[LEAPSECS] A New Year's Eve parable

[Rob Seaman]

Another New Year's Eve and the familiar talking points continue to swirl around on the Leapsecs list.  Let's try something different, a gedanken experiment.

Definition:  "The [SI] second is 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 cesium 133 atom."

Posit a cranky physicist who absolutely refuses to use this definition, but rather demands that a second be defined as some completely different and non-denumerable value, say, 50,000,000,000 periods of this transition.  Timers are built to match this new definition.  Experiments are performed.

The result?  Precisely the same physical laws are determined.  Interval time scales.

Now, posit a cranky astronomer.  The astronomer is only a little crank.  Instead of shifting the definition by 40807368230 periods (444%) , the astronomer just wants to shift the definition of a "day" by 1 unit, from 86400 "floosblats" to 86401 floosblats (0.001%)

The result?  In ten years the day's zeropoint has shifted by 3652.5 floosblats, more than an hour.  For instance, at midday the Sun is 15 degrees away from the meridian.

Civil timekeeping is cumulative.  Tiny mistakes posing the problem will result in large and growing permanent errors.

Best wishes for all in the coming year!

Rob