NavList:

De: NavList@fer3.com [mailto:NavList@fer3.com] En nombre de FrankReedCT@aol.com
Enviado el: martes, 12 de diciembre de 2006 22:32
Para: NavList@fer3.com
Asunto: [NavList 1893] Re: Astronomical Algorithms C++ source code

"A collection of freeware C++ classes which provide an implementation of the algorithms as presented in the book "Astronomical Algorithms" (2nd Edition) "

I just want to mention again that this *may be* the long way around to do this sort of thing in the year 2006. The JPL ephemeris data, the result of extraordinarily detailed and accurate numerical integrations, can easily be stored at one arcsecond accuracy in 40 kilobytes per year of data, maybe less if you put your mind to it. The only code you need to access it is trivial interpolation code. You need actual data only at one or two hour intervals for the Moon, once a day for Jupiter, etc. The cost to cover longer periods of time is strictly linear and the accuracy (at least from the point of view of ephemeris time) remains the same for thousands of years. This is not the case with algorithmic approximations to the orbits such as those published by Jean Meeus or the VSOP87 files. And by the way, the official Nautical Almanac also uses the direct JPL ephemeris.

Note that the above only applies to Solar System objects. You'll still want to apply precession, nutation, aberration, parallax, and proper motion to the stars themselves in code. Fortunately, that's not a big job.

Of course, there's nothing intrinsically wrong with using algorithms, like those published by Meeus. If you are coding for a device where storage space is in very short supply and you intend to cover more than fifty years of data, then algorithms like those will probably save you some storage.

-FER
42.0N 87.7W, or 41.4N 72.1W.
www.HistoricalAtlas.com/lunars