I spent last week at the Physizkentrum Bad Honnef on the Rhine, near Cologne, at conference with the name and wide-ranging remit “Astrophysics, Clocks and Fundamental Constants“. We were all there to talk about the ideas and technologies that relate those disparate fields: Can we measure what we think of as the fundamental constants of the universe (for example, the speed of light, c, or Newton’s constant of gravitation, G)? Here on Earth, this becomes the discipline of metrology, the science of measuring things. Traditionally, we would do this by comparing lengths to some standard meter stick, or masses to a standard kilogram. In fact, there is a kilogram in a vault outside of Paris, against which all others are meant to be compared, but which seems to be drifting away from the other standard kilograms elsewhere!

So the more modern way is to define our units of measure in terms of well-understood physical systems. As of 1967, the second has been defined as “the duration of 9,192,631,770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium-133 atom”.

Although the metrologists worldwide aren’t quite there yet, the eventual goal is to define all units this way, or by defining the physical constants to have specific numerical values. For example, the speed of light is defined to be c=299,792,458 meters per second. When combined with the definition of the second in terms of caesium atoms, this, in turn, defines the meter. We can then cascade our way to the kilogram, the ampere, and the other fundamental units by suitable combinations of physical systems and defined constants. In fact, the latest revision of the constants was released last week. When this is all done, in a few years, this will make up the “New SI” system of units.

The meeting was full of careful, painstaking physicists who are trying to make these measurements to a precision of a few parts in 100 million or better, including some synchronising atomic clocks which reminded me of my very first work at an observatory doing laser-lunar-ranging to measure the distance to the moon and laser-satellite-ranging as an alternate way to synchronise clocks. In some contrast, I talked about cosmology, where we are happy to refer to measurements of the density of the Universe to 1% as “precision cosmology“.

However, there is an exciting middle ground in a few areas of astrophysics. Millisecond pulsars, spinning neutron stars which I wrote about in my recent reminiscence of Don Backer, are clocks nearly as good as the best yet built in a laboratory on earth.

There was also a series of talks about attempts to measure the variation of these so-called constants over cosmological distances and times, a variation not present in most of our theories, but not logically impossible. There was a group representing a contingent at the University of New South Wales who have consistently, with very careful measurements of quasar absorption lines, found changes in the fine structure constant (α≃1/137), most recently and very unexpectedly varying across the sky. I admit that I am nearly certain that this result will turn out to be incorrect. But, as the chair of a session at last week’s meeting, I did my best to encourage a robust dialogue between the UNSW group and their interlocutors who claim to have found no such variation. Needless to say, no one’s mind was changed, but I think the rest of us understand the disagreement at least a little bit better. And I admit that I am dissatisfied with both sides’ explanations — the disagreements come down to the understanding and treatment of so-called “systematic errors”, the often-unavoidable errors that creep into these complicated measurements due to inaccuracies of calibration, and of the way the instrument is built. After listening to a few hours of discussion, my meagre understanding is that the Aussies may be slightly too cavalier in their treatment, but the other group seems far too conservative, essentially defining their result as contaminated and therefore unsuitable for any sort of further statistical treatment. I look forward to the next generation of experiments so that we can reduce these errors so that we can get or refute the extraordinary evidence required for these extraordinary claims.

Finally, on an unrelated note, here’s one of my favorite rooms from my visit to Köln’s Museum Ludwig of Modern Art:

(Not as spectacular as Gerhard Richter‘s stained glass in the Cathedral, but easier to photograph.)