Very cool! Would a day/night cycle be too difficult to implement on this simulation (like heat absorbtion with different constants, due to air humidity)? By heating and cooling, i think it can generate some interesting fenomena.
The model I'm using here is at constant temperature. It should not be that difficult to add a temperature field, though it would slow down the simulations a bit. But it would make the simulation more realistic, so I might try it.
