So is refraction a cause for why the bottoms of distant objects are obscured? I've always been a little confused as to what actually causes the "hidden" area to exist on a flat earth. And if refraction is solely a flat earth phenomenon.
For me, I just can't help referring back to the Rainy Lake experiment which is pretty rigorous in terms of eliminating, or accounting for as many variables as possible to give flat Earth a fair chance:
http://walter.bislins.ch/bloge/index.asp?page=The+Rainy+Lake+ExperimentRefraction, if we are to all agree is a phenomenon we observe, happens in both RET and FET to largely the same extent for the same reasons. The refractive index of air at different densities and the refraction coefficient on average is pretty well known. Over water especially, and where there are higher temperature differences between the surface and the air, you tend to get more refraction that (typically) causes light to bend down towards the Earth, and in some cases follow the curvature (i.e. causes you to see things much further away than you would do normally, making it appear that the Earth could be flat).
In a flat Earth model, you could reasonably argue that light from the bottom of a tall tower could get refracted down towards the surface of the Earth before it reaches your eyes, making it appear as though the bottom of the tower is indeed below the horizon. This effect would be compounded on a round Earth due to curvature and refraction. Without knowing what the atmospheric conditions were, pressure, temperature etc., and not knowing the refraction coefficient, it is hard to say in those images how much is refraction and how much is curvature.
This is why I return to the Rainy Lake experiment, because there they are on a frozen lake, known conditions, measured refraction coefficient to account for it, with targets of accurately known heights and shapes that are set where refraction should be minimised.