Hi,

On 2021-03-30 05:38, Hal Murray wrote:
> https://www.npr.org/2021/03/29/982417680/researchers-are-one-step-closer-to-red
> efining-the-second
> Audio is 2:54
>
> Researchers with the Boulder Atomic Clock Optical Network Collaboration are 
> one step closer to replacing the current atomic clock and officially 
> redefining the second.
>
>
> KENNEDY: Ultimately, the goal is to redefine the second in terms of a more 
> accurate and more precise standard.
>
> SHAPIRO: Kennedy, Ye and their colleagues compared three of those 
> next-generation atomic clocks, which use atoms of aluminum, strontium or 
> ytterbium.
>
> CHANG: The goal was not to see which atomic clock was the best ticker per se 
> but to show that these types of clocks can be compared from one lab to 
> another, perhaps from one country to another.
>
> KENNEDY: We want to make sure that a clock built here in Boulder is the same 
> as a clock built in Paris as in London as in Tokyo.
>
There is multiple problems to solve.

One is the repeatability in building a standard. You do not want subtle
differences to have a huge impact on the result of your clock, then you
can't really have a primary for this purpose. You want many labs to be
able to build their own or buy commercial devices. It cannot be limited
to a selected few, as that would result in a situation similar to the
kilogram weight issue.

Secondly, clock species comparison is a key to the path-way for
selection. Already at 8FSM conference we could see the aggregation of
comparison, but it was also obvious is was unsatisfactory in that
confidence intervals did not fully agree and sort things out. That has
improved. It was still not satisfactory last time I saw it, but
improving. I think that has progressed since.

Third, the Cs-optical comparison remain the hardest to push down, as you
have roughly two degrees in difference in noise and systematics. As you
make the transition, the error here will be inherited into history. By
doing inter-species comparisons, the Cs-optical comparisons done can be
used to ensemble-access the Cs in terms of the optical transitions. The
actual improvements in direct comparison and the inter-species
comparisons have improved the state of things.

Fourth, you want the standards to have good operation ratio.
Essentially, for EAL and TAI contribution you essentially want
continuous operation. It used to be far from that. I know multiple labs
have been looking at it, and in particular USNO has that as an explicit
concern. Just getting a optical clock to work have been quite a bit of
hand-tuning and taking time. That used to exist also with the PTB
optical chain that was locked to cesium, it could take days to start up,
and then you had maybe a few hours of lock time. It was not
satisfactory, but you could do some important measurements with it. It
was a pain, which is what the PTB old-timers tells me.

A number of different approaches to new SI definition of Hz (was second)
was put forward, by among others me, and that range from a single
species to actually use a ensemble of species. The later would have a
benefit in using the advancement of several species and allow a larger
range of clock realizations. Errors made in the early assignment could
be somewhat reduced. Others have rejected the idea, thinking it is too
radical. Regardless, I think there needed to be a richness to be
discussed, so that is why I offered multiple models for how to do it
along with concerns.

While optical clocks still is laboratory devices, the exclusiveness of
them is slowly decaying, and while we have yet to see the full-blown
commercial products, a lot of things shows that the world is heading
that way. The cold rubidium clocks is a step in the right direction.
It's a great technology and to some degree share the teething problems
with the optical clocks. Still, it will be a bit far before I see these
things in my basement.

It's a bright future! (Pun not intended, but then again it was accepted)

Cheers,
Magnus
_______________________________________________
time-nuts mailing list -- time-nuts@lists.febo.com -- To unsubscribe send an 
email to time-nuts-le...@lists.febo.com
To unsubscribe, go to and follow the instructions there.

Reply via email to