A solar pond is a way of capturing thermal energy from the sun. It is basically a pool of water, with two layers:
-a high density salt water layer at the bottom
-a lower density, non-salty layer at the top.
The bottom water heats up from the sun’s rays, and the top layer stays cool. You can then run a heat engine on the temperature difference between the layers. My question: why don’t convection currents mix the layers?
I am by no means an expert, but some quick googling led to this page on Wikipedia. According to the article, there are actually 3 layers, the top-most is fresh water, the middle has a salinity that increases as the depth increases, and the bottom layer has a uniform salinity. The middle layer acts as a barrier to keep the convection mixing from occurring.
The mixing is prevented because the salinity gradient makes the water denser as the depth increases, while the temperature gradient tries to make the density decrease as the depth increases. These 2 effects cancel each other out, preventing the mixing from occurring.
In solar ponds there is also a salinity gradient, but in the opposite direction of what you mention. You want the high salinity water on the bottom, because it is more dense, and under the right concentration of salt and temperature it will not lose enough density to cause a convection current. Your top layer is your low-salinity layer because likewise, it will not sink due to a higher density than your bottom layer.
There’s also a third layer between of intermediate salinity, that helps to separate the two.
… ETA: yes, what Mines Mystique said.
Plagiaristically like Sherlock Holmes.
August Derleth shoulda been sued.
While I understand the basic concept I was always kinda surprised that diffusion wouldn’t render the things unoperable in relatively short order. The perfectionist in me couldn’t help but put a thin layer of clear plastic between the the different layers of water.
My theory is they work like magnets 
Curse you! I wanted to be the one to make a Solar Pons joke!