We often think of our Sun as a typical middle-aged "garden variety" star,
quietly shepherding its planetary companions through billions of years of stable
evolution. Such benign conditions have allowed life on Earth to evolve to its
present diversity of forms. However, a new study of Sun-like binary star systems
may indicate that some of these "typical" stars are not as benign as we once

Dr. Brian D. Mason, an astronomer at the U.S. Naval Observatory (USNO) in
Washington, DC, and his collaborators have gathered compelling evidence that
these star systems tend to break apart over time. These results are being
presented today to the 203rd Meeting of the American Astronomical Society in
Atlanta, GA.

Dr. Mason worked with collaborators William I. Hartkopf (USNO), Todd J. Henry
(Georgia State University, Atlanta, GA), and David Soderblom (Space Telescope
Science Institute, Baltimore, MD) to observe more than 2,300 stars that are
similar to our Sun, but spanning a range of ages of about 10 billion years. The
ages of these stars were determined by Dr. Soderblom, who used indicators of
magnetic activity in the stars’ chromospheres, similar to the mechanisms
responsible for spots, flares, and powerful magnetic disturbances on our Sun.

Roughly two out of three stars in the night sky are not solitary travelers
through space; they have companions, forming systems with two or more stars
orbiting a common center of mass. In this context the Sun is an anomaly, since
it is accompanied by a rag-tag assemblage of small bodies (planets, asteroids,
and comets) whose total mass is far less than that of even the feeblest star.

Young solar-type stars show much higher levels of activity in their
chromospheres than our Sun currently does, and there is a steady decline in such
activity as these stars age. The observations to measure this activity were made
at Kitt Peak National Observatory near Tucson, Arizona and the Cerro Tololo
Interamerican Observatory in Chile. Both are operated by the National Optical
Astronomy Observatories for the National Science Foundation.

The key observations in this study were led by Dr. Mason and were done to
determine how many and what kind of companions the observed stars have. The
technique used is called "speckle interferometry", and involves taking a very
rapid series of images of bright stars to record individual pinpoints of light
produced by distortion in the Earth’s atmosphere that move too fast for the
human eye to see. These pinpoints are correlated by a computer into a single
image, allowing full use of a telescope’s capabilities. A typical image from a
ground-based telescope would appear as a blur of starlight, but the correlated
images produced by the speckle camera can reveal the presence of close companion
stars within the circle produced by the blurring distortion. The speckle
interferometry camera is optimized for detecting binaries at the resolution
limit of the telescope, so it is well suited for examining these stars.

The results from this study are being used to examine the details of the kinds
of companions that Sun-like stars have, but the most important finding is shown
by taking the full sample and placing these stars into four age bins based on
their chromospheric activity. As shown in the graph (Figure 1), there is a clear
and continuous drop in the frequency with which stars have companions. In the
youngest bin, 18% of stars have companions; this falls to 10% for the next bin
(age about 1 billion years), then to 7% at about 4 billion years, and to 3% for
the oldest stars (6 billion years and older).

This decline is unambiguous, but the reason for it is less clear. The oldest
stars could have formed with fewer companions in the first place, but there is
no basis for believing that. However, astronomers are aware that there are
factors that disrupt clusters of stars in our Galaxy, leading to them being torn
apart with time. The observations reported here suggest that the same or similar
factors cause binary companion stars to be lost over time. Serendipitous to
these investigations was the discovery of eleven new binary systems, one of
which is shown in Figure 2.

The detection of doubles by speckle interferometry is limited to companions of
similar brightness and will not detect doubles which are quite close. However,
these limitations apply to all data bins. The error bars on multiplicity are
rather large for the very active and very inactive samples due to the small
number of stars in those respective bins, though the other two bins are quite
well populated.

This work has been supported by NASA and the SIM preparatory science program
through NRA 98-OSS-007. We gratefully acknowledge this support.

[NOTE: Images supporting this release are available at
http://www.usno.navy.mil/pao/press/Stars%20Lose%20Their%20Companions.pdf ]