CSIRO astronomer George Hobbs and colleagues in the UK, Germany and Canada report that they have taken a big step towards solving a 30-year-old puzzle: why the "cosmic clocks" called pulsars aren't perfect. "We now have a more fundamental understanding of how pulsars work," Dr Hobbs said.
"We've shown that many pulsar characteristics are linked, because they have one underlying cause."
Armed with this understanding, astronomers will find it easier to compensate for errors in their pulsar "clocks" when they use them as tools - for instance, in trying to detect gravitational waves, which is something Dr Hobbs is doing with CSIRO's Parkes radio telescope.
The work is based on observations of 366 pulsars collected over several decades with the 76-m radio telescope at the Jodrell Bank Observatory, run by the University of Manchester, and grew out of work George Hobbs did for his PhD thesis.
Pulsars are small spinning stars that emit a beam of radio waves. When the beam sweeps over the Earth we detect a "pulse" of radio waves. The rate at which the pulses repeat, fast or slow, depends on how fast the pulsar spins and so how often its radio beam flashes across the Earth.
Each pulsar generates a cocoon of magnetic fields around itself - its magnetosphere.
The astronomers found that a pulsar's magnetosphere switches back and forth between two different states.
"We don't know exactly what happens," Dr Hobbs said.
"But one idea is that from time to time there is a surge of charged particles-electrons, for instance-whirling through the magnetosphere. Such a surge could apply the brakes a bit to the pulsar spin, and also affect the pulsar's radio beam." The change in a pulsar's magnetosphere shows up both in the shape of the radio pulses recorded on Earth and the pattern of the pulses' arrival times.
"Pulsars are very stable timekeepers, but not perfect," University of Manchester, George Hobbs' PhD supervisor and lead author of the Science paper Dr Andrew Lyne said.
"They have what we call 'pulsar timing noise', where the spin rate appears to wander around all over the place. This had baffled people for decades."
One of the aims of Dr Hobbs' PhD thesis was to find an effective way to filter out this 'timing noise'.
"We worked out how to do this, and along the way we were prompted to think hard about the nature of the timing noise," Dr Hobbs said.
The key advance was noticing that when the pulsar timing changed, so did the shape of the radio pulse. "This ran against accepted thinking," Dr Hobbs said. "Everyone had said they were unrelated. But we've shown they are."
Now astronomers can compensate for 'timing noise' by looking at the pulse shape change to spot when the pulsar magnetosphere has changed its state: this will show when the pulsar spin rate has also changed.
Dr Hobbs says there is no explanation yet as to why a pulsar's magnetosphere flips from one state to another.
"The switching seems random in some pulsars and regular in others," he said.
"We haven't solved all the mysteries yet."