NavList:

If you use Skyfield for python then you can load the JPL ephemeris. In fact, you can load several different versions of JPL's ephemerides depending on your needs. Here are some of them from the docs:

Ephemeris	Size	Years	Issued
de405.bsp	63 MB	1600 to 2200	May 1997
de406.bsp	287 MB	−3000 to 3000	May 1997
de421.bsp	17 MB	1900 to 2050	February 2008
de422.bsp	623 MB	−3000 to 3000	September 2009
de430.bsp	128 MB	1550 to 2650	February 2010
jup310.bsp	932 MB	1900 to 2100	December 2013

It's pretty easy to build an almanac generator with this package although you do have to spend some time thinking about time to get it right.

Ken Muldrew.

On 2017-10-16, at 7:33 AM, Bill Lionheart wrote:

I have been using the python ephem package for ephemeris and it is
based on code from Xephem, with a discussion of accuracy here
http://www.clearskyinstitute.com/xephem/

I have seen other discussions of accuracy of ephemeris software on
NavList and I am sorry if this has been covered before

1) Where can I find an analysis of the effect of errors on the
intercept  method? Hence the answer to the question "How accurate does
ephemeris need to be for celestial navigation?" . I know we are
limited by time and angular resolution in taking sights but we can
average over a series of sights to reduce this error, but after errors
in reading the sextant and chronometer is the next biggest error going
to be variations in the atmosphere? Suppose we are using celestial
navigation at a fixed point on land and we can average over different
atmospheric conditions.

2) Is there a "gold standard" ephemeris used in astronomy? It strikes
me in meteorology the gold standard must be to check predictions with
observations. Presumably there are observatories making high precision
measurements with minimal errors (for example when objects are near
the zenith to reduce refractions).

--
Professor of Applied Mathematics
University of Manchester
http://www.maths.manchester.ac.uk/bl