NavList:

What do you exactly mean by "calculating almanac data"?
Can you explain how you calculate Moon's coordinates
using 1448 bites??

The formula for the Moon that I am find in Meus contains
hundreds of terms!

I used Brown's theory from Astronomy on the Personal Computer, Astronomical Formulae for Calculators, N.A.O. Technical Note Number 48, and the Explanatory Supplement to the Astronomical Ephemeris.

 

I am looking at Astronomical Algorithms, 1991, p 309.  If I were using it as the basis, I would begin by doing three things.  First, I would change the summed terms of the form "Asin (D+M+M'+F) + Bcos (D+M+M'+F)" to the alternative form "Csin (D+M+M'+F) + E" and lump the Es to save a lot of keystrokes.  Next, I would cut the coefficients down to about 0.0001 degree and toss out the terms smaller than that.  Why have so many significant digits ?  Third, for the slower varying terms (those which don't change much during my lifetime or maybe between 1900 and 2100) I would substitute a constant.  I thus lose accuracy that I do not need, but I end up with a significantly shorter algorithm. 

 

When computers were small and expensive, people used to worry about this sort of thing.  I guess they worried a lot more when they were doing the almanac calculations by hand under candle light.

 

The 1448 byte subroutine takes ephemeris time and calculates the moon's equatorial longitude, ecliptic latitude, and radius then converts them to S.D., H.P., R.A, declination, and phase.  In the TI-82 a byte is about a keystroke.

 

Bill Murdoch