I noticed this at some point--other people already knew about it. G.I. Taylor actually described the situation for a fully submerged circular plate in the journal of applied physics circa 1952). One source calls them Falaco solitons (the word soliton is controversial for some reason or other). Basically the surface of the water appears to have two opposing vortices stuck near to each other. What you can't see under the surface is basically half of a ring vortex like a smoke ring or something, which keeps the surface vortices from moving away from each other. This is caused when the circular disk (ceramic plate) is dragged through the water. The water in front of the plate is pushed forward, while the water to the sides of the plate stays where it is. At the boundary between these two relative flows there is a sheering force, a curl, which manifests as rotation of the little parcels of water in the area. So this invisible smiley face part of the vortex connects the two surface features in a single 3 dimensional soliton. The appearance of two separate features in both the 2-d surface of the water and the 2-d projection at the bottom of the pool is really neat in my view. The image on the floor is just the image of the sun projected through the lens formed by the surface of the water.
