NavList:

   

Reply

On 5/31/2013 4:43 PM, Geoffrey Kolbe wrote:
> I would be intrigued to know how Gary was supposed to account for leap
> seconds....? And why (for navigation purposes) he _should_ account for
> leap seconds...?

Navigation or not, he has to compare the watch's time to some standard.
One can view UTC on the computer or listen to shortwave broadcasts
(complete with a double tick for each 0.1s fast or slow). The easiest is
UTC or UT1. With DUT1 0.1 s off the last I checked, and the human
ability to catch differences limited to approx. 0.2 s, he can ignore UTC
vs UT1 now.

Why one can't ignore leap seconds? I'll try an example. Suppose a watch
tracks UTC exactly. Zero error if no leap seconds happen during the
observation period.

Now lets suppose a leap second occurs during that observation interval.
Several members of Navlist watched the last leap second happen on their
computers (nist.time.gov), or perhaps their radio controlled (atomic)
timepiece. The time went from Xm 59s to Xm 60s then to X+1m 00s if I
recall. The hypothetical perfect watch continues to tick away, oblivious
to the leap-second stutter step, and is now 1 second off UTC (It is
unlikely a $17 watch has RC).

Nit picking in this case to be sure, but I was doing a 60-day test on
several watches when the last leap second occurred, and it did make a
significant difference.

In many circles I am the Grand Poobah of OCD, but doubt I'm even an
honorable mention in this discussion group :-)

Bill B

   

Reply