Assuming a simple model with an atmosphere of uniform global temperature, and assuming that the layer of sea ice accommodates all of the temperature difference between the surface/air and the still unfrozen water which is I assumed to be isothermal at near freezing, with simplifying assumptions about the properties of ice as an insulator and heat flux from the mantle,

I related the heat flux from the cold surface/air layer to the heat flux from the mantle, assuming the isothermal ocean had reached a state of constant temperature

$h(T - t_0 ) = Q_g$, where

• $h$ is a thermal property of the ice ($k/d$ where $d$ is the ice thickness)
• $T$ is the constant water temperature of near 275 K
• $t_0$ is the external temperature
• And $Q_g$ is the heat flux from the mantle

From this, I calculated that in this very simple model the required temperature of the air to create a layer of sea ice of thickness d can be found with this expression:

$t_0 = T - (Q_gd)/k$

Using a thickness of 4000 m, which would be a layer of sea ice that fills most of the ocean, I guess that the surface temperature of the Earth would have to be around 120 K for the oceans to freeze solid.

I don't think that I am right- obviously much is neglected in this model and some of my assumptions may not be warranted. Clearly, the Earth would have to be much colder than any time in its history for the oceans to freeze at great depth, but my question is would it have to be even colder than I have guessed, or could the oceans freeze completely or almost completely at warmer global temperatures than 120 K?

这很难，而且不经常发生。以前整个地球都是冰冻的，但那是因为大部分大陆都在赤道，风化速度非常快。< / p >

"Temperatures on a snowball Earth are estimated to have reached minus 58 degrees Fahrenheit (minus 50 degrees Celsius). As the ice spread, more heat was reflected back into space rather than absorbed by the planet, dropping temperatures down in a runaway effect that sped the formation of ice." (Space.com)

Certainly for the ocean to freeze entirely, I think first the entire ocean would have to cool to near freezing even in the tropics.

This technically really hasn't ever happened before in geological history, due to the sun and heat influx from volcanos. (Considering there was landmass in the equator). If the ice somehow created a runaway affect, this would mean that the ice reflected back more heat/sunlight than it received. P.S: If your not taking land into account then the oceans would for sure freeze at some point (not now though) :)

Now for the grand question: (Drumroll)

How cold would the Earth have to be for the oceans to freeze entirely?

So obviously there weren't any thermometers back in the day so we can only estimate. It's estimated it was minus 58 degrees Fahrenheit (minus 50 degrees Celsius) was the temperature for snowball earth (Plus a major redistribution of continents.)

So that wraps up our answer for today, please let me know if my answer is not in line with your question.