I've done several sunrise/sunset programs. Refraction is a significant
factor and a source of significant uncertainty. The best of the
astronomical computation references that I've found says that computing
and displaying sunrise/sunset to closer than the nearest minute will be
wrong most of the time. That long path through the atmosphere is least
predictable because the refraction depends on moisture at all elevations
along the path, as well as the height to the top of the atmosphere.
Surely humidity varies all over the world different every place and day
and elevation. Then the depth of the atmosphere depends on pressure at
the surface, and the most recent solar insolation. Greater solar
activity also affects the depth of the atmosphere. This also shows up in
changing orbits of low earth orbit satellites, that are sometimes
affected by drag from the atmosphere after the atmosphere is puffed up
by solar bombardment. Makes the low earth orbit satellite arrive early
compared to the predicted computation.
There's another very significant uncertainty for most locations except
on the sea and that's the elevation angle to the horizon. Its rarely
exactly zero unless swimming with the centers of the eyeballs at the
surface of a calm sea. Everywhere else that elevation varies according
to the terrain and elevation of the observer with respect to the local
and distant terrain. Hills setting the horizon make for late sunrise and
early sunsets.
My latest version (used on weather pages at www.weather.net) checks to
within 20 seconds plus or minus from central Iowa with my western
horizon set by a "ridge" about 100 feet higher in elevation out at 7
miles. Earlier versions including one I translated from Pascal in a Byte
magazine written by a astronomer was up to four minutes off.
73, Jerry, K0CQ
--
Entire content copyright Dr. Gerald N. Johnson, electrical engineer.
Reproduction by permission only.
|