Ross McCluney wrote:
.....
> EOT = .170 sin (4*pi*(J - 80)/373) - .129 sin (2*pi*(J - 8).355)
> 
> where the arguments of the sine functions are in radians and J is the
> number of days since December 31, and EOT is given in decimals hours.
> .....


Dear Ross,

I received your formula for the EoT as above.

I assume the last term must be ( J - 8 ) / 355.

I was curious about this formula, because I didn't expect the numbers
373 and 355 on those places. I expected the number of days in a year.
( 365 or 365.25 ).

So I calculated the results of the formula and my computer compared them
with the results of my procedures.
The + and - differences are shown below for the years 1997 - 2012.
When you accept this accuracy you may use this formula during a long
time.  
( My value - your value )

1997  19.38 sec at day 1.5   -45.34 sec at day 365.5
1998  26.37 sec at day 1.5   -38.26 sec at day 365.5
1999  33.37 sec at day 1.5   -31.17 sec at day 365.5
2000  40.39 sec at day 1.5   -24.30 sec at day 366.5
2001  19.12 sec at day 1.5   -45.61 sec at day 365.5
2002  26.10 sec at day 1.5   -38.54 sec at day 365.5
2003  33.10 sec at day 1.5   -31.45 sec at day 365.5
2004  40.12 sec at day 1.5   -24.57 sec at day 366.5
2005  18.86 sec at day 1.5   -45.89 sec at day 365.5
2006  25.83 sec at day 1.5   -38.82 sec at day 365.5
2007  32.83 sec at day 1.5   -31.74 sec at day 365.5
2008  39.85 sec at day 1.5   -24.84 sec at day 366.5
2009  18.59 sec at day 1.5   -46.16 sec at day 365.5
2010  25.57 sec at day 1.5   -39.10 sec at day 365.5
2011  32.56 sec at day 1.5   -32.02 sec at day 365.5
2012  39.57 sec at day 1.5   -25.12 sec at day 366.5

Lateron in the century the max error isn't always at the first or last
day but can fall on another day. See example for 2097.


2097  26.41 sec at day 71.5  -52.29 sec at day 364.5

The change in the error from year to year shows the change in the real
EoT, because your formula always give the same values.

Don't look at the decimals, the computer just printed them as well.

 
Fer J. de Vries.

Reply via email to