NavList:

I was thinking as I was trying to fit a horizon line to a digital photograph about the incertainty in the horizon. On the photo we were looking at on an earlier thread the horizon certainly had an ambiguity of a pixel or so and I was not sure what line to choose. Its not so much the waves, as fFank said they are fairly random, but when you look at it with a camera what line best corresponds to the appriximation we make with dip angle. When viewing this through a sextant we have the extra complication that the human visual system has an "edge enhancing filter" (just look at the edge of a piece of paper on a dark background to see this), and it could be our visual system makes a choice and sees a much more distinct line than is there.

What is the angle subtended by the uncertainlty in the horizon even under the best atmospheric conditions, and is what we are looking at really the tangent to the average surface (averaged over waves)?