Judith Curry, a professional climate scientist who writes the Climate Etc. blog, is apparently blown away by a recent paper in Nature. The paper is by Yu Kosaka and Shang-Ping Xie and is called Recent global-warming hiatus tied to equatorial Pacific surface cooling. The paper seems to be suggesting that the recent “hiatus” in surface warming is due to variations in the sea surface temperature of the central to eastern tropical Pacific. It seems that Judith Curry is blown away because the paper seems to be suggesting that natural variability could be partly responsible for global warming.
I don’t think that that is what this paper is actually suggesting. Judith is, however, a professional climate scientist so I’m more than willing to accept that I could be the one who is wrong. The paper is really only talking about how the sea surface temperature in the Pacific can modulate the rate at which the surface (sea surface, land and atmosphere) warms. Global warming is a consequence of rising atmospheric greenhouse gas concentrations that act to trap outgoing long-wavelength radiation and increase the amount of energy in the climate system. This energy is then distributed across the climate system, with the majority (90% or more) going into the oceans. How this energy is distributed across the climate system will, however, depend on the properties of the different parts of the system. It seems reasonable that there will be times when a larger fraction goes into the oceans and other times when this fraction reduces, increasing the fraction heating the surface and atmosphere.
All that this paper is showing, I think, is how the sea surface temperature in the Pacific can influence how the excess energy associated with global warming is distributed within the climate system. The paper actually ends with
We conclude that the recent cooling of the tropical Pacific and hence the current hiatus are probably due to natural internal variability rather than a forced response. If so, the hiatus is temporary, and global warming will return when the tropical Pacific swings back to a warm state. Similar hiatus events may occur in the future and are difficult to predict several years in advance owing to the limited predictability of tropical Pacific SST.
I think the above conclusion from the paper is probably consistent with how I’ve interpreted their work. Again, happy to be corrected by those who know more than me.
Having said the above, the paper doesn’t seem to do a particularly good job of distinguishing between the warming of the surface and overall global warming. I can see how some might be confused by what the paper is presenting. A comment by Bernard J over at HotWhopper explains this quite nicely, in my opinion. Bernard says (and I hope he and Sou don’t mind me quoting part of it here)
To be blunt I think that the paper suffers from poor editing.
The term “hiatus” as used implies a physical pause in the warming of the planet, but all it really is is a statistical variation in the overall surface temperature trajectory – one that results from the complex interaction of physical phenomena. There remains, however, a similarly physical top-of-atmosphere energy imbalance and hence heat content below is still increasing, irrespective of whether it is masked in the surface temperature record by the interaction of heat-transporting phenomena.
I, like Bernard, was also surprised by some of the terminology used in the paper. I would have expected people working in the field to realise that papers like the one being discussed here could be mis-interpreted by those who don’t realise that surface temperatures are not a particularly good indicator of global warming. I would have expected authors of papers like this to have been a little more careful in how they described the work so as to not introduce further confusion. I appreciate that there was a period (maybe 10 or more years ago) when surface temperatures were typically associated with global warming. This was, I believe, because the surface temperature record was the longest and most reliable record available. Now we have an extensive and reliable record of ocean heat content which shows that most of the excess energy associated with global warming goes into the oceans and that this indicates that global warming has continued despite the “hiatus” in surface warming. The ocean heat content is therefore not only more representative of global warming, but is also a better indicator of global warming than the surface temperature data.
Anyway, this seems to be a very interesting paper which is probably consistent with what many have expected (the hiatus in surface warming is simply a natural variation associated with ocean cycles, and not an indicator that global warming has stopped). I’m just a little surprised that the author, reviewers and editor didn’t put more effort into distinguishing between global warming and warming of the surface, given that many would probably (and some have) interpreted the paper as suggesting that natural variability can be playing a significant role in global warming. Just to be clear, I’m not suggesting that it doesn’t play a very important role in determining how the excess energy is distributed through the climate system. I’m simply pointing out that the reason we’re undergoing global warming is not because of natural variability. It’s because we keep adding CO2 to the atmosphere through our continued use of fossil fuels.
Wotts Up With That Blog 
Yup, you pretty much have it nailed. The oceans are warming due to the increased Greenhouse Effect. Sunlight, in the form of shortwave radiation, penetrates the surface ocean. This warms the surface which is typically warmer than the overlying atmosphere. The flow of heat is therefore from the (warmer) ocean to (cooler) atmosphere.
The ocean-atmosphere interface is a boundary. The cool-skin layer is the less than 1 mm thick uppermost layer of the ocean where turbulence breaks down and molecular forces dominate. Conduction is therefore the key determinant in the rate of heat loss from the ocean. Lower the thermal gradient through the layer and the rate of heat loss from the ocean decreases.
Increase the concentration of greenhouse gases in the atmosphere and more longwave (heat) radiation is directed toward the ocean surface. Longwave radiation cannot penetrate into the ocean, but it can and does warm the upper portion of the cool-skin layer and, therefore, lowers the thermal gradient through the layer. The end result is that the oceans grow warmer over time as more heat is trapped beneath the surface.
The eastern pacific cooling mentioned in the paper is a way of diagnosing of the mean state of the wind-driven ocean circulation. When the easterly trade winds intensify the system tends to spin-up. We observe strong poleward transport of surface water out of the tropics and strong convergence in the centre of the subtropical ocean gyres. This convergence pumps surface water downwards (Ekman pumping) leading to the deep ocean warming we have recently observed.
The surface flow of warm water out of the tropics and down into the centre of the subtropical gyres is balanced by a subsurface return flow back toward the equator, where it upwells. This bring cold water to the surface, as it is also pulling up cold water from deeper layers in the equatorial Pacific.
This is the cooling Pacific sea surface the authors are referring to. When the wind-driven ocean circulation is strong the upper layers of the eastern Pacific ocean in particular are anomalously cool. During such times there is anomalous export of warm surface water out of the tropics and also strong vertical pumping of surface water downwards in the subtropical oceans.
Over long time frames, and in a stable climate, the system would more or less balance out to zero. The deep oceans are warming because greenhouse gases are persistently trapping more heat in the surface ocean, and this is being transported downwards when the circulation is strong – as it now is. It is, however, fitfully winding down from a peak around 2003-2004. The current diminished rate of surface warming will not last.
Thanks Rob. Very interesting. What you mention are details about which I’m only starting to learn. Very useful to have it explained so clearly.
The authors of that Nature paper have written a good article for The Conversation about it – http://theconversation.com/warming-slowed-by-cooling-pacific-ocean-17534
They say:
Is Judith Curry blown away by their conclusion that the “planet-wide march towards higher temperatures will resume with vigor” or has she misunderstood their conclusion or perhaps I don’t understand why she is blown away?
In fairness, I realised that she actually said my mind has been blown which I interpreted as meaning blown away. Maybe there is an alternative explanation though 🙂
This is what blows Judith’s mind:
“Compare the temperature increase between 1975-1998 (main warming period in the latter part of the 20th century) for both POGA H and POGA C:
■POGA H: 0.68C (natural plus anthropogenic)
■POGA C: 0.4C (natural internal variability only)
I’m not sure how good my eyeball estimates are, and you can pick other start/end dates. But no matter what, I am coming up with natural internal variability associated accounting for significantly MORE than half of the observed warming.”
But 1975 has a big dip, and 1998 has a big and well-known spike, so they are obviously cherries.
If you instead pick 1965-2000, you get a cooling of 0.2 from natural internal variability only. With natural plus anthropogenic you get a warming of 0.6.
That natural variability plays a role in global warming is nothing new, like I have pointed out:
That’s me quoting the 2007 IPCC report. So results like this should not come as a surprise and are in line what we already know on this subject.
Shang-Ping Xie makes the same observation.
“1975 was a La Nina year, and 1998 followed the strongest El Nino in the instrumental record.
…
1998-2012 minus 1950-1964 temperature difference between the current and previous hiatus events: 0.66 C in POGA-H, and 0.08 C in POGA-C.”
But then Judith’s mind is easily blown from side to side anyway.
Like a feather…
What Judy fails to mention is that the ENSO index (or MEI for that matter) doesn’t provide ground truth when it comes to the estimation of longer-term trends. While there is no problem with the amplitude, the base line depends on the external forcing! For example, strong external (negative) aerosol forcing (a.k.a. global dimming) inevitably shifts the index towards more negative values (unless the forcing would explicitly be considered when computing the index – which it is not!). That’s what (very) likely happened between 1950-1980. Global brightening in the 1990s reversed this trend with the index to be more positive.
However, one thing that does make a difference is the number of La Nina and El Nino events in a certain period of time. Clearly, the 1970s were strongly dominated by La Nina, which – undeniably – amplified the sulfate aerosol cooling. In contrast, frequent El Nino events occurred in the 1980s and 1990s, amplifying the warming due to the reduced sulfate aerosol load (brightening period). The unresolved scientific question is whether the external forcing can trigger a preferred ENSO state. There is more than anecdotal evidence that this is the case for strong (tropical) volcanic eruptions (El Nino in response), but the response is thought to act in a counterbalancing fashion, such that it actually buffers the external forcing. Strangely, this hasn’t been the case for aerosols. It appears less strange if one takes its pure extratropical NH location of this particular forcing into account, which led to strong interhemispheric temperature differences. And yes, the NH temperature trend and the sulfate aerosol loading is correlated. Strongly so! It is no coincidence! But please, don’t tell Judy that aerosols even exist! Make also sure, that all those observations indicating decreased incoming SW radiation (later reversed) are non-existent.
While there might also be a trend issue with regard to ENSO index for the recent decade, it is small when remaining natural effects (namely volcanoes and solar variations) are accounted for (see Foster and Rahmstorf 2011). Moreover, the fact that we have seen more La Nina events in recent years imposes a cooling trend – undeniably so!
Plus, isn’t this simply surface heating/cooling which really doesn’t tell us whether or not global warming is actually happening. So, it’s fine to talk about a natural contribution to surface warming, but it still doesn’t imply a natural contribution to overall global warming.
So, in a sense, the paper seems to be saying that ocean cycles can modulate surface heating even while global warming continues, which – as far as I can tell – is kind of what people have been suggesting for quite some time now.
Lars has captured why Curry’s mind has been blown. I’ve got a post on this paper next week that briefly discusses her blown mind, so perhaps I can clarify as well.
In short, Curry is doubly cherry picking. First, she’s focusing exclusively on surface temps, as you and several others have noted. In its Extended Data section, the paper also shows that the POGA model does not simulate any slowing in ocean heat accumulation. Nor has any been observed, of course.
However, Curry has argued that natural variability may be responsible for most of the observed *surface* warming. She (wrongly) believes this paper confirms her arguments. She believes this because of a second cherry pick – only looking at data from 1975 to 1998, during the ‘warm’ Pacific cycle. Her comment on this cherry pick is funny:
“I’m not sure how good my eyeball estimates are, and you can pick other start/end dates. But no matter what, I am coming up with natural internal variability associated accounting for significantly MORE than half of the observed warming.”
No matter what? Really??
Actually if you look at the full record (1950 to 2012), or even 1970 to 2012 (the authors comment that the Pacific SST data is most reliable since 1970), in either case natural variability accounts for 20% of the surface warming trend. For 1975 to 1998 it is indeed about 50% (note that Curry screwed up the math when eyeballing the chart as well, when she put it at 60%). In fact you can pick virtually any other start/end dates, as Curry suggested, and you’ll find natural variability contributing to LESS than half of the observed warming.
This is precisely why scientists don’t rely on their senses. It’s too easy to see what you want to see. Curry wanted to see natural variability accounting for more than half the observed surface warming, so she (perhaps subconsciously) cherry picked the data to see what she wanted to see. As a psychological exercise it’s a pretty interesting case study.
Thanks, Dana. Very interesting. Presumably this whole natural versus anthropogenic heating of the surface is a little bit of a red herring anyway. As I think the authors of this paper have themselves commented, The IPCC conclusion applies to centennial warming from 1880. Much of the 0.8 C warming since 1900 is indeed due to anthropogenic forcing, because natural variability like PDO and AMO has been averaged out over this long period of time. So, if the general picture that is building is correct (as I suspect many think is the case) then there will be short periods where one could divide the surface warming into a natural and an anthropogenic component, but over a sufficiently long time interval, the natural component averages out, leaving only the long-term anthropogenic trend.
Well, that’s certainly my understanding. But believe me, Wotts, trying to explain this to “sceptics” is invariably a recipe for frustration.
Dana, even if you cherry pick the intervals as Judith did, the numbers in the POGA C experiment (see Lars) are still NOT attributable to natural variability alone. As I was trying to point out in the previous comment (probably not very clearly stated), the ENSO index (or the tropical Pacific SSTs for that matter) is (are) very likely contaminated by anthropogenic sulfate aerosol forcing! In forcing the model to follow the observed evolution of the equatorial (eastern) Pacific SST anomalies (POGA H/C), one passes the anthropogenic aerosol footprint on to the model! We don’t know the magnitude of this effect, but it is almost certainly not negligible. In order to make sure I expressed myself correctly this time: The +0.4C between 1975-1998 in POGA C are thus not 100% due to internal variability!
Otherwise, of course I agree entirely with your comment.
John N-G also called out Judith’s misunderstanding:
http://blog.chron.com/climateabyss/2013/08/learning-from-the-hiatus/
“Curry’s mind-blowing reading of the paper is incorrect. What she missed is that POGA-C is not “natural internal variability only”. It’s “natural plus forced in the El Niño region of the Pacific”. There’s no “natural variability only” run to compare to. The other two runs are “forced everywhere” (HIST) and “anthropogenic everywhere plus natural variability in the El Niño region of the Pacific” (POGA-H).”
Thanks Marco! Terrific posting from John! Hadn’t seen it before. He took it not two, but at least three steps further and provides some “hard” numbers. On top of that, he has nicely been trying to pull out a TCR estimate with an amazingly plausible result. I still don’t get my head around the fact that a honest atmosperic scientist of Judith Currys stature could miss such important bits! It’s sad to watch her 😦
John N-G is wrong about a few of things – his characterization of this whole topic for starters, but perhaps the most egregious is the notion that natural variability did not play a part in the rapid warming of global surface temperatures between 1975-2002.
Between 1976 & 1999 the Interdecadal Pacific Oscillation (IPO) was in a positive phase – meaning that the poleward transport of surface water out of the tropics was weak. As part of this whole spin down of the tropical-subtropical cell, the upwelling of cold water (driven by the now weakening easterly trade winds) shuts down. The sum effect is strong warming of the surface ocean in central and eastern tropical Pacific – the very region affected by upwelling (Ekman suction) because of the role of the meridional overturning cells and the tilting of the Pacific equatorial thermocline.
There are simply scores of papers on this topic dating back to the 1980’s at least, but two very useful papers for readers are:
1. McPhadden & Zhang (2002) – Slowdown of the meridional overturning circulation in the upper Pacific Ocean.
2. McPadden & Zhang (2004) – Pacific Ocean circulation rebounds.
So yes, natural variability, and perhaps global brightening (a reduction in industrial sulfate pollution), did play a part in the rapid surface warming during the late 1970’s to late 1990’s.
Rob, after having read John N-Gs article again more careful, there might indeed be an issue with the methodology (or rather, the way he has done it), given that a negative 1975-2002 natural contribution seems unlikely. I think that simply subtracting HIST from POGA-H does not tell you much with regard to the ratio of natural/anthropogenic contribution, since HIST does not know about the actual ENSO timing. But I’m not quite sure whether I am on the right track here, as I haven’t reproduced the results step by step. I might well be wrong …
Regardless, both papers don’t provide a proper explanation for why this regime shift has happened. Personally, I would find it extremely remarkable if strong interhemispheric forcing differences (such as imposed by anthropogenic aerosols) do not have an effect on trade winds and hence the timing of these regime shifts (if they had occurred at all). On the other hand, it does not matter what initially caused the shift as it ultimately leads to additional (natural) warming/cooling at the timescale of one or two decades, although still superimposed by external forcing which complicates the determination of its exact warming/cooling share.
Karsten – As far as I’m aware no one knows why the wind-driven ocean circulation bounces back-and-forth between sluggish and intense phases, although there are two leading hypotheses. The links I provided were simply for the readers benefit – showing that the circulation wound down and then spun up.
On sulfate aerosols, the changes in the phase of the Interdecadal Pacific Oscillation through the 20th century show a rather coincidental timing with global brightening and dimming trends. It is rather interesting…..
Pingback: Watt about doing revolutionary science? | Wotts Up With That Blog
“there will be short periods where one could divide the surface warming into a natural and an anthropogenic component, but over a sufficiently long time interval, the natural component averages out, leaving only the long-term anthropogenic trend.”
Exactly, and Curry has cherry picked one of those short periods to try and justify her desired conclusion.
Pingback: Why isn’t it getting colder? | quakerattled
Pingback: Watt about a science-based rebuttal to global warming alarmism? | Wotts Up With That Blog