Atomic clock with 2x10E-18 uncertainty in fractional frequency units

[CharonY]

The authors report to achieve a three-fold improvement over previous clocks.

 

http://www.nature.com/ncomms/2015/150421/ncomms7896/full/ncomms7896.html

 

 

[John Cuthber]

Well, thank goodness for that!

Six parts in a million million million just wasn't good enough.

[StringJunky]

Here's the dumbed down version for the rest of us. :http://phys.org/news/2015-04-atomic-clock-accuracy.html

[swansont]

Standard disclaimer: it's not a clock, it's a frequency standard (real clocks run continuously). But I'm happy to explain details as best I can, if anyone has questions.

[John Cuthber]

Cool. given the nature of time dilation with gravity, how good an altimeter s that clock?

(i.e. how big a change in altitude, near the Earth's surface, gives rise to a change in the rate of passage of time of 2* 10^-18?)

[swansont]

Cool. given the nature of time dilation with gravity, how good an altimeter s that clock?

(i.e. how big a change in altitude, near the Earth's surface, gives rise to a change in the rate of passage of time of 2* 10^-18?)

 

The gravitational effect is about 10^-16/m near the earth's surface, so this would be good to ~2 cm. It's at the level where the question of "where is this clock" can become ambiguous if the atoms aren't well localized.

 

A clock like this should be able to measure solid-earth tides (several tens of cm) if it ran long enough, and if it could be compared with a clock that was fixed, or possibly measure the diurnal fluctuations of the position in the sun's gravitational field (varying by some fraction of the diameter of the earth, depending on where the second clock was). I don't recall that number off the top of my head, but they were talking of possibly doing this measurement at NIST and somewhere else. The optical fiber link comparisons are getting good enough to do this.

 

A couple of years back they compared two of these standards side-by-side, but with one optical table hoisted up by a~30 cm (IIRC), and measured the frequency difference.

[John Cuthber]

So, since their lab seems to be in Boulder Colorado at about 1500 metres above sea level, and most people live in cities near sea level, it looks like they have the world's best clock, but it runs at the wrong rate from the point of view of nearly everyone on the planet.

Remarkable.

[swansont]

So, since their lab seems to be in Boulder Colorado at about 1500 metres above sea level, and most people live in cities near sea level, it looks like they have the world's best clock, but it runs at the wrong rate from the point of view of nearly everyone on the planet.

Remarkable.

 

You can and do adjust for height above the geoid, if this were a primary standard reporting to the BIPM. It's just another offset term.

 

For NIST's primary standard http://tf.boulder.nist.gov/general/pdf/2560.pdf(pdf) list the gravitational redshift as +179.15 x 10 ^-15 which is ~1650 m if you use 9.80 m/s² for g (Boulder is listed at 1655 m elevation)

 

There are a bunch of offset terms, because nobody can run their clock in the environment that defines the second (absolute zero, no fields, etc.). The analysis the standards labs do every so often is to measure these offsets and assess the uncertainties associated with them.

[CharonY]

A more general question, what is the impact of this improvement, i.e. is it something that you would see to be immediately useful for something potentially novel or rather e.g. for improving some current measurements?

[swansont]

A more general question, what is the impact of this improvement, i.e. is it something that you would see to be immediately useful for something potentially novel or rather e.g. for improving some current measurements?

 

The impact of these improvements will take time to manifest, since these are not primary standards (i.e. they don't use Cesium, that defines the second) but improvements in timekeeping have continued to reveal new applications over the years. From a basic physics standpoint, there are tests of General Relativity that can be improved with better clocks and with clocks that use different atoms, as they would respond differently to changes if some aspects of relativity were violated. As I mentioned above, sampling the gravitational potential in various positions (either from rotation or the orbit throughout the year) is one experiment people have done. Variations in the fine structure constant is something else that could be probed at a more precise level. Precise time is also crucial for very long-baseline interferometry.

 

There are also collateral effects, since the improvements in the laser systems needed for these clocks will improve other kinds of experiments. I was a talk on doing massively parallel spectroscopy with an optical frequency comb, so you can probe many wavelength at once, rather than scanning — one could imagine a device that could scan for volatile chemicals very quickly.

 

In a more practical arena, improvements in GPS will be one result of better timing (that's already "in the system", as it were, just waiting for the satellites to be deployed — the ground-based clocks for that are what I work on), since timing error translates to position error. But further clock improvements will find their way into this. As with a lot of basic research, that maturation to commercial (or otherwise deployable) technology will take a while.