"Roger Bailey" <[EMAIL PROTECTED]> asked on Monday, April 18, 2005 5:50 AM
Subject: Sonne Software for Shepherd Dials
<snip>...

> My only problem so far is the equation of time correction. I could not
find
> this option for this type of sundial. Has it been included? Did I miss it?
> Has any one else included this option in a similar program to give
"standard
> time" on a cylinder dial?
>
> Regards,
>
> Roger Bailey
> Walking Shadow Designs
> N 48.6  W 123.4

I'm not aware of any cylinder dial with this correction, but, by
coincidence, I have been investigating adding Equation of Time correction to
a totally different type of dial which has exactly the same problem. That
problem is that the same line is normally used to indicate two hours, one
before noon, the other after, and on two dates when the sun's declination is
the same. So, if you move the line to adjust for EoT before noon, the
adjustment will be in the wrong direction for the same hour after noon. And,
worse still you would think, it will be wrong for the other date with the
same declination.
So, instead of having hour lines for just half the day (the same line being
marked, for example, both 11am and 1pm) and half the year, you really need
separate 'am' and 'pm' scales, and a whole year date scale. That needs four
times as much area as a standard pillar or shepherd's dial.

But I can suggest a way to halve the area needed, if you can accept an error
of up to two minutes. This takes advantage of the symmetry of the analemma
and of the sun's altitude in the morning and afternoon. Whatever the sun's
declination, the EoT values on the two dates with that declination are
generally very similar, but with opposite sign. For instance, at the
equinoxes the EoT is about -7:30 in spring, +7:30 in the autumn. So the 11am
hour line in the spring is almost identical to the 1pm hour line in the
autumn.
So, while you do need separate am and pm scales for any date, these can be
switched half way through the year. The 'am' scale for Jan-Jun is also the
'pm' scale for Jul-Dec. The 'am' scale for Jul-Dec is also the 'pm' scale
for Jan-Jun. Using the average absolute value of the EoT for any declination
is never as much as 2 minutes wrong, and usually less than 1, which is
better than you can probably read the dial to anyway.

Let me know if you need further clarification.

Chris Lusby Taylor
51.4N, 1.3W

-

Reply via email to