Things that go bump in the night
January 21, 1998: Centaurus X-3 has been acting up, but until recently, no one knew about it.
Right: An artist's concept of a star feeding matter to an accreting pulsar.
"Instead, what we find is a very strange behavior where the frequency gets faster at a constant rate, then gets slower at a constant rate," Finger said. "And it's not correlated to anything else we've observed."
Other stars like Cen X-3 - accreting pulsars - have been acting strangely, too. The spin rate for 4U 1626-67 was increasing smoothly, then took a dive. 2S 1417-634 rose like a rocket, and GS 0834-430 sputtered on and off and ever upward.
Finger, Dr. Robert Wilson, also of NASA/Marshall, and several other scientists recently published their findings from watching accreting pulsars with the Burst and Transient Source Experiment (BATSE), launched aboard the Compton Gamma Ray Observatory (right) in 1991.
While BATSE was designed to watch for bursts of gamma ray - radiation with more energy than X-rays - it observes anything that releases enough energy in the 20 to 70 keV range (visible light has an energy of about 1 eV). BATSE sees anything brighter than 15 millicrabs (15/1000ths the brightness of the Crab Nebula). Special computer codes let BATSE scientists extract the true signatures of bursts. They also can extract information from a number of non-burst sources, including pulsars and even terrestrial thunderstorms.
Radio astronomers discovered pulsars - pulsating stars - in 1967. They were stunned by the clocklike regularity of the objects. That's what makes BATSE's pulsar observations so unusual.
"Radio pulsars tend to be regular," Wilson said. "Accretion pulsars are much more complicated."
The emissions from radio pulsars are powered by the star winding down from the drag of their magnetic fields. Accretion-powered pulsars are driven by mass transferred from their larger stellar companions which continue to evolve and change.
While you can almost use the Crab Nebula and other pulsars as a time standard (it is used as a brightness standard), the accreting pulsars observed by BATSE behave like cheap alarm clocks, gaining and losing time. While most of the data graphs hint at a pattern, no one can yet figure the inner workings to see what sets the patterns.
The pulsar retains the star's magnetic field (right; from NASA/Goddard) which now is more intense - about a trillion times stronger than that of Earth - because it has the mass on the sun crammed into a volume about 20 km (12 miles) in diameter. The surface gravity also is equally intense, and anything that hits gives off an incredible flash of energy, usually from the magnetic poles where the magnetic field lines emerge from the surface.
(The Earth's magnetic field generating the aurora borealis is basically the same mechanism, but far weaker.)
As a neutron star rapidly rotates, the hot spot at the magnetic pole shines like a lighthouse (check out our pulsar tutorial!) giving off everything from radio waves to visible light to X-rays. Accretion-powered pulsars rotate too slowly for this to happen. Material falling onto the the tiny collapsed object can give enough energy as it falls into the deep gravitational well of the neutron star and interacts with the magnetic field lines, matter already crowding the polar region, and then the star itself.
Now the story gets complicated. Before BATSE, observations with other instruments indicated that accreting pulsars were not as regular as radio pulsars, but the phenomenon was not widely known: the data were spotty, with just a few hours of observations here and there across two decades. As telescopes became more powerful, no one could afford to dedicate months or even weeks to observe one object. Most astronomical studies are snapshots compared to the long, complex lives of astronomical bodies.
Everybody do the twist (or anti-twist)
A good example of the unexpected behavior is Cen X-3, the third X-ray source found in the constellation Centaurus and one of the most studied X-ray pulsars before BATSE became available. It is a binary system with an orbital period of 2 days and accompanied by a visible companion called Cen V779.
Slight changes in Cen X-3's rotational period of 4.82 seconds (12.4 rpm) are caused by Cen V779 dropping matter onto Cen X-3. The challenging part is that Cen X-3 doesn't just spin up. It also spins down. That means that the pulsar is subjected to negative torque, or a twisting force against its spin.
Wilson added that, "It should lead to a new understanding of these systems, but the results are just too new."
Right, an enlarged view of how an accreting pulsar might look to a nearby telescope. (Courtesy Dr. Malcolm Coe, University of Southampton, U.K.)
Several models have been suggested and none have been accepted. One was that the accretion disk - the collection of gas flowing outward from Cen V779 (right) - collapses onto Cen X-3, then reforms in the opposite direction. Another tries to work out what happens when materials are simply orbiting the pulsar and then run into the effects of the magnetic field lines which sweep around space every 4.8 seconds in time with the pulsar. If the matter is going too slow, the magnetic field can propel the gases into space, and the star loses some speed.
Another mystery is 4U 1626-67 (the numbers refer to the fourth source catalog of the Uhuru satellite, and the location in the sky) whose spin rate was steadily rising for all of the last 15 years. Since the start of the BATSE observations, it has declined.
"This demonstrates the danger of extrapolating data," Wilson said.
"It's strange because it's frequency changes are very smooth and regular," Finger said, "almost like a radio pulsar. Yet you had this switch." The system is small, with the pulsar spinning every 7.7 seconds and its companion, a lightweight dwarf called KZ TrA, orbiting every 41 minutes.
"The system is so small that it's hard to understand how it has this incredible long-term memory and then just changes," Finger wondered.
|This plot summarizes all outbursts of transient pulsed sources seen by BATSE since launch of the CGRO (using optimal analysis techniques and longer observation intervals than 1 day where necessary, not directly obtained from standard daily Monitor).|
In total, Finger, Wilson, and others have observed 15 of 39 known accreting pulsars with BATSE data, and discovered another 5 for a total of 44 in the astrophysicists' bestiary.
The data set is unique because it is the first long-term set of observations of these stars. It will help astronomers in figuring out exactly what goes on in accreting pulsars, including the important question of how viscous (or, thick) is the plasma in the accretion disk as it swirls around the pulsar and then funnels down the magnetic field lines to the poles.
"This paper is causing a renaissance in the study of these systems," Finger said. "We think it's had a tremendous impact in the study of these systems."
Observations of Accreting Pulsars. Lars Bildstein (University of California at Berkeley), et al. The Astrophysical Journal Supplement Series, 113: 367-408, December 1997
(return to story)
Author: Dave Dooling
Curator: Bryan Walls
NASA Official: John M. Horack