QUESTION: remote sensing
How do you use the HST to determine the size, mass, composistion,
temperature, density, state of matter, etc., of the planets and other
objects in space?

ANSWER from David Soderblom on March 28, 1996:

This is an excellent question, as it's not exactly obvious how those
conclusions are drawn.  But please also realize that you're almost asking me
to explain all of astronomy.  I'll put my answers in fairly simple terms,
but please ask for more if I don't go far enough on a subject.  I have also
indicated between asterisks (e.g., *black body*) an entry you should be
able to find in an encyclopedia, so that you could get more information that
way.

1. Size, by which I believe you to mean "how many kilometers across is
Jupiter" or the like.

To get a size directly, we have to be able to resolve the object with our 
telescope.  That means we can tell it has a finite size and is not a point
source.  You measure the angular size the object appears to be (for instance,
the moon is about 1/2 degree across).  That and the distance are all you
need: you have a very skinny triangle with the Earth at its point and the
object forming the base.

So how do we know distances?  For solar system objects we compute their
orbits, using observations of their apparent position in the sky and
*Kepler's laws*.  For some of the nearer stars (out to about 200 light years
now) we can measure distances directly from the *parallax*, which is the
tiny amount the stars appears to move on the sky relative to more distant
objects (which don't move at all because they're too far away).  It's the
same trigonometry of the skinny triangle with the Earth's orbit as the
base and the parallax being the angle of the point.

For more distane objects we have to bootstrap our way through the universe.
That means inferring distances of more distant objects in our own Galaxy
by knowing that they're in a cluster, for example, and we know the intrinsic
brightness of some of those cluster stars because we find examples of them
close to enough to get a parallax.  From these clusters we can work to the
nearer galaxies, and then develop other indicators that let us estimate
the distances of remote galaxies.

The whole issue of the distances of remote galaxies is at the heart of the
current debate over the age of the Universe, their age and distance being
related by the Hubble constant.

2. Mass.  To determine the mass of an object you have to be able to directly
measure its gravitational effect on another body.  In other words, you need
a good quality orbit for it and something that goes around it.  Most planets
have *satellites*.  Once you have the orbit of the satellite around the
planet (distance and position versus time), Kepler's laws will tell you the
mass.  For the planets, they are satellites of the Sun, and we use them to get
the Sun's mass.  To get the masses of the satellites of the planets is
harder, and involves estimating a density and knowing its size.

Getting the mass of a star likewise requires seeing it in a situation where
it is orbiting another star.  The masses of galaxies are estimated in
essentially
the same way: we see how the galaxy influences the motions of the stars in it,
and that leads to the mass.  It is by this means that people have postulated
the existence of "missing mass" in galaxies.  We can see how much gravitational
effect there is, and then we can add up the mass of stars producing light,
and see that there's a deficit: not enough light-producing stars to yield as
much gravity as appears to be there.

3. Composition and temperature.
This is both simple and very subtle.  At its simplest, you
take the light from the object and pass it through a spectrograph, which is
essentially like using a prism to disperse the light into its colors, only
better.  In the spectrum you record (photographically, 20 years ago, but
now with solid-state devices), you will see a broad distribution to the
light.  Some objects will put out more red light than blue, and others more
blur than red.  That is quantified as the color of the star (or planet or
galaxy).  In the case of stars, that color indicates the temperature in that
blue objects are extremely hot while red ones are cool, with a range
inbetween.  In the case of planets, the distribution of light is the same
as that of the Sun (mostly) because we're seeing reflected light.

If you look at the spectrum in more detail you will see dark bands where
light has been absorbed.  In stars, these absorption lines are almost all
due to elements, such as iron, titanium, manganese, and so on.  We know a
particular line is from iron by work done in laboratories.  The presence of
an iron line obviously tells you that object has iron, although that doesn't
mean you know how much.  Determining how much iron is there requires a
detailed look at just how much light got absorbed and comparing that to
calculations of how the iron atom behaves.

In very cool objects (like planets) you'll see absorptions from molecules
like methane and ammonia.  Again it's easy to say the molecule is there
but a lot harder to say exactrly how much of it is there.

The strength of the absorption depends both on the amount of the element or
molecule present but it also depends on the temperature, which affects the
state of the atom or molecule and hence its ability to absorb light of a
particular color.

So the short answer is that the spectrum of a star or planet tells us about
both the temperature and composition in a form that requires detailed analysis
to deconvolve.

4. Density.
Density is almost always an inferred quantity, in the sense that we estimate
the size and mass of an object and then calculate what its density must be.
It's not as simple a calculation as you might think.  A very large object like
the Earth or Sun has a lot of self-gravity, and that leads to the central
regions being a lot denser than the surface.  That's why nuclear reactions
can take place in the center of the Sun: the atoms are jammed so close
together they start running smack into each other.

For such objects we cannot measure density in any direct way, so we construct
models of what it must be like.  We know the composition of the Sun, for
example, and we know the laws of physics that apply to it.  That enables us
to calculate self-consistent models that tell us how how and how dense the
interior regions are.  In recent years these models have been tested because
we can detect waves that travel all the way through the Sun, and those waves
are influenced by the local density at every step of the way.  Detailed
analysis of those waves has borne out the models in great detail, confirming
that we know what was going on.

In the case of the Earth we have the advantage of measuring seismic waves
(from earthquakes) that tell us about the interior structure.  That information
gets folded back into the models to improve the results.

The models also tell us when matter is a gas, liquid, or solid.  Again, in
most cases we cannot actually see that the core of a planet is a liquid, say,
but we can infer what the density structure is from models and know what
the composition must be from the spectrum (in part).

I hope this information helps.