A new observational constraint on the ECS

I wanted to just quickly write a post about something Dana highlighted in his recent Guardian article, global warming is unpaused and stuck on fast-forward. The article discusses the recent paper by Trenberth & Fasullo (2013) that considers whether we really have had a global warming pause. Although I had already read bits of the paper, until I read Dana’s article, I hadn’t appreciated that the paper had re-estimated the equilibrium cimate sensitivity (ECS), using some updated estimates of the system heat uptake.

In case anyone who reads this doesn’t know, the ECS is a crucial number in climate science in that it is the increase in global surface temperature resulting from a doubling of atmospheric CO2 [Addendum 12/12/2013 : See Steve Bloom’s comment below – I should have made clear that what I’m talking about here is the equilibrium temperature determined only by fast feedbacks. There are likely slower feedbacks that will result in an equilibrium temperature even higher than that determined only from fast feedbacks.]. It essentially tells us how much warming we should expect. There are a number of ways to estimate the ECS. One is to use past climate history, and another is to use detailed climate models. Recently, however, there have been some estimates based on recent observations (Otto et al. 2013, for example). What was interesting was that these observationally constrained estimates produced values that were quite a bit lower than those from other methods. Past climate history and climate modelling suggests an ECS around 3oC. The observationally constrained values are closer to 2oC.

In truth, the uncertainties on the different estimates are quite large, and so there is a large overlap in the possible ranges, but the difference between the methods is likely why the IPCC reduced the lower limit of its ECS range and didn’t produce a best estimate. The way one can estimate the ECS, observationally, is to use

ECS
where ΔQ2x = 3.7Wm-2 is the adjusted anthropogenic forcing after CO2 has doubled, ΔT is the change in global surface temperature over the time interval considered, ΔQ is the change in adjusted forcing over the time interval considered, and H is the current rate at which the system is gaining energy (i.e., it is the current total energy imbalance). Essentially, the equation is telling us that if H = 0, then the ECS = ΔT and the system is already in equilibrium. If H is positive, then the system is not yet in balance, is still accruing energy, and to reach equilibrium the temperature will have to continue rising (i.e., the ECS is greater than ΔT).

Otto et al. (2013) did all their calculations relative to 1860-1879 and for the period 2001-2010, their values were ΔT = 0.75oC, ΔQ = 1.95 Wm-2, and H = 0.65 Wm-2. If you put these numbers into the above equation, you get 2.1 Wm-2. It’s reasonably well accepted (Karsten could likely put me right) that anthropogenic aerosols are providing a not insignificant negative forcing. When Otto et al. considered the impact of aerosols, they reduced ΔQ to 1.73 Wm-2, giving an ECS of 2.6oC, so still lower than other estimates (I think I’m using 3.7 Wm-2 for ΔQ2x, instead of 3.44 Wm-2, but it doesn’t change things much).

What’s interesting about the Trenberth & Fasullo (2013) paper is that they suggest that the system heat uptake rate, H, should be 0.91 Wm-2, quite a bit bigger than the 0.65 Wm-2 used by Otto et al. Using this value gives an ECS of 2.7oC (no aerosols), and 3.4oC (with aerosols). With this value, this estimate is much more in line with the other estimates. So, why am I writing this? Well, I suspect most climate scientists suspected that the observationally constrained estimates were on the low side. They don’t cover the full time interval, they’re sensitive to short-term variations, and the uncertainty of the current aerosol forcing likely influences their robustness. So, it’s no great surprise that a more recent paper is bringing the observationally constrained estimate more in-line with other estimates.

To be clear, it doesn’t mean this new paper is right and I suspect this won’t be the last word. However, this is kind of how you expect science to work. Someone comes up with a new way to determine something which maybe initially disagrees with other methods. That could imply that the other methods are wrong, but could also imply that the new method needs refinement. As time goes on, either people discover that results from the new method are superior, or they discover that as the new method is further tested, its results come into line with the other methods. We still don’t know that that the case yet, but this is how the process is meant to work.

Advertisements
This entry was posted in Climate change, Climate sensitivity, Global warming, IPCC and tagged , , , , , , , . Bookmark the permalink.

95 Responses to A new observational constraint on the ECS

  1. What I meant to add to the post is an acknowledgement that I am ignoring the uncertainties here. The possible range is quite large, but that doesn’t change that Trenberth & Fasullo’s work has increased the likely ECS value. Also, as Ed Hawkins points out on Twitter, even what I’ve called the observationally constrained ECS still requires a model, in that the adjusted forcings are from model calculations.

  2. dana1981 says:

    “I think I’m using 3.7 Wm-2 for ΔQ2x, instead of 3.44 Wm-2, but it doesn’t change things much”

    Trenberth told me that this change is significant, and would bring ECS from 2.5°C (if using his OHC) to 2.8 or 2.9°C. I’m not sure why they used 3.44 rather than the widely accepted 3.7 W/m2 doubled CO2 forcing.

    Kevin Cowtan also says that using his surface temperature estimates by itself would bring the Otto ECS up from 2 to about 2.25°C. So when you put all these factors together, the result is right in the ballpark of 3°C ECS.

  3. dana1981 says:

    And yes, the uncertainties on this ECS estimate are quite huge; we’re just talking about most likely values here.

  4. Thanks, Dana. The reason I put that comment in about 3.7 rather than 3.44 was just to explain why my numbers were very slightly different to those in Otto et al. I also have no real idea why they chose 3.44. They do mention it in the paper. Given the range, maybe it doesn’t matter that much. Given how much people focus on the likely value, it probably does. It does seem, though, that everything is heading towards the different estimates being largely consistent.

  5. Steve Bloom says:

    Anders, because ECS is a somewhat misleading term to those unfamiliar with it (since the suggested equilibrium is unphysical due to slow feedbacks kicking in too quickly), IMO an article like this needs to also mention ESS and the difference between the two.

  6. Steve Bloom says:

    Hmm, still moderating?

  7. Steve Bloom says:

    Anders, ICYMI see here re ECS/ESS. To avoid the confusion I mentioned, as in the linked article ECS is often referred to as just plain CS, or sometimes as the Charney Sensitivity to give credit to the committee that originated it.

  8. Steve, yes you’re quite right. I should probably have made that distinction. When I saw your comment I thought of the report by the Geological Society, but I see that’s what the article you link to is about.

    For those who don’t know the distinction, the issue is basically that what I’ve called the ECS is really the equilibrium temperature determined only by fast feedbacks (for example, increases in atmospheric water vapour). There is now a recognition that there are likely also slower feedbacks (reduction in snow and ice cover, changes to vegetation) that can cause the equilibrium temperature to rise higher (by maybe 2 – 3 degrees) than that expected from fast feedbacks only.

    I’ll elaborate a little more on why I think Steve made that point, and Steve is welcome to correct me if he thinks I’ve got it wrong. There is now (as this post was trying to illustrate) multiple lines of evidence to suggest that fast feedbacks will result in a rise in equilibrium temperature after a doubling of CO2 of about 3oC. Depending on our future emissions, we could double CO2 – relative to pre-industrial times – by the middle of the 21st century. This means that it is possible that within 50 to 60 years, we will have essentially locked in 3 degrees of warming relative to pre-industrial times. Slow feedbacks, however, could cause this to rise even further, possibly reaching 5 to 6 degrees relative to pre-industrial.

    Essentially, if we choose to do nothing, within 50 to 60 years we could have essentially guaranteed a long-term warming of 5 to 6 degrees. It may still take centuries to get there, but there’s virtually nothing we can do about this without actually removing CO2 from the atmosphere. There’s also absolutely no evidence whatsoever that such warming would be beneficial, and plenty of evidence that it will be extremely damaging. It’s for this kind of reason why some people get extremely frustrated when others argue that we should just wait and see. One could argue that we could rely in geo-engineering, but that – I think – is likely to be even more expensive and more technologically challenging than putting some effort into developing alternative energy sources.

    Anyway, that’s enough about policies for the moment 🙂

  9. Brandon Shollenberger says:

    Doesn’t methane have a large enough effect and short enough residence time that we could reduce emissions of it reduce future warming at just about any point? I imagine the economics of such of doing so would be incredibly burdensome, but I think it should be possible.

  10. Brandon, I don’t understand what you mean. Why is methane relevant to this discussion? We’re talking here about CO2, not methane.

  11. Brandon Shollenberger says:

    You said:

    Essentially, if we choose to do nothing, within 50 to 60 years we could have essentially guaranteed a long-term warming of 5 to 6 degrees. It may still take centuries to get there, but there’s virtually nothing we can do about this without actually removing CO2 from the atmosphere.

    I was suggesting there is something we could do to combat that warming – stop methane emissions. CO2 has a long atmospheric residency, but methane does not. Stopping methane emissions would cause methane levels to fall back to pre-industrial levels within decades. That would introduce a cooling effect. The net result would be enhanced thanks to the loss of the expected warming from rising methane levels.

    The idea is even if we reach a point where we can’t stop a certain amount of warming via CO2, we could still introduce a cooling effect by cutting methane emissions. That’d result in less overall warming.

  12. BBD says:

    Brandon Shollenberger

    What do you think the major sources of anthropogenic methane are?

  13. Brandon Shollenberger says:

    BBD, I have no idea why you ask. As such, I’m not going to bother researching the matter. Going off memory, I believe natural gas, agriculture and landfills are the largest sources. For agriculture, I believe it is primarily from farm animals and rice paddies.

    I’m sure Google can give you plenty more information.

  14. BBD says:

    Brandon

    Your suggestion is that we reduce CH4 emissions. It seems… odd to me that you are clueless about their major sources and when asked, aggressively refuse to check. As usual, I discount your commentary.

  15. BBD says:

    The reason we need to be clear about the major sources of aCH4 is that we would need to estimate the relative cost of reducing them sufficient to produce a climatologically significant offset to ever-rising CO2 levels for BS’s suggestion to have even marginal merit.

    This looks very like a non-starter to me. Since the thread is about T&F and the reappraisal of “observational” under-estimates of S, it’s also rather off-topic.

  16. Brandon, I think you have to be careful with the methane argument. As you say, methane decays, so all that matters is the direct forcing due to methane today. This is about 0.7 Wm-2. Given that methane decays, this shouldn’t rise much with time as it depends only on the amount of methane released recently. When we reach the point at which CO2 concentrations have doubled, stopping all methane emissions will presumably only reduce the forcing by 0.5 – 1 Wm-2. Not insignificant, but not enough to induce a cooling trend. Furthermore, as I think BBD is trying to point out, simply stopping methane emissions is not particularly trivial. Why choose that option over, say, investing in alternative energy technologies? Having said that, this post wasn’t really meant to be about policies but about updates to recent climate sensitivity estimates.

  17. Brandon Shollenberger says:

    I, uh… whuh?

    Your suggestion is that we reduce CH4 emissions. It seems… odd to me that you are clueless about their major sources and when asked, aggressively refuse to check.

    I didn’t suggest we do anything. I suggested it’d be possible to do something. I think aggressive cuts to methane emissions would be a terrible idea due what would be required for it. I was just thinking out loud that it’d be possible.

    As for being “clueless about [methane’s] major sources,” I believe I nailed the top three contributors, even managing to get the subcategories for agriculture right. And I didn’t “aggressively refuse to check” anything. I said I didn’t know why I was being asked a question, and as such, I wasn’t going to take the time to research it. That’s not refusing, much less aggressively refusing to do anything. Had I known a reason to check my memory, I would have.

    As usual, I discount your commentary.

    It appears you discounted my commentary so much you didn’t even read what it said.

  18. I didn’t suggest we do anything. I suggested it’d be possible to do something. I think aggressive cuts to methane emissions would be a terrible idea due what would be required for it. I was just thinking out loud that it’d be possible.

    Sure, lots of things are possible and it is likely that we will be forced to consider something drastic in the future. None of that is a particularly compelling argument for choosing to do nothing now though.

  19. Brandon Shollenberger says:

    andthentheresphysics, I think methane’s forcing right now is closer to 0.5 Wm^2, actually. I’m not sure when projections estimate CO2 will double or what methane levels will be at that time, but I’m sure it wouldn’t be enough to change the planet’s warming to cooling (that’s why I said cooling effect instead of cooling trend).

    I was just thinking if we reach the point where CO2’s effect locks us into, say, five degrees of warming, we might not actually be stuck with five degrees of warming. We might be able to knock off as much as one degree. That might be enough to mitigate some damage.

    As I said in my first comment, I think it’d be incredibly burdensome. It just seems like an idea that, if the situation got bad enough, might help out. Sometimes you have to pick between two bad options.

  20. Brandon, the latest forcing diagram seems to show about 0.7 Wm-2, but it’s not that important. So, sure, you’re probably right that stopping all methane emissions could reduce the climate sensitivity by some amount. But that falls into the category of lots of things we could if forced to do so. Again, it’s not a particularly compelling argument for a “wait and see” strategy.

  21. Brandon Shollenberger says:

    To be clear, I agree with this:

    None of that is a particularly compelling argument for choosing to do nothing now though.

    I only mentioned my idea because I saw the comment there might wind up being “virtually nothing we can do” without removing CO2 from the atmosphere. I think reducing methane emissions would be significantly easier than removing CO2 from the atmosphere in levels great enough to reduce the amount of expected warming.

    If we reach the point where we seriously consider geoengineering solutions to try to pull CO2 from the atmosphere, I think severe cuts to methane emissions will be a better option.

  22. I think my comment still stands, because I was referring to what would happen were we to continue emitting until we double CO2 concentrations. Once we’ve done that, there is virtually nothing we can do to prevent the system from continuing to warm towards the new equilibrium without removing GHGs (mainly CO2 since CH4 has a short residence time). There are possibly geo-engineering solutions, but the point was mainly that without something drastic, a doubling of CO2 would likely – due to fast feedbacks – lead to around 3 degrees of warming, and – due to slower feedbacks – potentially 5 – 6 degrees of warming on longer timescales.

  23. BBD says:

    So:

    – No cost discussion of CH4 emissions reduction relative to reducing CO2, no idea at all, just handwaving

    we might not actually be stuck with five degrees of warming. We might be able to knock off as much as one degree. That might be enough to mitigate some damage.

    – No methane feedbacks with ~4C warming???

    This isn’t worth further discussion. Can we get back on topic please?

    Sometimes you have to pick between two bad options.

    Sometimes you have to face the facts: CO2 emissions need to be reduced sharply starting now.

  24. Tom Curtis says:

    Brandon, Ander’s estimate for the doubling time for the CO2 concentration accords with that in RCP 8.5 for CO2 alone. If it were for CO2 equivalents (CO2e), your point would be correct, but as it is for just CO2, the forcing would in fact be greater than indicated.

    Never-the-less, there is reason to suspect we will not face the full Equilibrium Climate Response (ECR) of the peak CO2 concentration. In essence, The CO2 burden in the atmosphere takes about two hundred years to equilibriate with the deep ocean. Consequently, if we were to raise CO2 levels to 560 ppmv and then drop to zero net emissions, the CO2 level would fall to about 350 ppmv over the next two hundred years. The time constant for that drop is approximately equal to the time constant for the increase from the Transient Climate Response (TCR) to the ECR. As a result, temperatures would approximate to the TCR of the peak CO2 concentration if we have zero net emissions thereafter.

    After the first two hundred years, CO2 is drawn down at a much slower rate, a rate that approximately corresponds to the time constant of the Earth System Responce (ECR), ie, the response due to slow feedbacks. Consequently for the scenario with no net emissions, the TCR of the peak CO2 concentration turns out to be a fair estimate of the expected temperature response for the next several thousand years. That is approximate, of course, for there will be some rises or falls depending on the exact rates of the processes involved.

    This is not as comforting as we would like to think. Specifically, even very low ongoing net emissions, equivalent to less than 1 GtCarbon per annum will prevent the initial drop in CO2 concentration. After the first two hundred yeas, such low onging net emissions will result in a steady rise in CO2 concentrations into the future. Given that some CO2 emissions are unavoidable, to benefit from this we need to explicitly sequester CO2 to bring net emissions to, or below zero; or we will still face the full ECR to the peak CO2 concentration.

  25. Tom, yes you – as usual – make a very valid point that I rather glossed over (ignored may be a better term – I wrote this when rather exhausted, so should probably have taken more care 🙂 ). If we were indeed to stop emitting after doubling, then CO2 concentrations would drop and my statement, that there’s nothing we can do to stop further warming is a bit extreme. We wouldn’t reach the ECS/ESS associated with that doubling, we’d – as you say – more likely equilibrate at the TCR of the peak.

    A bit sloppy on my part, but that’s what the comments are for. As you point out though, even modest emissions would prevent this drop in CO2 concentrations and then we would face the full ECS/ESS unless we were to do something like sequester CO2.

  26. OPatrick says:

    To be fair I thought Brandon’s initial point was perfectly reasonable and worth saying. Without active attempts to remove CO2 from the atmosphere we are tied in to centuries of higher temperatures, but that isn’t true for methane. It’s an interesting question as to whether this would be a cheaper and easier option than geoengineering. It’s no magic solution, but there wasn’t any suggestion from Brandon that it is.

  27. OPatrick, is that really true? Methane is a very effective greenhouse gas, but in terms of anthropogenic emissions, it’s currently about 1/4 of the total forcing and if we were to reduce anthropogenic aerosols, it would probably be about 20%. Also, it’s got a short residence time which would suggest that the forcing due to anthropogenic methane should not increase that much with time (or at least, should increase more slowly than CO2) hence it’s contribution should reduce. Stopping (or removing all anthropogenic methane) would therefore, I think, only have a small effect. I think the same is also true for natural methane emissions at the moment. It might be 30 times more effective as a greenhouse gas than CO2, but the rise in methane has been about 1 ppm, while for CO2 it’s been 120 ppm.

  28. OPatrick says:

    Sorry Anders (note this is in the vein of Aggers and should not be taken to imply any gender, or Nordic, specific nomenclature), I didn’t mean to imply any opinon on if it was right or not, just that I thought it was a reasonable point to raise and worthy of debate.

  29. BBD says:

    Since the largest contributory factor (~33%) to aCH4 is fossil fuel (natural gas; coal), reducing FF usage will be the most efficacious means of reducing aCH4. Source.

  30. Tapani L. says:

    Regarding methane, I think the current understanding is that there are some relatively easy changes in agricultural practices that would yield considerable benefits in reducing methane emissions, and that these changes should be made as soon as possible to mitigate short-term global warming. I think I read somewhere that it could be one of the most cost-effective ways to combat global warming in short timescales, basically, to buy us a little more time to tackle the really hard stuff (CO2 emissions).

    In this light I was a bit baffled by the harsh reaction to Brandon… There’s good science being done on this topic and I’m eager to see what IPCC WG3 is going to say about this when they get the report out next April.

    But this is rather off topic.

  31. verytallguy says:

    On the TCR/ECR/ESS issue, I’m sure I saw a paper (by Hansen?) recently which showed this nicely with a plot of time vs temperature and the three shown as big blue circles. Can’t find it – anyone else recognise this?

    (It’s possible I’ve imagined it…)

  32. Tapani L., there’s a little history that may have influenced how Brandon’s comment was received. It may have deserved better, but my desire to defend someone here is mediated somewhat by whether or not they’ve – in the past – chosen to imply that I’m a fool. I have no objection to being called a fool (it may even be appropriate), but the person choosing to do so then has to accept (without objection 🙂 ) how I choose to mediate any discussion in which they involve themselves on this blog 🙂

    You make a valid point though. There’s certainly no reason not to improve aspects of agriculture (and other processes) that could reduce anthropogenic emissions of greenhouse gases.

  33. Brandon Shollenberger says:

    OPatrick, Tapani L., thanks. I didn’t think it’d turn into a thing when I said it. It was just an idea I had thought about before, and it seemed worth bringing up.

    Anders (I guess that’s a new name for our host?), I don’t see any inherent reason a short residency time would mean forcing from methane should increase more slowly than forcing from CO2. That’s definitely a factor that would decrease methane’s growth rate, but that’s only one factor. I don’t think we can conclude one way or the other without looking into emission rates. I don’t recall any projections giving reason to believe there’d be significant changes in the ratio between CO2 and methane forcing, but I may have missed something.

    Anyway, since this is still being discussed, I should point out another interesting aspect of targeting methane. The more methane there is in the atmosphere, the longer its residency time becomes. The reason is the more saturated the atmosphere is with methane, the slower the reactions which remove it become (as there are fewer appropriate particles, proportionally speaking). That makes projections of methane effects more interesting/challenging.

  34. Tom Curtis says:

    vtg, somebody recently posted a link to a post by Isaac Held that discusses the issue.

  35. Rachel says:

    Following on from what Tom and Anders (This name feels right. Good suggestion OPatrick) have said about CO2 falling once emissions have stopped. I didn’t really understand the significance of this until I read Hansen’s paper from last week. He gives three very interesting scenarios:

    * If we stop emissions in 2015, CO2 will decline to 350ppm by the end of the century
    * If we wait 20 years before stopping emissions, CO2 will return to 350ppm at around 2300
    * If we wait 40 years before stopping emissions, CO2 will return to 350ppm sometime after 3000

  36. I don’t recall any projections giving reason to believe there’d be significant changes in the ratio between CO2 and methane forcing, but I may have missed something.

    Maybe I’m missing something, but isn’t this obvious. If methane emissions have a short residency time and CO2 emissions have a longer residency time, then surely the ratio between the forcing due to methane and the forcing due to CO2 should decrease as we increase emissions. Maybe I’m wrong, but I can’t quite see how this isn’t a relatively obvious conclusion.

  37. Tom Curtis says:

    Rachel, Hansen’s way of presenting it is interesting and to the point. Part of the reason for the extending interval to reach 350 ppmv is that the more CO2 we put in the atmosphere, the greater the fraction that remains after equilbrium is reached. Thus with 1000 GtC total emissions, around 25% remains in the atmosphere, but as that pushes out 5000 GtC that rises towards 35%. So, by delaying action, not only do increase the total carbon load, but we reduce the effectiveness of the ocean in saving ourselves from ourselves.

    For those who are interested in this issue, I recommend (in addition to various papers by Archer and others on the perturbed carbon cycle, and those linked by Isaac Held), the stripped down version of Geocarb made available by David Archer at the University of Chicago. Using it, for example I was able to determine that ongoing net emissions as low as 0.09 GtC per annum will out weigh the long term draw down in CO2 by weathering in the long term. That is only relevant after the faster processes of drawing down CO2 have finished (ie, after 10,000 years), so probably not relevant to immediate policy. It shows, however, that anything short of zero emissions is only a way station for future policy.

    More immediately relevant, 0.9 GtC ongoing emissions following an initial 5000 GtC release (approximately BAU) prevents the CO2 concentration falling below 1500 ppmv at anytime while the emissions continue.

    Finally, those discussing the effects of methane may wish to play with the slugulator to inform their intuitions.

  38. @ATTP: Re aerosols in Otto et al, I’m afraid you got that one wrong. What they did is to take the central AR5 estimate of the (total anthropogenic effective) aerosol forcing to obtain their principle numbers. In an additional experiment, they increased it to the referenced ACCMIP model study (Shindell et al 2013) in which the effective aerosol forcing is stronger by 0.3W/m2. I tend to think that it is indeed better to use the lower principle estimate (which yields ΔQ = 1.95 W/m2) as it includes aerosol effects already! Might be a bit too optimistic, but that’s another matter.

    @dana and ATTP: Given that they used effective (i.e. adjusted) forcings throughout the manuscript, they had to use effective forcing for GHG doubling as well. The thing is, the current non-adjusted forcing would otherwise simply be higher. In essence, the end result wouldn’t change much as long as GHGs are concerned (starts to change when it comes to aerosols). In any case, using Cowtan and Way (for temperature) and Balmaseda et al 2013 (for OHC), we are getting to an ECS of up to 2.7 already (with the lower aerosol estimate). Given that this is an ECS estimate which discounts non-linearities, one shouldn’t expect values higher than 2.5 anyways. So we might not even need to rely on Balmaseda. If we, for example, stick with [Lyman and Johnsons latest work], we are still in the right ballpark as far as I can see.

  39. @Karsten, thanks. The reviewers have been hard at work on this post. I clearly misunderstood/over-estimated the influence of the aerosol forcing. I also didn’t, to be honest, appreciate why they were using 3.44 Wm-2 rather than 3.7 Wm-2. Correcting for these errors, though, still gives and ECS around 2.5 Wm-2 or even slightly higher if we include the global surface temperature change suggested by Cowtan & Way. So, as you say, essentially everything is tending towards the same kind of ballpark.

  40. BBD says:

    There does seem to be no getting away from ~3C. And I speak as a former lukewarmer 😉

  41. RB says:

    I’ve been discussing this OHC update by Trenberth with Troy Masters and he makes what seem to me to be good points as outlined in his post here . Specifically, Loeb et al. postulate that due to the large uncertainties, there possibly is no missing heat pre-2005 prior to Argo. Further, the 0.91 W/m^2 of Trenberth (0.84 W/m^2 for just global heat accumulation as opposed to including energy for melting ice etc). principally owes to ~1.2W/m^2 in the pre-Argo data which then drops to ~0.38 W/m^2. Troy argues that there is no such relative change seen in the satellite data (even though absolute satellite data is noisy). So, it seems to me to be a reasonable case to be sceptical of Trenberth’s 0.91 W/m^2 calculation. Any thoughts on this?

  42. RB says:

    ATTP,
    Also, as I understand it, ECS is the sensitivity that you get when you run the model for 1000 years. For the next 100 years, what is relevant is the TCR. Otto’s calculation for that is 1.3C (neglecting the ocean uptake). So, I don’t believe you are correct in expecting 3C warming over the next 50-60 years as calculated by Otto. For the next 100 years, Otto’s most likely estimate would be <1.5C.

  43. RB, I did not say that. I can’t respond in detail as I’m on my phone, but maybe read what I wrote again and a little more carefully.

  44. RB says:

    Hi ATTP, sorry I did misread you since you said 3C from pre-industrial times. BTW, per Nic Lewis , the 3.44 W/m^2 comes from F-2X multi-model mean as computed by CMIP5. So, the discrepancy affects delta-F similarly. So, we have
    (a) 2C as computed by Otto using 3.44 W/m^2
    (b) Any increase from Cowtan/Way
    (c) I read Karsten Haustein as saying that aerosol numbers and Levitus OHC numbers are (Lyman/Johnson of 0.56 W/m^2 for the most recent period) reasonable?

    I think this indicates lower sensitivity than previous estimates. As I remarked earlier, Annan said that 2C would also be consistent with paleo. Higher estimates would presumably depend on (a) OHC updates (b) aerosol (particularly indirect forcing) updates.

  45. @RB: Troy Masters raises many interesting points, but he has a tendency to be dismissive against mainstream science. His scepticism is one-sided. Rob Painting is making many good points in the discussion (note that Rob is an ocean expert, while Troy is not) which you have certainly noticed. Nothing much to add. The main point is, whether you use Lyman, Levitus or Balmaseda data, over the past 30-40 years they show a very consistent picture. Balmaseda shows highest variability, which is however very plausible from a ocean surface (wind) dynamics point of view.

  46. BBD says:

    TCR is the estimated response at the point of doubling CO2, so could be later this century, not necessarily in 100y. Arguing about TCR/2xCO2 is to ignore the full implications of BAU over the next few centuries, which is too partial or myopic a viewpoint for comfort. Also, although full equilibrium might take ~1ka, most of the warming will happen centuries before that, so again, this long timescale thing is not a framing I am comfortable with at all.

    Finally, this is all predicated on stopping dead at ~550ppm CO2. At present, that is very far from something I’m confident will happen. So talking down the transient response to 550ppm is a further step in an intellectual direction I do not feel we should be following here.

  47. BBD says:

    As I remarked earlier, Annan said that 2C would also be consistent with paleo.

    He said it would be the lowest value consistent with paleoclimate behaviour. Values around 3C are a better fit.

  48. @RB: You’re saying: “I think this indicates lower sensitivity than previous estimates.”

    I can’t quite follow. I actually pointed out in my comment that using the OHC estimate from Lyman and Johnson would still produce an ECS value which is perfectly consistent with current wisdom. That is, no (realistic) chance to stay below ECS of 2.5 if non-linear feedbacks are considered.

  49. RB says:

    Hi Karsten,
    I did read Rob Painting’s points there as well, I haven’t had the chance to read the literature he describes. I also agree that non-linearities due to slow feedbacks (Armour et al. etc) are likely to raise ECS further than from linear feedback calculations. But then again, as Tom Curtis remarks and Isaac Held state, temperatures are likely to level off as indicated by TCR since CO2 levels also start dropping. So, while it could be true that ECS=3C for doubling, you won’t see an actual 3C change due to doubling.
    Also, from Lyman/Johnson I see 0.56 W/m^2 from 2004-2011, so I couldn’t directly compare with Otto of 0.65W/m^2 for the 2000s. Perhaps I didn’t understand you well there, so could you clarify if Lyman/Johnson OHC estimate raises ECS estimates above Otto’s using the same linear analysis? I agree that if Cowtan/Way places ECS at 2.25C, we could see ECS>2.5C based on non-linear feedbacks.
    Finally, it seems to me that it is TCR where the big debate is going to be from a policy perspective and for that it seems that non-linear feedbacks and OHC are likely to play secondary roles while aerosols are more important to estimate the true impact.

  50. @RB. Let’s something first. Nowhere did I suggest we’d have 3 degrees of warming relative to pre-industrial times at the moment CO2 has doubled. I used the term locked in to imply that we would reach the ECS eventually, not at that instant. As Tom points out, however, if we were to stop emitting at that time, CO2 concentrations would drop at a rate similar to the rate of temperature rise and so we would likely settle at the TCS of the peak rather than the ECS. That, however, would require that we emit virtually no CO2 beyond that point. Even emitting a few percent (if I’ve done my numbers right) would maintain the peak concentration.

    Like Karsten, I’m confused by your comment about Otto et al. If you use their numbers you get an ECS of 2 degrees. If you use Trenberth & Fasullo’s system heat uptake rate, you get 2.5 degrees. If you use Cowtan & Way, it goes up a little more. I’m not suggesting that this proves that it is higher than Otto et al’s initial estimate. However, the indications are that as we gather more information this estimate will rise to be more consistent with other estimates. Also, as Karsten points out, that it doesn’t capture some of the non-linearity we might expect it to be slightly lower anyway. Basically, it would appear that an ECS best estimate of around 3 degrees would be reasonable. Of course, there is quite a large uncertainty range, but the different methods appear to be becoming consistent.

  51. But then again, as Tom Curtis remarks and Isaac Held state, temperatures are likely to level off as indicated by TCR since CO2 levels also start dropping. So, while it could be true that ECS=3C for doubling, you won’t see an actual 3C change due to doubling.

    Except, as Tom Curtis very clearly points out, this is only true if we were to emit less than 1GtC per year after that point. If Tom’s number refers to carbon rather than CO2, this is around 7 times less than we’re emitting today. So, yes, if we were to double atmospheric CO2 and then virtually stop emitting CO2, the TCR would be the relevant number. If we continue emitting even a relatively small amount of CO2, the ECS/ESS is the relevant number. Which would you regard as the one we should consider?

  52. RB says:

    ATTP,
    Not sure where the confusion is regarding my comment about Otto. However, I was seeking clarification from Karsten whether using Lyman/Johnson instead of Levitus would raise ECS above Otto’s calculation (without taking Trenberth’s update into account).

  53. @RB, I guess, my confusion stems from the fact that you appear to have – in that comment at least – dismissed the 0.91 Wm-2 from Trenberth & Fasullo, which is kind of what the post was about.

  54. RB says:

    I’ll have to defer more conversation for now, but my position is that while I’m open to the numbers from Trenberth, I would also like to wait to see more papers in the future that confirm Balmaseda/Trenberth before considering it to be an update that places ECS estimates closer to the conventional wisdom of 3C.

  55. @RB, indeed and that was kind of the point of the post. I’m not suggesting (as I think I said quite clearly) that this is the last word. I was simply pointing out that it is interesting (and maybe not surprising) that more recent numbers are pushing the Otto et al. estimates up towards the other estimates. That’s really all I was suggesting. This may change again, so I am not claiming that this proves that the different estimates are now consistent, but it is suggestive. As always, more work will clarify the situation.

  56. Tom Curtis says:

    RB:

    But then again, as Tom Curtis remarks and Isaac Held state, temperatures are likely to level off as indicated by TCR since CO2 levels also start dropping. So, while it could be true that ECS=3C for doubling, you won’t see an actual 3C change due to doubling.

    Anders has already mentioned this, but whereas he gives a figure of 1/7th of current emissions, I will note that at 0.9 GtC emissions, CO2 levels will not fall, and indeed will rise slightly over time. That is 9.3% of 2012 anthropogenic emissions, or about 1/11th. To get a fall in CO2 concentrations with time, we would need to drop emissions to at least half of that. That is a very tall order, and is probably not achievable without active carbon sequestration from the atmosphere. Indeed, increased “natural” emissions generated as a feedback to warmer temperatures could potentially exceed that level.

    This is particularly concerning given that current policy targets being negotiated tend to aim for a reduction to 20% of current emissions. Indeed, it is implicit in the approach of most economists that there will be ongoing emissions into the future.

    Further, as I take care to mention, the plateau in temperatures depends on the balance of two uncertain terms, so that in practice, even with zero emissions we could still be facing greater than the TCR to peak concentration, or less. Uncertainty is not our friend.

    Therefore, simply assuming that temperatures will level of is not warranted. Neither is assuming that it will rise to the ECR or the ESR of peak concentration. It depends critically on what we actually do. Given that we cannot yet even persuade ourselves to stop increasing emissions, I am not optimistic.

  57. RB says:

    Therefore, simply assuming that temperatures will level of is not warranted. Neither is assuming that it will rise to the ECR or the ESR of peak concentration.

    I agree with this – it depends on how much we put in vs how much decays out and the timeframes involved. Current emissions trajectories suggest roughly doubling CO2 over the next 100 years with an attendant TCR’s worth of warming. I’m sure we will have technological progress over that timeframe. But I don’t know if any country will find it in its self-interest to look that long-term and locally raise the cost of energy. We should be glad if TCR turns out to be on the low end of the IPCC range to buy time for the transition to alternatives.

  58. Steve Bloom says:

    RB: “temperatures are likely to level off as indicated by TCR since CO2 levels also start dropping”

    It’s peculiar to see people making such statements as if there were no contrary implications from the inability of present models to replicate 1) *observed* polar amplification, 2) known mid-Piacenzian climate conditions and 3) known past fast climate transients. The assumption that a GHG-induced 2-3C warming spike over the next century or two could relax back to prior conditions without triggering major feedbacks in the Arctic and perhaps the Antarctic seems to me to be the most absurd sort of fantasizing. ECS and TCS are needed diagnostics for making progress in modeling, but relative to real-world climate they are essentially fairy dust.

  59. @RB:
    Re Lyman/Johnson, I would either have to crunch the individual OHC numbers myself or to ask Alex. For 0-700m depth it is Church et al. 2011 (based on Domingues et al. 2008) and for 700-2000m depth it is Levitus et al. 2012. The total in their paper (Otto et al. 2013) is however the total heat budget of the Earth. To get there, you’d have to add deep ocean OHC changes and energy uptake by the atmosphere and the cryosphere. Moreover, you’d have to subtract the preindustrial ocean heat uptake (which is a rather awkward thing to do as pointed out in my response to Nic Lewis, but that’s another matter). That’s how they get to 0.65W/m2 for the 2000s.

    Given that Lyman is higher than Levitus for the deep ocean, but probably lower for the upper ocean in the last decade, things might balance after all. What I’m saying is, I expect Otto et al’s current estimate for the 2000s to change only barely using Lyman/Johnson. But that’s really only for the 2000s. Might well change for earlier decades. The point with Balmaseda et al – and in fact with Kevin Trenberth’s whole argument – is, that (1) one decade doesn’t tell you much than it comes to ECS estimates, and (2) there is no “hiatus” if one were to look at the entire energy budget of the Earth. Perhaps one more thing to note: While Balmaseda’s OHC data produce higher ECS in the last decade, it produces lower estimates in the 1990s instead. Again, things have the tendency to balance.

    I hope these explanation aren’t too confusing.

    So yes, Cowtan/Way brings the linear ECS to at least 2.25K, with no-linear ECS >2.5K. I tend to think that it is indeed more useful to focus on TCR. The most appropriate way (for the time being) to get a meaningful ECS estimate is to convert the TCR into ECS on the basis of the TCR/ECS ratio from CMIP5 (approximately 0.55). With a TCR of around 1.5 (currently my best estimate, which coincides with Otto et al. using Cowtan/Way), we end up with an ECS just above 2.7. Pretty much the consensus estimate. Whether we ever see a temperature change of 3 degrees or not, is entirely dependend on our CO2 emissions.

  60. RB says:

    Karsten,
    Thanks – yes, the TCR/ECS ratio multi-model mean is ~0.55 . On the lower side, GISS-E2R is TCR/ECS of 1.5/2.1 and GISS-E2H is TCR/ECS of 1.7/2.3. As Nielsen-Gammon showed in his posts discussing Kosaka/Xie, even after you fixing SSTs to observations in the Pacific thus replicating the relative flattening in the recent decade, model mean diverges from observations. So to me, it is something to keep in mind whether there is something else that is being missed in models.

    Steve Bloom: I found this graph interesting leading me to wonder whether part of the explanation for the Arctic might lie in the south-to-north heat transfer not being modeled properly as is known from observations and papers discussing the meridional heat transfer .

  61. Joseph says:

    http://www.reuters.com/middle-class-infographic

    “or the first time in history, a truly global middle class is emerging. By 2030, it will more than double in size, from 2 billion today to 4.9 billion. Brookings Institution scholar Homi Kharas estimates that the European and American middle classes will shrink from 50 percent of the total to just 22 percent. Rapid growth in China, India, Indonesia, Vietnam, Thailand, and Malaysia will cause Asia’s share of the new middle to more than double from its current 30%. By 2030, Asia will host 64% of the global middle class and account for over 40% of global middle-class consumption.”

    The biggest unknown is how much more fossil fuel consumption we will need to maintain people living more advanced world lifestyles. If we can’t give the developing world more clean energy alternatives we won’t be able to reduce carbon emission by much (if at all) .

  62. RB says:

    sorry, i should have said the kosaka/xie GFDL model (with a TCR of 1.5) POGA-H output, not model mean

  63. Steve Bloom says:

    RB, that a shortfall in meridional heat transfer is involved is pretty much definitional. The big question for the models is what’s missing that results in the shortfall. Some of the suggested mechanisms have scary implications, unfortunately.

  64. Use the observational data of change in temperature with change in CO2 and you get 2C for TCR and 3C for ECS. ECS is based on land warming because it has no heat sink to get in the way.

  65. Rob Painting says:

    RB – “So, it seems to me to be a reasonable case to be sceptical of Trenberth’s 0.91 W/m^2 calculation. Any thoughts on this?”

    Of course. The OHC estimate is but one amongst a handful now, and they all come up with different values using different analysis methods. But the sharp uptake in OHC in the early 2000’s (as indicated by Balmaseda [2013]) is largely consistent with other observations, and physical expectations.

    It is quite reasonable to suspect that the change over from XBT’s to the ARGO system may have been responsible for this sharp increase, but multiple research groups have looked at this and the ‘spike’ in OHC still remains. Indeed, Balmaseda (2013) state that deep ocean warming remains even after removing the ARGO measurements entirely, although the trend is reduced.

    A compelling observation in support of the spike in OHC is the global sea level trend. Although it is complicated by factors such as the huge year-to-year variation in continental water storage (predominately La Nina & El Nino), sea level rise peaked over this period too. Sea level rise has continued it inexorable upward march, but at a reduced rate compared to the late 1990s-early 2000’s. Due to the thermal expansion of seawater when warmed, it is reasonable to expect a surge in sea level if that much heat was being mixed into the ocean (even after accounting for cabbeling – the increase in sea water density after mixing). So qualitatively, the early 2000’s spike in sea level rise is consistent with Balmaseda (2013).

    Multiple observations show that the wind-driven ocean circulation greatly intensified during that period too – the most obvious tell-tale being the spin-up of the subtropical ocean gyres. When that happens cold deep water upwells in the eastern tropical Pacific and spreads westward – cooling the equatorial surface ocean in the process. The reduced evaporation diminishes cloud cover and allows more sunlight to reach the tropical surface ocean. In the early 2000’s the anomalously cool sea surface temperatures in the tropics are quite remarkable. The strong export of surface water out of the tropical surface ocean and mixing down into the subtropical gyres, when the circulation intensifies, provides a mechanism for getting heat down into the deep ocean.

    So whilst one can be legitimately sceptical of the value of the imbalance, there is no reasonable basis for proclaiming that the large uptake in OHC during the early 2000’s is spurious.

  66. Rob, thanks for the comment. Very interesting. I was following some of the discussion between yourself and Troy Masters, but – to be honest – some of the subtleties escaped me.

  67. Rob Painting says:

    RB – “As Nielsen-Gammon showed in his posts discussing Kosaka/Xie, even after you fixing SSTs to observations in the Pacific thus replicating the relative flattening in the recent decade, model mean diverges from observations. So to me, it is something to keep in mind whether there is something else that is being missed in models.

    That makes sense. The CMIP3 models as a whole tended to underestimate the variation in winds which drive the wind-driven ocean circulation (not sure if this remains in CMIP5). As a result the multi-model mean exhibited reduced decadal variability when compared to actual observations. Since 2000 the Interdecadal Pacific Oscillation (IPO) has been in its negative (cool) phase – meaning the winds are mixing more heat down into the ocean via the subtropical ocean gyres (confirmed by ARGO observations and spatial sea surface height variations). If the models still underestimate the wind variation they should show greater warming than the observations.

  68. Thanks Rob! Your comments are always very enlightening, particularly for ocean noobs like me.
    Perhaps one additional remark re Kosaka/Xie: Most of the remaining mismatch is due to the recent cold winter anomalies over Eurasia. It’s substantial and something no model seems to be reproducing. Whether sporadic or indeed a temporary negative feedback caused by retreating Arctic sea ice remains yet to be seen.

  69. RB says:

    So whilst one can be legitimately sceptical of the value of the imbalance, there is no reasonable basis for proclaiming that the large uptake in OHC during the early 2000′s is spurious.

    That’s reasonable, Rob, thanks for the insights.

  70. RB says:

    Rob, while I acknowledge the phenomenon you describe, I must say that I was also left with Troy’s question from an energy balance perspective. That is, satellites measure E_in – E_out. The absolute value of this has a pretty large error, but the error in relative change from one year to another must be lower. Correct me if I’m wrong, but you are arguing that E_in shows a big jump (from the sun) while E_out doesn’t change much (related to surface temp). That difference is going into the oceans. But then, that implies that E_in – E_out should be showing a big jump (the 0.8W/m^2 change seen in Balmaseda) which on a relative basis, the satellites apparently don’t. So, while the effect might be there and small enough to escape detection at the TOA due to satellite accuracy for relative changes, I too am left with the doubt of whether the magnitude could be comparable to a 0.8W/m^2 drop which cannot be heat drawn from the atmosphere.

  71. RB, to be honest, I too am slightly confused about what’s discussed in Troy Masters’s post. It seems, from what Rob says, that the enhanced rate of uptake in the early 2000s is consistent with other observations. However, I’m still unclear as to why the change in rate seen around 2005 doesn’t appear to be reflected in the satellite measurements. I should probably work through Troy Masters’s post in a little more detail.

  72. RB says:

    ATTP,
    I was going to invite you to chime in. In my opinion, what Rob describes is a good explanation for why the surface temps don’t rise much because of processes that cause mixing into the deep ocean, but net heat going into the oceans (which take up 93% of the energy) is pretty much indeed the TOA imbalance from energy conservation.

  73. RB, yes I agree. That does indeed make sense to me too. However, I’m still slightly confused as to what Troy Masters seems to be highlighting – the sharp change in rate around 2005. If he’s correct and the difference is around 0.8 Wm-2 (i.e., 1.23 Wm-2 for the early 2000s and 0.38Wm-2 for the latter 2000s) then that seems so large that it would be hard for such a difference to be absorbed by some other part of the climate system. Hence one might expect it to be reflected in the TOA measurements. On the other hand, I realise these measurements are very uncertain and there are uncertainties associated with the OHC data, so could those be masking the change. Is the change really as big as Troy Masters estimates? Are we going to make a big deal out of short trends? Am I completely missing the point 🙂

  74. RB says:

    ATTP,
    Satellite accuracy is +/- 4W/m^2 but this paper points to two references (including Loeb et al) which use relative changes to interpret the energy budget. The 1.2W/m^2 rate for the early 2000s is confirmed by Trenberth/Fasullo (page 8).

  75. RB, yes I’m aware that the satellite accuracy is only +-4 Wm-2. Doesn’t the 1.2 Wm-2 in Trenberth & Fasullo refer to the whole 2000s, rather than simply the early 2000s, or have I misread that page?

  76. RB says:

    Hmm… you are right.. but Troy is also referring to numbers for the TOA imbalance and the average imbalance matches i.e 0.5*(1.23+0.38) is nearly the same as Trenberth’s 0.84 (0.7*1.2) for TOA imbalance. So, we are only arguing about the change point ~2005 calling into question the pre-Argo 2000-2004 numbers.

  77. RB, not really arguing, I would say 🙂 Yes, I think you’re right though. The issue is why the sudden change in the rate of ocean heat content uptake in about 2005 isn’t reflected in the satellite measurements and what that implies about the rates determined from OHC data. I don’t have any real sense of how to explain this or even the significance of this.

  78. RB says:

    ATTP,
    to me, the significance is that it raises questions about the ORAS-4 reanalysis estimates of OHC which look to be on the higher side of Levitus and Lyman thus leading to higher ECS estimates than Otto. So, a few more years of Argo data should settle this debate, I think.

  79. RB, sure, I agree. I certainly wasn’t suggesting here that it’s all done and dusted. It’s fairly clear that things will change as we collect more data and refine the analysis.

  80. Rob Painting says:

    RB – “I too am left with the doubt of whether the magnitude could be comparable to a 0.8W/m^2 drop which cannot be heat drawn from the atmosphere.

    Yes, that was Troy’s mental stumbling block too. The heat isn’t drawn from the atmosphere, it is in the ocean. It got there in the form of shortwave radiation and never much warmed the global surface in the first place. That’s why the thermal expansion component of sea level rose abruptly over that period despite increased cabbeling – there was anomalous heat input into the ocean that more than compensated for densification associated with strong mixing. That heat has not suddenly departed from the ocean – it would heat the atmosphere on its way out to space, and we would notice a drop in the ongoing rate of sea level rise. None of things have happened, so it hasn’t disappeared, as Troy seems to think would be the case.

    On a smaller scale a similar thing happens every La Nina – heat is taken up by the oceans and stored in the subsurface layers. There is only a small reduction in surface temperatures during La Nina because of compensating behaviour in the atmospheric circulation which takes up the slack in transporting heat poleward.

    The period after the monster 1997-1998 El Nino would have been an exceptional time. There was an enormous amount of heat lost from the surface ocean in that El Nino. The cool tropical surface waters afterwards (which show up in satellite observations) would have primed the tropical surface ocean (where most heating occurs) to take up a lot of heat. The combination of the intense wind-driven circulation (cool phase IPO), cool surface water (due to equatorial upwelling), and the strong reduction in tropical cloud cover should have seen an greater-than-normal uptake in ocean heat as intensified surface currents transport heat out of the tropics and down into subtropical gyres.

    Why don’t the satellite observations detect these changes at TOA? I don’t know, it’s not an area I’ve researched. I find it disappointing that no research group that I’m aware of has looked into this whole abrupt OHC uptake issue. From my perspective it would be a fascinating subject to delve into.

    So, we are only arguing about the change point ~2005 calling into question the pre-Argo 2000-2004 numbers.

    As stated earlier, when the authors of Balmaseda (2013) remove the ARGO data entirely the trend, albeit reduced, still remains. Then there’s the whole spin-up of the subtropical ocean gyres to consider, as well as the thermosteric sea level issue. Any person wanting to truly understand what went on would not simply ignore these observations.

  81. RB says:

    Hi Rob,
    All of the points I want to make have been made by Troy and I can’t disagree. Nobody is suggesting that the heat is drawn from the atmosphere. The only reason we made any statement regarding the atmosphere is that the extra heat, if Trenberth/Fasullo are right and it’s not detected by the satellites, has to come from somewhere. We also know that the land/atmosphere could not have supplied 0.8W/m2 to the oceans in one year.

    In any case, for illustration, let’s say the sun produces 340 units of energy and all of the stored energy in the system is stored within the ocean and look at the energy balance as follows.

    Time period Re-radiated Stored TOA imbalance
    Before 2000 339.4 0.6 -0.6
    2000-2004 338.8 1.2 -1.2
    2005-2010 339.4 0.4 -0.4

    So, the incoming solar hasn’t changed throughout the entire period, but during 2000-2005, the amount absorbed (stored) by the oceans has increased. This has to show up as a TOA imbalance! While we may not be able to detect within absolute units as small as 1.2, we can detect the relative changes (1.2-0.4=0.8) in year 2005. I think Troy, ATTP and I are all in agreement regarding this from an energy balance perspective. So, if satellite TOA is to be believed for relative changes, after accounting for measurement uncertainty, the conclusion must be that there was no 1.2 units in the prior period and 0.4 after, most likely there was a steady 0.6ish for the entire period, just getting delivered deeper into the ocean.

  82. RB says:

    Table re-do:
    Time period /Re-radiated /Stored /TOA imbalance
    Before 2000/ 339.4 / 0.6/ -0.6
    2000-2004 /338.8/ 1.2/ -1.2
    2005-2010/ 339.4/ 0.4/ -0.4

  83. RB says:

    BTW, Rob, the observations and phenomena you describe are interesting. Perhaps there was more heat absorbed in the tropical waters as you describe which was counter-balanced somewhere else so that we don’t see it in the global TOA imbalance measured by satellites. Or, the satellite data is wrong.

  84. Rob Painting says:

    RB – “The only reason we made any statement regarding the atmosphere is that the extra heat, if Trenberth/Fasullo are right and it’s not detected by the satellites, has to come from somewhere. We also know that the land/atmosphere could not have supplied 0.8W/m2 to the oceans in one year.

    Indeed, but the oceans aren’t heated by the atmosphere. I know it is a common misconception, but it is a misconception nevertheless. So stating that the atmosphere could have supplied any heat to the ocean is just plain wrong. That is not part of the physics of ocean warming due to the increased greenhouse effect.

    Just a question here; where do you think all that heat going into ocean during La Nina is coming from? The atmosphere?

    So, if satellite TOA is to be believed for relative changes, after accounting for measurement uncertainty, the conclusion must be that there was no 1.2 units in the prior period and 0.4 after, most likely there was a steady 0.6ish for the entire period, just getting delivered deeper into the ocean.

    This idea is rebutted by the trend of thermosteric expansion in the ocean – there was an increase over this period, which tailed off thereafter. It is also rebutted by the intensification of the wind-driven ocean circulation, which must have mixed more heat into the ocean up to 2004, and which then weakened slightly afterwards. The Interdecadal Pacific Oscillation is still in a negative (cool) phase, but certainly not as intense as it was in the early 2000’s. The spatial change in sea surface heights is a bit of a giveaway here – water strongly piled up in the subtropical gyres as Ekman transport intensified in the early 2000’s.

    There isn’t any physically-based reasoning to support the notion of a steady OHC uptake during recent times, and there’s a compelling argument (based on fundamental oceanographic principles) that the general trend should match that of Balmaseda (2013), although the magnitude is, as I said earlier, debatable.

    I don’t think that operating from the perspective of personal incredulity is very helpful. It may well be that Balmaseda/Trenberth etc are wrong, but then you have to wonder why these other observations support the strong uptake of heat by the ocean in the early 2000’s.

  85. RB says:

    So stating that the atmosphere could have supplied any heat to the ocean is just plain wrong.

    No, Rob, neither I nor Troy are trying to make that point or suggestion. Thanks for your insights though, there are a bunch of things I learned that I will follow up on.

  86. Rob Painting says:

    Okay then. My bad. I misinterpreted your earlier comments.

  87. Rob, thanks. Just to clarify, what you’re suggesting is that the change in trend in around 2005 (which suggests a faster uptake in the early 2000s) is supported by multiple lines of evident even though it’s not evidence in the satellite data?

  88. Rob Painting says:

    Anders – that would seem to be the case for the TOA satellite data. Of course it’s not just Balmaseda that show this surge in ocean heat uptake, Lyman’s work does too.

    Lyman’s methodology for infilling missing or spurious data differs from Levitus in that Lyman infills the data gaps with anomalies from adjacent grids. Levitus fills the data gaps with the averaged value of the anomalies – which has the effect of reducing anomalies toward zero. Warming, or cooling, should be reduced with the Levitus method, and that seems to be the case – Levitus’ dataset exhibits a more steady rate of ocean heat accumulation.

    At the moment we’re left with multiple observations which support the rapid uptake of ocean heat in the early 2000’s versus the TOA satellite data which appear not to. Either all these independent satellite, buoys, XBT’s, and ARGO floats observations are wrong, and it would suggest back to the drawing board for oceanography and the conservation of angular momentum, or the TOA satellite data is wrong, or at least is being misinterpreted.

  89. Either all these independent satellite, buoys, XBT’s, and ARGO floats observations are wrong, and it would suggest back to the drawing board for oceanography and the conservation of angular momentum, or the TOA satellite data is wrong, or at least is being misinterpreted.

    Thanks, that was what I had thought. Would be amazing if conservation of angular momentum turned out to be wrong, but that would seem unlikely 🙂

  90. tallbloke says:

    There is no ‘basic physics’ here. ECS is a flawed concept relying on a nest of unwarranted assumptions.

  91. Dunno what we did in a previous life to deserve this, but for future reference, next time I’d rather be reincarnated as a cockroach. #CoveringAllTheSidesOfPascalsWager_Part_1_of_INFINITY

  92. There is no ‘basic physics’ here. ECS is a flawed concept relying on a nest of unwarranted assumptions.

    Yup, that’s solid science there tallbloke. When you don’t like what a particular calculation is suggesting, claim that it is a flawed concept with unwarranted assumptions. That’s how we do it. Don’t bother with actually showing why, or discussing which of the assumptions are flawed, or how one might do it differently; just sound convinced that you’re right and brook no further argument.

  93. Steve Bloom says:

    Unless he’s removed it, TB has a post advocating for the ether hypothesis. Yes, *that* ether hypothesis. I rest my case.

  94. Steve, it does appear as though posts about the ether hypothesis are still present on tallbloke’s site. Fascinating.

  95. BBD says:

    I’m sure it’s all just an elaborate practical joke and we have been taken in. Of course TB couldn’t possibly be being serious.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s