Exosat: Europe s New X-ray Satellite
France A. Cordova, Los Alamos National Laboratory, and Keith O. Mason, Mullard Space Science Laboratory, England
THE EARLY MORNING mist en-shrouding Vandenberg Air Force Base in California on May 26, 1983, was sud-denly parted by a rocket's dazzling glare. Thundering skyward, this particular Delta launcher bore aloft the European Space Agency's X-ray astronomy satellite rL-Exo-sat — and with it a decade of work by sci-entists and engineers. In the months that followed, they put the craft through an ex-tensive period of trials and calibrations. And now, about one-third of the way through its observing program, Exosat has already contributed substantially to our knowledge of neutron stars, white dwarfs, and other exotic objects.
Exosat is unique in the scope of the X-ray investigations it can tackle. Sources can be studied continuously for as long as 72 hours and events as short as 10 microseconds can be discerned, over an energy range stretching from 40 to 50,000 electron volts (corresponding to wavelengths be-tween 300 and 0.25 angstroms). The satellite can map extended objects and locate point sources to within a few arc seconds. It can also provide spectral information to determine the temperature, density, and chemical abundances of the emitting gas. All of this capability has not come about easily, however, as outlined in the box on the next page.
During its planned three-year lifetime, Exosat will be operated much as ground-based observatories and astronomical satellites (like the International Ultraviolet Explorer) are. For the mission's first nine months, observers were chosen from among the European astronomical community, but now the satellite observatory is available to astronomers worldwide, under the direction of project scientist Tony Peacock.
Exosat's inherent flexibility paid scientif-ic dividends in the first two months after launch. Throughout this calibration period, its instruments observed objects that had been examined repeatedly with previ-ous X-ray satellites. Among these was Hercules X-l, a low-mass binary system that consists of a neutron star (spinning once every 1.2 seconds) in a 1.7-day orbit around its F-type, optically variable com-panion HZ Herculis. X-rays are emitted as gaseous material is drawn from HZ Herculis onto the neutron star.
Astronomers were already aware that the X-ray source stays bright for about 10 days, then dims for another 25 days. The popular explanation of this on-off cycle is that the X-ray emitter is periodically hid-den from us by an accretion disk sur-
rounding the neutron star. In this model the disk is tilted with respect to the orbital plane of the binary system and precesses, or wobbles like a spinning top, with a 35-day period (see diagrams on page 399).
However, Exosat's observations soon showed that even old favorites, seemingly understood, could offer a few surprises. The spacecraft was pointed at Her X-l in June, July, and October of last year when the X-ray source should have been "on." Instead, the instruments recorded émissions characteristic of the "off" state — the fîrst known instances of the source fail-ing to switch on as predicted.
Normally, X-ray émission from the neutron star heats the adjacent hemisphere of its companion, causing it to appear bright-er in visible light than the other hemisphere. It is the alternate présentation of these two faces that causes the optical var-iability. However, several times this cen-tury (and most recently from 1949 to 1956) HZ Herculis has remained optically faint because, it was thought, X-ray production had stopped. And after Exosat's observations, the attending astronomers thought such was again the case.
Soon thereafter, the optical companion was examined by members of the Ameri-
can Association of Variable Star Observers (AAVSO), who found it still bright. Con-sequently, accretion must still have been going on, with the F star continuing to "see" X-rays even though Exosat did not. Most likely, the accretion disk has thick-ened to the point that it continually hides the neutron star from Exosat. Results ob-tained in early March showed that Her X-l had switched back on. The implications of this behavior are not yet clear.
In contrast, the neutron-star component of another binary, the 5.2-hour-period 4U 2129+ 47 (the désignation is based on its coordinates), apparently has stopped accreting. Wolfgang Pietsch from the Max Planck Institute and his colleagues discov-ered last October that this system's X-ray flux had hit an unprecedented low. But unlike HZ Herculis, the visible component was also unusually faint, indicating that X-ray production and subsequent heating of the normal star had ceased.
IDENTIFYING A RECURRING X-RAY TRANSIENT
As demonstrated so vividly during earlier surveys by satellites such as Vela 5A and 5B, Ariel 5, and Uhuru, the X-ray sky continually flickers with the "light" of transient sources. In some instances these stars are, for a few days, among the
Author France Cordova uses sophisticated image-processing software to display and analýze Exosat data at the Mullard Space Science Laboratory, University College, London. All photographs are by the authors unless otherwise credited.