At the poles, there is no velocity due to the rotation, of course. For any intermediate latitude, the velocity of rotation is (2R)/(Period of Rotation), where R is the distance from the axis of rotation to the surface of the planet--something like the dashed lines in the figure above.

Rising material from the equatorial zone that moves north is going faster than that the surface under it. A similar thing holds for rising material that moves south. So the flow arrows on both the north and south edges of the Equatorial Zone point to the right (which is the assumed direction of rotation of the planet).

Material flowing from the South tropical zone into the South equatorial belt (thence down) is going more slowly than the material of the belt, because the belt is closer to the equator. Hence the flow pattern is to the left. Material from the South tropical zone that flows (south) into the South temperate belt (thence down) is deviated to the right, because the South tropical zone is going faster than the South temperate belt (to its south).

Check the lectures for a diagram. If there is time, I'll put one up here.

Figure 32 - 2: Distortion of the Cyclone Pattern by Rotation

(see lecture for drawing)

The envelopes of Saturn, Uranus, and Neptune resemble that of Jupiter, though not in detail. We will not describe them separately.

The Galilean Satellites

The heat output from Jupiter may have influenced the densities of the bodies that formed near it, in particular, the Galilean satellites. Here are some figures. The ages of the surfaces are determined from crater densities.

Table 31 - 1: Galilean Satellites


Object                   Io      Europa     Ganymede    Callisto     Mercury
Density (water=1)        3.5      3.3        1.9         1.8          5.4
Composition              rock    rock      rock+ice     rock+ice    Metal+rock
Diameter(km)             3630    3128       5262         4800         4878

                              8         8          9         9            9
Age of surface (yr)   few x 10  few x 10   3.5 x 10    4 x 10       4 x 10

Of the Galilean satellites the innermost Io is still active, with volcanos and lava flows. Most of the craters on its surface have been covered. The energy that drives this activity comes from tidal interaction between Io and Jupiter. While Io is some 5.8 Jovian radii away, it is only slightly over twice the Roche limit (see Lecture 31), and is therefore subject to strong tidal forces. We think that the orbit may have been more eccentric in the past, and this would have enhanced the tidal heating.

Europa is not now active, but must have been so in the not too distant past. Its crater density is quite low. Recent images of the surface of Europa suggest that there have been extensive outflows of water from its interior. It is possible that some tidal heating has been sufficient to maintain a zone of liquid water within the body. We shall soon discuss why the presence of liquid water always raises the possibility of life forms.

By contrast with Io and Europa, Ganymede and Callisto are heavily cratered.

Possibly the heat from Jupiter, during the formation of the satellites accounts for the density decrease, from inner to outer. Certainly, this mimics the densities of the planets themselves. This is still an open question.

It is entirely possible that the tidal heating drove off an icy fraction of Io, either when it was forming or afterward. Perhaps to a lesser extent, the same thing happened with Europa. This might account for the high densities of these inner satellites, without invoking the heat from the central body. In much the same way, some people have attempted to account for the density gradient of the planets by ad hoc (special) causes. Mercury, they say, is dense not because the rocky component could not form at its position, but because it was blasted away, after the core had formed.

Figure 32-3: Terrestrial Planets and Large Moons

The Jovian planets have lots of satellites. The 1998 Astronomical Almanac lists 16 for Jupiter. We have discussed the major, Galilean satellites of Jupiter. These have radii of some 1500 to 2400 km. Thus, they are relatively large bodies, ranging in size from slightly smaller than our own Moon to about 50% larger.

The remaining Jovian satellites are all an order of magnitude (or more) smaller. Four of these small satellites are nearer to Jupiter than Io. Their orbits, like the Galilean satellites, are closely confined to Jupiter's equatorial plane, and their eccentricities are very small. Generally speaking, interactions of accreting satellites, collisions among the fragments, are expected to circularize the orbits and to confine them to the equatorial plane of the planet. This happened not only for the Galilean satellites, but for four smaller inner satellites.

Outside of Callisto, the outermost Galilean satellite, we could group the satellites into two categories. There is a group of four, some ten times the distance of Callisto from Jupiter. The orbits of these four show a definite ``relaxation'' of the close coupling to Jupiter. They are somewhat eccentric, and somewhat inclined. Another set of four satellites is about 20 times further than Callisto. All four have retrograde orbits--they revolve opposite the direction of the inner satellites. Speculations are that the outer four satellites, and perhaps the outer eight were captured by Jupiter.

Satellites of Saturn, Uranus, and Neptune

Saturn has 18 satellites listed in the Astronomical Almanac. Only one of them, Titan, is comparable in size to the Galilean satellites of Jupiter. This satellite has an atmosphere, mostly of nitrogen gas (N2),with small admixtures of argon, methane, and hydrogen. Only one of the 18 satellites orbits retrograde, but some of the outer satellites show modest eccentricities and orbital inclination. They have either been perturbed or perhaps captured.

Uranus is the fascinating case, because its axis of rotation is tipped more than 90o to the pole of the ecliptic. We could say the planet rotates in a retrograde sense, but the angle of the polar axis to the plane of the ecliptic is only about 8 o. The rotation axis is almost in the plane of the orbit. We call upon the planetary deus ex machina, the collision, to account for this strange rotation, just as we must in the case of venus. Somehow, 15 satellites knew to line up in the tilted equatorial plane of the planet. Presumably this means the satellites formed after the putative collision that affected the rotation.

Neptune has eight satellites. The two outermost ones have clearly disturbed orbits. The outermost, Triton is about 20% smaller than our Moon. It's retrograde orbit was once the basis for speculation that Pluto was an escaped satellite of Neptune. The two objects are of comparable size. Nereid is about one tenth the size of our Moon. It's orbit has the highest eccentricity of all the satellites, and its semimajor axis is larger by more than 10 than that of Triton. Surely these satellites were either captured, or their orbits were severely perturbed by some large planetesimal.

Summary

Energy from the interiors of the Jovian planets is carried by convection. At the surfaces of the planets, convection currents manifest themselves in bright zones of rising material, and dark belts of sinking material.

The Galilean satellites of Jupiter show a density decrease that might be due to heat from the planet, though this is not certain. Io is currently active, and Europa may have been active recently. It may also still have liquid water below its surface. Ganymede and Callisto have very old surfaces.

The satellites of Jupiter and Saturn fall into two categories, inner and outer. The inner ones all have very low eccentricities, and orbital inclinations. The outer ones have larger eccentricities as well as inclinations. In addition, the outer ones may orbit in a retrograde fashion. The outer satellites may be captured asteroids.

Satellites of Uranus are generally constrained to the plane of it's tipped equator. Their orbits are generally ``well behaved,'' that is small inclinations and eccentricities. Neptune has four well-behaved inner satellites and two outer ones with highly perturbed orbits.

Lecture 33 - Planetary Rings and Satellites

Until the space age, the only planet that we knew had rings was Saturn. The rings were poorly, though definitely seen in Galileo's crude telescopes. These rings can be clearly seen in good quality telescopes of small aperture, but the images are small and often disappointing. Larger telescopes, say 30 inches or more in aperture, provide quite breathtaking views of Saturn, but these telescopes are usually available to the public only at the times of ``open houses'' at observatories. Beautiful telescopic photographs of Saturn, many in color, appear in popular and textbooks on astronomy. The first Saturn fly-by's, the Voyagers (1979), returned many images of Saturn's rings as well as some of its smaller satellites. Many new aspects of the ring system were seen, and some presented interesting challenges to the astronomer to account for them.

Jupiter, Uranus, and Neptune, all have ring systems. Jupiter's ring was discovered by the Voyager mission. The rings of Uranus and Neptune were discovered by ground observations, but confirmed by the Voyagers.

Tidal and Chemical Forces: the Roche Limit

The nineteenth-century French astronomer Edouard Roche (1820-1883) provided a general explanation for ring systems. He was able to show that small bodies within a certain distance from a planet would not be able to grow in size because of tidal forces. The critical distance is now known as the Roche limit.

Consider the following question. Suppose there are two spherical masses 'm', in contact with one another, and lined up with the center of a much larger mass 'M'. Let the distance of the first little mass from the center of the large one be 'd', and let the radii of the little masses be 's'.

What will the distance d be such that the difference in the gravitational force on the inner and outer small mass is equal to the gravitational attraction of the two masses to one another?

Figure 33-1: The Roche Limit


.(see figure from class)
.
.
.
.
.
.
.
.

We can write this out:


Force     Force of planet  Force of planet
between   on closest       on outer
2 masses  little mass      little mass
  |          |                |

Gmm         GMm              GMm
-----  =   ------     -     -------         (1)
    2         2                    2
(2s)         d               (d+2s)

The difference in the forces on the right-hand side of (1) is a tidal force. Some algebraic manipulation, along with approximations when 2s << d show that the tidal force depends on the inverse cube of d. Thus the tidal forces increase rapidly when d gets smaller. Equation (1) allows us to find the border where the tidal force and gravitation are just in balance.

Let us assume that the densities () of the large and small masses are the same. Then it follows from (1) that the tidal and gravitational forces are in balance when


 d
--- = 2.520...              (2)
 R
where R is the radius of the mass M. The distance 2.5R is known as the Roche limit. If the distance is less than 2.5R, tidal forces would not allow a body to be held together by gravitation.

None of the major satellites revolve around their parent planets within this Roche limit, but all of the ring systems do. We therefore conclude the rings did not form because of tidal forces.

Tidal forces do not tear the Space Shuttle apart--why? The answer is that the Roche limit only compares tidal forces to gravitational ones. Gravitational forces are much weaker than chemical forces which are responsible for the shapes of objects we are familiar with. On the other hand, if we compare chemical and gravitational forces for large objects, the story is different.

Consider a square block of matter that we can conceptually divide in half. If each side of the block is s, say, then one half of the block will attract the other gravitationally with a force that is equal to Gm2/s2. There is a factor , that depends on the geometry. If we took two balls in contact, then would be a little different than for blocks. We will just set = 1, here. It is of order unity, and it's exact value will not affect our argument.


                  --------------------------      Cross section
                  |           .            |      of a block of
                  |           .            |      material.  Dots
                  |   m/2     .    m/2     |      represent a plane
                  |     x<-- 0.5s-->x      |      dividing one half
                  |           .            |      of the block from
                  |           .            |      the other.
                  |           .            |
                  --------------------------
                  <------------s------------>

What are the chemical forces that hold one half of the block to the other? There will be some chemical force per unit area that depends on the nature of the substance. The total chemical force will be related to the number of atoms or ions on the left surface that attract those on the right surface of the contact plain, indicated by the vertical dots. The total chemical force holding the halves of the block together will depend on the area. That area depends on s2. We can actually make a rough estimate of how strong these forces are, because we know the strengths of typical chemical bonds. Let's save this exercise for the moment.

The masses m/2 depend on s^3 and a density. This means the gravitational force will depend on s6/s2 or s4. The ratio of the gravitational to the chemical forces in any body thus depend on s4/s2 or just s2. So the bigger the body, the more important the gravitational force.

Remember the four forces of nature. Gravitational ones are the weakest. Electromagnetic forces, which are basically what chemical forces are, are much stronger. But for bigger and bigger bodies, there will come a time when gravitational forces are more important for the body as a whole. Locally, chemical forces will always be more important.

Bodies in the rings are thought to be no larger than some tens of meters across. So this must be about where gravitational and chemical forces are in balance for the material in these ring systems. There is a distribution of sizes from the maximum (tens of meters) down to dust.

A question that is related to the concepts of the Roche limit is the following: A at what point gravitational forces make cosmical bodies round. That turns out to be a much larger size, of the order of 10 to 100 km or so. Small satellites and asteroids are lumpy, large ones are round. After the mass of an object is sufficiently large, the gravitational forces simply crush the material into a spherical form. The reason for this is that the sphere is the geometrical figure that contains the greatest volume inside the smallest surface area.

Resonances and the Ring System of Saturn

Figure 33 - 2: Saturn's Largest Rings

The Voyager space probes revealed a more complicated ring system than had been seen from the ground. Photographs from earthbound telescopes typically show two rings, called A and B. The gap between the A and B ring is named after the astronomer Cassini, and is called Cassini's division. The A ring is itself divided by a less prominent gap known as Encke's division.

Cassini and Encke were both important figures in the history of astronomy. Cassini (1625-1712) was a contemporary of Newton. Perhaps his most important contribution to astronomy was the determination of the distance to Mars, which was for many years the basis of the astronomical unit (Lecture 4). Encke was nineteenth century, and may best be known for the discovery of a bright comet that bears his name.

Inside of the B ring, there is a C ring (dark in Figure 32-2) and a D ring. These, and outer rings are not usually seen from the earth, but they do show up on Voyager images. Indeed, the Voyager images show so much structure within the rings, and even in Cassini's division, that the traditional names are not particularly descriptive. These images also showed markings or shadows on the rings themselves. Some of these features have been described as ``spokes.'' The origin of these features is currently being studied. According to some theories, the features arise from gravitational forces, such as those that cause spiral arms in galaxies. It has also been speculated that they may be electromagnetic in origin, somehow related to Saturn's magnetic field.

It has been known for a long time that particles that would orbit in the Cassini division, would have a period that is very closely half that of the inner satellite Mimas. Therefore, it has been argued that this resonance is responsible for the gap. This explanation is still in the textbooks, and must therefore represent our best guess! Recall (Lecture 27) that the 1:2 resonance in the asteroid belt is a gap, although the 2:3 is not.

The Voyagers showed unanticipated interactions of the Saturnian satellites with rings. A newly discovered ring outside of the A ring seems to be maintained by what are now called shepherd satellites. One of these satellites revolves just inside the F ring, and another just outside it. The effect of gravitational pulls from these two satellites maintains the ring as a narrow feature. See if you can understand how this might work with the help of the description in the following paragraph.

A shepherding satellite of an entirely different nature is thought to be responsible for the Encke division. In this case, the satellite revolves within the division itself. The effect is not due to a geometrical sweeping up of material, although this may play some role. Particles just inside the Encke division move faster than the satellite. They therefore give up energy to it. This causes them to move toward Saturn, while pushing the satellite away. Particles just outside of the division have the opposite effect. They will gain energy from the satellite and move away from Saturn. The satellite, which will have lost energy will move toward Saturn. This process has apparently reached a virtually stationary state.

The Equatorial Plane

The earth, Mars, and the Jovian planets are flattened as a result of their rotation. This flattening causes the gravitational field in which their satellites revolve to depart slightly from that about a point particle with the planet's mass. The classical two-body problem, with a solution that is valid for all time, applies to two point masses. It would also apply to two perfect spheres, since it can be shown that they attract one another as if all of the mass were concentrated at a point at the center.

If departures from the ideal conditions are small, they are said to give rise to perturbations. We have discussed them before. A perturbed orbit will precess. Consider an orbit with an inclination i to some planet's equatorial plane. It will cross the plane at two points, called the nodes of the orbit. As a result of perturbations, the line connecting the nodes will rotate in space. It is straightforward to visualize that orbits of individual particles will collide with one another, and that these collisions can continue as long as these particles have different inclinations to the equatorial plane.

Figure 33 - 3: The Equatorial Plane is Favored

Consider a family of circular orbits, with different inclinations to the equatorial plane of some planet. If all of these orbits have the same semimajor axis, then it is straightforward to show that the ones orbiting in the plane have the greatest (negative) potential energy. This arises simply because they are closer to the equatorial bulge, and get a stronger gravitational tug.

We see that the equatorial plane is favored because (1) collisions between particles are reduced and (2) the orbits are energetically favored.

All of the ring systems are in the equatorial planes of their parent planets, and the inner Jovian satellites also are constrained in this manner. This is true even for Uranus, whose rotational axis is nearly in its orbital plane. The more distant a satellite orbit, the less likely it is to orbit in the equatorial plane. This is entirely reasonable, since the forces responsible for this constraint, diminish with distance.

Orbital eccentricities are subject to changes that will usually drive them toward zero. Consider a small particle with semimajor axis `a' and eccentricity `e' that is orbiting within the Roche limit. Its distance from the central planet would change from a(1-e) at perigee to a(1+e) at apogee. In other words, it would oscillate, in and out, and in the course of doing this, it would move through the orbital positions of other such particles. There would thus be collisions which would diminish if all of the particles achieved circular orbits.

The forces we have described seem to have influenced not only planetary rings, but also inner satellites, and perhaps also the orbits of the planets about the sun. Orbits with large inclinations and eccentricities are usually attributed to ad hoc or special causes, the favorite being interplanetary collisions.

Summary

Small satellites are incapable of growing within the Roche limit, because of tidal forces from the central planet. Satellites will grow as smaller bodies stick together because of chemical forces. These become less important relative to gravitational forces--tidal forces are gravitational--as a body grows. All of the Jovian planets have rings, and they are all within the Roche limit. Saturn's rings are the most developed. Of these the most prominent are an outer A and an inner B ring. Cassini's division, which separates them, is thought to be due to a 1:2 resonance with an inner satellite. The A ring has a gap known as Encke's division. It is maintained by a small shepherding satellite. Shepherding satellites also maintain a thin F ring outside the A ring.

The equatorial planes of rotating planets are favored for orbits of satellites. Orbits with low inclination are also favored. Such orbits are the rule for ring particles as well as for inner satellites. Satellites with large orbital inclinations and eccentricities are likely to be captured or severely perturbed.

Lecture 34: Saturn, Uranus, and Neptune; Pluto and Charon

Distant Worlds

Jupiter is about 3.5 times farther from the sun than Mars. According to Bode's law, the distance to the "next" planet nearly doubles, so that Saturn is about twice the distance from the sun as Jupiter, and Uranus is twice the distance from sun as Saturn. The Bode law breaks down at Neptune, which is only about 1.5 times as far from the sun as Uranus. All told, these are distant bodies, on the scale of the inner solar system, and their exploration has only begun.

Most of our ignorance of the outer solar system comes from space probes launched in the 1970's.

Space Missions to the Outer Planets

The earliest missions to the Jovian planets were Pioneer and Voyager fly-by's. Both were "double" missions, in the sense that they consisted of two space probes, launched separately. The first of these missions, Pioneer 10, flew by Jupiter in 1973, and continued on a path that will take it some 75 AU from the sun in the year 2000. Pioneer 11 flew by Jupiter in 1974, and Saturn in 1979.

The two Voyager missions reached Jupiter in 1979, and Saturn in 1980 and 1981. Voyager 1 was then directed out of the solar system, while the Voyager 2 mission continued for fly-by's of Uranus (1986) and Neptune (1989).

The early missions to Jupiter have been largely superseded by the Galileo orbiter, which arrived in 1995, releasing a probe which entered the giant planet's atmosphere. The orbiter continues to function, and in spite of a crippled antenna has returned a wealth of new information.

Beyond Jupiter, we still rely primarily on the information gained from the Pioneer and Voyager missions, and from some old-fashioned "look-see" astronomy by the Hubble Space Telescope Telescope. This situation will be remedied when the Cassini mission reaches Saturn in 2004. This mission is on schedule, and has received the first of several energy boosts in a fly-by of Venus in April of 1998.

Resume

We have already discussed many of the important aspects of the three large planets beyond Jupiter. For example, we have discussed the rings and satellites of Saturn, and the concept of the Roche limit, within which only small objects can exist as a result of tidal forces from the main body. Roughly, if a body is small enough to be lopsided, it is small in this sense. Can you say why?

In many ways, the Jovian planets are scale models of one another. Their compositions are close to that of the SAD. There is no solid surface, and the atmospheres are dominated by molecular hydrogen gas. This gas is thought to be in convective equilibrium, bringing heat from the inner parts of the planet to the atmospheres, where the heat is radiated away into space.

Although the material within the interiors of the Jovian planets is often described as liquid, it would be better to call it a fluid, a term that may designate either a gas or a liquid. In fact, it is almost certain that the material in these planets never actually liquefies, as in the condensation of water vapor to liquid water, where a distinct phase boundary occurs. The temperatures are always too high for this to happen. But the densities within the fluids become comparable to those of many liquids, and the properties of the high-pressure fluids emulate liquids. It is common, therefore, for writers to refer to material in the interiors of the Jovian planets as liquids. This is probably done sometimes in ignorance, and sometimes with the license of an expert to use a common term in a specialized way.

As the pressures in the interiors of these planets becomes high enough, first one, and potentially the second electron of the molecular H2 become able to move more or less independently of the protons. The fluid then becomes an efficient conductor of electricity or heat, and is often called ``metallic.'' Current models of Jupiter and Saturn attain this metallic phase, but Uranus and Neptune do not.

The Jovian planets, with the possible exception of Uranus, radiate more energy into space than they receive from the sun; We discussed this in detail in Lecture 29.

The densities of Uranus and Neptune are somewhat higher than those of Jupiter and Saturn, and this is consistent with our picture that the former objects have relatively larger rocky cores. The relatively larger rocky cores are not the result of more rock being available further from the sun, but of the depletion of the hydrogen gas toward the outskirts of the solar nebula. Presumably there was less mass available further out to collapse gravitationally on the rocky cores of Uranus and Neptune.

Figure 33-1: Jovian Planets

We discussed the peculiar orientation of the axis of rotation of Uranus, which is nearly in its orbital plane. A major impact, early in the history of the formation of the planet is the current consensus among planetary scientists as an explanation for this.

Astronomical Discovery

Until the end of the 1700's only seven planets were known. Herschel's discovery of Uranus, in 1781 led, eventually to the discoveries of Neptune (1846) and Pluto (1930). In our own time, minor bodies are being discovered beyond the orbits of Pluto and Neptune, while major planets have been found orbiting nearby stars. These have all been major discoveries and all have interesting histories.

William Herschel, the discoverer of Uranus, was a typical example of an amateur astronomer who turned professional. There are many examples, including The U of Michigan's own Professor Robert McMath. He was a successful engineer who built a major solar observatory, founded the Kitt Peak National Observatory, and eventually became president of the American Astronomical Society.

Herschell's success as a performer and teacher of music gave him the opportunity to indulge in his hobby. He made his own reflecting telescopes and used them to make a variety of major discoveries, some of which we have described elsewhere (cf. Lecture 8). Naturally, the discovery of the first planet of historic times was a major event. Herschell's fame and future were secured, and he was able to devote full time to his hobby. But we must leave the pursuit of this remarkable figure to those with an interest in the history of astronomy.

From the time of the discovery of Uranus, astronomy of position, and the mechanics of the solar system were in full bloom (Lecture 9). Most of the famous mathematicians of the eighteenth and early nineteenth century, made significant contributions to the theoretical astronomy of their day. Euler, Laplace, Lagrange, Jacobi, and Gauss, refined and extended the Newtonian edifice to such a high degree that it became possible to speculate that the future of our world might be predicted for all time on the basis of Newtonian principles (see the discussion on reductionism in Lecture 3).

By the middle of the century, enough evidence had accumulated to show that the orbit of Uranus was incompatible with models of the solar system as it was then known. Two brilliant mathematical astronomers, one in England and one in France, were able to reconcile the observations of Uranus by postulating the distortion of its orbit by an additional planet, still further from the sun.

The story of the discovery of Neptune is one of the favorites in the history of astronomy. The British theoretician, John Couch Adams, was a young student, making his first venture on the scientific scene. He had no access to a major observatory, and in any case, was theoretically inclined. Adams wanted to turn the search for the predicted planet over to an individual with observational talents and experience. To that end, he sought help from a major figure in British astronomy. Alas, that great man moved slowly to implement the requests of the youth, and in the end, they were scooped by the team on the continent of Europe.

The European team was led by the French theoretician Urbain Leverrier. Unlike Adams, Leverrier was an established figure, and had published the results of his calculations. Nevertheless, French observers had not rushed to follow them up. Eventually Leverrier wrote to colleague in Berlin, who had just completed a new set of star maps. We quote a translation of letter:

Direct your telescope to a point on the ecliptic in the constellation of Aquarius, in longitude 326o, and you will find within a degree of that place a new planet, looking like a star of about ninth magnitude, and having a perceptible disk.

On 23 September, 1846, the German astronomer found Neptune within a degree of the predicted position--after only a half hour's search. In the meantime, Adams, whose prediction was made somewhat earlier than Leverrier's, had finally enlisted the help of an astronomer at the Royal Greenwich Observatory. That notable, however, had decided to make maps of the sky over several nights, and only then made a comparison with existing maps. It turns out he had actually observed Neptune prior to his rival, but because of his cautious technique, failed to realize it.

National rivalries were as intense in the mid-1800's as they are now, and there was much second guessing over the failure of the British team to follow up on its early lead. In the end, Adams and Leverrier are recognized as codiscoverers; both had long and distinguished careers.

Neptune's discovery was a triumph for Newtonian theory and for science. It is one thing to be able to account for known phenomena, or even to predict recurrant phenomena such as eclipses. Ancient astronomers were able to do such things. But the prediction of a new planet from perturbations of the orbit of another was well beyond the scope of a system of careful extrapolation such as that of Ptolemy.

The American astronomer Percival Lowell is know primarily for his obsessive preoccupation with the planet Mars. But he was also keenly interested in repeating the triumphs of Adams and Leverrier by finding a trans-Neptunian planet. Irregularities in the orbits of Uranus and Neptune could, in principle, be made to yield yet another member of the solar system, which Lowell called "Planet-X." Lowell made extensive calculations of the position of the new planet, but the search was primarily carried out in the years after his death. The discovery was actually made in 1930 by the American astronomer Clyde Tombaugh, working at the Lowell Observatory in Flagstaff, Arizona.

It is ironical that both the discoveries of Neptune and Pluto have been questioned as serendipitous. While both planets were found near their predicted positions, separate workers have questioned the rigor of the predictions. History has generally concluded in favor of Adams and Leverrier, but against the hapless Lowell. At least some of the orbital irregularities accepted by Lowell turned out to be due to observational errors. Probably because of this, Lowell's Planet-X was predicted to have a mass six times that of the Earth. We now know Pluto is only about 0.002 Earth masses.

Triton, Pluto and Charon

Over the years there has been much speculation over the origin of Pluto. Quite early on, it was known that it was not simply a more distant version of the Jovian planets. This was clear from the fact that the orbit was significantly inclined to the ecliptic (17.14 degrees), and was eccentric (e = 0.249). A large inclination or eccentricity is a good indication that an object is out of a "natural position." This could be for a variety of reasons.

One popular idea is that Pluto was an escaped satellite of Neptune. This idea arises from the fact that Pluto's orbit apparently intersects that of Neptune. In the 1980's and 90's Pluto was actually closer to the sun than Neptune, and only in March of 1999 has it become the most distant planet again. If Pluto had actually been a satellite of Neptune, then it should be possible to trace its position back in time, and find some epoch when these two planets were very close to one another--the orbits should intersect. There was an additional piece of evidence that reenforced this argument.

Only a month after the discovery of Neptune, its major satellite Triton was discovered, and found to be in an irregular or retrograde orbit. Could this irregularity be a consequence of whatever interactions caused Pluto to leave the system? With no positive evidence of the contrary, it all seemed likely. However, the discovery of Pluto's satellite Charon in 1978 this scenario became unlikely. If Pluto had been torn from an orbit about Neptune, one would hardly expect to have taken a companion with it. Moreover, there are no known moons that orbit one another.

It is now common to call attention to the physical distance of Pluto's orbit from that of Neptune. Although projections of the orbits on the ecliptic plane intersect, the real orbits do not, because of the difference in their inclinations.

Over the last decade, numerous small objects have been identified with semi-major axes between that of Neptune comets of the distant Oort cloud (Lecture 34). Many astronomers feel it is reasonable to classify Pluto (and Charon) as a member of this class of objects, and demote it from planetary status. This will probably never actually happen, largely because too many people are sentimental about Pluto as a planet.

Neptune's satellite Triton is actually slightly larger than Pluto, but it is probably quite similar in structure. A recent tabulation gives its density as 2.05 (water = 1), which is indistinguishable from that of Pluto. The same reference, The New Solar System 4th Ed, 1999, gives Pluto a density of "(2.0)"--a little uncertain. Both Pluton and Triton have surfaces with reflectance spectra indicating frozen methane (CH4), a little of which evaporates to provide a thin but detectable atmosphere.

It is entirely plausible that Triton was originally a Kuiper-belt object that was captured by Neptune. It might have been captured, lost, and captured again.

In our discussion of the origin of the Earth's Moon, we said that capture was very improbable, because pure two-body capture was impossible. For capture, some third body would have to take up the excess energy of the infalling Moon to prevent it from returning to "infinity." However, a moon and a major planet do not comprise a two-body system, but what is known as a restricted three-body system. Figure 33-2 shows an example of what we shall call "restricted three-body capture". In this calculation, the moon, or infinitesimal body leaves an orbit about the sun, orbits the planet, and returns to orbit the sun again. This cycle is repeated for the duration of the calculation.

Figure 33-2: Restricted Three-Body Capture

If the planet were surrounded by a system of moons, it is possible that these objects could subtract energy from the newcomer, so that it could be permanently captured. This could certainly have happened in the case of Neptune and Triton. Could Pluto have played a role?

The present orbits of Pluto and Neptune rule out the possibility they were ever close. But if Charon were formed in an impact, Pluto might once have had quite a different orbit.

Figure 33-3: Charon's Orbit

Pluto and Charon orbit one another in a plane more nearly perpendicular to the ecliptic than in it. They are tidally locked, so their periods of rotation and revolution are equal--6.4 days. These periods were determined from observations of the brightness of the system as a function of time. The reflected light from the system changes as the bodies rotate because the surfaces are not uniformly reflecting. Much additional information on the satellite size and orbit was obtained when the bodies eclipsed one another.

The orientation of the Pluto-Charon system is reminiscent of Uranus, and our only explanation for this strange phenomenon is to postulate an impact. Could such an impact have also modified Pluto's original orbit so that it no longer intersects that of Neptune? It seems unlikely that an object of Charon's current mass, about a tenth that of Pluto would have done the trick, but one of equal mass might have. If most of the debris were lost, then the current Pluto-Charon system could have settled into its current configuration. Is this likely? Probably not. If it happened, the irregularity of Neptune's satellites are then a natural consequence of interactions with Pluto.

It is more probable that Triton became a Neptunian satellite by an independent, restricted three-body capture.

Neptune has several "regular", inner satellites that were discovered by the Voyager mission. An outer satellite, Nereid, is in a highly inclined, distant orbit, that could result from perturbations by Triton.

This leaves the Pluto-Charon system to have been formed as the result of an impact, probably of two (large) Kuiper-belt objects. If we say the system belongs to the Kuiper belt, the strangeness of the orbit about the sun--its inclination and eccentricity--is readily understood.

Summary

After the discovery of Uranus in the 1700's an outer planet was predicted from perturbations of its orbit. The celebrated discovery was made independently, and very nearly at the same time, by British and European teams in the mid 1800's. The calculations of the position of Neptune are regarded as sound. Pluto, however, was sought, and eventually discovered, by calculations now regarded as invalid.

The planets beyond Jupiter were investigated by Pioneer and Voyager space probes in the 1970's and 1980's. None of these missions reached the Pluto-Charon system. Our knowledge of this outermost "planetary" system comes from observations with ground-based telescopes and the Hubble Space Telescope. Pluto's satellite orbits nearly perpendicular to the ecliptic plane, and planet and satellite are locked in synchronous rotation. It is possible that the disturbed orbits of the outer satellites of Neptune are the result of interactions with Pluto, whose orbit was then altered by a collision so that it no longer intersects that of Neptune. It is more likely that the Neptunian satellites Triton and Nereid are irregular because of the capture of Triton, and that the Pluto-Charon system formed independently from a collision of Kuiper-belt objects.

Lecture 35: Comets, Their Nature and Origin

Comets must have been known since prehistoric times. It isn't clear how comets fit into the view of those ancients who thought the heavens reflected some kind of perfection. Since the time of Halley, we have known that some comets are periodic, and that their heads obey Newton's laws, while their tails do not. Fast computers, and myriad orbital calculations have brought heightened interest of astronomers in comets. What is the true nature of these objects? How did they originate? What is their history? Above all, how might comets influence our future through a possible collision with the earth?

Cometary Orbits

Do comets belong to the solar system, or do they pervade interstellar space? We can approach this question with the help of the statistics of orbits of comets.

Figure 34 - 1: Statistics of Orbits

Figure 34-1 shows a plot of orbital inclination vs. eccentricity for planets, asteroids, and comets (open circles). Notice that the planets generally have small eccentricities and inclinations. Most of the comets have eccentricities near unity. Unit eccentricity strictly means the orbit is not an ellipse at all, but an open curve, the parabola. Likewise, if the eccentricity is greater than unity, the orbit is hyperbolic.

Consult Figure 10-3 to help interpret this plot. Low eccentricities mean a small change in the separation, r, from the sun to the orbiting body. As the eccentricity increases, the body oscillates (in r) from a minimum (perihelion) to a maximum (aphelion). For these cases, the effective potential energy (EPE) is always negative. At the points where r changes direction, the r-component of the velocity is momentarily zero, and here EPE equals the total energy, which is also negative.

An eccentricity of exactly unity occurs when the total energy is exactly zero. For such an energy, a body can come in from infinity, reach some minimum distance from the sun, and then go out to infinity again. In this case, as r approaches infinity, the velocity becomes arbitrarily small. Crudely, we can say the velocity is zero ``at'' infinity.

Eccentricities larger than unity correspond to cases where the velocity ``at'' infinity has some finite value. Then an incoming particle would approach the sun to some minimum value of r that corresponds to a point on the curve for EPE greater than zero. This EPE would be exactly the r-component of the kinetic energy ``at'' infinity.

Figure 34-1 shows a number of comets with ellipticities less than, say 0.9. These comets will be periodic, although their periods will depend on their semimajor axes which are not shown. These comets all have inclinations less than about 50o, so they are prograde--they revolve in the same sense as the planets.

There is a cluster of points for comets with eccentricities near unity. A few comets have eccentricities just a little less than unity and a few have eccentricities a little more than unity. Note that there are no points for eccentricities much greater than unity. Therefore, the interpretation that is usually given to Figure 34-1 is that the cluster of points along the e=1 line really corresponds to comets with very large elliptical orbits rather than hyperbolic orbits.

It is entirely plausible that orbits with very large semimajor axes would appear nearly parabolic when just a small part is observed when the comets are in the inner solar system. Then those few comets with e slightly larger than unity are presumed to also have large elliptical orbits that appear to be parabolic because of errors.

The argument runs that if truly parabolic orbits occurred, there ought to be at least a scattering of orbits well beyond e = 1. Modern observations have revealed a few genuinely hyperbolic orbits, but no more than could be accounted for by perturbations from Jupiter.

Now suppose that comets pervaded interstellar space. They would be moving with typical relative velocities of stars near the sun. These are a few tens of kilometers per second, not very different from the orbital velocity of the earth. If such objects entered the solar system, they would have positive total energies, and would be observed to have genuinely hyperbolic orbits. However, no such objects have been observed.

Observations of comet orbits have thus led to the picture summarized in the next section.

The Canonical Picture--The Oort Cloud and Kuiper Belt

The celebrated Dutch astronomer Jan Oort suggested a general model that explains most of the observations of comets. In this model, most of the long period comets reside in a kind of reservoir that is spherical in shape, and extends as far as 50 to 100,000 AU. Now 105 AU is 1.5 x 1018 cm, or about half a parsec! Since the nearest star, Proxima Centauri is at only 1.3 parsecs, we might expect comets in the Oort Cloud to be influenced by passing stars, and this is just what was postulated. Perturbations from passing stars would jiggle objects in the Oort cloud, and occasionally send one into the inner solar system, where it would be observed as an object with nearly unit eccentricity.

Of course, most of the comets within the cloud were somewhat closer than 105 AU. A modern model puts some 1012 comets in a spherical, outer Oort Cloud that stretches from about 10 to 100,000 AU. An inner Oort cloud, between 300 and 10,000 AU contains 1013 comets. Both clouds are spherical, so that stellar perturbations would bring comets in with random orientations to the inner solar system. This explains nicely why the orbital inclinations scatter randomly along the axis with e = 1 in Figure 34-1. These cometary orbits show no ``memory'' of the geometry of the system of planets.

The Oort cloud model does not account for those comets with eccentricities in the range 0.2 to (about) 0.9. All of these orbits are prograde, and the inclinations are not too different in their distribution from those of the asteroids. To account for these comets, it is necessary to postulate a closer dynamical coupling to the planetary system. This region is now called the Kuiper belt, shown schematically in Figure 33-2.

Figure 34 - 2: Oort Clouds and Kuiper Belt

Comets, like the planets, obey the Kepler law that P2 = a3. This means that comets at the inner periphery of the Kuiper belt (50 AU), will have periods in years of 354 years. If an average comet with this period were observed today, the previous viewing would have been in the mid-1600's, pre-telescopic times. It would be difficult to verify any comet seen then with a comet observed today, since the old position could not have yielded a reliable orbit.

In the case of the famous Halley's comet, sightings have been recorded going back several hundred years BC. In this case the period is somewhat shorter, 77 years, and the comet has been quite bright.

The Masses and Origins of Comets

Cometary masses are not well known, because we have not observed anything orbiting about them. So mass estimates usually come from a size and a guess at the density. In the case of Halley's comet we have reasonable measurements of the size. It was observed from space probe flybys in 1986 to be roughly 16 x 8 x 8 km in size. Mass estimates are near 1017 grams. The 1013 comets of the inner Oort cloud dominate the cometary mass. If we use the estimate of the mass of Halley's comet as typical, then the total mass of comets belonging to the solar system is some 1030 grams, not far from Jupiter's mass.

One picture of the place of comets in the history of the solar system had them as leftovers from the general formation. An interstellar cloud collapsed, but left some material on its periphery. This material eventually became the comets. This accounts for the Oort cloud, and its apparent spherical symmetry. But it is not clear how this material would have collected to cometary masses, given the very low densities of material far from the center of the solar nebula. Additional assumptions must be made to account for the comets that apparently come from the Kuiper belt.

Over the last decade or so, a variety of small bodies have been discovered in orbits beyond Neptune. The general belief is that they were once much closer to the sun, and were ejected by larger planets, primarily Jupiter. Oort himself proposed that the comet cloud was once much closer to the sun, but was ejected by perturbations from the Jovian planets. This may be the favored hypothesis today.

Professor Gary Bernstein recently of the U of M, along with his students have been searching for new Kuiper Belt Objects. You can find much more information about their work, and a very neat movie here.

If the comets were ejected from the inner solar system, they would initially have its preferred orbital direction. It is then necessary to assume that passing stars smoothed out the distribution to its present state of symmetry. It is generally assumed that this smoothing could have taken place.

An intriguing question concerns comets that would have been ejected altogether form the system. The assumption is that myriad objects were scattered into the Oort cloud, with total energies very, very close to zero. Surely the scattering would not have had a sharp boundary, and some comets would have been ejected altogether. If this process took place around typical stars, then there must be many comets in interstellar space. If this were true, then there should be hyperbolic comets. Recall that the absence of truly hyperbolic comets was the basis for the assumption of the Oort cloud in the first place.

We do not know how many comets would have been ejected totally from the solar system, but we can make an estimate that it might have been comparable to the total number of comets in the Oort cloud. A rough calculation shows that there might be 4 hyperbolic comets per century, within Jupiter's orbit. The calculation is surely uncertain by at least one and perhaps two orders of magnitude.

We conclude that it is entirely plausible ejected comets from nearby stars do fill interstellar space, but are not dense enough for us to have detected them as hyperbolic objects. Would such comets dim the light of stars? Yes they would, but not very much. In the general plane of the Galaxy, the effect of these putative comets would be less than that of interstellar dust by many orders of magnitude.

What Are Comets?

Comets are believed to be large, dirty snowballs. Large comets may be some 1018 grams, and be 20 km across. They ought to be similar in nature to the materials that formed the outer Jovian planets--mostly ice, but with a sprinkling of silicate material, or dirt. When these bodies come into the inner solar system, heat from the sun heats, and drives off volatile, icy materials. These volatiles would be mostly water ice, but include CO2 ice as well as frozen methane and perhaps ammonia.

When the volatile materials vaporize, and enter the low density interplanetary medium, the molecules dissociate, and often ionize. The solar wind pushes the gas and dust away, so that the tail of the comet always points away from the sun. The ionized gas usually points directly away from the sun, while a dusty tail may form that points generally away from the sun, but with some curvature.

What is the difference between a comet and an asteroid? Generally speaking, if it grows a tail, it's a comet. We have seen that asteroids generally live in a belt between Mars and Jupiter, while most comets live in distant reservoirs known as the Oort clouds and Kuiper belt. Nevertheless, there are undoubtedly a number of objects--somewhere between Jupiter and Neptune--that haven't grown tails that might do so if they were to come closer to the sun. If they did, we might call them comets rather than asteroids.

Since we think most comets belonged to the outer solar system, we are keen to bring some of their material into our laboratories and analyze it. This should give us a better grasp of the primordial composition of the solar nebula, that basic composition we now approximate with our tables of the SAD.

Meteor Showers

There are some meteor showers that are known to be associated with the orbits of burned-out comets. Perhaps the most reliable are the Persieds, which may be seen on 10 August. These meteors-- shooting stars--are the result of small particles that have slowly distributed themselves along the orbit of the comet known as 1862 III.

The meteors all seem to radiate from a point in the sky. In the case of the Persieds, it is the constellation of Perseus. The individual paths do not actually diverge from a point, but seem to do so as the earth moves rapidly into the volume of space occupied by the debris from the comet. The divergence of the meteors is caused by perspective, and is somewhat similar to the convergence of railroad tracks as one follows them into the distance.

Unfortunately, no material from a meteor shower has ever been definitively recovered. Probably, the meteoroids (the bodies in space) contain only small silicate particles, which do not fall to the ground, or if they do, are easily lost.

General

Bright comets are fun to watch. Over the last several decades, a few comets did not reach their expected brightnesses, and were a disappointment to those expecting a brilliant display. Comets got to be known to the public as things that fizzled. Fortunately, in the summer of 1994, the comet Shoemaker-Levy 9 collided in a spectacular manner with the planet Jupiter. The collision was photographed by the Hubble Space Telescope, and images were on the nightly news, for all to see. Then, in the spring of 1996, Comet Hyakutake put on a impressive display for naked-eye and small telescopic observers. So comets became generally respectable again.

Comets clearly contain a high percentage of volatile materials, and must have made some contribution to the volatiles of moons and planets. Just how large--percentage wise--that contribution may have been is not yet certain. Keen interest has recently been focused on the question of polar ice on the Moon and mercury.

Doomsday movies and TV shows have increased public awareness of the significance of comets. We have seen that the impact of a body with a mass of 1013 grams would have the effect of 144 megatons of TNT (second hour quiz). This is a larger explosion than that of the largest hydrogen bomb (about 40 megatons). This mass is far less than that of a typical comet.

Summary

Comets are dirty snowballs that begin to evaporate as they approach the sun. The solar wind and radiation pressure push their tails away from the sun. Cometary material may be very primitive, and therefore useful for the theory of the history of the solar system. The comets may have made a substantial contribution to volatile materials on the inner planets and their satellites.

Most comets reside in reservoirs known as the Oort clouds and the Kuiper belt. Perturbations from passing stars dislodge them, and cause them to enter the inner solar system. Kuiper belt objects may be very long period comets. They have prograde orbits, and are generally in the ecliptic plane. Oort cloud comets show no preferred orbital orientation. Possibly one Jovian mass is contained in some 1013 comets, mostly in the inner Oort cloud, between 300 and 104 AU from the sun.

It is not yet certain whether the comets originated at large distances from the sun, or were formed slightly beyond Neptune's orbit, and ejected by planetary perturbations. The latter theory is perhaps the most favored today.

Lecture 36 - Part I: Meteorites and the SAD

- i -

There is an old story that recounts Thomas Jefferson's reaction to the news that meteorites had been seen to fall by two farmers. The incident was later reported by two Harvard professors, and Jefferson is supposed to have remarked it was easier for him to believe two Yankee professors would lie than that stones would fall from the sky.

Jefferson's remark was supposed to have been made in 1807. Only a few years earlier, in 1803, the French Academy had decided to make a thorough investigation of these reputed stones from the sky. They sent the distinguished physicist Jean Baptiste Biot to investigate reports of a shower of stones that had fallen at the town of L'Aigle. Biot's report convinced a skeptical world that the objects we now call meteorites really are of extraterrestrial origin. After that time, Thomas Jefferson notwithstanding, meteorites began to be collected and preserved in museums.

The Michigan astronomer/geologist Dean McLaughlin wrote in his 1961 textbook:

Accounts of meteorite falls occur in very ancient writings. Before the Christian era it was well known that stones fell from the sky, but during the seventeenth and eighteenth centuries skeptical scientists discounted this belief as superstition. The German physicist Chladni in 1794 argued strongly for their extraterrestrial origin, on the basis of consistent accounts by witnesses of falls, but skepticism continued until 1803, when a completely convincing shower of stones fell at L'Aigle in France.

This skepticism of the 1600's and 1700's is quite fascinating. It corresponds in time to the so-called Age of Enlightenment in Western thought, also sometimes called the Age of Reason. Perhaps we can forgive the zealots of this period, who were busy escaping from the myth-ridden ages that preceded them.

- ii -

Meteorites have been mentioned at a number of points in these lectures. It has been generally sufficient for us to know that these materials were samples of matter from beyond the solar system. In the next lectures, we take up their study in some detail.

Let us now consider a few points of terminology:

Stones, Irons, and Stony-Irons

There are three broad classes of meteorites: stones, irons, and stony irons.

Figure 35 - 1: The three basic meteorite types, stones (left), stony-irons (center), and irons (right).

Iron meteorites contain genuinely reduced metal, sufficiently well refined by nature that the cut and polished face of an iron is highly reflecting. It has been suggested that the existence of these stones gave early man the notion of refining metal from ores. There are records of iron meteorites being used as ploughs by farmers--after appropriate treatment by a blacksmith.

The reduced metal in iron meteorites is actually one of two alloys, of iron and nickel. These alloys are usually considered among the minerals classified as native elements (Lecture 14). The minerals differ in their nickel content, with one of the alloys containing as much as 65% nickel, and physical properties. These differences make it possible to tell something about the freezing history of the meteorites from the intergrowth of crystals of the two minerals.

On crystallization, a mixture of the two nickel-iron alloys will grow together in a distinct patterns that can be revealed by polishing the surface of the meteorite, and then etching it with acid. There results a crisscrossed pattern known for the inventor of the technique, Widmanstatten. Measurements on this pattern reveal the cooling history of the meteorites. It has been found that some of them cooled very slowly, at rates of only a few degrees in a million years. This is surely a clue to the manner in which these materials formed. How and why might such cooling occur? We shall address this below.

Stony meteorites are of two types. Traditionally, a division into one category or the other was made by the presence or absence of small inclusions called chondrules, shown at the right. Most chondrules igneous in nature, but we are not sure how the melting took place. One theory is that it did so in the solar nebula, because of electrical discharges, or lightening. These melted small bits of material with rocky composition, making droplets which froze to form the rounded chondrules.

The most common form of chondrule is rounded, and we know that it is igneous in nature from geological thin sections. Other inclusions, sometimes also called chondrules, are irregular in shape, and may never have been melted. In some cases, it seems that rare kinds of chondrules may be solids that formed before the birth of the solar nebula, and were never mixed with it in a vapor phase. We will generally use the terms ``fragments'' or ``inclusions'' rather than chondrules for the non-igneous components of stony meteorites.

The word ``chondrule'' comes from the Greek root meaning cartilage--it isn't a very helpful mnemonic. Meteorites that contain chondrules are called chondrites. Those without chondrules (with one important exception) are called achondrites. Most of the achondrites are mafic or ultramafic (no quartz or feldspar) igneous rocks. We are convinced that some achondrites came from the Moon and from Mars. Others show almost the identical reflection spectrum of the minor planet Vesta.

Stony irons are, as one might anticipate, a combination of the two kinds of meteorite. The cut face of a stony iron might reveal a polished metal framework, with green crystals of olivine in little cells where the metal was missing.

Most of the meteorites in museums are irons. This is because they are the most easily recognized as something distinct from terrestrial rocks. The majority of these irons are `finds.' Most `falls' are chondrites. We may conclude that the majority of the meteoroids are rocky rather than metallic in nature. More than 80% of the falls are chondrites.

The Origin of Meteorites

The falls of several meteorites have been recorded photographically, so that their orbits could be determined. These are shown schematically in Figure 35-2.

Figure 35 - 2: Meteoroid Orbits

The figure shows orbits of Mercury and Venus (inner circles), Earth, Mars, and Jupiter. These are shown in blue. The green ellipses are meteoroid orbits corresponding to five meteorites. The most recent of these, Taglish Lake fell in January 2000 in British Columbia. The figure shows close spacial association of these objects with the asteroids, whose orbits are concentrated in the region between Mars and Jupiter.

Meteoriticists had long discussed the possibility that meteorites were debris from the failed planet or planets that were once in the asteroid belt. Such a hypothesis would account for the existence of igneous material, and core-mantle differentiation in a putative parent body would account for the iron meteorites.

There is one problem with the postulate of complicated igneous activity on asteroids. They are small bodies, and could lose their heat quite rapidly. Consequently, it is unclear that they would have retained enough heat to form cores, and igneous melts. What source or processes might have supplied the required heat?

Among the nuclides that have been investigated in physics laboratories there are a number that have half-lives much shorter than the heat sources we discussed in connection with the earth and Jovian planets. Let us explore why such radioactivities are important for the asteroids.

Table 35 - 1: Half-Lives of Some Radioactive Isotopes (years)

 Long     Half-life       Short     Half-life

 K-40      1.28E09        Al-26     7.3E05
 Th-232    1.4 E10        Ca-41     1.03E5
 U-238     4.47E09        Fe-60     1.5E06
 U-235     7.04E08        Pu-244    8.0E07

The sources we have discussed--uranium, thorium, and potassium--have long half-lives and supply heat very slowly. Now if a body is to heat up, it is necessary that the energy not leave the body at a rate comparable to that with which it is supplied. How fast does a body cool?

Let us assume that any energy flux (energy per unit area per sec) that reaches the surface will be radiated away. In other words, we assume the surface temperature will rise until it reaches a value such that T4 equals the energy flux from below (joules meter-2sec-1 or ergs cm-2sec-1). Then the rate of cooling of a body will depend on how fast the transport mechanisms can carry the energy to the surface. This depends on the nature of the transport mechanisms, but for purposes of argument, we assume the relevant mechanism is conduction.

If we have two bodies with equal conductivities, it is possible to show that the cooling times will depend on the squares of their characteristic dimension. This means that a spherical body with radius R would cool 4 times faster than a similar body with radius 2R.

Detailed calculations show that small bodies would cool too quickly for the long-lived radioactivities to heat them to the point where core formation could occur. We can make up a little table based on the cooling time for the Moon.

Table 35 - 2: Cooling Time Estimates for Bodies of Different Sizes

characteristic length     typical body      cooling time (years)
or radius (km)                             E9 means x ten to 9th

2000                      Moon                1. E9
 200                      large asteroid      1. E7
  20                      small moon          1. E5

We see from Tables 34-1 and 34-2 that the long-lived radioactivities are capable of reheating the Moon--and causing the impact basins to flood. This takes place on a time scale of the order of half a billion years. Even for asteroids, the cooling times (1 x 107 years) are 3 to 4 orders of magnitude shorter than the half-lives of 232Th, 238U, and 40K. Perhaps 235U could be relevant. For smaller bodies, the shorter half-life isotopes are clearly important as heat sources.

We shall see in the next lecture that there is good evidence that these short-lived isotopes were present in the early history of the solar system.

What about the collisions themselves as heat sources? Surely this is relevant, but most discussions have not taken them into account. Perhaps good quantitative estimates have not yet been made.

We conclude that meteorites come from chemically differentiated bodies primarily in the asteroid belt.

Carbonaceous Chondrites and the SAD

Among the chondritic meteorites there is a class known as the carbonaceous chondrites (CC's). These meteorites consist of chondrules and a great variety of mineral fragments imbedded in a matrix that is rich in carbonaceous material. The carbonaceous material is heterogeneous, and includes soot and organic polymers. Considerable interest focused on the presence of amino acids in CC's, but there is little basis to conclude that these simple organic compounds are the product of life.

Astronomers have known for most of the present century that the elemental composition of chondrites mimicked that of the sun for non-volatile elements. From about 1950 through perhaps 1980 there was a competition among meteoriticists to find which among the meteorites most closely resembled the solar composition. This ``race for the Holy Grail'' will be described in the next lecture.

Eventually, the Swiss-American cosmochemist Edward Anders won the prize with his advocacy of meteorites known as Type I carbonaceous chondrites. We will call these special meteorites the CI's, for short. These meteorites are an exception to the rule that chondrites contain chondrules. Indeed, they appear to be quite similar in their entire composition to the carbonaceous background matrix of other carbonaceous chondrites. Roughly speaking, if you took the chondrules and some hard mineral fragments from a carbonaceous chondrite, and compressed the remains into a rock, you would have a Type I carbonaceous chondrite.

It has been important to establish the Holy Grail (SAD) because of the role it plays in theories of the origin of the nuclides. Since the late 1950's it has become clear that the majority of the chemical elements and their isotopes (nuclides) were manufactured in the interiors of stars. These processes have left their signature in the abundance patterns in the SAD. Just as the geochemist can read the history of rocky materials from their mineralogy, the astrophysicist can read the nuclear history of matter from its abundance patterns. We shall have more to say about this in the ensuing lectures.

Summary- Lecture 35 Part I

The Age of Enlightenment had a blind spot for meteorites. In the nineteenth century their extraterrestrial nature was again recognized, and productive research could proceed. The three broad classes of meteorites, stones, irons, and stony irons correspond roughly to planetary mantles, iron cores, and core-mantle boundaries. Meteorites that have been seen to fall are called falls, while those subsequently discovered are called finds. Most finds are irons, because they are easy to recognize. Most falls are chondrites. The achondrites are mostly igneous stony meteorites. A number have been identified with a planet, satellite, or asteroid.

Orbits of several meteorites indicate an origin in the asteroid belt. We believe they came from differentiated minor planets that were broken by collisions. Short-lived radioactive isotopes can provide the necessary heating for core-mantle differentiation.

The carbonaceous chondrites, and particularly the Type I Carbonaceous Chondrites (CI's) provide the best guide to the SAD and the nuclear history of matter.

The Meaning of the SAD

The meaning of the term SAD (standard abundance distribution) changes subtly from one usage to the next. This is a general problem with language, and it is why dictionaries typically give more than one meaning for individual entries. Let us examine some of the various meanings of this term. The SAD may mean:

  1. A table of relative abundances of the chemical elements. Astronomers arbitrarily set hydrogen at 1012 in their tables of the SAD. This is because hydrogen is by far the most abundant element in the universe. Geochemists, including meteoriticists, have used 106 for silicon. Rocks and meteorites, unlike stars, contain more silicon than hydrogen. Tables of this kind are in a state of constant revision, so it is necessary to give a reference to the source of the numbers.
  2. Relative abundances of the elements in the sun.
  3. Relative abundances of the elements in the primordial cloud from which the sun and solar system formed.
  4. A reference system of relative abundances that describes average abundances in our Galaxy in the neighborhood of the sun, that is, within a few hundred parsecs of the sun.

Meaning (1) refers to a specific set of numbers assembled by some individual or group of people. These workers attempt to find the best estimates of the solar composition--meaning (2). The problem is that solar abundances can not be determined as accurately as those in meteorites. Experience has shown that as the solar determinations improve, for most elements the relative abundances in the sun and the Type I carbonaceous chondrites (CI's) agree with one another.

The elements that fail to agree in the sun and the CI's are what we have called the supervolatiles: hydrogen, helium and the noble gases, carbon, oxygen, and nitrogen. With the exception of these few elements, there is no basis to say that the relative solar and CI abundances differ from one another. This situation could change, as the accuracy of solar abundances improves, but for the present, we have the rather bizarre situation that certain special rocks (meteorites) tell us what the sun is made of better than the sun itself!

We assume that because most of the mass of the solar system is tied up in the sun, the sun should be the most representative object of the primordial cloud from which sun and its planets originated. Thus we have definition (2) for the SAD, and we see that we are really using the sun as a means to get at definition (3)---astronomers want a representative set of abundances for our location in the Galaxy. It is possible that the sun is special in some way, and that its abundances are not ``typical.'' For example, it is generally known that the abundance of oxygen generally measured in hot interstellar gas is lower than assumed for the sun.

Oxygen is a super volatile, so that the value that appears in our table of the SAD does not come from the CI meteorites. For oxygen and some other super volatiles, it comes from spectroscopic determinations of the solar composition. Generally, these are much less accurate than meteoritic determinations, but for the supervolatiles, we must use them. From the best information we have, it appears that oxygen in the sun is rather higher than the average in the Galaxy near the sun. So if we used definition (4) for the SAD, we would have a lower oxygen abundance than for definition (2) (and by default, definition (1)).

The SAD as the Holy Grail

Qualitative chemical analyses of the sun and stars date from the early use of the spectroscope more than a century ago. In qualitative analysis, one determines the presence of substances, but not how much is present. When we measure how much, we are doing quantitative analysis. Quantitative analysis of the sun and stars dates from roughly the 1930's, and even in those times it was known that the relative abundances of many elements in the sun and meteorites were similar.

It is interesting that the theory of stellar spectra had finally progressed to the point where it was possible to say that the composition of most stars were remarkably similar. Now the spectra of stars are not similar. They look very different from one another.

Stars with the surface temperature of the sun are full of lines of neutral iron, while the spectra of the bright stars Sirius and Vega are dominated by strong lines due to hydrogen. In the first decades of the 20th century, astronomers thought the sun was made mostly of iron, and Sirius and Vega were mostly hydrogen. In the 1930's it became clear that the huge difference in the spectra of the sun and these two bright stars was caused largely by their different temperatures, and not by their composition.

It was a triumph of the theory of stellar atmospheres to be able to explain the diversity of stellar spectra in terms of physical conditions. Two basic principles were involved--the excitation and ionization of atoms. The average energy per particle--(3/2)kT--is not enough to excite hydrogen to the levels that give rise to the strong lines in Sirius and Vega. On the other hand, by the time the temperature has reached that in the atmospheres of Sirius and Vega, most of the neutral iron has been ionized away.

There were still a few stars whose spectra clearly indicated very different chemical compositions from the sun. However, these objects were apparently set aside in the minds of astronomers and knowledgeable geochemists. The notion of a universal abundance distribution began to appear in the scientific literature. As early as 1930, the term ``cosmic'' was used to describe what was thought to be the universal abundance pattern. Interestingly, the coiners of the term were analytical chemists whose specialty was meteoritics.

If we assume that the abundances of the chemical elements are the same everywhere in the universe, it is natural to seek one event where they were all made. This led the incomparable George Gamow and his coworkers to the early theory of the big bang. They sought some way in which all of the chemical elements could be created in the initial explosion that gave birth to our universe. Their work, which was done in the 1940's and 1950's had only limited success. They could never make elements heavier than helium! However, they predicted the cosmic background radiation long before its discovery. In addition, we now accept a nearly universal helium abundance, and that it emerged from the big bang.

Nucleosynthesis is the name given to the processes that made atomic nuclei. After stellar nucleosynthesis (Lecture 7) became accepted, is was clear that abundances in stars must vary. These variations are not large for the majority of stars, but important and occasionally quite large variations do occur. For this reason it is no longer appropriate to speak of a `cosmic' abundance pattern. Nevertheless, the four meanings of SAD given above are all useful concepts. The French nuclear astrophysicist Jean Audoze and the American astronomer Beatrice Tinlsey suggested use of the phrase `standard abundance distribution' (SAD) as a replacement for the anachronistic `cosmic abundances,' which is still occasionally used.

Even though the SAD may not be ``cosmic,'' it is still of keen interest, and the question naturally arose of how best to get at it. In principle, spectroscopic analyses of the sun's atmosphere would be the best way to proceed. Unfortunately, these determinations were notoriously inaccurate, and they gave almost no information at all on isotopic abundances. Meteorites thus persisted as being the best overall guide to the SAD, but there are many and varied kinds of meteorites. Which ones would be the best guide to the SAD? An American cosmochemist compared this quest with that for the Holy Grail.

Nuclear and Nonnuclear Patterns

Possibly since the heyday of ``cosmic abundances'' we have assumed that the abundance pattern we seek should reflect nuclear rather than chemical properties. That's because the nuclei had to be made by nuclear processes, and the results should reflect nuclear rather than chemical properties.

We have seen a plot of the SAD in which relative abundances vs. the atomic number Z are displayed. Nuclear astrophysicists have found a similar plot somewhat more informative. We show that here. Abundances are plotted vs. the mass number A rather than Z.

Figure 35 - 3: SAD Abundances vs. Mass Number (A)

We will not go deeply into the theory of nucleosynthesis that accounts for the details of this plot, but we shall outline the salient points. They are generally accounted for in terms of nuclear rather than chemical properties. Generally speaking, those nuclides that are more stable are the more abundant.

We understand why the bumps occur on the abundance curve for heavy nuclei. It is closely related to the stability of nuclei. This stability derives from shell closing, that is similar to the closing of electronic shells or subshells. However, the number of nucleons that belong to closed shells is different from that of atomic shells because the nuclear forces are different. For a long time, physicists did not know why nuclei with certain special numbers of nucleons were very stable. One famous physicist said it was ``magic,'' and the term stuck. Thus, the bumps on the abundance curve are due to nuclear magic numbers.

Let us briefly note what we mean by a nonnuclear abundance pattern. A good example would be the earth's crustal abundances (Figure 19-2), with their depletions of the siderophiles, and severe depletions of the noble gases. These patterns exist because of chemical properties of elements. In this section we have mostly discussed abundance patterns that exist because of the properties of atomic nuclei.

Volatilities Provide the Key to the Holy Grail

As our knowledge of the structure of the atomic nucleus improved, it could be integrated into the search for the most representative meteorites. The search for these objects, improvements in our estimates of the SAD (at that time, the cosmic abundances), and our knowledge of nuclear structure and systematics went on simultaneously.

In the late 1960's, the Swiss-American cosmochemist Edward Anders and his colleagues narrowed the search for the Holy Grail to the Type CI carbonaceous chondrites. The strongest piece of evidence in favor of these meteorites was their high volatile content. Remember that volatiles are defined as materials that remain in the vapor phase of a cooling gas. We will simplify the argument given by Anders and his colleagues in the following paragraph:

Suppose you had a box full of the primordial gas of the solar nebula, and you cooled it until all of the vapor had condensed as solid material. Surely, this would be the long sought cosmochemical Holy Grail. Any subsequent heating of this solid would drive off sequentially the most volatile materials. The greater the heating, the further the composition would depart from that of the primordial nebula.

The most volatile-rich meteorites known are the Type I carbonaceous chondrites (CI's). They lack the igneous materials, the chondrules, which have undergone heating and melting. This would drive off the volatiles components of the chondrules. The overall volatile content of these meteorites varies inversely with the igneous fraction. The CI's contain no chondrules. CM's do contain this igneous material, and the CV's have still more. Consequently, the latter types have fewer volatiles (See Figure 35-4).

We assume the CI's have lost the least of their original material, and therefore choose them as the best approximation to the primordial composition. Of course, even the CI's have lost their supervolatiles: hydrogen, helium, nitrogen, (most of their) carbon and the noble gases.

Figure 35 - 4: Volatilities in Carbonaceous Meteorites

In Figure 35-4 we compare the volatilities of two carbonaceous meteorites, called CM's and CV's, with the CI's. Logarithms of ratios of elemental abundances are plotted on the lower part of the graph, with the scale to be read for the y axis on the left. All three meteorites have approximately the same abundances for Mg and Si. The abundance ratios are therefore unity, and its logarithm is zero. Volatile elements plot as negative numbers. For example, zinc (Zn) is about half as abundant in the CM's as the CI's. The ratio is therefore 1/2, and the logarithm -0.3. This is plotted on the solid line. The ratio is about 1/4 for the CV's divided by the CI's, so the logarithm of the ratio is about -0.6 (dashed).

The logarithm of the condensation temperature is plotted in the upper part of the diagram, with the scale on the upper right. High condensation temperatures fall high on the plot, and low ones, low. This curve was generated with the help of the thermodynamic properties of real materials--compounds rather than elements--likely to condense from the gas of the solar nebula.

First, we see that the theory and the meteorites tell the same story about volatility. It isn't perfect, but the correspondence is too close to be accidental. Also, we see that the CM's and CV's agree about what is volatile, and with a few exceptions The CM's have about half the volatile depletion of the CV's.

Tests of the Holy Grail Candidates

Arguments that the CI are the most volatile rich, and that they have not been heated would not be worth much in the absence of two conditions: