Arguments
Unprecedented trade wind strength is shifting global warming to the oceans, but for how much longer?
Posted on 10 February 2014 by dana1981
Research looking at the effects of Pacific Ocean cycles has been gradually piecing together the puzzle explaining why the rise of global surface temperatures has slowed over the past 10 to 15 years. A new study just published in Nature Climate Change, led by Matthew England at the University of New South Wales, adds yet another piece to the puzzle by examining the influence of Pacific trade winds.
While the rate of surface temperature warming has slowed in recent years, several studies have shown that the warming of the planet as a whole has not. This suggests that the slowed surface warming is not due as much to external factors like decreased solar activity or more pollutants in the atmosphere blocking sunlight, but more due to internal factors shifting the heat into the oceans. In particular, the rate at which the deep oceans have warmed over the past 10 to 15 years is unprecedented in the past half century.
Research led by Masahiro Watanabe of the Japanese Atmosphere and Ocean Research Institute suggests this is mainly due to more efficient transfer of heat to the deep oceans. Consistent with model simulations led by Gerald Meehl, Watanabe finds that we sometimes expect "hiatus decades" to occur, when surface air temperatures don't warm because more heat is transferred to the deep ocean layers. A paper published last year by Yu Kosaka and Shang-Ping Xie from the Scripps Institution of Oceanography found that accounting for the changes in Pacific Ocean surface temperatures allowed their model to reproduce the slowed global surface warming over the past 10 to 15 years. However, the mechanism causing these Pacific Ocean changes has remained elusive.
The new study published by Matthew England's team helps explain how and why more heat is being funneled into the deeper ocean layers. The study indicates that a dramatic acceleration in equatorial trade winds, associated with a negative phase of a cycle called the Interdecadal Pacific Oscillation (IPO) has invigorated the circulation of the Pacific Ocean. This has caused more heat from the surface to be mixed down into deeper ocean layers, while bringing cooler waters to the surface. The combination of these two processes cools global surface temperatures. Like the rate at which heat is accumulating in the deep oceans, the recent strengthening of the trade winds is unprecedented, as the bottom frame in the figure below shows.
Top frame: Global surface temperature anomalies. Bottom frame: Pacific wind stress anomalies. From England et al. (2014).
Not only is this acceleration of trade winds unprecedented, but it also far exceeds anything captured by climate models. Hence they have difficulty reproducing the recent slowdown in surface warming. The catch is that oscillations eventually change phases, so as England notes, the strengthened trade winds and faster rate of ocean heat accumulation are only temporary.
"the heat uptake is by no means permanent: when the trade wind strength returns to normal - as it inevitably will - our research suggests heat will quickly accumulate in the atmosphere. So global temperatures look set to rise rapidly out of the hiatus, returning to the levels projected within as little as a decade."
The study estimates that by shifting more heat into the oceans, the strengthening trade winds can account for 0.1–0.2°C cooling of surface temperatures over the past 10 to 15 years. This would account for most of the slowed rate of warming, especially when combined with a recent study showing that the global surface warming slowdown is not as large as previously thought. The lead author of that paper, Kevin Cowtan said of this study,
"I think Professor England has uncovered the biggest piece in the puzzle of recent temperature trends"
In the figure below, the England study compares observed surface temperature changes (black) with IPCC model projections (red), and projections made by models that incorporate these changes in trade winds (green and blue). The models including trade winds can reproduce the surface warming slowdown. However, once the IPO cycle shifts and winds return to previous levels, the models see an accelerated warming at the surface, and temperatures start to catch back up to the IPCC model projections.
Annual (grey bars) and a five-year running mean (black solid line) global surface temperature measurements. Model projections are shown relative to the year 2000 and combine the CMIP3 and CMIP5 multi-model mean (red dashed line) and range (red shaded envelope). The cyan, blue and purple dashed lines and the blue shading indicate projections adjusted by the trade-wind-induced surface cooling estimated by the ocean model (OGCM), under three scenarios: the recent trend extends until 2020 before stabilizing (purple dashed line); the trend stabilizes in year 2012 (blue dashed line); and the wind trend reverses in 2012 and returns to climatological mean values by 2030 (cyan dashed line). The black, dark green and light green dashed lines are as per the above three scenarios, respectively, only using the trade-wind-induced SAT cooling derived from the full coupled model (CGCM). Shading denotes the multi-model range throughout. From England et al. (2014).
A consistent picture is emerging in the climate research;
@HK #49:
My original post didn't mention the sun. It put forward a relatively simple concept: if the IPO can suppress warming, it possibly can enhance it too. Hence some part of the warming of the late 70's to 2000 might have been driven by the +phase of the IPO. Whether this cycle in the IPO is driven by indirect solar effects (I don't think it could be TSI), I don't know. Certainly some scientists like Nicola Scaffeta think it is somehow related to solar cycles, even though he does not specify a hypothesis of the mechanism.
However, the cycle exists, whether we understand the mechanism or not. That is the real topic of this blog, and my point remains: could the +IPO have enhanced the warming between 1975 and 2000?
Last year, Tamino estimated the influence of ENSO with a 2-box model, NASA forcings and SOI data for ENSO.
Could someone explain me what's the difference with this tremendous new study ?
Klapper I have responded on aa index in a more appropriate thread, here.
Klapper, "could the +IPO have enhanced the warming between 1975 and 2000?"
Firstly, the IPO is an index rather than process and would appear to be a metric recording the superposition of a number of oceanic processes. If you restate the question, "could internal variability enhanced the post 1975 warming, (esp 1992 to 2006)" then yes, as noted it in Rahmsdorf 2007.
Klapper 51:
Of course a +IPO can enhance surface warming just as –IPO can enhance surface cooling as more heat is released from or stored in the oceans, but there are very good reasons to believe that its impact on the long-term climate is close to zero compared to anthropogenic greenhouse gases.
Look at the bottom frame in figure 1 in the blog post.
The IPO was negative until 1920, positive until 1945, negative until 1978 and then positive until 2000, in other words a typical oscillation around an average value without a clear trend.
The temperature in the top frame shows a very clear trend over the same period, so whatever caused that, the IPO cannot have been a factor when we consider the entire period from about 1910 to 2000.
If IPO was as important as you seem to think it is, the negative IPO should have had a cooling effect between 1945 and 1978. In that period, the global sulphur emissions tripled, from about 25 to 75 million tonnes, so the negative IPO got a lot of “help” from global dimming.
Why was the temperature more or less flat despite two large negative impacts? Maybe because one of them wasn’t as large as you seem to think?
The next period (1978-2000), with a positive IPO and a very clear warming trend, coincides with a continued growth of the greenhouse gas forcing and a moderate reduction of the sulphur emissions (global brightening). It’s worth noting that the oceans continued to accumulate a lot of heat, so the main driver of the surface warming must have been an external forcing even if the positive IPO gave a small contribution.
Regarding the period after 2000, I find it interesting and revealing that it takes “unprecedented” trade winds and likely unprecedented heat uptake in the oceans to slow down the surface warming, even with some help from a weak sun and (possibly) Chinese aerosols. And maybe the surface warming hasn’t slowed down that much after all.
Greenhouse gases are still the most important factor controlling the Earths climate!
mgardner and ianw01:
Thanks for the feedback. Perhaps it could be said that "all explanations are wrong, but some explanations are useful".
I do indeed manage to muddy the waters a bit by switching back and forth between force and shear stress, which have different units. SkS does not allow one to edit posts - perhaps a good thing as everything that is said is there for posterity. Shear stress is a concept that is difficult to put into simple words. I won't even try to go into strain.
..but to clarify mgardner's comment about the friction not depending on area - yes, as is easily demonstrated (if your brick isn't a cube) by turning the brick onto the smaller side (less area in contact with the ground) and seeing that friction (as measured by the pull on the rope) has not changed. The shear stress between the brick and the surface has changed, though. If the area is now half, then the shear stress has doubled - but with only half the area, the (shear stress X area) is constant... (to first order, as Ianw01 says).
Another place where shear stress shows up is on hill slopes. Stability depends on how large the shear stress induced by gravity is compared to the shear strength of the soil. As soon as the stress exceeds strength, the soil "breaks" and failure occurs.
We now return to our regularly-scheduled program.
Papy:
The 2-box model that Tamino uses has no "horizontal" in it, and very limited "vertical", so it can't include much in teh way of a physical process. When he incorporates an El Nino/La NIna process via the Souther Oscillation Index (SOI - a pressure indicator), all he is saying is that he can adjust heat transfer between his two ocean boxes (stacked vertically) according to the SOI. Including this adjustment in the model improves the fit, so that he had conclude that something (not sure what) about SOI/El Nino/La Nina cycles provides statistical explanation of the global surface temperature variation. It's a statistical correlation, not a physical explanation.
In contrast, the new study examines the physical process in much greater detail. It includes measurements of actual wind (not just a pressure index), reanalysis data (based on GCMs), and consideration of the physical processes. This is a much stronger methodology/confirmation than the statistical correlation at Tamino's.
Bob Loblaw - Tamino's model is even simpler than that. The two boxes are representative of the atmosphere and the oceans, not different levels of the oceans. What he's doing is a (relatively) simple statistical correlation of forcings and temperature with a 6-month lagged ENSO index (lag from the lowest residual result as per Foster and Rahmstorf 2011) as part of the multiple regression.
The fit of even this simple energy balance correlation is quite remarkable.
[Source]
@ scaddenp #54:
If by "internal variability" you mean some random process, I disagree. The IPO cycle doesn't look random. And if it is not random, then what drives it?
Klapper - The IPO is generally considered to be strongly driven by (among other things) teleconnections to the long term ENSO variations, as per Newman et al 2003.
Given a strong influence such as that, the IPO may wll be an add-on effect to the ENSO variations. And since over the long term ENSO variations cancel out, reddened ENSO teleconnections will too. Remember that correlation is not causation - variations in the IPO may very well be longer term lagged ENSO effects, rather than causes of atmospheric temperature variation in and of themselves.
As others have noted, you appear to be jumping around from thread to thread, topic to topic, in search of support for pre-existing convictions - rather than following the physical evidence to see where it leads. Once a particular unsupported assertion is shown not to hold (it's the sun!, the models are bad!, ENSO, now IPO) you jump to something else that might give your views support. The term for this behavior is confirmation bias.
I would recommend applying some of your skepticism to your own views, rather than making additional unsupported assertions.
@ HK #55:
"Of course a +IPO can enhance surface warming...."
I think you're starting to understand my argument. I accept there is an anthropogenic signal underlying the IPO cycle. However, the models have trouble with the IPO. They can't duplicate it so they ignore it and project that it should be warming right now at 0.20C/decade. After all, it's already warmed at over 0.18C/decade on a 30 year timescale and the CO2 level is higher then ever so it should be warming even faster than 0.18C decade, right?
The problem is that 0.18C/decade trend was boosted by the +phase of the IPO. The underlying anthropogenic signal is not that strong.
"...likely unprecedented heat uptake..."
That statement is probably not true since a rolling 30 year trend on the Church & White tide level sea level dataset shows an early peak of 2.13mm/year in the period 1933 to 1963 and 2.09 mm/year in the period 1980 to 2010. So if there was unprecedented heat uptake there should have been unprecedented sea level rise, unless you believe that the sea level rise before 1963 had higher levels of glacier melt than in the period 1980 to 2010.
Obviously those rates of sea level rise span both + and - phases of the IPO. However, if you look at shorter time spans (10 years) you will also find that current heat level gain is likely unprecedented, or not by much if it is. The sea level gain 1943 to 1953 was at 3.5mm/yr. 2000 to 2010 was 3.2 mm/yr., so there is nothing unprecedented about the last decade (Church & White ends at 2010 so I can't compare trends later than that to the '50s).
How much does this extra heat dumped into the ocean affect sea level rise?
Klapper - 1976-1999, the last positive phase of the IPO, was interrupted by the large tropical volcanic eruption of Mt Pinatubo in 1991. - which cooled the Earth for several years and made an enormous impact on heat uptake by the ocean. See the graph below from Balmaseda (2013). This would have to be taken into consideration.
@Rob Painting #63:
Pinatubo would affect my 30 year trend, but not the 10 year (2000 to 2010). Mind the 10 year trends get a bit wobbly as opposed to the 30 year. However, looking at 10 year trends which avoid the Pinatubo problem, I can see that the current rate of sea level rise from tide guages is comparable to the rate of rise from '43 to '53. The complicating factor is net runoff/meltwater input into sea level rise during the recent and mid 20th periods. Because the planet is warmer over the last 10 years than in 1950, one would expect a higher rate of sea level rise from net melt runoff. What that means is there must have been a very large heat gain post mid '40's. Unfortunately the heat content data is pretty sparse back then and even into the '50's.
Klapper - you made this comment @61 - "The problem is that 0.18C/decade trend was boosted by the +phase of the IPO. The underlying anthropogenic signal is not that strong."
You neglected to account for the strong cooling of both the ocean and global surface from the light-blocking effects of sulfate aerosols ejected into the atmosphere by the 1991 Mt Pinatubo eruption. It goes to the heart of your claim highlighted above.
@KR #60:
"As others have noted, you appear to be jumping around from thread to thread, topic to topic..."
My motives are irrelevant to whether my arguments have merit or not. The facts determine that. Read my post number 61, which ties my thinking together somewhat. The the IPO, ENSO, models, solar etc are not disconnected arguments, they are components of an overarching idea, namely that there is a climate cycle overprinting a secular warming trend, and the warm phase of that cycle has confused the tuning of feedbacks in the AOCGCMs. Not really my idea though is it?
Indirect solar effects may be a factor in driving this cycle, I don't know. What do you think it is? Roy Spencer just treats it as an oscillation, nothing more. Whatever drives it, saying it doesn't exist because you don't know the mechanism doesn't make logical sense.
@Rob Painting #65:
My 30 year 0.18C/decade trend was from 1978 to 2007 inclusive. Because Pinatubo lies right in the middle of this period, it has no leverage on the trend either way. El Chichon does have an effect on this trend however, making it stronger than it would be otherwise by introducing a cold deflection at the beginning of the record.
Klapper - Variations overlying an underlying trend signal? Certainly, and in fact that is exactly what this particular thread is about. So are papers such as Foster and Rahmstorf 2011, which look at identifying just how much influence different variations have on the global temperature trends.
However, statements such as "...there is a climate cycle overprinting a secular warming trend, and the warm phase of that cycle has confused the tuning of feedbacks in the AOCGCMs" are neither correct nor helpful. There is no single climate cycle overlying anthropogenic influences, rather there are many sources of variation, global circulation models are not tuned to their overall outputs but rather are composed of multiple components driven bottom up by physics constrained by observations, and incidentally the word 'secular' is inappropriate as there is no religion involved.
If as you say your major point is that you feel climate models are mistuned, I would suggest taking that to an appropriate thread, rather than hopping from influence to influence looking for a significant factor - as you have (solar, IPO) in this one.
Addendum to my previous post - 'secular' is used in many papers, and is valid in it's less common definition of "existing or continuing through ages or centuries". However, in most cases 'long-term' would be more appropriate for the durations discussed, and the the religous linkage is by far the most common usage.
Just my personal opinion - I dislike the use of that term in science.
Klapper - "My 30 year 0.18C/decade trend was from 1978 to 2007 inclusive. Because Pinatubo lies right in the middle of this period, it has no leverage on the trend either way"
Now you are descending into complete nonsense. This demonstrates you cannot be taken seriously.
Shorter Klapper: LINEST(1,2,3,4,5) = LINEST(1,2,-327,4,5)
@ Rob Painting #70:
@ DSL #71:
I think DSL is saying I am in fact correct about the trend with Pinatubo in the centre. It affects the intercept of the regression equation but not the slope.
Klapper #61:
Sea level rise from thermal expansion is not an accurate proxy for ocean heat uptake without a more careful analysis.
Take a look at this graph in Wikipedia. (the water’s density for each degree from 0°C to 100°C is listed further down the page) If you convert density change to expansion you get this for some temperatures:
4-5°C: +0.001%
6-7°C: +0.004%
10-11°C: +0.01%
15-16°C: +0.016%
20-21°C: +0.021%
Heating water from 10 to 11°C makes it expand ten times more than if you heat it from 4 to 5°C! So the amount of thermal expansion doesn’t only depend on how much heat is being added, but where it is added, since warm water expands more than cold. These numbers are for fresh water. Salt water continues to contract down to its freezing point, but the principle is the same.
And where has the ocean heating taken place?
If you study this and this and this graph thoroughly, you will find that a rising fraction of the heat is accumulated in the deeper and colder parts of the oceans.
With a caveat for quite uncertain pre Argo data I got this result for the fraction of the heating taking place deeper than 700 meters:
1957-1994: 25%
1994-2011: 38%
2005-2013: 49%
The tendency is quite clear: More of the heat accumulation happens in the deeper, colder parts of the oceans where each unit of energy cause less thermal expansion than in the warmer, upper layers.
And finally, this graph shows that the warming in the upper 100 meters have been almost zero for the last ten years. Each unit of energy added here would produce at least 10-20 times more expansion than in the deep and cold parts of the oceans!
Conclusions:
1. You can’t translate thermal expansion to heat accumulation without knowing where the heating takes place.
2. Much of the SLR from increased melting of ice sheets in the 2000’s has been offset by decreased thermal expansion because more of the warming happens in colder water. That probably explains the apparent lack of acceleration seen here.
Not really, Klapper. I'm pointing out that you can throw an absurd number in the middle of a simple point-to-point linear analysis and get a conclusion that has absolutely no informational content with regards to physical reality. I should have made the series (1,1,1) and (1,4000,1).
@ DSL #74:
I'm trying to figure out what your point is. I didn't throw an absurd number in the middle of anything. I'm doing linear regression, not point to point, of the SAT record. Pinatubo has no effect on the warming rate of a trend from 1978 to 2007, although I didn't think about that until Rob Painting (incorrectly) pointed out it was affecting the trend. Maybe for some periods, but not the trend I picked which was basically to show what the maximum warming rate had been in the late 20th, early 21st.
HK @73
i was aware of the 10 fold difference in coefficient of thermal expansion of water at different temperatures.
Given that the models did not predict the heat going into the cold lower sections of the ocean over the last decade, caused by the trade winds, resulting in no appreciable surface temperature increase over the last decade when surface temperature increases were expected by the models, then have the models got the sea level rise predictions correct?
ie if the models didn't have that the heat was going into ocean depths over the last decade, do the models have the correct amount of heat going into the oceans in future prediction, and more importantly given the 10 fold difference in coefficient of thermal expansion, if the models predict heat going into the wrong part of the ocean the sea level rise predictions could be dramatically out?
Do the models need to be revised, if not why not?
Klapper, what I'm saying is that you can swim in statistics and never get physically wet. You dismiss Pinatubo because you can make it seem to effectively disappear by running linear regression across a specific period. Does Pinatubo actually have no effect on the temperature element of climate over the next 5-10 years? Of course it doesn't! If less solar energy reaches Earth's surface, global mean surface temperature is going to drop relative to what it would have been without the decrease in solar. Further, that drop is going to alter OHC uptake, as Rob pointed out. That will further affect GMST and ocean-atmosphere dynamics. There's probably a feedback from stratospheric warming as well, though I haven't read the literature on it. Pinatubo is written into GMST in complex ways. For you to claim that "Pinatubo has no effect on the warming rate of a trend from 1978 to 2007" is for you to claim that Pinatubo effectively didn't exist where climate is concerned. If Pinatubo doesn't occur, do you think we have precisely the same GMST monthly values from 2000-2007? Do we have the same shaped annual time series? Does ENSO follow the same path post, say, 1995?
Knapper @67
What you say about 1 data point (Pinatubu) on the middle of a linear regression not changing the trend is partially true. It will not change the gradient but will lift the overall line up or down.
However with Pinatubo the question is what effect it had on the years following 1991? It probably had an effect and thus would also change the gradient but I am unaware of any analysis quantifying the effect for what you are looking at. It probably exists.
So it is difficult to rule out Pinatubos effect without more data.
The comment by RPainting@ 70 is simply offensive name calling.
[JH] Your assertion about Rob Painting's comment is patently false. Please cease playing the "victim card."
Markoh,
Scientists continualy review all their models. Your suggestion that they are not is simply false and demonstrates that you do not understand how science and models work. There is a question of how (and if) the physics in the models needs to be adjusted. Some of the models handle ocean heat well, others not so well. Scientists are trying to determine if this is because of random weather or climate. Adjustments are always made when new physics is learned.
If Klapper really thinks that two different data sets are equivalent because the slope of the linear regression is the same, then perhaps he should review Anscombe's Quartet - four different data sets that have identical (to a certain precision);
- mean X and mean Y
- standard deviations (or variance) of X and Y
- same linear regression (slope and intercept)
- same correlation coefficient
...so, from the point of view of several common statistical tests, the four data sets might be thought of as "the same".
Yet when you graph them, you get (from the Wikipedia page linked to above):
KR:
Thanks for the correction on Tamino's model.
Michael Sweet @79
i ask a question and you respond with insults, which is unacceptable. However moving forward, you made statements of how models work. Is this knowledge from modelling you have personally done? If not can you please cite a source for the explanation?
[JH] What is unacceptable is your mischaracterization of responses to you. Please cease and desist.
Markoh @76:
Honestly, I’m not a model expert, so I don’t know how much deep ocean warming they have predicted in the future. If they have predicted less deep ocean warming than we are seeing now and this pattern continues, it will certainly reduce the expected SLR from thermal expansion.
But it’s also worth noting that the models have missed the accelerated ice loss from Greenland and Antarctica. The annual contribution to SLR is now between 1 and 1.5 millimetre per year from Greenland (figure 56 in the link) and maybe about 0.5 millimetres from Antarctica (360 gigatonnes of ice = 1 mm). Add the melting mountain glaciers, and we’ll find that melting land ice is now the dominant cause of the SLR.
HK @ 83
Thanks for that. It makes sense. I am not an expert by any means, but I do know from non-climate modelling that the assumptions into the model are as important as the results out.
Good point about the accelerated ice loss. Suggests the earth is more effective/has more mechanisms for transporting heat to the cold reaches than models anticipated.
Markoh,
I do not see the insult that you are so concerned about. Clearly you do not understand how the climate models work since your questions are inappropriately phrased.
There have been a number of new posters lately (especially you) who have been extremely sensitive about responses to their questions. By the same standard, you have insulted me by your response. I only answered your question. Please point out what I said that was insulting.
I teach High School science and scientific models are discussed in class. The models are continually revised as new knowledge is gained. This statement is a general statement that is common knowledge. There is a meme at denier sites that models are not corrected as new knowledge is learned. That is false. Read the posts on Realclimate about climate models if you do not understand them. Which statement do you question that I made ? Your statement is to vague to respond specificly to.
To support what michael has written, climate modellers have conferences, where they discuss ways to improve their models (you need to know where the problems lie if you want to improve them, so talking about the failings of models is an important activity). Climate modellers also have journals where they publish papers explaining how to improve modelling of climate (again, you can't explain how to improve models without discussing the failings that justify the need for the improvement). Modellers also take part in Model Intercomparison Projects (e.g. CMIP3 and CMIP5, but there have been many others focussed on specific topics).
Your comment "However with Pinatubo the question is what effect it had on the years following 1991? It probably had an effect and thus would also change the gradient but I am unaware of any analysis quantifying the effect for what you are looking at. It probably exists." demonstrates that you are not familiar with the work that has been done on modelling, as the sucessful forcasting of the effects of Pinatubo was a useful exercise in evaluating climate models. It is not an insult to point this out; we all start knowing very little and learn more by listening to those more expert than ourselves. However looking for insults where none was intended is not going to help you learn, so you need to take such comments as useful advice.
January GISS L-OTI is out (unadjusted): 0.72C.
DJ MEI = -.318
DEC PDO = -.41
CT SIE = lowest in the satellite period
CT SIA = 2nd lowest in the satellite period
Global SIA = 8th lowest in the latellite period
Still, DM @ 86, perhaps some slack should be cut here. I mean, I did a Google Scholar search and only came up with ~8000 hits. Perhaps I over-limited my search string (pinatubo global mean surface temperature volcanic).
Markoh: Following up on comments by Michael and Dikran--the U.S. National Research Council wrote a whole report in 2012: A National Strategy for Advancing Climate Modeling.
DSL, slack-cutting is always a good idea, I was aiming for a concillatory tone (obviously I was rather off-target ;o), I was just trying to show that michaels comment was was not unreasonable and not an insult.
Just performing an ordinary google search for "Pinatubo climate model" works pretty well, and a Google Scholar search for "pinatubo climate model" brings up Hansens 1992 paper as the first hit for me. The real point is though that anybody that had looked into the reliability or otherwise of climate models should be particularly aware of the effects of Pinatubo as it provided a chance to test the predictive power of the models, and so has been very widely discussed.
There is nothing wrong with not knowing things (there is *plenty* I don't know, which is why I read much more than I write here), but in order to learn, you need to be able to admit that you don't know and not get upset if others tell you (this is a good guard against the Dunning-Kruger effect, so they are doing you a favour in the long run, even if it isn't pleasant at the time).
DK @86
i believe if you read my post @78 more carefully you will see my focus was the mathematical consequence of outlier data on a linear regression.
And when I say "I am unaware" that is what I mean, and further pontifications are not necessary.
[JH] Please lose the snark.
Sorry Markoh, I am not going to rise to the bait ("pontificating"), you may be here just to argue for the sake of it, but life is just too short. If you want to know why posters here are being somewhat abrupt with you, read your post above and consider why that might be.
Markoh:
Saying another poster is "pontificating" is insulting. You have already insulted me on this thread. If you want to discuss science with others here you need to stop insulting the other posters.
[JH] Markoh is skating on thin ice and is close to recusing himself from posting comments on SkS.
Markoh:
Please note that posting comments here at SkS is a privilege, not a right. This privilege can be rescinded if the posting individual treats adherence to the Comments Policy as optional, rather than the mandatory condition of participating in this online forum.
Please take the time to review the policy and ensure future comments are in full compliance with it. Thanks for your understanding and compliance in this matter.
@HK #73:
I agreed in theory with HK's post, but posted a rebuttal on how well it fits with numerical analysis of the NODC data. For some reason this didn't pass the censors so perhaps the moderator could explain why.
Thankyou.
Klapper,
Your post was deleted for sloganeering (I cannot delete posts, I am not a moderator).
In your post you discussed looking at data, but you did not show any results. You reached a conclusion that was unsupported, since no data was shown. Reaching unsupported conclusions several times in a row is sloganeering. My impression of your posts is that you are long on conclusions and short on data that supports your claims. In a scientific discussion you must support your claims with data, preferrably peer reviewed data. People used to posting on other blogs sometimes have difficulty with the concept of referring to peer reivewed data, since the style there is everyone just says what they think without any data. You might want to consider asking more questions and see what type of response you get.
It is much better to cite a peer reviewed publication to support your position than to make a calculation. Obviously a calculation you (or I) do is not peer reviewed. If a peer reviewed publication does not support your claim you might want to consider if it is reasonable. Can you really do these calculations when you cannot deal with gridded data (I doubt it)? Why should I believe your calculation?
While HK had calculations in his post, he showed the data and had 4 links to other graphs. His conclusions were pretty basic. (I read his conclusion as the problem is complicated. It requires detailed [peer reviewed] calculations to reach a conclusion).
moderator: delete my post if it is inaccurate.
@Michael Sweet #96:
My calculations weren't peer-reviewed, but then neither was anything in HK's post. In fact I used data used to produce all his graphs come from. However, that was on a different computer a long way's from me now and I note that currently links to ascii data on pentadal data seem to be broken, plus I can't find the composited pentadal data I used anyway, only gridded data.
I'll try to get the excel file sent from the computer's owner and post my graphs and numberical results.
@HK #73:
I've analyzed your results using a rolling delta OHC anomaly using data from these sources:
LINK
LINK
In your analysis, which you don't describe or source explicitly you compared a 37 year delta, to a 17 year (exclusive), to an 8 year (overlapping). I'm not sure what the significance of these periods are as you don't elaborate.
I used a rolling 15 year delta OHC (NODC ZJ anomaly, creating the 700-2000m delta from the 0-2000 and 0-700 datasets. Following is a graph of the results:
As you can see from the graph, there were 15 year periods ending in the early 70's where the deep ocean heat gain was over 50% of the shallow ocean heat gain; certainly more than the last 15 years where the % of deep ocean is only +35% or so. I'm not sure how meaningful this comparison since the data are very sparse for the deep ocean, but I don't agree there has been an unprecendented amount of heat gain by the deep ocean relative to the shallow in recent years. Your theory might still be correct if you compared 500 to 300 metre data, or some other depth ranges, but what I have easy access to does not have the resolution.
[RH] Hotlinked URL's that were breaking page formatting.
@HK #73:
Graph legend is in error running delta's are 15 year not 20.
I have a couple of questions concerning the model projections:
1. Can anyone tell me the relationship between what the authors call the AR4+AR5 model projections and the model projections shown in figure 1.4 of the latest IPCC report?
2. Also, does anyone know why the authors begin their model projections in 2000 and not 1990 as in the figure I referenced in #1.
Thanks for any information!