[LEAPSECS] Testing computer leap-second handling

[Rob Seaman]

On Jul 9, 2012, at 8:24 AM, Poul-Henning Kamp wrote:


> In message <46D53F0B-BF98-46C0-A485-4A1494E2CE88 at noao.edu>, Rob Seaman writes:

> 

>> More deeply engrained yet is the simple fact that "day" on any

>> planet, dwarf planet, or (spheroidal) moon means the synodic day.

> 

> Yes +/- 4 hours or so.



The rotational period of Mercury is 58.6 earth-days.  A day on Mercury is 175.9 earth-days.  (The irony of using that unit is not lost on me.)  Mercury is in a spin-orbit resonance.  There are 1.5000 rotations per year, but 2 years per day. The meaning of the word "day" is not unclear on Mercury.  It is the same meaning as on Earth.

A rotation of Venus is -243 earth-days (it rotates retrograde).  Day on Venus is -117 earth-days.  The Sun rises (in the west) and rises again 117 earth-days later.  Were there a civilization on Venus, its many diurnal dependencies would be tied to the mean solar day, not Venus' rotational period.

Closer to home, the Moon rotates in 27.3 earth-days.  Day on the Moon is 29.5 earth-days.  Planning for the Apollo missions depended on both, but the astronauts' time outside the LEM was scheduled by the mean solar (synodic) day.  When we say it's full moon, we mean it's noon on the Moon (at the terminator).  Full moons arrive "like clockwork" 29.5 days later.  There are twelve months in a year, not thirteen.

Meanwhile, apparent time is a red herring.  A static or periodic offset from the local mean solar time drops out in the rate, and the rate is another thing that the ITU proposal gets wrong.

The Martian day is an analogue to the Earth's, both roughly 24 Earth-hours long. Where the mean solar day on Earth is 4 minutes longer than the sidereal rotation period, the Martian day is about 2 minutes longer than Mars' rotation period.  The equation of time, however, has about three times the amplitude on Mars as on Earth due to Mars' more eccentric orbit.  The periodic effects are larger, but the mean solar clock ticks on steadily nonetheless.

No day is 24 TAI hours long:

	http://en.wikipedia.org/wiki/File:Deviation_of_day_length_from_SI_day_.svg

The length-of-day on the y-axis of that plot is the mean solar day.  This may seem inconvenient, but it is a simple fact.

Rob