I’ve mentioned I’m at a conference — it’s the 7th Symposium of Frequency Standards and Metrology being held near Monterey. It’s a bunch of scientists getting together every ~7 years to discuss the state-of-the art in frequency standards, clocks, and precision measurements, and float ideas for future experiments. The last one was in St. Andrews, Scotland in 2001 (unfortunately it spanned 9/11/2001, which was a bit of a distraction, to say the least.)

There have been a lot of talks that I couldn’t possibly distill into coherent summaries, but I’ll try to do one or two when I get the chance. I’ve got one for now, though, that doesn’t require as much heavy lifting.

Dan Kleppner gave the first talk (Is Gravity Ruining Time?) as a sort of introduction, and gave some perspective on timekeeping, since he has been doing physics from before the development of the hydrogen maser (making him, as he put it, prehistoric). Two main things came out of this talk: an appreciation of a limitation on how we define the second, and a story about I.I. Rabi.

The second is defined as 9,192,631,770 oscillations between the hyperfine states of an unperturbed cesium-133 atom, but this definition does not explicitly mention anything about relativity, of which gravity is a part. It’s basically taken by convention that we use devices at rest on the geoid (an idealized surface of the earth, basically what it would look like without tides) but devices have reached the point where this may not be good enough. The gravitational redshift is given by gh/c^2 near the earth, and this is about a part in 10¹⁶ per meter change in height. Clocks need to be adjusted for their altitude/elevation, and this has been necessary for some time; the effect has been measured in the Pound-Rebka experiment and in the rocket launch of a hydrogen maser by Robert Vessot, and is accounted for in GPS and every other satellite carrying a clock. But ground-based clocks are now getting to be good enough to where sub-meter changes in height will need to be taken into account. And since the geoid can only be determined to several cm and it changes with time (and clocks move with respect to the geoid via earth tides of about 30 cm), this will soon become a significant term in the error budgets of frequency standards. So the point of the talk was that gravity is going to take a more prominent role in frequency and time measurements, and may in fact require a redefinition of the second, though it would not impact “everyday” time.

The story he told about Rabi went something like this: Rabi didn’t like writing articles, so there is no formal writeup of his proposal to use an oscillator tuned to a hydrogen transition as a time measurement device — the idea that would eventually become the hydrogen maser and used in other atomic clocks. But in 1945, after he had the Nobel prize, he gave the Richtmyer lecture to the American Association of Physics Teachers on the topic of using a hydrogen magnetic resonance measurement as a potential timekeeping device, and it was written up by the New York Times science correspondent, William Laurence, in an article called ‘Cosmic Pendulum’ For Clock Planned, in which he gives a very basic summary of the principles Rabi had explained. So the cutting-edge science was “formally” proposed in the Times rather than a science journal. In the AAPT’s list of Richtmyer lectures, Rabi’s is one of the few from that era that were not written up and presented in the American Journal of Physics.

(The Times article is here but the archive is paywalled)