What of “The Pause”?

I have previously shown how global temperatures rose in three distinct and abrupt steps from the 70s to the 00s – one in 1979, one in 1988 and one in 1998 – and at all other times, not at all. These three steps occurred relative to the SSTa curve of the NINO3.4 region in the equatorial zone of the central-eastern part of the Pacific Ocean. Before, between and after the three steps, global temperatures appear simply obediently to follow NINO3.4 without any sign of a continued slow, but steady upward drawing away as if from a ‘steady rising background forcing’:

Warming steps

Figure 1.

My opinion on the much talked about “Pause” or “Hiatus” in ‘global warming’ still said to be going on (the considerable final, level stretch of the upper blue curve in Figure 1), is thus naturally coloured by this understanding of how global temperatures normally progress through time, as exemplified by the period from 1970 till today.

Within this perspective, the “Pause” is but one of many temperature ‘plateaus’ between sudden steps up or down (the last time it went down was back in 1964, before the ‘modern warming’). The relevant questions are: When did the last step occur? When will the next one take place? And will it go up? Or down?

At the present time, I would still maintain that the last well-established step in global temperatures happened in 1998, following directly in the wake of the mighty 1997/98 El Niño. Simply because not enough time has elapsed to be able to say anything for certain about more recent events.

But there are definitely a couple of things at work today that deserve some close attention.

Firstly, I have already in earlier posts pointed out the direct equivalence between the events playing out both in the tropical East Pacific and globally as a response during and after the El Niño of 2009/10 and those following the El Niños of 1997/98 and 1987/88, 12 and 22 years prior. These three El Niños are simply events of equal standing (together with a fourth one – the 1972/73 event) in a particular ENSO sequence having repeated itself nearly four times now since 1970. They are the only ‘solitary’ El Niños since 1970, meaning, the only distinctive warm ENSO events both preceded and succeeded by major (and mostly multi-year) La Niñas (distinctive cool ENSO events).

And this is significant. This particular succession makes a difference.

You need only look at the temperature timeseries to see that something major happens globally, not just after the 1987/88 and 1997/98 El Niños, but also (the exact same thing, in fact) after the 2009/10 El Niño. Global temps do not cool adequately during the succeeding La Niña. And in both 1988 and 1998 this led to an upward shift in mean global temperature level, one step up from the NINO3.4. It would seem then, the mechanism being the same, that history should repeat itself post 2010.

Well, the course of events does start off in familiar fashion – the global cooling response to the 2010/11 La Niña, directly on the heels of the great El Niño, seems to be highly opposed by something, significantly reducing the total effect, exactly like in 1988 and 1998, but very much unlike what happened during the La Niñas of 1999-2001 and 2007-09 (Figure 2 below). Already in mid-2011, however, at the entrance to the second consecutive Niña season, the global curve falls back in line, once again latching on to the safe lead of the NINO3.4. So this time around, a new and higher mean level is never established post the original suppressed La Niña cooling. And there ends the equivalence to the global events of 1988> and 1998>.

But why this change? What’s different?

Who’s to tell? It has been suggested that a weaker solar cycle could indirectly be to blame. I will not immediately give my assent to such a proposal, because I have not seen any direct evidence backing it up (it’s not enough just pointing out that the current cycle is weaker than the previous ones, after all). But it’s an interesting idea …

The story doesn’t end in 2011, though.

All through 2012 and halfway into 2013, the global and NINO curves can be observed to be firmly back on track. And so, there is no hint of any gradual ‘background trend’ (‘global warming’) deviation from 1998/99 and all the way down to this point. The two curves are as tight-hugging in 2011-13 as they were in 1999-2002. Global temps are no doubt kept on the shortest of leashes by NINO3.4.

But then we reach May 2013 (light green vertical line in Figure 2):

Gl vs. NINO 97-14

Figure 2.

And all of a sudden something new and big is evidently starting to happen. The global SSTa curve soars up, high above the NINO curve. This highly unusual development has now kept up for more than a year. But what’s actually going on?

Here’s what’s going on:


Animation 1. The global SST anomalies evolving from January 2012 to September 2014 according to satellite-based NOAA (Reynolds) OI.v2 (anomaly baseline: 1997-2013). Note the first appearance of ‘The Red Blob’ in the northeastern Pacific basin in mid-2013 and how things, after a short intermission, completely take off during 2014.

The same period graphed:

2012-14 flight

Figure 3. What you will notice here (and which is confirmed in the animation above) is that the 2013 peak in global temps, although quite a bit lower than the 2014 peak, comes at a time when the equatorial part of the East Pacific (NINO) is actually quite cool, while the 2014 peak correlates to a warming NINO region.

If you wonder – even after having watched Animation 1 above – whether or not this current global peak in SSTa, high above the corresponding NINO3.4 curve, might in fact be an actual ‘global’ phenomenon, I’ll advice you to take a look at this map:

Jul-Sep 14 SSTaFigure 4.

Here is depicted the global mean sea surface temperature anomaly for the most recent three-month period (July-September 2014) to showcase the location of the current ‘extra-ordinary heat’. This peak is clearly not ‘global’ in scope. And that’s why this new development is so intriguing. How will it all pan out …?

A slight digression at this point: The recent ‘extra-NINO’ heat of the global ocean is definitely there, but is the NOAA (Reynolds) OI.v2 SSTa dataset overdoing this particular ‘global warming’ peak somewhat?

There are other global SST datasets out there. Two of these seem to correlate especially well with the OI.v2. They are the HadISST1 and the ERSST.v3b. A third one, that could have been included, is the famous HadSST3 dataset, but this does not correlate well with any of these three over the period in question. This is something that I might come back to in a later post.

Anyway, here is global OI.v2 vs. HadISST1 and ERSST.v3b from Jan’97 to Sep’14:

OI.v2 vs. ERSST.v3b & HadISST1Figure 5.

There are sections of this graph where the three curves do not fully agree – note for instance the stretch between the end of 2004 and mid-2006. But that particular discrepancy appears to be a fairly random one; each curve follows a course of its own. What happens after 2012, though, going into 2013 (marked by the light green vertical line), seems different. While the HadISST and ERSST curves continue to track each other ever so closely, the OI.v2 curve all of a sudden shifts up relative to the other two by about 0.025-0.03 degrees … and stays there. I have no way of telling the reason for why this happens, but it clearly does. And I also cannot say which one of the three datasets displayed here will in the end prove to be the one giving the ‘correct’ rendition of the real situation. I’m only observing what’s going on between them. And I feel it’s worth addressing …

Now back to the main discussion.

The recent abnormal ‘global’ heat is actually not global at all. It is very much confined to the North-Pacific, mostly to the northeastern part. We don’t see it in any other subset of the global ocean. We don’t see it on the global land masses, where people actually live. We only see it in the North Pacific.

As it turns out, it’s also more or less confined to the surface, even in the North Pacific itself. Have a look at this:

NP sfc vs. tropoFigure 6.

What you see here is the anomalies at the surface compared to the corresponding ones in the lower troposphere just above. (I have deliberately limited the extent of the region examined simply because widening it too much would cause substantial portions of continental land masses (in Asia & North America) to be included in the tropospheric signal, which would take away from the overall significance of the correlation.)

It’s hard to find a very direct connection here between surface processes and tropospheric response, like you do in the tropics. But there is a general correspondence to be seen. Note especially how there is no trend to speak of from 1997 to 2013. And then, out of nothing it seems, the surface temperature anomaly simply shoots up in two (or three) major spikes, the final one even considerably higher than the first one. It’s indeed a strange development.

And what about the troposphere resting on top of this sector of ocean surface? Well, there is a response, isn’t there? But it seems a bit reluctant and definitely nowhere near proportional to the surface signal.

So what’s going on?

The answer is probably to be found in the fact that the coupling between surface and troposphere is very much convective in nature. Winds, solar surface heating and evaporation help bring surface energy up into the troposphere by way of bulk air movement. In the tropics, this coupling is generally incredibly strong and tight, thanks to the fluctuating, but still relatively constant band of deep moist convection circling the entire globe. Outside the tropics, like in extratropical North Pacific, the coupling is much less predictable. It’s there, but it’s clearly weaker, susceptible as it is to the whims of the capricious Ferrel Cell weather systems. Convection/evaporation is simply not as strong and persistent outside the tropical zone. So changing pressure systems, wind patterns, jet streams and cloud cover all affect how readily the signal of surface processes is transferred up into the air column above.

The northeastern part of the North Pacific during the past year, on the face of it bears the signature of a stagnant or locked weather situation where the winds of the region are not adequately able to couple the surface and the atmosphere above, a situation maybe (most likely) manifesting itself in what has been dubbed the ‘Triple R’ (‘Ridiculously Resilient Ridge’) of 2013-14.

Well, anyway – this lack of an adequate tropospheric response to the recent surface goings-on in the North Pacific goes a long way to explain why global lower troposphere temperature anomalies do not appear to follow global surface anomalies up, why they rather appear still to be tracking NINO3.4 to this day, when the surface hasn’t since mid-2013.

Let’s verify that.

First of all, a definite disparity between the two major providers of lower troposphere temperature data – UAH and RSS – has been thoroughly noted and discussed. And it is easy enough to confirm:

UAH-RSS trend

Figure 7.

What is probably not as well-known, is that this striking disagreement arises from one single departure alone. It all happens in the latter half of 2005, an artificial step from one month to the next. Amend it, and the fit between the two timeseries suddenly becomes near-perfect:


Figure 8.

The Aug-Sep ’05 step separated the two timeseries by 0.1 degree in one go. As you will observe from Figure 8, there is no gradual drift in either of the two curves relative to the other before or after this particular shift. The entire discrepancy results from this one irregularity.

In Figure 8 I have moved the RSS curve 0.05 degrees up and the UAH curve 0.05 degrees down post Sep’05. This move might seem arbitrarily ‘egalitarian’, but it does in fact hold some merit. UAH does tend to run slightly too warm post 2006, while RSS similarly tend to run slightly too cool:

UAH just vs. hav gl

Figure 9.

UAH lower troposphere, after downadjustment, measured against the average of the three global sea surface temperature datasets referred to above. A pretty tight match all the way from 1998 to the end of 2013. Next up, RSS:

RSS just vs. hav gl

Figure 10.

Also after adjustment, only this time up. We see a similar good fit, but here the final drifting apart takes place already in the latter half of 2013, like with NINO3.4.

There is definite sense in finding that the evolution in global lower troposphere temp anomalies should be and is constrained like this by the global ocean surface, but even more sensible to suggest that, moreover, it would favour the tropical part of it. Why? Because of the all-important convective connection. The troposphere is, after all, by and large heated by the transfer of energy from the surface in the form of so-called ‘latent heat of vaporisation’. This process takes place mainly through the mentioned moist deep convection in the tropical oceans:

Rain vs. heating

Figure 11. From the JRA-25 Atlas.

Since, at this point, it’s hard to choose one satellite series over the other after they have both been adjusted to fit neatly with each other and with the global ocean, the best thing to do would probably be to just take their average and present that as the ‘Satellite Mean’:


Figure 12. (Note, the black horizontal line is NOT a computed linear trend line. The linear trend would go up across this period. I simply drew it as a visual tool helping to verify the extent of “The Pause”.)

Watch now as we directly superimpose the NINO3.4 curve from Figure 2 on this one:

Satellitt gl vs. NINO3.4

Figure 13.

Quite an astounding fit (from 1998/99 on), I dare say!

And by this, we have verified what we set out to verify: The global lower troposphere continues to slavishly follow NINO3.4 to this day, even as the global surface does not.

In conlusion:

What of “The Pause”? Well, it’s not a ‘pause’. It’s just an expression of the natural progression of things.

On the other hand, it’s foolish to talk about ’18 years without warming’. There was massive warming in 1998 – that’s how we came to be at the elevated plateau where we are today in the first place. There is no point (if you know the processes behind) going further back in time than 1998/99 to claim ‘a halt in warming’. After this point, global surface temperatures once more stabilised, returning back to the habit of simply following NINO3.4 (just like they always do outside the actual steps). And they continued doing so, dutifully and loyally, up until mid-2013. For the last extended year they’ve been tracking higher. But this is all centred in the North Pacific. We don’t see it anywhere else in the global ocean, we don’t see it on land and we don’t see it in the troposphere.

So how long will this situation last? The North Pacific sure doesn’t seem to be in balance with the rest of the world as it is now. Could this stark segregation of heat really just go on like this and sustain a new elevated ‘global’ temperature level all by itself?

I feel that at some point, something’s gotta give …

6 comments on “What of “The Pause”?

  1. David A says:

    Very good and interesting discussion. A couple of thoughts. You have established that the mean of the RSS and UAH GAT, follow the Nino 3.4 region very well, and you have maybe fairly well established the overall divergence of the satellites from the SST due to the RRR.

    Have you confirmed this by comparing the satellite mean, minus the RRR with the SST, minus the RRR? Another way of looking at this is to see if the surface GAT, minus the area of the RRR, matches the satellite mean, minus the RRR. I do not know if it is possible to do this.

    Currently the 2014 satellite mean anomaly is about 1/2 of the 1998 satellite anomaly, while the surface GAT is near the all time high. The pacific RRR should fairly rapidly transfer heat to the troposphere; I would think through conduction and convection and increased evaporation, just not at the same rate as the tropics, but especially after over a year of relatively stable high isolation in that area.

    Have you looked at the area maps of the anomalies, both surface and pacific RRR, to confirm if that is where the separation from the record takes place? Oh, I remember where I saw this..http://stevengoddard.wordpress.com/2014/09/18/out-of-africa/

    Now here I see the RRR well reflected in RSS anomaly map. The differences I see are in RSS cool central Africa, and Gavin spreading anomalous warmth to central Africa where he has no data, and in very warm Gavin surface anomalies in Antarctica, even in areas of record sea ice. This potentially leads me to give credence to some of Goddard’s articles pointing out the adjustments to the surface record as not matching the historic highs, the raw data, the satellites, and the increased sea ice and snow cover in both hemispheres, whereas RSS to raw USHCN is a very good match.


    • okulaer says:

      Hi, David.

      No, I haven’t gone into such detail as you suggest here. But feel free to try it out. Removing the northeastern part of the North Pacific from the global mean (after area-weighting) should prove to bring the global much more in line with NINO3.4.

      Thanks for the links 🙂

  2. David A says:

    I think I failed to enter my info, but it said it went to moderation?

    • okulaer says:

      Sorry. All comments go into automatic moderation. And I tend to be a bit slow at times … But they will show up eventually 😉

      • David A says:

        No Problem . Did you look at the out of Africa Goddard post, showing the anomaly map for RSS and for GISS? To my eyes it appears that RSS reflect GISS fairly well, but not in central Africa and those very hot GISS anomalies in Antarctica. Your thoughts?

  3. okulaer says:


    I did look at the Goddard post. It simply goes to show that Schmidt should stick to the 250km smoothing and discard the whole 1200km product. It’s a disgrace. In Antarctica and the Arctic there are many single data points (like one station surrounded by hundreds of miles of wilderness) whose anomalies are being smeared across vast tracts of land (AND, significantly, wide stretches of ocean). Those land areas don’t really have surface data (they could, however, anomalywise be compared with the satellites, like Goddard did for Africa), but for the oceans, there IS data available, GISS simply choose to ignore it. And get warming in return. Have you read this (old) post by Bob Tisdale? Quite an eye-opener:


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 )

Connecting to %s