Sometime between 2002 and 2005, while working with Joerg Hipp at Caltech, I discovered a phenomenon that must have been seen by ancient astronomers already, but which I have so far failed to find described: the fact that Earthshine (the dimly lit part of the Moon lit by light reflected from Earth) appears to be part of a circle of smaller diameter than the sunlit, bright part of the Moon.

Below you can see the phenomenon proved:

The blue circle above and below is the diameter of the Moon shown by Earthshine, which is clearly smaller than that shown by the sunlit part of the Moon.

It's not clear that this phenomenon is always present. It seems to happen only when the sunlit part of the Moon covers a small minority of the Moon's surface, but not when it's about half of it.

I do not know the cause of this, which could be due to an astronomical phenomenon or a perceptual illusion. Yet I hypothesize that Earthshine falls below the level detectable by the human brain as the curvature of the Moon becomes too large to reflect enough light back to Earth. In contrast, the sunlit part of the Moon receives much more light, and from a different angle than the Earth, and so can reflect enough light to be perceived up to closer to the edge.

Another possibility I considered is "bleeding", where the light bleeds into surrounding space in the sunlit portion, for example because of reflections in stellar particles. But this phenomenon appears to happen in addition to and distinctly from the difference in the diameters of the fully lit circle, even excluding the bled halo.

-Alex Bäcker, Ph.D., Maui, October 4, 2019