[LEAPSECS] LEAPSECS Digest, Vol 51, Issue 7

[Warner Losh]

On 02/02/2011 12:23, Rob Seaman wrote:

> The word "mean" here doesn't refer to averaging statistically independent observations.  Mean solar time is simply time based on the synodic day, that is, the underlying sidereal rotation period of the Earth adjusted by one day per year from lapping the Sun.  It is the sidereal rate, measured against "fixed" stars or cosmologically distant quasars, that is varying due to tidal and other geophysical effects.  The synodic day, and thus "mean solar time", go along for the ride


Well, what you are describing is an averaging over time.  There are so 
many revolutions of the earth as it goes from a given point in its orbit 
until the next time it is at that point.  These revolutions would be 
around 366.23<mumble> times for a year.  If we subtract one, we get the 
average length of the day for that year.  This average would estimate 
the true mean for true mean for that year, but would be different from 
what was computed 100 years ago or 100 years from now.  It would also be 
different if you looked at 10 years instead of 1 year to come to your 
answer.

That's what I'm asking when I ask about the mean solar time's averaging 
period.  It has to be specified, explicitly or implicitly, to be meaningful.

Warner