It could fit on your desk, and it’s made mostly from
parts bought at a camera shop, but two scientists believe
their new instrument will help them find a slew of large
planets orbiting stars in our Milky Way galaxy.

“An amateur astronomer could do this, except maybe for
the debugging of the software, which requires several people
working 10 hours a day,” said Dr. David Charbonneau of the
California Institute of Technology in Pasadena. “But it’s
easy to understand what’s going on and cheap to build the
equipment. That’s why everyone thinks it’s an ideal project,
if it works.”

The assembly of the new instrument is a cooperative
effort between Charbonneau and Dr. John Trauger of NASA’S Jet
Propulsion Laboratory in Pasadena, which is managed by Caltech.
“David’s approach promises to locate new planets orbiting distant
stars. The instrument is simple and straightforward, taking advantage of
spare parts and computer code we already have on hand at JPL,
and we hope to have it up and running in a few months,”
Trauger said.

Charbonneau and his colleagues will soon use their gizmo
to begin a three-year survey for extra-solar planets at
Palomar Observatory in San Diego County. The instrument is
based on a standard telephoto lens for a 35-millimeter camera.
It will sweep the skies, looking for “hot Jupiters,” or large,
gaseous planets, as their fast orbits take them in front of
other stars, into the line of sight between a star and Earth.
Astronomers will watch for the “wink” from the star as an
orbiting planet partially blocks its light.

Charbonneau, a recent import to the Caltech astronomy
staff from the Harvard-Smithsonian Center for Astrophysics,
Cambridge, Mass., is a leading authority on the search for
such “transiting planets.”

The new instrument uses a standard 300-millimeter Leica
camera lens, with a charge-coupled device, or CCD. The CCD, which costs $22,000,
will be mounted in a specially constructed camera housing to
fit at the back of the lens. The entire device will be fitted
onto an inexpensive equatorial mount, available at many stores
carrying amateur astronomical equipment.

“Basically, the philosophy of this project is that, if we
can buy the stuff we need off the shelf, we’ll buy it,”
Charbonneau said. The project costs $100,000, a fraction of
the cost of most large Earth and space-based telescopes.

The Palomar staff will provide a small dome for the
instrument, and the system will be automated so it can be
operated remotely. The new telescope will be linked with an
existing weather system, which will monitor atmospheric
conditions and determine whether the dome should be opened.

Charbonneau will be able to photograph a single square of
sky about five degrees by five degrees. About 100 full moons
or an entire constellation could fit in that field of view.
With special software Charbonneau helped develop at Harvard-
Smithsonian and the National Center for Atmospheric Research,
he will compare many pictures of the same patch of sky to see
if any of the thousands of stars in each field has “winked.”

If the software reveals a star has dimmed slightly, it
could mean a planet passed in front of the star between
exposures. Repeated measurements will allow Charbonneau to
measure the orbital period and size of each planet. Further
work with the 10-meter (33-foot) telescopes at Keck
Observatory at Mauna Kea, Hawaii, will provide spectrographic
data, and thus, will infer more detailed information about the
planet.

Weather permitting, Charbonneau will gather up to 300
images a night. With 20 good nights per month, about 6,000
images would be gathered each month for computer analysis.
The ideal time will be in the fall and winter, when the Milky Way
is in view, and an extremely high number of stars can be squeezed
into each photograph.

“It’s estimated that about one in three stars in our
field of view will be like the Sun, and one percent of Sun-
like stars have a hot Jupiter, or a gas giant that is so close
to the star that its orbit is about four or five days,”
Charbonneau said. “One-tenth of this 1-percent will be
inclined in the right direction so that it will pass in front
of the star, so maybe one in 3,000 stars will have a planet we
can detect. Or if you want to be
conservative, about one in 6,000.”