Guest post by Bob Tisdale
This is a deviation from my typical presentation of a subdivided dataset. Usually, I divide the dataset in a way that is intended to illustrate how and why natural variables can explain the warming over the term of that data. In this post, I’ve broken it into subsections that allow the data to show behavior that cannot be explained by anthropogenic global warming, and I’m leaving it to the proponents of manmade global warming to explain, through their own data analyses of the five subsets, how those five subsets show continuous and continued warming, when clearly they do not.
Believe it or not, the NODC’s ocean heat content data for the depths of 0-700 meters contain a couple of hockey sticks—that is, no warming for 4 decades and then, presto, there’s warming. One of the datasets is relatively small, but the other is quite large, representing about 39% of the surface area of the global oceans.
FOREWORD
The National Oceanographic Data Center (NODC) Ocean Heat Content data is only available to the public in an easy-to-use format through the KNMI Climate Explorer, where it is available only for the depths of 0-700 meters. The NODC recently released its new dataset for 0-2000 meters but it’s available only to the public in limited subsets and it is smoothed with a 5-year filter, which makes it useless in attribution studies. Regardless, this doesn’t stop proponents of anthropogenic global warming who repeatedly and nonsensically claim only greenhouse gases could have caused the warming and that the warming continues.
We know the NODC’s ocean heat content data for depths of 0-2000 meters are available on a monthly basis because the UKMO uses it in its EN3 ocean heat content dataset. The NODC and UKMO apparently do not want KNMI to provide the public easy-to-use access to UKMO EN3 data (in unadjusted form) because by KNMI removed it from their Climate Explorer only a day or two after my first post that included that data. Refer to the post here.
With that in mind, please don’t ask me why I did not use the NODC ocean heat content for 0-2000 meters in this post. That will save me the time of suggesting to you that you read the post instead of looking only at all the pretty pictures.
USING A GLOBAL DATASET TO REPRESENT GLOBAL WARMING IS MISLEADING
It sounds odd, but it’s true.
By looking at a dataset on a global basis, one can only assume greenhouse gases play a role in the warming. As I’ve noted in numerous previous posts, dividing the dataset into smaller subsets allows the data to present how it truly warmed.
That is, global temperature (and related) metrics show evidence of global warming. These include sea surface temperature, lower troposphere temperature, combined land+sea surface temperature and ocean heat content for depths of 0-700 meters. See Figure 1 for the NODC global ocean heat content anomalies for depths of 0-700 meters. While each of those datasets show warming has occurred, for more than 3 ½ years, I have illustrated and discussed here and in cross posts at WattsUpWithThat how the warming over the last 3 decades can be attributable to natural factors, primarily strong, naturally occurring El Niño and La Niña events. I’ve also published an ebook in pdf form that explains the natural processes that cause the warming. It’s written for those with and without technical backgrounds.
Figure 1
I’ve divided the global oceans into 5 subsets for this presentation. See Figure 2. As noted earlier, I’m taking a change of tack for this post. I’m presenting the data so that it shows how it contradicts the hypothesis (fancy word for guess) of manmade global warming.
Figure 2
But in this post, as also noted earlier, I’m leaving it up to proponents of anthropogenic global warming to explain, based on their data analyses, not climate models, how and why they find evidence of continuous and continued anthropogenic global warming in all 5 of the following subsets.
LOW-TO-MID LATITUDES OF THE NORTH ATLANTIC
The ocean heat content anomalies of the low-to-mid latitudes of the North Atlantic (0-45N, 80W-20E), Figure 3, would be ideal for proponents of anthropogenic global warming if it wasn’t for the fact that it stopped warming in the early 2000s. With its excessive trend (0.215 GJ/m^2 per decade) versus the global trend (0.075 GJ/m^2 per decade), this portion of the North Atlantic exhibits signs of the ocean heat content equivalent of the Atlantic Multidecadal Oscillation, but with this dataset, it has already started to cool.
Figure 3
This subset clearly fails to illustrate “continued recent warming”.
NORTHERN NORTH ATLANTIC
Figure 4 shows our first ocean heat content anomalies subset with a hockey stick-like curve. Ocean heat content anomalies for the Northern North Atlantic (45N-90N, 80W-20E) cooled significantly for 40+ years, from 1955 to 1996, a time period when manmade greenhouse gases were increasing at accelerated rates. Then, magically, in 1997, ocean heat content anomalies there skyrocketed. Notice also how the ocean heat content anomalies for the Northern North Atlantic peaked in the early 2000s and have been cooling since then.
Figure 4
This subset definitely does not show “continuous warming”.
SOUTH ATLANTIC
As clearly shown in Figure 5, since 1960, the ocean heat content anomalies for the South Atlantic (90S-0, 70W-20E) warmed in 1981 and over the 2-year period of 2004 and 2005. For the multidecadal periods before and between, and for the short period after, the South Atlantic exhibits no evidence of warming. In other words, the South Atlantic ocean heat content anomalies only warmed during the three years of 1981 and 2004/05. I don’t believe greenhouse gases can pick and choose which years they’ll impact and then sit idly by for the other 50+ years.
Figure 5
The South Atlantic does not pass the test for “continuous warming”. The same can be said for the next subset.
EAST PACIFIC
Figure 6 presents the ocean heat content anomalies for the first of the two major subsets. The East Pacific (90S-90N, 180-80W) covers about 33% of the surface area of the global oceans. There are a number of papers that discuss the impact of the 1976 Great Pacific Climate Shift on the sea surface temperature of the East Pacific. It also appears to have had an impact on the ocean heat content of the East Pacific. The data also exhibits an upward shift in 1990, immediately after the 1988/89 La Niña event, which was the strongest single season La Niña event in recent history. If not for the upward shifts in those two years, the East Pacific ocean heat content anomalies show no evidence of warming for the decadal and multidecadal periods before, between and after them.
Figure 6
INDIAN-WEST PACIFIC
The Indian-West Pacific (90S-90N, 20E0180) is the largest of the subsets presented in this post. It represents about 39% of the surface area of the global oceans. Curiously, it is the only subset to exhibit warming in recent years. Note also how the ocean heat content anomalies for this region failed to warm from 1955 to 1997, even though greenhouse gas emissions were increasing over those 4 decades. If anything, they cooled slightly. Then in response to the 1998/99/00/01 La Niña, ocean heat content shifted upwards. That upward shift actually makes sense, though we might have expected to see other less-notable shifts in the past. What really looks awkward is the continued warming in response to the pair of double dip La Niña events that followed the moderate-to-strong El Niño events of 2006/07 and 2009/10. They weren’t super El Niño events by any stretch of the imagination, but they caused unusually strong ocean heat content rises according to the data.
Figure 7
This is when I wish we still had access to the UKMO EN3 ocean heat content data through the KNMI Climate Explorer. That dataset presented the ARGO-era ocean heat content data without the NODC’s constant adjustments. Could it be that those adjustments are the only reason the ocean heat content data in this region continues to exhibit warming? Do we assume that when corrections are made they’re made equally across all ocean basins? They may not.
Regardless, the Indian-West Pacific dataset fails to provide the continuous warming one would expect from anthropogenic greenhouse gases.
CLOSING
Any takers?
If you’re a proponent of anthropogenic global warming and if you choose to present your data analyses, please do so using data available on a gridded basis in a reasonably easy-to-use format, from a source such as the KNMI Climate Explorer, as I always do in my blog posts so that anyone can verify results. What we’re not looking for are claims to the effect of, “oh, that’s caused by aerosols.” You’ll need to supply the data source to accompany your claim, to show cause and effect. If you’re a modeler and you’d like to discuss your models, please ask KNMI to add to their Climate Explorer the outputs of your ocean heat content simulations that exist in the CMIP3 and CMIP5 archives.
Please also explain, as part of your analyses, how anthropogenic forcings are responsible for the disparity in the trends, as shown in Figure 8. Don’t forget the data to accompany your claims.
Figure 8
If you’re a regular visitor to SkepticalScience, please don’t waste your time and present the gif animation The Escalator. That would clearly indicate you haven’t a clue what you’re talking about.
SOURCE
The data presented in this post is available through the KNMI Climate Explorer.
WHEN did the ARGO BOUYS come on line?
WHAT is the basis for this alleged “data” prior to that point?
Garbage IN, Garbage out… The king has no new clothes. This is a contrived data set (except for the ARGO data, which is MEANINGLESS.
Pitiful.
Max
Thanks for the answer and I completely understand. I will suck it up and download the pdf. Oh and great work. Please keep it up. Those out there that truly appreciate real science, not the voodoo political science we see all too often in climate science, really appreciate what you are doing.
A quick look at N. Scaffeta’s paper “A shared frequency set between the historical mid-latitude aurora records and the global surface temperature” shows something corresponding to the short cycle in aurora data 1872-1966 (figure 4D).
There is a twin peak of (visual read-off) 3.7 and 4 years. This is one of the strongest peaks in relation to his tests for significance against red noise. Though his paper deals more specifically with the longer cycles, the fact that this is found in aurora data corroborates my suggestion that this frequency is of extra-terrestrial origin.
Good writing is in itself a pleasure, of course this goes much further.
Bob says: The anthropogenic GHG effect is represented as being continuous and persistent.
That does not spell linear.
Even if there is a linear component you are not going to be able to fit it until you have a stab at removing the bigger pseudo cyclic signals.
Yabut, if the solar explanation is correct why with a grand solar thingy from 1920-2000 take until 1980 to show up? And if there’s a 60 year build up there’s another 60 years until things get back to 1920 data. The temperature may not increase anymore, but it is still far above the 1920 values. A look at any of the 5 regions show them to be far warmer than 1955 data, anyone can see that. The CO2 theory is not reasonable, but they have the world pretty much eating out of their hands, so the emphasis is on us to provide a compelling reason for the warming, not to yap that CO2 couldn’t do it. After all they say they couldn’t find any other means for the AGW theory so we must provide that and not be so negative about CO2. We cannot disprove AGW by crying about CO2, we need to provide evidence that can be used to prove there is a method of warming that doesn’t rely on CO2.
vukcevic says: “I distrust the ‘teleconnections’ hypothesis since contradicts physics as I understand it:
“If the heat energy has to to ‘teleconnected’ i.e. transported from Pacific to Atlantic, it has to be via atmosphere, but we know that difference in the heat capacity between water and air makes that next to impossible.”
There is no transport of heat through the atmosphere as part of a teleconnection. You’ve misunderstood the term. To save myself 5 minutes, let me quote from “Who Turned on the Heat?”:
We already know how and why the trade winds weaken in the eastern tropical Pacific during an El Niño. Those changes in atmospheric circulation in the Pacific, in turn, cause the trade winds in the tropical North Atlantic to weaken, too. The slower trade winds blowing across the surface of the tropical North Atlantic Ocean don’t cool the surface waters as much as they normally would; there’s less evaporation with the slower trade winds; so the sea surface temperatures warm in the tropical North Atlantic. That’s only part of the explanation. With trade winds in the North Atlantic at their normal strength, cool waters from below the surface are pulled up to the surface there. In other words, where upwelling occurs, it is occurring at its normal rates when the trade winds are at their normal strengths. When the trade winds weaken during an El Niño, there is less cool water being pulled up to the surface, so the tropical North Atlantic warms as a result of that weakening process also. Refer to Wang (2005) ENSO, Atlantic Climate Variability, And The Walker And Hadley Circulation for a more detailed discussion.
HHHH
Because the tropical North Atlantic consistently warms in response to El Nino events, the sea surface temperatures there and in the NINO3.4 region are said to be teleconnected.
Even allowing for some lumpiness, isn’t there too much lumpiness? That’s what Bob should have been attacking, and that core of his argument remains after its excesses have been pruned.
Ian W says: “I know that it is fashionable to use ‘anomalies’ rather than actual heat content. However, is there anywhere where total ocean heat content is available?”
I write ocean heat content anomalies out of habit. According to Levitus et al (2012), their ocean heat content data is presented as anomalies so my adding anomalies is redundant. Refer to their paragraph 3:
http://data.nodc.noaa.gov/woa/PUBLICATIONS/grlheat12.pdf
So the short answer to your question is, nope.
One question I have is, from looking at the graphs is, are only anomalies being studied? If so, this seems kind of short sighted since eventually there should be no anomalies. We’ll also never know the quantity (5 days out 365) or whether or not the anomalies out weigh normality.
Have I missed something or am I looking at the wrong part of the data?
eric1skeptic: I understand what you’re saying, but the sudden shift looks odd. Also, why then aren’t volcanic eruptions more evident in the other subsets? And are the apparent dips in 1982 and 1991 in the Indian-West Pacific subset actually responses to ENSO and not volcanic eruptions? It’s very difficult to find a volcanic aerosol signal in OHC data.
phlogiston says: “Ever heard of thermo-haline circulation (THC)? It takes seawater about 1000 years to ‘slosh’ around the planet.”
On the surface and “near surface”, the time is measured in months and years. See my October 14, 2012 at 7:24 am reply to Denis Rushworth. But it depends on how we’re defining “sloshing”.
Regards
P. Solar: With respect to your comment on my earlier reply to Mike Jonas I presented a follow-up:
Update: Maybe the sudden shift in the response to ENSO during the last decade or so is real. If we look at the difference in sea level pressure between the East Pacific-Atlantic subset and the Indian-West Pacific subset [(EP-A 90S-90N, 180-20E) MINUS (I-WP 90S-90N, 20E-180)] smoothed with a 121-month fliter, we can see that there was a recent change in delta SLP.
http://i49.tinypic.com/2cru97c.jpg
Or maybe it’s a combination of that and the under sampling that took place prior to the late 1990s.
And again, P. Solar, there is nothing wrong with my investigations or analyses. The fact that you want more simply indicates that YOU should carry the ball farther. You’d never be happy with what I’ve done anyway. Have fun.
RHS says: “One question I have is, from looking at the graphs is, are only anomalies being studied? If so, this seems kind of short sighted since eventually there should be no anomalies.”
OHC data is only presented as anomalies. Refer to paragraph 3 of Levitus et al (2012):
http://data.nodc.noaa.gov/woa/PUBLICATIONS/grlheat12.pdf
They write: “We use the term ‘ocean heat content’ as opposed to ‘ocean heat content anomaly’ used by some authors because ‘ocean heat content’ is an anomaly by definition. OHC is always computed with a reference mean subtracted out from each temperature observation. Otherwise the OHC computation depends on the temperature scale used.”
Bob says: It’s very difficult to find a volcanic aerosol signal in OHC data.
It;s not that hard, in fact that was one of the main things that I found straight away by looking at rate of change of OHC. as posted above: http://wattsupwiththat.com/2012/10/14/tisdale-wheres-the-anthropogenic-global-warming-signal-in-the-nodc-ocean-heat-content-data-0-700meters/#comment-1109491
NeedleFactory says: “Aren’t sloshings different from waves?”
It depends on your definition of sloshing, doesn’t it? Refer to Bill Kessler’s ENSO FAQ webpage:
http://faculty.washington.edu/kessler/occasionally-asked-questions.html
A Kelvin wave is the initial part of the sloshing he describes as part of ENSO.
“After all they say they couldn’t find any other means for the AGW theory so we must provide …”
What? The fact that they can’t find anything else (mainly because they’re not looking) is really such a stupid argument to put forward in a scientific context. It certainly does not put the onus anyone else to provide an alternative explanation.
” I haven’t found anything else” is not a scientific argument and does not need any rebuttal other than to point out how stupid it is.
Can you explain why circa 1950 is the starting point on these data sets? Why not 1900 or 1930? What would the trend lines show with these starting points?
Steven Mosher says: “Err Bob.
“The theory is not that GHG will warm the ocean continuously or homogeneously…”
Err Steven: Anthropogenic global warming is said to be “systematic and persistent”. Refer to paragraph 32 of Trenberth and Fasullo (2012):
http://www.cgd.ucar.edu/cas/Trenberth/trenberth.papers/TF_RHW_JGR_2012JD018020.pdf
I could also find a couple of quotes in AR4 if you’d prefer, but you could do the same.
Also, climate models suggest a relatively uniform warming among the ocean basins, inasmuch as the zonal-mean trends of satellite-era sea surface temperatures are basically the same for the ocean basins:
http://i56.tinypic.com/t4wpys.jpg
While in the real world they are not:
http://i53.tinypic.com/24zf4f9.jpg
Those two graphs are from the following post:
http://bobtisdale.wordpress.com/2011/04/10/part-1-%e2%80%93-satellite-era-sea-surface-temperature-versus-ipcc-hindcastprojections/
Bob says: And again, P. Solar, there is nothing wrong with my investigations or analyses. The fact that you want more simply indicates that YOU should carry the ball farther. You’d never be happy with what I’ve done anyway. Have fun.
I have carried the ball, done the analysis and posted a graph and detailed comments. Maybe you missed it 😕
Never happy, not really, I think I was quite complementary about your showing how ENSO could actually be a means of getting more solar energy into the system , rather than just being an internal cycle. I also pointed out some faults in your argument and suggested how you may improve it. You were as refractory as ever, but I did say it looked like an important finding.
You will notice I did not need to detrend the data to get this graph.
http://i48.tinypic.com/zx1d9k.png
The volcanic rebound is similar to what I suggested you may see in your ENSO work if you avoided the detrending.
Since the presumed volcanic cooling is essential to exaggerated CO2 warming I think the question of volcanic rebound is highly significant.
Without volcanic cooling, climate models would have to apply the true physically provable CO2 warming without jacking it up by a factor of 3. They may then start to behave a little more like the climate they are purported to model.
The modellers will probably have to be dragged though hell, backwards before they accept that but you never know.
[sourcecode]
BTW, for completness, here are the NLLS fitting results for the model shown in the graph
Final set of parameters Asymptotic Standard Error
======================= ==========================
x1 = 2003.33 +/- 0.8568 (0.04277%)
a1 = -0.0297402 +/- 0.004427 (14.89%)
p1 = 21.807 +/- 0.7037 (3.227%)
x2 = 2000.29 +/- 2.773 (0.1386%)
a2 = -0.0172149 +/- 0.04718 (274.1%)
p2 = 3.72418 +/- 0.03225 (0.866%)
a = 0.0010536 +/- 0.0001978 (18.77%)
c = 0.0411344 +/- 0.004228 (10.28%)
[/sourcecode]
P. Solar quoted Lightrain without attribution: “After all they say they couldn’t find any other means for the AGW theory so we must provide…”
And then in the same comment you quoted again without attribution, “I haven’t found anything else…” and in this instance that statement appears nowhere else on the thread. To whom are you attributing the quote, when no one wrote it?
Bob Tisdale on teleconnection
….
Yes, trend winds idea at first instant sounds plausible, but as far as I can see it 3 to 9 months delays make a questionable proposition.
also Roger Pielke Sr quotes example paper:
Impact of Atlantic sea surface temperature anomalies on the summer climate in the western North Pacific during 1997–1998
https://pielkeclimatesci.wordpress.com/2008/11/18/what-is-a-teleconnection-why-are-teleconnections-important-in-climate-science/
I have my doubts, but thanks anyway.
P. Solar says: “It;s not that hard, in fact that was one of the main things that I found straight away by looking at rate of change of OHC. as posted above:”
And as I noted in the comment upon which you are responding, it could be argued you’re seeing a response to ENSO. As you may not be aware, the tropical Pacific was under-sampled prior to the installation of the TAO project buoys, which were completed in the early 1990s, so one would not expect the OHC data to respond similarly to the 1997/98 and 1982/83 El Nino events.
If you like, why don’t you study the impact on the variability of the increasing samples with time? Also, while your studying that, you’ll need to consider that the Indian and West Pacific OHC data vary inversely with ENSO and that the variations in the East Pacific and Atlantic are related positively, which would affect their perceived impacts.
Paddy says: “Can you explain why circa 1950 is the starting point on these data sets? Why not 1900 or 1930? What would the trend lines show with these starting points?”
There’s too little source data at these depths prior to the 1950s. In fact, in the 1950s, the sampling at depth was pretty much limited to the North Pacific and the Atlantic.
Nah. Just because “they couldn’t find” anything else to explain warming, etc., etc., does not establish CO2 as the Null (default). They would have had to have PROVEN that there is no other “forcing variable” for that argument to have weight. They haven’t come close. The CO2-AGW cartoon hypothesis is simplistic compared to any real analysis, so the lazy may prefer it, of course.