I still find it slightly puzzling that during the winter there’s only a few millimeters/centimeters of ice on a lake, but not more.

I simply find it interesting how little heat is lost between the ice and the (liquid) water. Maybe I should sit down some afternoon, look up all the relevant constants and compute the actual heat transfer rate to figure out, after how many weeks/months/years with an air temperature of -5° a “standard” lake would turn into a solid block of ice.

Of course, if the water was constantly being “stirred” (as is the case in most rivers), then the temperature distribution in the lake would be more even and it would take much longer to freeze over, but when it does it should probably turn into a single block of solid ice pretty much instantly. (It might also be fun to figure out, how much of the kinetic energy of the river is turned into heat. If it wasn’t for this dissipation the river would accelerate more and more towards the sea and would flow into the sea at a velocity of several hundred kilometers per hour.)