by Jeremy Tatum
A project supported by The Planetary
From the May/June 1988 issue of The Planetary Report
Astrometry, the practice of measuring stars' positions, may sound
like some tedious activity of nineteenth-century astronomy, far removed
from today's exciting frontiers of black holes, quasars, pulsars
and planetary exploration. But there is something deeply satisfying
and exciting about striving to measure precisely the positions of
asteroids and comets, to calcu- late their orbits, and to predict
where they will go next. Keeping precise track of interplanetary
bodies is vital to today's and tomorrow's explorers of the solar
The tremendous successes of the spacecraft trips to Comets Giacobini-Zinner
and Halley depended greatly on ground-based astrometry. Future exploration
of comets and asteroids, such as NASA's proposed Comet Rendezvous
Asteroid Flyby (CRAF) mission, will also need accurate measurements of
their targets. The Planetary Society has provided us with a generous
grant to assist with this work, so I'd like to share with members
some of the excitement we feel about our project.
Let's begin with a brief review of the types of objects we're
observing. There are 130 or more short-period comets, those
whose orbits are known and whose return can be predicted years in
advance. In addition, every year some dozen or more long-period comets
are discovered. They move across the sky for a few weeks and
then disappear, not to return again for tens of thousands of
years. Most of the approximately 3,700 minor planets or asteroids
have fairly well known orbits in the asteroid belt between
Mars and Jupiter. A few, called the Trojans, are in almost the same
orbit as Jupiter, 60 degrees ahead of or behind the giant planet.
An important and exciting few have eccentric orbits that sometimes
take them close to Earth. Because these Earth approachers seem to
move so rapidly across the sky, we often call them "FMOs," short for fast-moving
objects. These asteroids with well-known orbits have a permanent number
attached to them, and most of them have names; 46 Hestia and 433 Eros are
just two exam- ples. They are often called "numbered asteroids."
Every month astronomers observe additional very faint asteroids that
they cannot identify. Many thousands of these "unnumbered asteroids" exist,
quite a few of them FMOS. They receive a temporary designation-usuafly
a year plus two letters, such as 1987 HR-and most of them are lost shortly
after discovery. But a few are observed over a sufficiently long time span-maybe
a few weeks-to enable astronomers to compute at least rough orbits.
Sometimes two apparently distinct objects seen in different years prove
to be the same object, and then we can find precise orbits. If an unnumbered
asteroid is well observed at three or more apparitions, it may become the
proud possessor of a permanent number or even a name. We are particularly
interested in recovering unnumbered objects and following them long enough
so that they can receive a permanent number. Some 20,000 asteroids have
at least rough orbits.
With so many objects to keep track of-most exceedingly faint and hidden
in the background of millions of stars-how can we find any given asteroid?
Or if we spot an asteroid in a pho- tograph, how can we identify it or
tell whether it might be a new one never spotted before? First we must
under- stand a little bit about orbits.
Each asteroid's orbit is approximately an ellipse, with the Sun at
one focus. Each orbit is characterized by six orbital
elements: the first two numbers describing the size and shape of the
orbit, the next three angles describing the orientation of the orbit in
space, the last telling the instant in its orbit when the minor planet
is closest to the Sun.
We can calculate the orbital elements from a set of astrometric measurements
of the object--that is, from a number of observations of the precise positions
of the object at different times. This is not an easy calculation. Nevertheless,
astronomers have calculated the six orbital elements for almost 10,000
asteroids, and these 60,000 numbers require little space on the disk of
a modem computer.
From the orbital elements we can then compute an ephemeris, an hour-by-hour,
day-by-day, or week-by-week prediction of exactly where the object will
appear against the starry background of the night sky. The most difficult
part of this calculation is the need to allow for Earth's position in its
journey around the Sun, a calculation made particularly difficult by the
perturbing effects of the Moon. Nevertheless, a modem computer can make
fairly short work of it and can include certain refinements such as al-
lowance for the position of the observer on Earth's surface or for the
time re- quired for light to travel from the asteroid to Earth.
We have stored in the University of Victoria's computer the orbital
elements as well as a program for computing the ephemerides of all the
asteroids. At the beginning of each month we decide which asteroids we
might be interested in observing. We simply type numbers on the computer
keyboard, type the single word ELLIPSE, and presto! Within seconds the
computer calculates for us an ephemeris, at half-day intervals, for the
entire month for all the asteroids requested.
Our telescope, a 10-inch-diameter Schmidt that is quite tiny by modem
standards, is a wide-field astronomical camera. Still, the stellar images
are so sharp and we can use so many of them for comparison stars that the
system has proved very effective for astrometric work.
We have a few extra tricks. We some- times use a colored filter to
increase the contrast between a faint asteroid and the sky's light pollution.
We also have a device that is especially useful for fast-moving objects.
The image of an FMO appears as a thin line on the photograph, while the
stars appear as sharp dots. With our electronically controlled system,
we can automatically move the crosshair (spun for us by a real, specially
trained spider!) of our guiding telescope at the right speed and
direction predicted for the asteroid's or comet's motion. The
stars then appear as short streaks, while the asteroid or comet is
a steady bright dot. This enables us to photograph objects so faint
that they would not be detectable at all if the image were allowed
to drift across the photograph during the exposure.
Most of our exposures are made near the ecliptic (the plane cut
by Earth's orbit about the Sun; most known solar system objects lie in
or near the ecliptic), so a photograph may show several or even
a few dozen asteroids. Of course, it also includes the images of
hundreds of thou- sands of stars. The asteroids are exceedingly small
and faint and can be seen only with the aid of a high-power microscope.
How can we possibly find and correctly identify these tiny images?
Our solution is to take two photographs a couple of hours apart. During
this interval, the image of each asteroid will have moved-perhaps
a distance comparable to the diameter of the im- age, maybe
just a few thousandths of a millimeter. Nevertheless, it has moved.
We put the two photographs into a great contraption called a blink
comparator, which enables us to look rapidly to and fro from
one photograph to the other, two or three times a second. The image
of each asteroid then appears to oscillate rapidly to and fro, and
this movement can be quickly spotted against the back- ground of numerous
but stationary stellar images. The search can be quite tedious, but we
experience a special duill each time we spot one of these tiny planets.
Another trick used by some observers is to look at the two photographs
through a stereoscopic viewer. The images on the two plates then overlap
completely except for any asteroid that has moved between exposures. The
asteroid then stands out fairly obviously. It is all
very well to find an asteroid, but how do we know which of many thousand
it is? We consult the computer again. This time we type in the astronomical
coordinates of the center of the photograph, the time it was taken, and
the single word PLATE. The computer reads through the entire list of orbits
of all the asteroids and works out the position of each object for the
time when the photograph was taken. It then calculates to see whether the
asteroid would have been in the telescope's field of view at the time,
and, if so, it prints out the asteroid's coordinates. As a bonus, it prints
an asterisk if the object is within a degree and a half of the center of
the field of view. Next comes the part that ought to
be tedious-the careful measurement of the position of each asteroid or
comet image. We have to compare the position of each object with that of
several stars whose positions are known precisely from catalogs. Fortunately,
the scale of the photographs closely matches that of a reliable star atlas,
so we can quickly identify several comparison stars. The positions
of the quarter of a million or more stars of the Smithsonian Astro- physical
Observatory Catalog visible from our latitude are stored in the com- puter's
memory; all we have to do is tell the computer the catalog numbers of the
stars involved, and it will read the posi- tions and other data ftom its
enormous memory bank.
Our Magic Measuring Machine is a microscope with a stage that can be
moved to left or right or up or down, or rotated at will. We place the
photograph on the stage and look at it through a microscope that has a
fine crosshair in its eyepiece. We try to bisect the image of each asteroid
or comet and the comparison stars with the crosshair by moving the stage
bearing the photograph beneath the microscope. Meanwhile, at the other
end of the table where we are working is a mysterious metal box with a
large display of neon numbers that change rapidly as we proceed. The numbers
indicate the precise position of the microscope stage to one thousandth
of a millimeter.
Most people know that often in magic it's all done with mirrors, and
the sharp- eyed may notice a tiny mirror on the side of the microscope
stage. A beam of laser light is reflected from this mirror, and the incident
and reflected beams form a system of standing waves. The mysterious metal
box is actually connected to the laser, and as the microscope stage moves,
the alternating light and dark bands of the standing wave system are electronically
counted, each band corresponding to just half a wave- length of the laser
The measurement over, the computer makes short work of some more calculations.
It checks to ensure that all the star measurements are consistent with
their catalog positions and wams us if any of our stars are misidentified,
if we have made poor measurements, or if the catalog position is inaccurate.
It will make corrections for the proper motions of the stars. The stars
are not fixed, but move by an appreciable fraction of an arcsec- ond, or
even more, per century, and it is essential to allow for this. The computer
corrects for the refraction caused as the starlight passes through the
Earth's at- mosphere. But finally, if completely sat- isfied, it produces
a position for the ob- ject we're investigating.
The computer may be satisfied, but will the Minor Planet Center in
Cambridge, Massachusetts be? We know that the demands made there are, properly,
very exacting and we don't want to risk our reputation by submitting observa-
tions that are in any way suspect. We know that ours will be scrutinized
very carefully in comparison with those sub- mitted to the Center by other
highly skilled and experienced observers. A final stage, especially with
newly discovered objects such as comets and Earth- approaching asteroids,
is to add our observations to those already available and to refine the
computation of the object's orbit and then see how far our measured position
is from that predicted by the refined orbit. At last, if all is satisfactory,
we submit our observation.
Despite many astrometric measurements, we have not so far discovered
many new objects, so our first asteroid discovery is especially exciting.
David Balam found the faintest of telltale streaks on a photograph made
by our colleagues Leopoldo lnfante and Chris Pritchet with the Canada-France-Hawaii
Telescope in April 1987. The original image was barely visible with a micro-
scope, but it has now been computer-en- hanced with an image processor
to pro- duce the rather impressive image ac- companying this article. The
object has been named 1987 HR. However, the chances of finding it again
are not great, and its overall astronomical significance should not be
exaggerated. Some observers routinely pick up lots of faint, previously
unknown asteroids, and Leopoldo "discovered" some quarter of a million
galaxies on the same photographs on which Dave found the asteroid.
Jeremy Tatum is a Professor of Astronomy at the University
of Victoria in Victoria, British Columbia.
This web page was prepared by Michael
Paine for The Planetary Society Australian Volunteer
Co-ordinators. Please excuse any OCR typos.
This article was written in 1988. The technology has changed substantially
but the task is just as difficult.
See also Predicting a NEO's orbit
Last update 7 November 1998