NavList:

Bill you wrote:

"My next query, will there be horizontal refraction as well?"

 

Luckily no. Consider: which way would it go? North, South? By the symmetry of the atmosphere alone, there shouldn't be any "sideways" refraction, it has to be all in the vertical direction (except under "weird" atmospheric conditions which would probably guarantee cloudy weather, too). Refraction lifts all stars. It compresses the constellations towards the zenith.

And you wrote:
"In Meeus's chapter on refraction he gives an example of the reduction of
the Sun's vertical observed measurement when near the horizon due to
refraction.  He then states, "...the horizontal diameter of the solar disk
is very slightly contracted by reason of the refraction.  This is due to the
fact that the extremities of this diameter are raised along vertical circles
that meet at the zenith."  He goes on to mention Danjon, "...writes the
apparent contraction of horizontal diameter of the Sun is practically
constant and independent of altitude, and  that this contraction is
approximately 0".6."

 

He's talking here about a small effect that arises from ordinary (vertical) refraction. Imagine modelling the Sun as a diamond of four stars. One at the top of the Sun, one at the bottom, and one each on the horizontal limbs of the Sun's disk. In the absence of refraction, this is a perfectly symmetrical diamond -- if the distance from top to bottom is 30 arc minutes, then te distance from side to side is also 30 arc minutes. OK, now throw in refraction. All four stars are pushed towards the zenith though by varying amounts depending on their distance from the horizon. The stars on the sides of the diamond are both raised by the same amount, but they're being pushed towards the zenith from slightly different azimuths. The horizontal distance across the diamond is decreased very slightly. Meeus quotes an amount, but you can confirm it yourself using your star-star calculation and this "diamond" of stars that I'm describing.

 

By the way, parallax, for the Moon, Sun etc. is also an entirely vertical correction so you can go immediately to the case of lunar distance calculations from you star-star calculations, if you want. There's one small issue: the Earth's oblateness yields a slightly non-vertical component to parallax but that can be dealt with separately.