My father taught us this as a simple conjecture of sorts that somehow always stuck with me: "in a frozen lake the warmest layer of water is at the bottom and it's at 4°C".
Now there might be a plethora of cases where this is not exactly true, as is often the case with simplified models, but this is how I would word it to my kids :)
It is a simplified model. The obvious next question is why doesn't the entire lake freeze from the top down? Why does the ice not get progressively thicker every day? The next-level is to model the thermal insulation of the ice. Heat from beneath is insulated from the cold air above. With a constant supply of heat from below but a reduced heat transfer due to the ice/snow, along with reduced evaporation, it is theoretically possible for the depths of a lake to get warmer once frozen over.
At low altitudes, temperature reduces with height (in daytime the atmosphere is heated by contact with the hot surface of the earth (as the atmosphere itself doesn't really get heated by the sun). The hot air rises, and expands as it ascends (pressure reduces as you climb, less air pushing down from above) and therefore cools (adiabatic expansion). Therefore, in the troposphere at least, temperature reduces with altitude due to all this vertical air movement and heating from below.
This is contrasted with the stratosphere, where the temperature begins to rise with altitude again and therefore is very stable and stratified without much vertical air movement.
Somewhat different physics at play, but funny how there are similarities.
As a child I used to main line OU lectures on Saturday mornings.
One was on Ecology, and had about thermal inversion in a lake in the Lake District when the temperature of the lake fell to 4ºC and the thermal circulation changed, between above 4ºC and below.
We already have the frozen and boiling temperature of water or any other liquid.
I remind you that boiling water is 100c and not higher as the energy is used to change phase. (at 1atm and close to the sea level)
You can use different materials for the temperature you are more interested in if you want more precision, we don't need to go under bodies of water to calibrate thermometers
I think they meant that hypothetically, presuming a case where we didn’t have access to better methods. The question might be rephrased as “would this be accurate and reliable enough to calibrate a thermometer if someone cared to try?”. I might be wrong, but it’s how I interpreted it.
