Over on Science of Doom there has been a long thread about radiative flux and ocean heating. Frank mentioned this in suggestions, so I’m reposting his comment here for discussion as S.o.D didn’t engage with this reply from Frank:
Frank said: November 11, 2010 at 5:48 pm
Tallbloke, Cynicus, SOD, et al: Is there one explanation that could satisfy all? An attempt to reconcile divergent views:
Cynicus wrote (11/7 8:11 pm):
The longwave heating of the top ocean layer has even been measured as reported by RealClimate: http://www.realclimate.org/index.php/archives/2006/09/why-greenhouse-gases-heat-the-ocean/
This RealClimate post contains experimental evidence showing that the temperature difference between the surface of the ocean and 5 cm below the surface varies with net long wavelength radiation. The greater the observed imbalance between upward and downward long wavelength radiation, the colder the surface is compared with 5 cm below the surface. These observations demonstrate ONLY that DLR warms the top few microns of the ocean. This experiment demonstrates just what Tallbloke asserts and nothing else! The RealClimate author (Peter Minnett) SPECULATES that:
“Reducing the size of the temperature gradient through the skin layer reduces the flux” [of energy from the ocean to the atmosphere].
This statement ASSUMES that we understand the main mechanism of heat flux through the skin layer of the ocean. IF conduction – which varies with the temperature gradient – is the main mechanism of heat transfer, then increasing DLR will: lessen the gradient, decrease heat flux through the skin layer, and warm the ocean. If heat transfer occurs by what SOD called forced convection (by wind, waves, eddies?), the steepness of the temperature gradient could be irrelevant. Free (buoyancy-driven) convection could increase with the steepness of the gradient, but surface tension might need to be overcome first. Therefore, Minnett has not PROVEN that the DLR warms anything other than the “skin layer” (top 10 um) of the ocean.
Where Tallbloke Gets It Right (with some caveats): Does it make any difference if DLR only warms the top 10 um (hereafter called the “skin layer”) of the ocean? The ocean is still being warmed, isn’t it? No, Tallbloke believes that the energy from DLR is immediately returned to the atmosphere – before that energy can be transferred to deeper layers of the ocean. Tallblock is certainly correct; the surface of the ocean is exchanging long wavelength photons with regions of the atmosphere that are colder (and have lower emissivity) AND losing additional photons directly to space when it isn’t cloudy. Therefore, long wavelength radiation MUST BE a net loser for the top 10 um of the ocean because this skin layer both absorbs AND EMITS ALL long wavelength radiation. There is not “too much” DLR being absorbed in the skin layer (with its miniscule heat capacity). There isn’t ENOUGH DLR to replace the energy lost through upward long wavelength radiation from the skin layer. In a previous post, I showed that shortwave radiation from the sun could deposit enough energy in the skin layer to negate the net loss at long wavelengths for a few hours around noon, but not when averaged over a whole day. This explains why Minnett’s data shows the skin temperature to be an average of 0.2 degC COLDER than 5 cm below the surface. There are a few data points on Minnett’s graph showing that the temperature difference is occasionally zero, which is [remarkably] consistent with my back-of-the-envelop calculations showing short periods of net warming via the small fraction of short wavelength radiation that is directly absorbed by the skin layer.
Equilibrium Considerations (aka Tallbloke’s Waterloo): Although the ocean’s temperature varies with day and night and the local weather, we can define an equilibrium temperature for the ocean as the temperature at which the AVERAGE net upward loss of energy (which increases with ocean temperature by oT^4 and evaporation) is exactly balanced by the AVERAGE downward energy input from radiation (both solar and DLR). When the ocean maintains an equilibrium temperature, the average net energy loss from the skin layer via long wavelength radiation must be supplied by energy from average short wavelength radiation. Short wavelength radiation is absorbed mostly in the top 10 m, with only a small fraction of that energy actually being directly deposited in the skin layer itself. (On the average, DLR directly deposits far more energy in the skin layer than does sunlight). We have been debating the mechanism of (and proper name for) the energy transfer from the top 10 m of the ocean to the colder skin layer, but most readers should recognize that such a transfer must take place: If it didn’t, the skin layer would continue to lose energy through long wavelength radiation to the colder atmosphere and space and the 10 m immediately below would get hotter and hotter. I don’t know whether this energy transfer occurs by conduction (Millett’s assumption), free convection, or forced convection – but the exact mechanism is irrelevant when we are discussing EQUILIBRIUM temperature. The mechanism IS relevant to how fast temperature returns to equilibrium after it is disturbed (for example by night and day or by GHGs), but it doesn’t change the equilibrium temperature needed to balance downward and upward flow of energy.
CONCLUSION: When increasing DLR from GHGs reduces the net loss of energy by long wavelength radiation from the skin layer, the EQUILIBRIUM temperature of the 10 m immediately below MUST warm: If some of the energy that previously went to the skin layer (and then to the atmosphere) doesn’t go anywhere else, it must increase the temperature of the 10 m immediately under the skin layer.
What Part of the Ocean Warms in Response to GHGs and How Fast? (Some of our disagreements may be arise because we are thinking differently about these parameters.)
a) The top 10 um: The best fit equation on Millett’s graph shows that a 4 W/m^2 increase in DLR – the forcing for 2X CO2 – is expected to reduce the difference in temperature between the skin layer and the water immediately below by a negligible 0.008 degC. So the temperature of the skin layer and the next 10 m should change in parallel.
b) The top 100 m (or the physically “mixed layer”): We know that seasonal changes in insolation cause seasonal changes in the temperature of roughly the top 100 m of the ocean. If increasing DLR reduces the need for energy flux from the top 10 m to the skin layer, the extra warmth in the top 10 m clearly can spread throughout the physically mixed layer within a year – even if the total warming is as small as the 0.02-0.05 degC/year projected by the IPCC.
c) The next few hundred meters: We are principally worried about climate change over the next century. Presumably some warming in the mixed will slowly penetrate to these depths by conduction or, more likely, by vertical mixing as ocean currents pass over obstructions on the sea floor.
d) Deeper: If we concern ourselves with only the next century, warming in the mixed layer of the ocean is probably not going to reach the deep ocean. With a circulation time of about 1000 years, the meridional overturn current (or thermohaline circulation) that conveys surface water to and from the deep ocean appears to be too slow to convey more than a small fraction of surface warming to the depths.
In summary, absorption of increasing DLR by the skin layer will reduce the need for energy flux from the top 10 m to the skin layer. This will warm the top 10 m and this warmth will be distributed throughout the top 100 m in less than one year. Increased DLR will not significantly warm most – but not all – of the remaining ocean in the next century.