Pulsars are rotating neutron stars that produce highly periodic bursts of radio waves. So accurate are pulsar signals that when they were discovered, astronomers gave serious credence to the idea that they were evidence of intelligent life elsewhere in the Universe because they were unmatched by anything physicists could make on Earth. This has lead to the widespread belief that pulsars are the most accurate clocks in the Universe.

40 years later, astronomers have yet to work out exactly how pulsars generate such accurate signals. But physicists on the other hand, have been working hard to find their own ways to better the performance of pulsars.

Today, John Hartnett and Andre Luiten at the University of Western Australia ask whether Earth-bound time pieces have usurped their astrophysical rivals as the best clocks in the Universe.

On the face of it, the answer is pretty clear cut to anybody who has followed the amazing advances in quantum optics in the last few years.

"The accuracy and stability of terrestrial clocks have improved more than an order of magnitude, on average, in each decade over the last 60 years," say Hartnett and Luiten. Today, the best optical lattice neutral atom clocks and trapped ion clocks have a frequency stability approaching one part in 10^17.

By contrast, as more pulsars have been discovered, their timing stability has improved by less than an order of magnitude in the last 20 years. The best millisecond pulsars have a stability of only one part in 10^15 at best.

That means that terrestrial clocks can rightly be crowned the best clocks in the Universe, say Hartnett and Luiten.

That's impressive but there is one other issue to consider before physicists in quantum optics labs can start popping champagne corks. This is the question of long term stability.

It's all very well to build a clock that can outperform pulsars for a few months or years but try it for a significantly longer period of time, say centuries or millennia, and a whole host of other issues raise their heads. The makers of the Clock of the Long Now have already studied this issue. They've asked how you can guarantee a stable power supply over such a period? How do you store spare parts or ensure that the knowledge to effect a repair survives? Can you even rely on the survival of the human race over these periods?

The answers to these questions suggest that it will be very difficult to run a wristwatch let alone a trapped ion clock over this time-scale. And yet in thousands of years pulsars will still be producing their regular heart beat.

Earth-bound clocks may be able to outperform pulsars over human time-scales but to do it over significantly longer time-scales is another challenge altogether. Terrestrial clocks may have stolen the crown for now. Keeping it will be much harder.

Ref: arxiv.org/abs/1004.0115: A Comparison of Astrophysical and Terrestrial Frequency Standards: Which are the best clocks?