NavList:

John, in the Martelli thread, you asked:
"Can anyone shed light on the accuracy of using an almanac that is exactly four years out of date? I know this is the basis for many long term Almanacs - HO 249, Bowditch, Geoffrey Kolbe etc."

Sun tables are the only thing that repeat in a four-year cycle. Even a single year's almanac can be used for the Sun in other years if you know a simple rule for shifting the time look-up. For example, you would get the Sun's declination by looking 6 or 12 or 18 hour earlier or later in a year that doesn't match the published year (in this case, GHA is a little more complicated than the old Dec/EqT system). The four-year repetition is reasonably good for three or four cycles (12-16 years) into the future (unless the years 1800, 1900, 2100, or 2200, etc. are in there since those aren't leap years). After that, you need a subsidiary table or a fresh set of Sun tables. Star tables are reasonably good for three to five years into the future and get progressively worse due to precession (and nutation) and on a longer time scale proper motion. Precession is a simple rotation of the whole celestial sphere, so a fix generated from older star data is just shifted some miles away, in some azimuth, from the correct fix. With suitable short tables (e.g. the P&N tables in Pub.249), one can correct for this. By avoiding fast-moving Arcturus and Alpha Centauri (Rigil Kentaurus), proper motion is not a real concern. Naturally planet and Moon tables are no good for any years except the published date.

-FER

----------------------------------------------------------------
NavList message boards and member settings: www.fer3.com/NavList
Members may optionally receive posts by email.
To cancel email delivery, send a message to NoMail[at]fer3.com
----------------------------------------------------------------