I know of no studies regarding aeration/thermoclines to cite here, so I can only throw SWAGs around.
I think 2 CFM through a decent diffuser would draw at least some of the waters up from the 10 foot depth mixing this 39 degree water with the surface waters. The way I think about it is that the diffuser is acting like a pump (about 700 to 1000 gpm for discussion) moving the water upward. The water going upward has to go somewhere and, for the most part, it goes outward along the water's surface (it cannot go directly back down against the upward current). The water going upward has to come from somewhere and that is from below. Most is from directly below, but there are no strong confines to keep it from drawing up some from the depths. Some will be drawn from the bottom. The unknown is how much. Now, set the diffuser 18" down with a large plate a foot or so below it and you have created a confine that keeps any water from being drawn upward from down deep. Water goes up and meets the water/air interface, spreads out along the surface, but must be drawn back from the sides at a shallower depth. This creates a micro boil, if you will, that stays up in the water column.
Would something like this server your purpose?
If you do not restrict the available water flow from beneath, I fear, that you will affect the thermocline to at least some degree.
I also think using multiple diffusers spread out with a lot less air throughput could better suit you purpose. Less air per diffuser means less "pump" power, means less water movement, which means it's less likely to draw water up from deeper depths. This could also be done by moving the diffuser closer to the surface, but it's less efficient than spreading that air out over multiple heads.
Bottom line is that I think 1 diffuser at full throttle will move water up from 10 foot deep. Key word is "think".
