Arguments
How Increasing Carbon Dioxide Heats The Ocean
Posted on 18 October 2011 by Rob Painting
Much like a heated kettle of water takes some time before it comes to the boil, it seems intuitive that the world's oceans will also take some time to fully respond to global warming. Unlike a kettle, however, it's not obvious how the oceans warm.
Adding further greenhouse gases to the atmosphere warms the ocean cool skin layer, which in turn reduces the amount of heat flowing out of the ocean. Reducing the heat lost to the atmosphere allows the oceans to steadily warm over time - as has been observed over the last half century.
Warming on sunshine
Sunlight penetrating the surface of the oceans is responsible for warming of the surface layers. Once heated, the ocean surface becomes warmer than the atmosphere above, and because of this heat flows from the warm ocean to the cool atmosphere above. This process is represented in the graphic below:
Figure 1 - simplified steps of ocean heating
The 'cool skin' layer
The rate of flow of heat out of the ocean is determined by the temperature gradient in the 'cool skin layer', which resides within the thin viscous surface layer of ocean that is in contact with the atmosphere. It's so named because it is the interface where ocean heat is lost to the atmosphere, and therefore becomes cooler than the water immediately below. Despite being only 0.1 to 1mm thick on average, this skin layer is the major player in the long-term warming of the oceans.
Curious behavior in the cool skin layer
The cool skin behaves quite differently to the water below, because it is the boundary where the ocean and air meet, and therefore turbulence (the transfer of energy/heat via large-scale motion) falls away as it approaches this boundary. No longer free to jiggle around and transfer heat via this large scale motion, water molecules in the layer are forced together and heat is only able to travel through the skin layer by way of conduction. With conduction the steepness of the temperature gradient is critical to the rate of heat transfer.
Greenhouse gas-induced warming of the ocean
Greenhouse gases, such as carbon dioxide, trap heat in the atmosphere and direct part of this back toward the surface. This heat cannot penetrate into the ocean itself, but it does warm the cool skin layer, and the level of this warming ultimately controls the temperature gradient in the layer.
Increased warming of the cool skin layer (via increased greenhouse gases) lowers its temperature gradient (that is the temperature difference between the top and bottom of the layer), and this reduces the rate at which heat flows out of the ocean to the atmosphere. One way to think about this is to compare the gradient (steepness) of a flowing river - water flows faster the steeper the river becomes, but slows as the steepness decreases.
The same concept applies to the cool skin layer - warm the top of the layer and the gradient across it decreases, therefore reducing heat flowing out of the ocean.
The ever-present effect of the cool skin layer
An important point not be be glossed over here, is that changing the temperature gradient in the cool skin layer by way of greenhouse gas warming is a worldwide phenomenon. Once the gradient has changed, all heat leaving the ocean thereafter has to negotiate its way through the layer. With the gradient lowered, the ocean is able to steal away a little bit more from heat headed for the atmosphere. It is in this ever-present mechanism that oceans are able to undergo long-term warming (or cooling).
Experimental evidence for greenhouse gas heating of the oceans
Obviously it's not possible to manipulate the concentration of CO2 in the air in order to carry out real world experiments, but natural changes in cloud cover provide an opportunity to test the principle. Under cloudy conditions, the cloud cover radiates more heat back down toward the ocean surface than happens under clear sky conditions. So the mechanism should cause a decline in skin temperature gradients with increased cloud cover (more downward heat radiation), and there should also be a decline in the difference between cool skin layer and ocean bulk temperatures - as less heat escapes the ocean under increased atmospheric warming.
This was observed in an experiment carried out in 2004, aboard the New Zealand research ship Tangaroa. Using intruments to simultaneously measure the 'cool skin', the ocean below, and the amount of heat (longwave radiation) reaching the ocean surface, researchers were able to confirm how greenhouse gases heat the ocean. It should be pointed out here, that the amount of change in downward heat radiation from changes in cloud cover in the experiment, are far greater than the gradual change in warming provided by human greenhouse gas emissions, but the relationship was nevertheless established.
Figure 2 -The change in the skin temperature to bulk temperature difference as a function of the net longwave (heat) radiation. The net forcing is negative as the atmosphere is cooler than the ocean skin layer, but approaches zero under cloudy conditions. See Real Climate post "Why Greenhouse Gases Heat The Ocean" by Professor Peter Minnett.
Greenhouse Gases: On duty 24/7
The effect of greenhouse gases on ocean heat isn't confined to daylight hours however, they toil away around the clock. The warming of the oceans by sunlight, makes the daytime surface waters more bouyant than the cooler waters below and this leads to stratification - a situation where the warmer water floats atop cooler waters underneath, and is less inclined to mix. At night much of the heat accumulated during the day is lost back to the atmosphere (the overling air still being cooler than the ocean), and this cooling leads to the stratified surface layers sinking and mixing with lower layers. This allows the remaining heat to be transported down deeper into the ocean, causing an increase in ocean heat content over the long-term. The typical diurnal (day/night) cycle is seen in the figure below:
Figure 3 - Schematic showing the upper ocean temperature profiles during the (A) nighttime or well mixed daytime and (B) daytime during conditions conducive to the formation of a diurnal warm layer. Image from Gentemann & Minnett (2008)
Warming in the pipeline
Given the atmospheric lifetime of carbon dioxide is many hundreds to thousands of years, we can now understand that long-lived greenhouses will also continue to exert a warming influence on the worlds oceans for a very long time. Indeed, climate models suggest that ocean warming will continue for at least a thousand years even if CO2 emissions were to completely stop. See below:
Fig 5 - Time series of the (modeled) climate response to a cessation of CO2 emissions. a) global mean thermosteric sea level anomaly (b) and zonal mean ocean temperature at 792.5mtrs, 66 S (the Southern Ocean). Green line = cessation of CO2 at 2010 & red line = cessation at 2100. From Gillett (2011).
Ocean warming not just skin deep
Because of their effect on lowering the temperature gradient of the cool skin layer, increased levels of greenhouse gases lead to more heat being stored in the oceans over the long-term. This ocean warming mechanism has been observed experimentally, and is also supported by numerical modeling.
So although greenhouse gases, such as carbon dioxide, don't directly warm the oceans by channeling heat down into the oceans, they still do indeed heat the oceans, and are likely to do so for a very long time.
trunkmonkey @50 would like an explanation as to why the Indian, North Atlantic and Arctic are warming, while "All the other oceans are flat or cooling". I can offer no explanation for his claim is simply false:
As can easilly be seen, the South Atlantic, South Indian, and North and South Pacific are all warming quite strongly. Of course, Levitus et al (2012), the source of the graph, just divide the worlds oceans up into three oceans, so that the Southern Indian, Southern Atlantic and Southern Pacific oceans by their division includes the Southern Ocean. Perhaps it is flat or cooling?:
trunkmonkey, it appears, wants to claim falsehoods, and then fault climate science for not explaining why those falsehoods are true. No doubt he will next disprove climate science by pointing out, quite correctly, that it has no explanation as to why the moon is made of green chease.
trunkmonkey @55, got it. Your argument is that you will ignore data you find inconvenient.
I'm hesitant to weigh in on this, as poor Kevin may well feel that there is dog-piling going on, but I wish to address a phrase of Kevin's that KR has already commented on in #58, regarding the air a water being at the same temperature. KR points out that the water and air are not at the same temperature, but a key element I think frames why Kevin is leading himself astray at this point:
1) at the surface, which separates the water and the air, there is only one temperature, and this surface temperature will be the same regardless of whether you are looking at it from the perspective of the water or the air. This may be what Kevin has in mind whe he says the water and the air are at the same temperature.
2) The surface isn't water, and the surface isn't air. The surface is the plane between the two. The water is below the surface, and the air is above the surface. The fact that the water and the air share the same temperature at the surface does not mean that the water below and the air above are at the same temperature. Thus, KR is right.
3) Saying that the water and air are at "the same temperature", and restricting the statement to the surface, is as useful as saying "the air temperature at 2m is the same as the air temperature at 2m". It essentially means nothing - except that a graph of temperature through that point will not have a sudden jump or discontinuity. What matters is the gradients, and calculus tells us that lines can have gradients (slopes) at a sinlge point (i.e., we don't need two points to calculate a slope), and that tells us that gradients can exist at the surface. So again, KR is right: there are gradients at the surface.
(As an aside, chances are that the gradient in the water and the gradient in the air close to the surface will not be the same, so a plot of gradient versus height/depth likely will have a discontinuity. This is not a mathematical difficulty, and arises from consideration of all the energy flows towards and away from the surface,)
And Kevin: this is not a personal attack, but an evaluation of what I see from what you write here - I agree with Tom Curtis that your level of understanding leaves you ill-equipped in this discussion. You have fundamental misunderstandings that you need to unlearn before you can provide constructive analysis of what is being said to you.
Hmmm. While preparing what now appears as #53, I think some pruning of Kevin's sloganeering (and responses to it) may have occurred, as suddenly there are several missing comments.
guinganbresil #34 I think you don't understand because your thermocline graph lacks the detail that's the entire topic of discussion, the ocean "skin" and "sub-skin" that's shown in the
posting, but without scales that might give a clearer picture of what is happening. Sample numbers below will make it clear, I'm quietly confident.
Minnett & Kaiser-Weiss GHRSST 12-Jan-2012 has a graph of ocean skin temperature variation. Here are my sample or average global ocean numbers, forced to give the Sun's 161
w/m**2 exiting but nonetheless illustrative as nearly as matters.
temperature:
1,000m depth temperature = 5C
thermal conductivity of seawater 0.58 W/mK
ocean-air interface = 17.000C
1.441mm depth temperature = 17.400C (the warmest spot in the ocean depth though the "few metres" of depth below it is only a miniscule bit colder, all warmed by Sun SWR)
this top 1.441mm depth is the "skin" and "sub-skin"
temperature gradient of top 1.441mm of ocean is 277.6 Celsius/metre
By conductivity, temperature gradient pushes 161.00 w/m**2 up from 1.441mm depth to ocean-air interface which precisely removes the Sun's 161 w/m**2 going into the top few metres
depth and leads to no ocean warming.
AGW increases downward LWR and air temperature directly above ocean with extreme rapidity of a few decdes, which warms ocean-air interface by 0.700C, so:
ocean-air interface = 17.700C
1.441mm depth temperature = 18.097C
temperature gradient of top 1.441mm of ocean is 275.5 Celsius/metre
By conductivity, temperature gradient pushes 159.79 w/m**2 up from 1.441mm depth to ocean-air interface which leaves 1.21 w/m**2 of the Sun's 161 w/m**2 going down into the ocean
below and leads to ocean warming of 13.8 ZettaJoules / year (the billions of atomic bombs in sks widget). So, the difference of 0.003C in the warming over the top 1.441mm of ocean
causes ocean warming that is 7.5 times as fast as the average post-glaciation ocean heat gain that moved the ecosphere from an ice age with glaciers down to New York State and
today's climate and ecosphere warming that is 4.5 times as fast including all the "ice-age" glacier melt that happened.
If the 1.441mm depth had warmed by 0.700C same as the ocean-air interface then oceans would gain no heat, but the massive colder oceans below will only let 1.441mm depth warm
by 0.697C and only when the entire ocean has warmed by 0.700C in a few thousand years will it let that 1.441mm depth warm the final 0.003C and stop heat gain with 4,100 ZettaJoules
of heat having been added to the oceans, enough to melt 13,666,666 cubic kilometres of ice. Of course, that will never happen because the ocean-air interface is going to keep warming
with the +CO2 that will keep happening and it's all going to accelerate.
Temperature gradient from 1.441mm (18.097C) to 1,000m (5C) depth is 0.01310 Celsius/metre
By conductivity, temperature gradient pushes 0.0076 w/m**2 up from 1.441mm depth to ocean
This is only 0.6% of the actual heat transport of 1.21 w/m**2 because 99.4% is transported down by water circulation, mostly natural with shark & whales & krill helping a bit.
Me @ 55. Nuts! Penultimate sentence should end "pushes 0.0076 w/m**2 down from 1.441mm depth to 1,000m depth."
Geoengineering might work well if the sun was blocked from hitting the oceans in areas where there is not much life. Also reducing sun light from hitting the Arctic might help sea ice stay in summer and increase its volume.
Gac73,
The simple answer to your question of how can a warmer atmosphere heat the oceans is this:
1) The ocean absorbs heat from the sun and is warmer than the atmosphere above.
2) According to the first law of thermodynamics, the ocean loses heat to the colder atmosphere. The atmosphere radiates the heat into space. The system is at equilibrium.
3) AGW causes the atmosphere to become warmer.
4) According to the first law of thermodynamics, the ocean now loses heat more slowly to the atmosphere, since the heat gradient is reduced.
5) The ocean is now losing heat more slowly than before, while still receiving heat from the sun at the same rate. The heat remaining in the ocean causes the ocean to warm until equilibrium is restored.
6) Winds and current affect where and how fast the additional heat is distributed through the ocean. Because the ocean is so huge, compared to the atmosphere, it retains the majority of heat from AGW.
Gac73... There's also a good site explaining ocean/atmosphere coupling here:
http://eesc.columbia.edu/courses/ees/climate/lectures/o_atm.html
@Rob Thanks for the post.
@Michael
Thanks for your prompt reply, however I feel that I may not have clearly communicated my question as the process that you describe above is how the atmosphere can alter the temperature of the ocean.
What I was wanting to know is if someone here could assist me with the following :
I understand how c02 is absorbed into the oceans.
But how can 'heat' IN the atmosphere move FROM the atmosphere INTO the ocean?
The Laws of Thermodynamics as applied to two thermodynamic systems not in equilibrium with eachother (the atmosphere and the ocean), says that net sum energy TRANSFER can only ever be in one direction. Clouds are the observable evidence that net sum energy transfer is from the ocean TO the atmosphere. So atmospheric heat cannot move FROM the atmosphere INTO the ocean. This would be a violation of Thermodynamics.
Academia states that the 'missing' forecast atmospheric heat has moved from the atmosphere INTO the ocean. i.e the ocean is a heat sink for atmospheric heat Moving directly FROM the atmosphere INTO the ocean.
What I would like to know is how is such a position tenable? i.e Trenberth - the 'missing' heat is transmitted into the oceans ; Tim Flannery - 90% of the heat (atmospheric heat) ends up in the ocean.
(I have used capitals to clarify the question).
Hoping I can get some further feedback on this.
Thanks
[Rob P] - I thought Michael Sweet did a great job of summarizing this, so am unlikely to make it any clearer - but I'll try.
Additional greenhouse gases cause more heat (longwave radiation) to be directed toward the oceans. This heat cannot penetrate into the ocean proper, but it does warm the uppermost surface and thus lowers the thermal gradient through the cool-skin layer. The normal flow of heat from the warmer ocean to the cooler atmosphere is reduced. As the oceans are heated by shortwave radiation from the sun, a reduction in the rate of heat loss through the cool-skin layer causes the oceans to warm.
So the Laws of Thermodynamics are quite safe - the net flow of heat still moves from the warmer ocean to the cooler atmosphere.
If said academia are phrasing it this way, that is incorrect. Citation required, please.
As the moderator notes, rising levels of CO2 act (via back radiation) to reduce the temperature gradient across the cool skin layer boundary at the ocean:atmosphere interface. This occurs 24/7/365, and slows the rate of energy transfer (via longwave [IR] radiation) from the oceans to the air. With the ensuing result of the oceans retaining a greater level of thermal energy than they did in a lower atmospheric CO2 environment. With the added note that the energy retained by the ocean is already in it via absorption of shortwave radiation from the sun, directly.
Is that more clear?
gac73, you seem to have been misinformed somewhere along the line.
For example, you seem to suggest that Trenberth's 'missing heat' has something to do with heat transfer from the atmosphere to the oceans. It doesn't. The 'missing heat' is simply the difference between the amount of additional heat satellite measurements suggest should be here (i.e. measured energy in minus measured energy out) and the amount of additional heat we have detected by measuring various parts of the climate system (i.e. increased air temperatures + increased water temperatures + heat to convert ice to water, et cetera). If satellite measurements show that energy came in and didn't go back out then logically it must still be here, yet all our measurements of the climate system come up with a smaller amount of heat... some of the heat is therefore 'missing'. Whether that is due to measurement errors, heat buildup in areas we are unable to measure (e.g. the deep oceans), or some combination of the two is still being worked out.
Likewise, this whole claim of massive heat transfers from the atmosphere to the oceans is a distorted/false representation. Greenhouse gases slow the release of heat from the atmosphere, which in turn slows the release of heat from the oceans. If you slow the release of heat while maintaining the addition of it (from the Sun) temperatures increase. No atmosphere to ocean heat transfer involved.
As to why most of the heat is in the oceans... the amount of heat required to raise the average temperature of the world's oceans 1 degree is much greater than the amount required to raise the average temperature of the atmosphere 1 degree. This is true both because of the size of the oceans and because water is simply a better heat sink than air. You can observe this yourself by putting two identical sealed containers in a freezer, one filled with air and the other with water. When you take them out (preferably before the water freezes) the container with the water will remain 'cold' long after the container with the air does.
Gac73,
I like CBDunkersons reply. The energy goes: Sun, Ocean, Atmosphere, space. Do you have additional questions?
Be careful when you look up the citations. It is easy to take a quotation from someone like Dr. Trenberth out of context to suggest that his comment is incorrect. Try to find the context and the question being answered.
At the risk of incurring a "pilling on" Moderator comment, can I just add this for Gac73
The explanations that Michael and CBDunkerson have offers, include a kernel of information that explains why your use of the Laws of Thermodynamics is flawed. The key point is that, since the Ocean is warmed by the Sun (as indeed is the atmosphere to a lesser extent), the Ocean + Atmosphere are not an energetically isolated system. They may be stable, but they are only so with a constant flow of energy through the system, and so are not at equilibrium. If they were energetically isolated, then you would be correct: for the very simple reason that there is nothing to heat the ocean except the atmosphere, and it cannot do so because it is cooler.
Rigourously thermodynamics only applies to energtically isolated systems. We can treat the Earth as a "nearly" thermodynamic system but only if we include the Sun, and outer space as a heat sink
Hope that helps.
Since the slope of the relationship is 0.002ºK (W/m2)-1 how does a 3 w/m^2 forcing, leading to a skin change of 0.006C acccount for a 0.7 C rise in SST?
matt sykes @66 refers to Fig 2 of the OP for the slope relationship. However, he is confused. First, a globally averaged forcing of 3.7 W/m^2 at the skin layer of the atmosphere results in a 5.4 W/m^2 increase in net downward longwave radiation at the surface. Ergo he has underestimated the forcing change at the surface. Second, that figure is for the forcing only. It does not include feedbacks which further increase the OLR.
Finally, the figure shown is for data collected over less than a month. It follows that, due to the large thermal inertia, the surface does not reach equilibrium in that data. Therefore the slope is not a slope of the equilibrium responce, or even the Transient Climate Response to the change in forcing.
Dont forget the SAGE COARE results used above (the .2C/100wm^-2 data) used clouds s the DLR source.
CO2 DLR penetrates water even less so wil have less warming effect.
[PS] I dont think you have understood the article. There is a 6 part series here which goes into the physics in exquisite detail. I suggest you read that first.
I have a query. In which case would the oceans likely warm more (most). If you have an atmosphere that has 300ppm CO2 and is 27°C. Or a situation where the atmosphere has 400ppm CO2 but 25°C? Or is it impossible to tell without knowing other variables?
Steve37341 :
Your question re 300ppm/27C versus 400ppm/25C cannot (IMO) have a short answer.
In rough figures : the oceans (which have average depth of 4000m) have a bottom temperature of 4C and a near-surface temperature of 15C. And the 15C is only an average ~ higher in the tropics, lower near the poles. Different temperatures above and below the thermocline. Dozens of years to dozens of centuries, for surface warming/cooling to mix into deeper layers.
So the question becomes modified to: Which parts of the oceans warm up, and by how much, and over what time span? . . . a year, or a thousand years (or inbetween). And do subsequent changes in CO2 solubility cause an alteration of the 300ppm or 400ppm levels . . . or will other climate variables come into play?
300ppm CO2 and 400ppm CO2 are figures which are now entirely in the past ~ and so your question is extremely hypothetical; as are the ocean temperatures you mention.
Perhaps you could express the thoughts and ideas that led you to compose your original question. That may lead to a better way forward in thinking about these sorts of climate aspects.
steve37341 @69,
As Elcectic @70 suggests, the question you pose concerns a complex and dynamic situation.
But if what you ask is reduced to the question of whether a +2°C in global surface temperature is a bigger kick to the climate (and thus the ocean temperature) than a 33% increase in CO2 (from 300ppm to 400ppm), the very basic response would be the +2°C is the strongest. A simple 33% incease in CO2 would result in a forcing of +1.54Wm^-2 which for an Equilibrium Climate Sensitivity of +3.2°C per 3.7Wm^-2 (the central estimate of ECS) would result in a surface warming of +1.33°C.
Adding more realistic assumptions would quickly impact that result. Thus, for instance, one early added consideration is that a surface temperature increase would be greater over land than ocean raising the question of whether your +2°C is a global figure or a surface (atmosphere) increase measured over the ocean.
steve37341 @69,
As others have said, the question is not sufficiently detailed. To add my little bit, I ask "warmed from what state?". Or are you talking about warming from the state you mention, to some other state?
It is possible to image a state where an ocean at 25C is in equilibrium with an atmosphere at 400 ppm CO2, and another state where an ocean at 27C is at equilibrium with an atmosphere at 350 ppm CO2.
Those states could have cooled from some other stable state, due to some change in conditions - or they could have warmed to those states from yet another state. Many factors affect climate (local or global).
Eclectic's suggestion is a good one - explain more of the scenario that you are imagining, to help us understand how you arrived at your question.
Eclectic @70, MA Rodger @71, & Bob Loblaw @72 thanks for each of your replies. I realized that I did not really give enough information about my scenario. The 25°/27° comparison was for atmosphere temperatures. Not average ocean temperatures. The average ocean temperature is the same in each situation. The only difference in the ocean variables is that in the example with the 300ppm CO2 and 27° atmosohere temperature, the sea level is 20 feet higher.
Also, the to get at the suggestion of where each situation is based on it's state, the example with the 300ppm has been steadily warming about 5000 years.
In the example of the 400ppm has been warming for about 8000 years and is still warming and the CO2 levels are continuing to increase, with the upper limit unknown.
I assume that the extra CO2 in the second example would eventually, is it stays higher long enough, would cause greater atmospheric temperatures and greater warming of the oceans. Or is this an incorrect assumption or could other variables come into play to alter that result(s)?
Another question. Assuming the mixing of the ocean waters in deep and shallow areas is relatively the same in each example, which is more important in heating more shallow areas? Higher atmosphere temperature, or higher CO2 level?
Steve @#37 ,
I must beg you for more details & more background to your questions. Are you basing your questions on real-world situations? If so, then the 25C or 27C are global average Surface Air Temperatures which have not been existent during the past 10 million years or so.
If you are talking of recent millennia, then coming from the latest "ice age" glaciation, the global SAT rose by around 5C and stayed at that Maximum (or "Optimum") for about 5000 years. Then during the past 4000 years, the SAT has gradually cooled by about 0.7C . (As you are likely aware, the PAGES12k studies indicate that the current [2020] temperature is nearly 0.5C higher than the Holocene Maximum.) During this time, the atmospheric CO2 level hovered around 280 ppm during the Holocene, and started rising during the past 2 centuries.
So your 300ppm and 400ppm scenarios are quite disconnected from the Earth's surface conditions during the past 10k , 100k , or million years.
Ocean warming is very slow, and it is the oceans which put a brake on SAT rise ~ more so than the other way around. The oceans take in more than 90% of the heat energy gained during total global warming periods, and the air itself represents only a few percentage points of the total. Which makes it difficult to get a good picture of what would happen in the abstract hypothetical situations that you pose.
steve37341 @73,
The 20ft difference in sea level suggests you are envisioning the melt-out of Greenland in the 300ppm/27°C scenario but not (or not yet) in the 400ppm/25°C scenario. This has profound implications as the absence of an icy Arctic in the first scenario is a situation requiring millennia for the second scenario to accomplish. Can you confirm this Greenland melt-out is what you intend to confirm with this 20ft SLR?
And can we be clear that the 5,000y "steadily warming" in scenario one has now stopped?
Finally, if the 400ppm CO2 level in scenario two is continuing to increase, the rate of increase is required, as is the history of this increase - when did it begin and from what level?