[time-nuts] Detecting gravity with optical atomic clocks

[Tom Van Baak]

"Paul Bicknell" --
> Any chance of a picture of your Synchronome pendulum clock and associated
> timing / logging equipment 

Yes, I'll update that page with more info and photos at some point.
Meanwhile there are some links to follow at the end of the page:
http://leapsecond.com/pend/synchronome/quake.htm

----

"Didier Juges" --
> Tom,
> I suspect something so sensitive gives you significant "false positives"
> when a delivery truck goes by. I assume you try to correlate your data with
> other enthusiasts nearby to resolve those discrepancies the way we do with
> our clocks?

The pendulum is directly bolted to large thick basement corner wall. Local door slams or delivery trucks have no effect that I've ever seen. But a 7.0 earthquake is truly massive and it creates ground motion at the many microns to mm levels even 1500 miles away. A precision pendulum clock is affected by this level of vibration, especially when it persists for many minutes.

Yes, the pendulum data correlates in time with professional seismometer stations here and around the northwest.

----

"Hal Murray", "Poul-Henning Kamp" --
> And can you reverse-engineer the local ground movement from the
> pendulum measurements ?

For quartz, rubidium, and pendulum clocks, it is possible to partially reverse-engineer effects of temperature, pressure, and humidity. These are scaler quantities and very slow moving processes.

Seismic effects on pendulums are a whole different problem. It's a 3D vector quantity. They are very dynamic (rapidly changing), with a complex power spectrum. And the interaction of ground acceleration with a swinging pendulum is extremely dependent on angles and on the instantaneous pendulum phase vs. seismic power relationship, which is changing every millisecond and lasts for minutes. Plus the pendulum reaction to some of these changes is non-linear. As Bob would say, lot's of fun.

In theory if you had several pendulums arranged around a circle, and used ultra wide band seismometers (or super high resolution accelerometers), and took measurements at 10 to 100 Hz, then I suspect computer simulations might be able to make some predictions out of the data. Which you could then back out.

/tvb