Ocean heat content and Earth’s energy balance

In Climate clippings 46 I thought the most important segment was the last, on Deep heat. I don’t think it attracted a single comment.

To recap, the world’s oceans have a total mass of 1.37 billion gigatonnes of water. A gigatonne of water equals a cubic kilometre.The average temperature is, I understand, 3.5C, so the capacity for storing energy in the oceans is truly massive.

Around 90% of heat trapped by greenhouse gases ends up in the ocean.

A post by Kevin Trenberth on the earth’s energy balance tells us of a study that will show that energy can easily be “buried” in the deep ocean for over a decade.

Skeptical Science
now has a post on this study, by Meehl (2011), including this graph showing periods of more than 10 years with the ocean heat content at 0-700m roughly static:

Figure 1: Ocean heat content

They think heat was sequestered in deeper layers of ocean during “hiatus decades”:

Figure 2: Ocean heat storage during 'hiatus' years

So far so good.

Now Gavin Schmidt at RealClimate has a post, and wonderfully clear it is but it raises some problems. Joe Romm at Climate Progress quotes Trenberth in emphasising that the heat is not lost in the deep and can come back quite fast to warm us at the surface. Schmidt tells us that he estimates the impact on the deep temperature at something less than 0.1 deg C or so, and:

Neither is this heat going to come back out from the deep ocean any time soon (the notion that this heat is the warming that is ‘in the pipeline’ is erroneous).

Of relevance here is a long paper by Hansen et al Earth’s Energy Imbalance and Implications. They say that that the deep heat is not well mixed and at least some of it can come back. What really counts is what is called the climate response function. They believe that climate models including the NASA GISS model assume a response which is too slow. They come up with this graph:

Figure 3: Climate response function

The phenomenon of decadal variance in ocean heat storage is a natural variation and the climate system doesn’t always produce the expected over short time-lines, which is what decades are. Hansen et al say that paleoclimate data on what happened since the Last Glacial Maximum indicates that deep ocean heat and surface temperature will reach an equilibrium within a millennium. They are confident that the range of 60 to 90% over 100 years represents the possible range and the mid-range of 75% “is plausible for climate sensitivity 3°C for doubled CO2.”

More importantly, though, Hansen et al do their own calculation of global energy imbalance and come up with different results from those of Trenberth and Fasullo (2010) (Panel A) which identified a “missing energy” problem. Here are the two graphs compared:

Figure 4: Earth's net energy budget

In Hansen et al (panel B) there is no missing energy. The graph represents a 6-year moving trend.

It needs to be emphasised that the graphs don’t indicate total heat content, rather the amount of warming measured in watts per square metre. So there is no cooling from 2003, merely a slowdown in warming. For the continued increase in ocean heat content, see here.

The thick red line represents where around 7% of the extra heat ends up, according to Skeptical Science:

Figure 5: Destination of heat from global warming

So it is easy to see that a small percentage change in ocean uptake can result in a large percentage change in the remainder.

From the posts it may be gathered that Trenberth and Fasullo only considered the heat in the top 700 metres. From reading the text, I take their graph to represent the top 2000 metres. Hansen et al have included the whole ocean, using sources they cite.

The top line in both cases represents the top of atmosphere (TOA) net energy balance.

One of the most concerning aspects is the issue of measurement. The Argo system only measures temperature in the top 2000 metres, whereas the average ocean depth is something like 4000 metres. Apparently Argo measures the top 700m better than the rest.

Hansen et al detail difficulties in other areas. The CERES (Clouds and the Earth’s Radiant Energy System) instrument measuring the TOA energy budget produced a result so implausible that it was calibrated using climate models. But they say it is simply not capable of producing the required accuracy to pick up small variations.

With aerosols the situation is even worse. They are not measured at all. Hansen et al:

We also must quantify the causes of changes of Earth’s energy imbalance. The two dominant causes are changes of greenhouse gases, which are measured very precisely, and changes of atmospheric aerosols. It is remarkable and untenable that the second largest forcing that drives global climate change remains unmeasured. We refer to the direct and indirect effects of human-made aerosols.

The Glory satellite mission would have measured the direct effects of aerosols as well as solar irradiance. Unfortunately the rocket launch failed earlier this year. A replacement will not be available until 2015-2016, but this will not measure the indirect effects of aerosols, the effect of clouds, for example. To do this you would need to make “simultaneous measurements of reflected solar and emitted thermal radiation fields” by looking at the same area at the same time. What’s needed is to implement a mission concept defined by Hansen back in 1992. As he describes it in Storms of My Grandchildren you actually need four co-ordinated instruments. He couldn’t persuade Al Gore and he’s still looking for funding, which is estimated at a mere $100 million.

Hansen has 10 pages of explanation detailing how they brought information together to arrive at their estimates. Suffice it to say that observations from particular studies as well as models, basic science and paleoclimate data are brought to bear. Climate change obscurantists will make much of the uncertainties, but the links between trace gases such as CO2 and other GHGs and such phenomena as surface temperature and sea level rise are quite robust and well-established from the basic science, from the observational record and from paleoclimate data.

Trenberth and Fasullo used two standard measurements and found an anomaly. Hansen et al gave solving the anomaly a red hot go. The measurement difficulties and the uncertainties are perhaps the main point. There is much to be done and we need funding for more measurement.

Gavin Schmidt tells us that Roger Pielke Sr claims that the Meehl paper ‘torpedoed’ the use of the surface temperature anomaly as a useful metric of global warming. Schmidt says no-one has claimed it is the whole story. In logic Pielke is right to emphasise the larger picture, but the Earth’s surface is where we live and grow our tucker. It’s the pointy end of the warming story and it’s where our story will be played out. Moreover, Schmidt points out that if “global warming” doesn’t mean “surface temperatures” then confusion will reign. Perhaps Pielke would be happy with that.

Finally, I want to comment on the surface temperature record of the last two decades. This is HadCRUT and NASA GISS from Skeptical Science:

<strong>Figure 6:</strong> HadCRUT vs NASA GISS

Two comments. First, HadCRUT and NASA GISS may be parting company. HadCRUT assigns the global average to the polar regions, whereas NASA makes an estimate based on the nearest measuring stations. Warming at the poles is multiples greater than warming at the equator. The difference may finally be showing.

Second, looking at NASA GISS there are two exceptional events in the 1990s which I’d suggest don’t contribute to the underlying trend, the Pinatubo effect and the El Nino of 1998. Take them out and you’ve completely lost the pause in warming.

The Meehl article shows us what can happen, according to the models when the heat goes deep (see Figures 1 and 2 above). It seems that the heat has indeed been going deep for the last few years. We’ll have to wait and see what happens next, but to truly understand it we will need to measure more than we are currently doing.

That’s how I see it, but this stuff is reasonably complicated for my aged but untutored brain, so I could be wrong.

57 thoughts on “Ocean heat content and Earth’s energy balance”

  1. This study confirms exactly what I’ve been on about – got to get my skates on before these bright sparks get all the credit! My paper resubmission that describes how this affects the warming signal has been slowed down by the third reviewer.

  2. Good post Brian but i am not sure how to translate it into stuff that the majority of voters understands.
    I would have thought that the temperature of deep oceans was controlled by the sinking of cold water in the arctic and antarctic regions. The temp of the sinking water should stay close to the melting point of water until the ice runs out then it will start to rise – so deep temperature may be stable for quite some time.

  3. Thanks, folks.

    John D, whenever I read stuff about ocean circulation, convection etc it just makes my head spin. I think they are looking for a change in the ordinary, and/or something not ordinary, or episodic like ENSO. The latter is powerful enough to have an impact on global surface temperature.

    It appears that they don’t really know how the heat is being moved to cause decadal variations.

  4. I wonder how these deep layers of sequestered heat might play out with the oceanic thermohaline circulation once they start to rise to the surface? Would they possibly lead to the shutdown of the Atlantic Conveyor, and consequent baking tropics and freezing higher latitudes while the average ocean surface temperature rises?

  5. tigtog, following your links I got to Woods Hole 2003 piece which spoke of the Thermohaline shutdown as a low-probability, high impact event. From what I’ve seen it is still regarded as such, but with a lower probability than previously thought.

    The Atlantic Conveyor is a prime example of how heat is transferred from ocean to land and atmosphere, as the warm tropical water moves north, evaporates and causes rain. As a result salinity increases, giving denser water which sinks and drives the Thermohaline. It could be interrupted or slowed by increased freshwater from northern rivers and melting Greenland ice.

    The Younger Dryas event was possibly caused by a giant freshwater flush. It’s thought that a repeat is unlikely. There isn’t as much ice around.

    There is information about the Thermohaline at Wikipedia and Rahmstorf with more about a shutdown or slowdown at Climate Research Unit and again Wikipedia.

  6. Thanks for the extra links, Brian. My understanding is that even without shutting down, the warming of the oceans overall weakens the oceanic conveyor effect, which is part of what makes for the so-called “paradoxical” colder Northern Hemisphere winters that so many people point to as “proof” that warming isn’t really happening. The effect is much stronger in the Northern/Land hemisphere than it is in our Southern/Water hemisphere because the patterns of warm-water currents are so very very different.

  7. Brian, I agree with you that this is a hugely important issue that didn’t get the attention that it ought to have last time you raised it (I think I had just had another stoush on one of your threads on which I thought I should let the dust settle, or I might have been tempted). Thank you for re-raising and elaborating on it.

    One key point, which you put less explicitly this time, is (to quote from your previous thread) “It should be noted that this study was done with models to explore possibilities. Trenberth says we now need observational studies, which they plan to do but are awaiting better datasets.” All this at this stage just represents a modelling exercise with no empirical data to support it, and we shouldn’t get too carried away with it yet.

    This does not make it wrong of course – there is no data to disprove it either. But what has happened in effect is that, because the GCMs on which the IPCC relies don’t balance the heat budget, another set of modelling has been done to try to explain why. Nothing wrong with that at least in terms of developing hypotheses for testing. As a physical scientist who was brought up on experimental observation and empirical data as the basis of everything, though, I do get uneasy at this sort of piling more modelling on top of earlier (inadequate or you wouldn’t have to pile it) modelling, thus getting another step further removed from real world observations.

    It is not least easily misinterpreted by the media as meaning more than it does, as the Trenberth paper has been in several quarters. The cynical may even wonder whether his media-centric approach is designed to encourage this.

    So far as existing observational data is concerned, Pielke Sr’s point seems valid to me: “If heat is being sequestered in the deeper ocean, it must transfer through the upper ocean. In the real world, this has not been seen that I am aware of. The Argo network is spatially dense enough that this should have been seen.”

    All this said, I agree that the case is at least open pending better empirical data being available.

  8. Brian I note incidentally you quote extensively from Skeptical Science again.

    I take it you are unaware of the latest the latest kerfuffle about Skeptical Science’s practices in rewriting history to suit the narrative it promotes? Cook has been caught red-handed changing articles in which he made incorrect assertions – and not only the article, but comments, which he deleted and/or edited in considerable numbers, misrepresenting the commenters, in order to hide his original errors and make it appear as if it had been the commenters, not himself who didn’t know what they were talking about. He has not denied the accusations either; they are well documented and it seems to me it would be very hard for him to do so. See eg:

    This is not the first time Skeptical Science has been pinged for this sort of thing either. It really is an unreliable source, lacking basic integrity and objectivity, and IMHO using it does not help the credibility of some of your pieces.

  9. It really is an unreliable source, lacking basic integrity and objectivity, and IMHO using it does not help the credibility of some of your pieces.

    I’d say it’s one of the most reliable sources, precisely because it is constantly referring to credible literature.

    There is uncertainty in all research and errors in all compilations. Parading uncertainty and error as invalidating is, frankly, anti-scientific.

  10. I once gave a talk at an energy conference a long time ago where I raised the question of the huge amount of energy that was stored in the top 10m of sea water with only a small rise in actual temperature. My point then was that this represented a massive energy input to an oscillating system (the weather), the inevitable result being an increase in the amplitude of the oscillations; more severe droughts and storms etc.
    I would really like to see if this speculative hypothesis was any-where near the truth. What your post made clear to me is that the coupling between the stored energy and the weather system is rather weak. As well it is or we might be in even more trouble.

  11. OK, Martin, so you believe that retrospective editing to cover up errors, including the unapproved rewriting of comments, all done entirely secretly, makes for “one of the most reliable sources” on climate change. Rather revealing about your view of what constitutes reliability and truth on the part of AGW promoters, but your privilege.

    You won’t mind, I’m sure, if I quote back at you this cavalier attitude about mere trifles like honesty, integrity and facts if we are arguing in the future.

  12. so you believe that retrospective editing to cover up errors, including the unapproved rewriting of comments, all done entirely secretly, makes for “one of the most reliable sources” on climate change.

    I said that it’s reliability on reporting climate change rests on its good referencing. I also said that there was error in all compilations. Would you really like to run a head-to-head credibility check between reporters like Cook and reporters like Watts?

    Unlike you, the sources that I look to for information on climate are first the conventional climatologists and second the contrarian climatologists.

  13. It is hard to fathom, why, anybody who ran a website and worked hard at attracting and nurturing an online community, would commit the most fundamental of indiscretions with his readers’ comments – deleting and moulding them at his own whim.

    zOMG! The horror of moderation…

  14. Just been out to see a band (rode home 20 km at 2 am ’cause I missed the last train), but after I get this paper off, I’ll elaborate on what I think is going on. It relates to this comment here about baths and jacuzzis. (Won’t take the permalink so you’ll have to scroll down).

    One thing I didn’t get clear enough. The radiation being absorbed by the ocean is long-wave radiation or heat. There are two heating processes, I reckon. One above the thermocline (which is the boundary between the mixed layer of water ~100 m or so and deep water) and one below. The Meehl et al paper works on the deep mixing. The stuff I’ve been working on is a result of interactions between the mixed ocean layer and the atmosphere. The interaction between the mixed layer and the deep ocean mixing is another puzzle.

    Have to wait for the Hansen et al paper. Sounds they might be onto it. I think Schmidt is wrong.

    Oh, and the 1997-98 El Nino – Brian there you missed it. That is the event. That El Nino emitted enough energy into the atmosphere to warm the globe by 0.29C. The temperature never dropped back (this has been missed by pretty much everybody). After that, deep ocean mixing has kept things reasonable stable.

    The models produce the same sort of hiatuses, but the statistical models used to analyse the warming curves tend to smear them out. Warming is fundamentally non-linear.

    Fantastic post though.

  15. Rog & others: I suspect that part of the problem is that despite the importance of the deep ocean currents, deep ocean mixing can also occur via eddy diffusive processes, which our models are only just becoming capable of resolving. Unfortunately even with fancy tools like Argo, we can only measure temperatures in the ocean, not heat fluxes, so despite Wozza’s disdain for models, we’ll have to plug along with them for the time being.

    One other interesting problem is much closer to home – the role of the Southern Ocean in regulating the heat budget of the ocean. One aspect of a warming world is the poleward shift of the mid-latitude westerlies is a corresponding poleward shift in the Antarctic Circumpolar Current. We can already observe a lot of warming throughout the depth of the Southern Ocean over the last half-century or so (see this paper (PDF) – there’s also an interesting discussion of measurement biases), although this isn’t due to a warmer atmosphere but probably a shift in the location of the ACC and possibly an increased eddy heat flux towards the pole.

    Point is, we’re getting warmer water closer to the Antarctic, which is kind of like blowing the insulation off our freezer, since the thermal isolation of Antarctica by the ACC is probably a big reason why temperatures have been so stable for the last little while.

    John D: Ocean basin topography is also important in determining the stratification in the bulk of the Atlantic, because the sinking plume has to flow out of the Arctic basin over the Scotland-Iceland-Greenland ridge to reach the Atlantic. As it does so it can interact with the warmer waters in the thermocline and this can change how much heat gets trapped in deeper waters. In fact, sea level can also play an important role, since if the sea level drops by about 80 meters the Scotland-Iceland-Greenland ridge is enough to separate the Arctic basin from the main body of the Atlantic. You can see some experiments designed to simulate this in the lab here.

  16. John D: Just found Kial has stuck some video of his experiments up on Youtube. Here’s one where the sill is close enough to the thermocline to isolate the marginal basin. Things are upside down here relative to the real world (as it’s easier to heat water than cool it) but the small partition to the left corresponds to the Arctic, and the rest of the tank to the Atlantic. The sill (black thing blocking the flow) corresponds to the Greenland-Scotland ridge.

  17. jess,

    With Arctic ice melting causing a local increase in sea level, how long does it take for increase to normalized over the global sea level?

  18. BilB: I think it might be difficult to identify a disturbance from melting ice – the timescale of the melting would probably be too long.

    That said, the physics of how the ocean communicates changes in pressure and how they propagate around the globe is pretty interesting. Most disturbances in pressure will propagate either as Rossby waves across the interior of an ocean (but only in a westwards direction), as a Kelvin wave near the equator, or around the boundary of the ocean as a boundary wave. The boundary wave is a Kelvin wave for vertical sidewalls but when things aren’t vertical the dynamics of the boundary wave get more complicated IIRC.

    You might like to take a look at this paper by Chris Hughes and Michael Meredith (open content) where they try to pull the effects of these boundary waves out of correlations in satellite altimetry data. There’s some really interesting stuff about how the subsurface pressure changes along the South American Pacific coast due to changes in wind forcing in the eastern Pacific, with links to other good references as well. The velocities for the boundary waves come out on the order of meters per second, so its relatively quick. But (as the authors emphasise) there’s still a lot that’s unknown about how these signals get around, especially because it’s difficult to separate the effect of these waves out from some kind of long-term coherence in the forcing.

  19. Bilb @ 20, of course the Arctic ice as such floats, so melting doesn’t affect sea levels. Greenland is a different matter.

    Jess, I haven’t had time to follow all those links, but I wondered what you thought of the following in relation to BilB’s question.

    The UNEP Yearbook 2009 Ch 3 says this:

    However, a new study of how a sudden release of meltwater, or its ice equivalent, from ice sheets into the oceans would influence sea levels shows that, in the first years, the rising waters would not flood with equal speed everywhere. It would take decades for a pulse of rising sea levels to spread around the world.

    From the Greenland ice sheet, most of the melted water would initially stay in the Atlantic (Figure 3). Fifty years after release, sea level rise would be thirty times greater in parts of the North Atlantic, including the Gulf of Mexico, than the Pacific. Similarly, the study found that water from a collapsed Antarctic ice sheet would swamp coastlines in the southern hemisphere, while being barely measurable in the northern hemisphere for at least 50 years (Stammer 2008).

  20. Brian: well a sudden release of meltwater is a bit different to gradually melting ice. Maybe we’re talking at cross-purposes here.

    But looking at the 2008 Stammer reference (abstract here), he suggests the same mechanisms for communicating the sea level change I mentioned above (Rossby, Kelvin and boundary waves). I guess the question I didn’t really answer was how fast they can propagate a signal of a certain size. I’ll have to have a read of the paper and let you know what I think.

  21. Great post Brian and interesting discussion folks, thank yous all.

    Jess, I always knew of Rossby waves as atmospheric events eg the meanders of mid latitude jet streams. For example, the ‘blocking event’ that triggered the catastrophic Russian wildfires and floods in Pakistan last year were attributed to a strong Rossby wave. Do we know if similar strong and solitary waves in the oceans can trigger major weather events? For example, do they have the potential to ‘dump’ warm water on our east coast?

  22. Wowie Jess,

    When you put it that way the total system hyper complexity is quite mindblowing, and yet it largely revolves around several key physical relationships, those being the energy absorbing properties of hydrogen and carbon. And when I think about it, hydrogen drives the oceans with evaporation (with heat) being the driver for the thermohaline circulation, and atmospheric moisture content (with heat amplified with Carbon in CO2) being the driver for the atmosphere, all of course energised by the earths 2 rotations, gravity and time. A few relatively simple relationships create this immensely complicated energy distribution effect. This jacuzzi of energy flow which we experience as weather and seasons.

    Nature is truly spectacular.

  23. Actually, Jess@18, I said that I was “uneasy” at ascribing too much significance to piling modelling on top of other modelling, in circumstances in which even the modeller acknowledges that the real immediate problem is lack of empirical data, not that I held modelling in principle in “disdain”. There is a difference.

    What seems to be happening to me in climate science is that, given the huge compexity of the atmospheric (and ocean) system and the consequent difficulty of getting adequate amounts of reliable real world data, it is just becoming so much easier, especially with advances in technology, to sit back in the lab and model. And, in the usual publish or perish environment and with no shortage of journals prepared to accept papers long on modelling and short on data, there is a real incentive to do so.

    I don’t know how you get round this, but if it is happening (and I accept that not working in the area I may be quite wrong) there are real dangers ahead for the discipline. Certainly, harking back to a couple of recent threads in which declining public support for climate change action has come up, a growing public perception that climate science is all hat and no cattle seems to me to be an issue even now.

  24. Yeah Woz, sorry for being a bit snarky earlier. I’m not an oceanographer either, although most of the others in my research group are.

    I think that we do have a lot of fundamental reasons for believing that the models are doing a good job, and the evidence comes from a comparison of their output to what we would expect from our theoretical understanding of oceanography and meteorology. It’s not just a matter of publishing whatever the model spits out, it’s really more about being able to explain what’s going on physically, and then tying that insight to the real world.

    So that’s why I don’t think that climate science will come a cropper so easily – I’ve sat through enough oceanography seminars to know there are more than enough obsessive geophysical fluid physicists who will come down hard on something that doesn’t agree with basic physics.

  25. Wozza @ 10, you are right, I wasn’t aware of the stoush about Skeptical Science’s moderation practices. There is no easy way of doing moderation, but it has to be done. You said that John Cook has

    deleted and/or edited in considerable numbers, misrepresenting the commenters, in order to hide his original errors and make it appear as if it had been the commenters, not himself who didn’t know what they were talking about.

    If he did that, then that is indeed unacceptable. In the examples you link to, the “deletions” are all too obvious, and I wouldn’t do those bulk crossings out either, but I can’t see that as unethical.

    If Cook has changed his original post to correct a mistake, he should mark and note this in some way. It is possible to get something wrong. But I’d agree with Martin B’s comment about SkS’s reliability in general in reporting the content of science articles.

  26. Wozza,

    your comments come back to the bath vs jacuzzi analogy I was making. There’s been a lot of science that shows the energy balance of the planet can be expressed as a set of reduced-form equations, without much loss of accuracy. So it’s relatively straightforward to estimate how warm the bath might get.

    However, the process of warming as a function of time is what matters when we need to understand how the climate changes. This doesn’t invalidate the bath model, though it shows its limitations.

    So the physics that explore energy balance are pretty well known, the thermodynamics are much more complex. I understand Jess is onto that between blog comments 🙂

    But the uncertainty surrounding how climate changes does not invalidate the bath model, but it does show that conceptualising warming, sea level rise and other processes as a smooth one is incorrect at that level of detail. It also says that the dichotomy between the anthropogenic component being smooth and the variability being interpreted solely as natural variability is incorrect. The anthropogenic signal within the earth system is also variable because of complex system thermodynamics. My work is showing that for SE Australia at least, this leads to risks being seriously under-estimated, not the other way around.

    You seem to have a limited appreciation of models – all information including empirical data is interpreted through models. Models are just concepts of how things work. Scientific models are a subset of that. The model false – data real construct creeping into your comments is a false one. (This is well understood within the philosophy of science but not appreciated by all scientists)

    You also don’t seem to appreciate the close links between global observation systems and scientific modelling. The two are now so tightly coupled they inform each other. Observation programs are organised through big international programs such as the World Climate Research Program. Meetings are held to ask questions about what obs are needed most. Programs are designed and the hat passed around to get governments to fund those programs. Australia funds a lot of the Southern Ocean obs, for example. Models are used to take data that is uneven in space and time, and used to build a 3-D picture of earth system processes. There are several big re-analysis projects globally and these are compared and contrasted to gain how processes should best be represented. The findings are used to improve how these processes are represented in the climate models (that are now earth systems models). They are also used to ascertain where new obs are needed most.

    One of the big issues is that earth observations are only satellite age and some even shorter. This means we have a limited insight into decadal scale processes. The surface meteorological observations then become the links to the past, and palaeoclimatic reconstructions the link to even further back. Over the past decade palaeo modelling is becoming much more prominent and it tries to link a sparse network of proxies within a dynamic physically-based simulation.

    So I don’t accept your premise that modelling has embarked on a data-free jaunt. Of course you will always hear people say they want more data. I’d like long-term time series of vertical ocean warming profiles, thank you very much.

    It also doesn’t stop modellers (or any other scientists) making over confident claims. This isn’t restricted to climate science – the overpromising of how mapping the human genome would reveal all is a prime example. This is why internationally reviewed programs and assessments are so important.

  27. Ootz: Rossby waves already dump large amounts of warm water on our eastern coast – they help keep the waters of the NSW South Coast nice and warm for when I go swimming. 🙂

    Basically long wavelength Rossby waves in the ocean propagate rapidly westward until they hit a boundary (the east coast of Australia for example), and their energy gets trapped on that boundary. So you get something called western intensification (basically all the build up of energy on the western arm of a subtropical gyre) which helps to drive an energetic current right along the east coast. In Australia this current is the East Australian Current (I just found a sweet pic of this from CSIRO Hobart is here), and the North Atlantic equivalent is the Gulf Stream.

    This is a uniquely oceanic phenomenon – since the atmosphere has no horizontal boundaries, the Rossby waves in the atmosphere can propagate right around the whole planet.

  28. A couple of questions, Jess.

    How does a Rossby wave, which is presumeably a vertical movement of water propogate a horizontal flow. Is this action caused by water flowing down the albeit very shallow but long face of the wave towards the trough?

    Are atmospheric Rossby waves the same as gravitational waves?

  29. BilB: Those are big questions! This might be a long post…

    Waves are generated from an initial disturbance by the action of a restoring force, that acts to bring the system back to an undisturbed state, and some kind of inertia, that causes the system to overshoot, so that we get an oscillation. The dynamics of the wave propagation are determined by the balance between the restoring force and the inertia of the fluid. Gravity waves are simply waves which have gravity as their restoring force. The very familiar example is the surface gravity wave, in which gravity acts to restore the surface of a fluid after an initial disturbance.

    The wave itself has energy which propagates in the direction of travel of the wave, both as kinetic energy associated with the motion of the fluid, and potential energy which is formed from deformation of the surface. Generally the particles of the medium will travel in orbits aligned with the direction of travel of the wave, with particles moving forward under a crest and backward under a trough. Then the energy of the wave will travel through a medium, but the the medium itself doesn’t move when averaged over a whole wave cycle.

    So we’re not really talking about a horizontal flow of fluid per se, but a horizontal flow of energy. There are other reasons why the ocean currents are generated (it has a lot to do with wind stresses and Ekman pumping) but the western intensification is the dumping of this energy by Rossby waves on the western boundaries of the oceans.

    In geophysics, we are (often) interested in flows in which the dynamic balance occurs between pressure gradients (which are generated by gravity acting on vertical disturbances of the straified ocean or atmosphere) and the Coriolis force, called a geostrophic balance. Flows whose dynamics are set by this balance are called geostrophic flows. Kelvin waves and Rossby waves are two examples of waves which use a balance between the Coriolis force and pressure gradients to propagate.

    A Kelvin wave is a rotational wave which propagates along a boundary (such as a coastline), and uses the effects of the boundary to balance the effects of the Coriolis force. Imagine a wave propagating northward along a coatline which is to its right in the northern hemisphere. As the crest passes the particles of the fluid all move northwards, but the Coriolis force also pushes them to the right, so that they pile up along the coast. This piling up continues until the pressure gradient caused by the raised sea level along the coast is enough to balance the Coriolis force. As a trough comes through the opposite occurs, particles move south, the Coriolis force pushes fluid away from the coastline and the surface is depressed until the pressure force can balance again.

    It turns out when you do the math that the only way that this can occur is when you have a coastline to the right of the propagating wave. So Kelvin waves propagate polewards on eastern boundaries, and equatorwards on western boundaries. It also turns out that the Earth’s equator acts as a virtual boundary, so you can get Kelvin waves which propagate around the equator also.

    A Rossby wave is a wave which uses the variation in the Coriolis force with latitude as a restoring force, and the disturbances are due to the changes of relative vorticity when you change the thickness of a volume of fluid. This is due to the fact that potential vorticity is conserved in these flows. I’m afraid I can’t really simplify things much further without the math so I think I’ll leave this long post there.

    As an aside, oceanic Rossby waves are bloody hard to spot since the surface amplitude is about 5 cm over a wavelength of hundreds of kilometers (although there is a corresponding shift in the thermocline of ~50 m). We really only managed to spot them properly after the Poseidon mission – here’s an image of passing Rossby waves in the Indian ocean observed by satelite altimetry.

    So to sum up: the differences are in the mechanism that the wave uses as a restoring force. Surface waves and internal waves use gravity with the effects of a vertical perturbation of a surface or density gradient, Kelvin waves use the Coriolis force and the effects of a boundary, and Rossby waves use the variation of the Coriolis force with latitude and the effects of conservation of potential vorticity. (Sorry for the long post Brian).

  30. Jess, no problem. I said back @ 5 that this stuff makes my head spin. The value, I think, is that it is giving some flesh to Roger’s jacuzzi comment of the climate system being a really complex system.

    Roger, going back to comment @ 16 I didn’t miss the 1997-98 El Nino, just misinterpreted it – completely. I’ve been bringing a mindset of share trading charts to temperature graphs. With shares I’d tend to ignore a one-day spike for example, due to irrational exuberance or whatever, and look at the underlying trend. Thing is that the default position for the share price the next day is zero, if new buyers and sellers don’t show up, which of course they do. But after a temperature spike like 1997-98 , the default position next year is the same as last year, because the heat has been added and is already there.

    Which means that there has indeed been a pause in warming since then, even on the NASA GISS record.

    Strikes me that 0.29C is a helluva lot of warming, maybe about 15 years worth in one year. Where did it come from?

    Presumably what’s coming in at the top of the atmosphere is much the same. I can’t see cloudiness making that much difference, because when it’s drier here it’s raining more in America.

    During a La Nina you have a lot of cold water upwelling as the wind blows the warm surface water westwards. That doesn’t leave a big hole down there, so it must be replaced by warmer water from near the surface. This could be assisted by the winds evaporating the westward fleeing water and making it denser.

    During neutral ENSO years you have some cold water upwelling. During La Ninas it just sits there and warms.

    If I’m anywhere near right that could explain some of the exceptional surface warming, but I’d reckon nowhere near all of it. Heat must be coming out of the ocean by some other mechanism.

    Anyway I’ve probably made an ass of myself a second time, but what the hell!

  31. Thanks Jess @31, fascinating stuff fluid thermodynamics. To study planetary scale energy distribution patterns must be truly one of the most challenging fields in science. We are much obliged to you and Roger for sharing your knowledge and particularly those informative links. I have a few more questions re periodicity and heat exchange in oceans as opposed to the atmosphere. However, as I just about get hit by a storm, better attend to things. Definitely an early and solid build up to the Wet for us. PS reports of some hail up here in the Tropical Tablelands.

  32. Jess, I got the impression from your earlier comments on Rossby waves that they occur at the density boundary layer between surface warm water and colder water at depth.

  33. BilB: Yes, that’s correct. The Rossby waves have a surface expression of
    about a 5 centimetre high disturbance, but that’s spread over a wave length of order hundreds of kilometers across. They move at about 10 centimeters/second, requiring months to years to cross the ocean. The surface perturbations are mirrored by perturbations in the thermocline, which are much larger, about 10-50 meters in height. When the surface goes up, the thermocline goes down. The point is how this change is restored – it’s not by gravity but by the Coriolis force acting on the potential vorticity change induced by the perturbation.

    I’ve just found a video of Rossby waves in the Pacific, taken from filtered data from the TOPEX satellite mission. The total time frame for the video is about three years – you can see just how slow these things are. The changes in velocity occur because the Rossby waves have a phase speed which varies with latitude.

    What’s really interesting about this video is that you can see the changes from one type of wave energy into another – the same waves at high latitude become unstable and break up into smaller eddies before they can completely cross the Pacific. Also, if you look at the fast Rossby waves at the equator which propagate westward, you can see that the wave peaks often split into two blobs north and south of the equator – this is from Kelvin waves which propagate eastward along the equator when a Rossby wave hits the western boundary of the Pacific.

  34. Brian,

    “If I’m anywhere near right that could explain some of the exceptional surface warming, but I’d reckon nowhere near all of it. Heat must be coming out of the ocean by some other mechanism”

    That is what I was trying to figure out in an earlier thread. I still don’t know.

  35. Thanks Jess, just a few branches down, friends copped a direct cg strike, hail further up the the Tablelands around where Wantok lives.

    Very hard to grasp the nature of the ‘Jacuzzi’ with major forces such as solar radiation, wind, gravitation, Coriolis driving it, never mind the shape of it. First, as Roger said we don’t know much about the workings of decadal variations. Is there any indication that the periodicity of these Rossby and Kelvin waves have anything to do with it? Second, Rogers comment that ” El Nino emitted enough energy into the atmosphere to warm the globe by 0.29C. The temperature never dropped back …..” has me baffled. Why is there no such effect during a la Nina event, since these are, as I understand, similar events just opposite? Where does that additional heat get transferred?

  36. Good to hear you survived ok Ootz. Was up that way a couple of months ago – dragged the missus along to look at the Undara Lava Tubes on our anniversary, lucky lady. It’s a lovely bit of the country.

    Regarding ENSO, I think that Rossby and Kelvin waves have a lot to do with it, since they are the predominant means by which the ocean responds to changes in forcing. You might be interested in this discussion from the Jet Propulsion Lab (Caltech), where they discuss the role of Rossby and Kelvin wave dynamics in the differences between the monster 1997-98 El Niño, and the much weaker one in 2002-03:

    In an El Niño, a breakdown in the easterly (east to west) trade wind system spawns two different kinds of waves: Kelvin waves and Rossby waves. Kelvin waves travel from west to east along the equator. Rossby waves move in the opposite direction from east to west on either side of the equator. These waves create a change in ocean circulation.

    When Hackert and his colleagues dissected ocean height satellite data to isolate the individual effects of the Kelvin and Rossby waves, they began to see differences in how the two El Niños developed.

    They found that the Kelvin wave component for the two El Niños was similar at first but then strengthened for the 1997 El Niño and weakened for the 2002 El Niño. Even more striking was the influence of the Rossby waves. They had little effect on the development of the 2002 El Niño but made a large contribution to the strength and duration of the 1997 El Niño.

    As to how ENSO develops in the first place, I’m not sure that anyone’s come up with a completely convincing grounding for it in terms of basic fluid dynamics. Whatever is driving it is coming from the linkages between ocean and atmosphere, and that’s a really hard problem to unravel.

    Roger might want to elaborate on your second question some more, but I would suggest that although we think of the two phases of ENSO as being opposites in some sense, the way in which they contribute to the atmosphere energetics isn’t necessarily symmetric – energy flows out of the tropical oceans near the equator no matter what ENSO is doing. So it’s not like a La Niña event can completely ‘undo’ the effects of an El Niño.

  37. By the way, I forgot to mention the more obvious point that Rossby and Kelvin waves are what help drive and end the different phases of ENSO: El Niños often start with a Kelvin wave propagating from the western Pacific over towards South America. Once an El Niño gets going it generates Rossby waves that drift slowly towards southeast Asia. After several months of travelling, they finally get near the coast and reflect back. The changes in interior ocean temperature that these waves carry “cancel out” the original temperature changes that made the El Niño in the first place. A complete ENSO cycle has two pairs of Kelvin and Rossby waves crossing the Pacific.

    There’s a nifty image of the different paths of the Rossby and Kelvin waves here which might help to visualise things a bit as well. Compare it to the video which I posted earlier from TOPEX-Poseidon – you can see the eastward-propagating equatorial Kelvin waves interacting with the westward propagating Rossby waves.

  38. Brian @35,

    you haven’t made an ass of yourself at all. This stuff is really complex and I keep making mistakes with it. Trawling around for a response and also to finish my paper I found more critical bits of information.

    First, back to 97-98. If you remember the sea surface temps (SST) from that episode, the eastern Pacific was up to 4 C warmer than usual. Most El Ninos have a warm spike. This one did too, but the return afterwards was 0.3C higher than before. This is the bit that has been overlooked by most analysts because they are just thinking generic climate variability.

    The Meehl et al paper talks about deep ocean mixing during the period after that explaining the hiatus. Today I found a paper in Physics Letters A, which is very exciting because it contains analysis related to my speculations on shifts. I’ve been looking at the surface expression (air temp). These guys have been looking at the top of atmosphere, ocean energy balance relationship and find switching at similar time periods to those I’m looking at. They are using ocean heat content to infer the top of the atmosphere radiation. It is negative from the fifties to the early 70s (global temps pretty flat), positive from the mid 70s to round 2000 (climbing), and negative in the 2000s (they stop at 2003). They seem to be regime changes, which is what I have been measuring in air temps. My feeling is that most of the warming occurs in switching between these modes. It implies tight coupling of climate variability and change – neither are smooth.

    Here’s my speculative mechanism. During the 1997-98 El Nino for example, there was a lot of warm water in that anomaly trapped above the thermocline in the eastern Pacific (it’s deeper than usual in that region during an El Nino). Just like a steam engine, if there’s enough energy it will do work – in this case, upwelling. But not every El Nino is associated with step change warming – in fact only a few are. Perhaps that is a response to build up of energy over time.

    Reasonably flat temperatures over the oceans during the 20th century punctuated by occasional steps suggest that not every episode of variability such as an El Nino causes warming (different timing in other ocean basins suggest they have their own mechanisms, so I may not always be ENSO). My suggestion is that most of the time the roughly 85% of energy absorbed by the ocean (the rest by atmosphere) is mixed deeper. When there is a build-up in energy, perhaps also a hiatus in deep mixing, the ocean does work and an upwelling episode warms the atmosphere. Afterwards, business as usual sees normal mixing going on. Air temps may flatten out and perhaps even decline for a bit. Have several of these and its easy to put a trend through it and call the wiggles natural variability. However, the warming is anthropogenic and it’s coming in bursts.

    During normal climate without GHG emissions, we would probably be experiencing both warmer and cooler episodes during such events driven mainly by the internal energy balance of regional systems. Now, the system is being dominated by upward step changes. The models simulate these too, so they do contain the physical mechanisms, but they are not well understood.

    I may be wrong about how this mechanism works, but the stats seem pretty clear. Understanding this is important, because most people think we can adapt by adjusting gradually to a smooth change, because that’s how they see climate information communicated.

  39. Thanks, Roger. I’ll have to let that settle in my brain.

    BTW the link to “a paper in Physics Letters A” doesn’t work. I went into ‘edit’ mode and found you neglected to post the url.

  40. This is what worries me the most that the masking effect of aerosols has led us into too delayed a response… and that it is now too late.

    That is: when our aerosol production drops, either by improved tech, economic collapse, fossil fuel depletion, etc, then temps will rise dramatically and very quickly.

    Then fast feedback effects accentuate it.

    One (scary) possibility is that we trickle along for another decade or two, with moderate (but still uncomfortable in many places) warming. This lulls us into even more delayed responses (business as usual). Then at about the 2040’s we get a dramatic reaction, with 0.5-1C increases a decade for several more decades. Possibly even into the ‘death zone’ of +5C. At that point it’s going to ‘lights out’ for many regions.

    And, of course, there is no way back at that point.

  41. OldSkeptic, the modelling that I did for the Garnaut Review Mark I suggested that fewer SO2 emissions in a high emissions scenario added about 0.5C before 2050 to what otherwise would have been. If we are reducing emissions growth, the acceleration process will be reduced, but requires faster reductions after peaking, all other things being equal, to minimise peak temperatures.
    The storing of energy in the oceans before release into the atmosphere is also a delay process. It shows the benefits of understanding the underlying science rather than following trends.
    It’s not too late for major benefits, but as to what’s achievable …

  42. Roger: Thanks for this – very interesting and illuminating thread. (Hooray for a good SNR…)

  43. Martin B, I was thinking the same thing myself. This is one of the most productive threads we’ve had.

  44. Ootz: Yeah the tubes were pretty sweet. I’ve been through the Kazamura tube on Hawaii which is only a couple of hundred years old and it was amazing how well preserved Undara was when you think they’re 190 thousand years old.

    Nice vids but I’d be careful of anything coming from Bob Tisdale re. ocean heat content though. He already has form…

  45. Sorry, yeah I did know Tisdale has form, should have noted it on the link. However, it was the illustration of the asymmetry that I was after.

    Credit to the private management and the relative high entry cost that preserved Undara, as there is very little impact from human visitation. It makes their visitation a truly stunning experience. Did you go onto the walk up the volcano cone from where one can get an impression of the monumental scale and extent of these lava tubes?

  46. Oh and I was going to ask Roger if he would have the time to elaborate his last sentence, as it gets to the crux of it.

    “It’s not too late for major benefits, but as to what’s achievable …”

  47. Jess says: “Nice vids but I’d be careful of anything coming from Bob Tisdale re. ocean heat content though. He already has form…” And you linked a response to one of my posts from Ben at WottsUpWithThat.

    If you’re not aware, Jess, Ben’s retort was comical, at best, and not in need of a reply. Ben also linked to Tamino’s complaints on my ARGO-era model-data Ocean Heat Content comparison graph. I went to great lengths to reply to each of Tamino’s complaints in my post:

    You should always check to make sure that Tamino’s complaints are sound. Sometimes they are and sometimes they are not. With respect to the posts he has written that are erroneous, he has written a couple of posts with complaints about my presentation of data and my discussions of natural variables and each time it has been easy to show that his complaints have been unfounded.

  48. Bob, I don’t have a lot of time to answer your response to Tamino point by point, but I will say this: you have no physical reason to call out a flattening out of thermocline heat content over a period of 10 years and say that the rate of ocean heat content increase is slowing. OHC trends do not have to be linear, and indeed we expect such deviations from long-term trends based on our understanding of ocean physics.

    So despite your efforts to persuade us that your 2003 start date isn’t cherry picked, you still have the fundamental problem that the timescale for the data you use is simply too short relative to oceanographic timescales to be statistically valid.

Comments are closed.