NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Skyfield
From: Paul Hirose
Date: 2018 Oct 25, 12:58 -0700
From: Paul Hirose
Date: 2018 Oct 25, 12:58 -0700
On 2018-10-23 17:15, Josh Paterson wrote: > Using the method in skyfield.almanac I get: > 2018-Dec-21 22:22:44.1118 UT Solstice at 22:22:44.1117 UTC per the SofaJpl demonstration program, JPL DE431 ephemeris, IAU 2006 / 2000B precession nutation model. The Sun moves about 1°/day in ecliptic longitude, or about 1e-5 °/sec. Thus, 1e-8 degree precision in longitude should allow single milliseconds of time to be resolved. I set that as the program's precision. It's phony precision, since IAU 2000B nutation is accurate to only about a millisecond of arc, or 3e-7 degree. Therefore, I believe my 3rd decimal place is valid within a few units, and the 4th is noise. The excellent agreement with Josh is probably coincidence. I'm not sure what I ought to expect, since Josh has not stated which ephemeris and precession / nutation model he's using. I assume his "UT" is UTC. (At the time of the solstice, UT1 and UTC will be only a few hundredths of a second different.) JPL Horizons puts the solstice at 22:22:45.246 UTC, about 1.1 second later than Josh and I got. It uses the same JPL ephemeris that I used, so I suspect the difference is due to precession and nutation.
