Bob Tisdale goes AMO-ing to a big chill - not!

Update: The University of Southampton has issued a clarification. It doesn't address the substantive issues raised here. There's also no mention of 0.5C.

Also see Gerard McCarthy's replies to my query to him via Twitter - here and here and here.

Sou 2 June 2015

---

There was a new paper that came out this week in Nature, which had a bit of coverage around the traps. Deniers rather liked it not so much because of what was in the paper itself, but because of what was in the press release.

The paper was by a team from the University of Southampton, led by Dr. Gerard D. McCarthy. The paper was about the Atlantic Multidecadal Oscillation or AMO and ocean heat transport, and related.

The AMO is used to indicate changes in sea surface temperature in the North Atlantic. It is of such long duration that there aren't instrumental records going back far enough to show multiple repeating patterns. It's probably not a regular cycle with a fixed period. Estimates seem to place the latest period at around 70 years (from the beginning of a warm phase in the 1920s to the end of a cool phase in the 1990s - see here). The latest IPCC report had this to say about it (page TS-25):

A number of studies have investigated the effects of the Atlantic Multidecadal Oscillation (AMO) on global mean surface temperature. While some studies find a significant role for the AMO in driving multi-decadal variability in GMST, the AMO exhibited little trend over the period 1951-2010 on which these assessments are based, and the AMO is assessed with high confidence to have made little contribution to the GMST trend between 1951 and 2010 (considerably less than 0.1°C). {2.4, 9.8.1, 10.3; FAQ 9.1}.

From the above paragraph I take it that the nature of the AMO is not all that well agreed upon - by some at least.

A new climatic phase? Perhaps, but not as this paper predicts.

But first, what did the press release say that got deniers all a twitter? Well, it's this - from the press release from the University of Southampton:

A new study, by scientists from the University of Southampton and National Oceanography Centre (NOC), implies that the global climate is on the verge of broad-scale change that could last for a number of decades.

The change to the new set of climatic conditions is associated with a cooling of the Atlantic, and is likely to bring drier summers in Britain and Ireland, accelerated sea-level rise along the northeast coast of the United States, and drought in the developing countries of the Sahel region. Since this new climatic phase could be half a degree cooler, it may well offer a brief reprise from the rise of global temperatures, as well as resulting in fewer hurricanes hitting the United States. 

I looked at the paper to see where and why the authors thought that there would be such a "broad-scale change" that could be "half a degree cooler". I couldn't find it. Not anywhere was there a mention of any "half a degree cooler". Let's see what that would look like - using GISTemp:

Data source: GISS NASA

And that's just because of an oscillating AMO? How likely is that? I'd say not at all likely. For one thing - what about the Pacific?  Though there's more to question than just the Pacific. There's the measure of the AMO itself.

So is the press release way wrong?

About the paper

The paper is about trying to prove a link between the AMO and ocean circulation. In fact in the press release the words are:

The study, published in Nature, proves that ocean circulation is the link between weather and decadal scale climatic change.

It's not often you'll see such a strong claim in any paper. Usually the strongest is something along the lines "the study supports..." or even "this research provides strong support for..." or more commonly "these findings suggest....". Prove is a very strong word. My antenna is quivering.

It seems the researchers used sea level changes on the east coast of the USA as a proxy for ocean circulation changes. The wrote that north of Cape Hatteras sea level fluctuations are linked to the fluctuations in the overturning circulation. South of Cape Hatteras down to Florida, it's changes in the Gulf Stream. So they've split their measurements at Cape Hatteras and developed what they called a circulation index. They wrote:

This index projects onto observed surface velocities during the satellite era in the intergyre region, with a positive index associated with more northwards flow and a more northerly path of this circulation (Extended Data Fig. 4). Similarly, in a high-resolution ocean model, over timescales that contain both the cool phase of the AMO in the 1970s25 and the warm phase of the 1990s26, the sea-level index projects onto a similar pattern of circulation, with a positive index associated with more northward heat transport (Extended Data Fig. 5). 

From there the authors go on to discuss heat transport, and say that although they didn't have direct observations, "we can relate our sea-level index directly to the heat content changes in the subpolar gyre since 1960".  After going through their reasoning, they conclude that the North Atlantic Oscillation "forces the ocean circulation and consequently the ocean heat transport into the subpolar gyre".

Where does the AMO and the supposed half a degree drop in global surface temperatures come into the picture? The former seems to have been calculated, like the NAO and their circulation index - by linear detrending.

How to estimate the Atlantic Multidecadal Oscillation (AMO)

You might recall a paper discussed here in the past by Byron Steinman et al (2015). In it they wrote about defining the AMO using a "target regression approach". As stated in that paper:

Prior methods used to define these internal variability components and their influence on Northern Hemisphere temperature include

• a simple linear detrending of the mean North Atlantic SST time series (17–21), 
• estimating the forced trend based on regression of North Atlantic SST against global mean SST and removing the forced trend to yield an estimate of the internal variability (16, 22, 23), and 
• defining the forced component as the mean of North Atlantic SST in an ensemble of climate model simulations and defining the internal variability component as the difference between the observed SST series and the multimodel mean (24, 25). 

These methods, as shown below, do not in general yield correct results. We estimated the Atlantic and Pacific-basin multidecadal internal variability components and their contribution to Northern Hemisphere temperature change on the basis of a new target region regression approach.

They argue that linear detrending does not properly partition the internal variability from the forced variability. As I understand it, linear detrending leaves some of the forced variability in the AMO, making it appear larger than it actually is. That means that they would be attributing to internal variability, some of the variability that should properly be attributed to, for example, aerosol cooling and/or greenhouse warming and/or volcanic cooling.

The same authors published a paper that argued along the same lines in GRL in May last year. That paper referred to the stadium wave hypothesis of Marcia Wyatt (that Judith Curry touted - and co-authored a related paper). In the abstract, Mann et al argue (my emphasis):

Using the synthetic temperature histories, we also show that certain procedures used in past studies to estimate internal variability, and in particular, an internal multidecadal oscillation termed the “Atlantic Multidecadal Oscillation” or “AMO”, fail to isolate the true internal variability when it is a priori known. Such procedures yield an AMO signal with an inflated amplitude and biased phase, attributing some of the recent NH mean temperature rise to the AMO. The true AMO signal, instead, appears likely to have been in a cooling phase in recent decades, offsetting some of the anthropogenic warming. 

The GRL paper explains that the researchers compared model-based estimates of forced temperature changes with observed NH mean temperatures over the historical era. In that way they were able to determine and isolate the internal variability component of the NH mean temperature. By contrast, a linear detrend of sea surface temperatures would leave a lot of the forced component in the AMO.

Nor were these authors the first to suggest this. Back in 2006, Kevin Trenberth and Dennis Shea wrote a paper about the Atlantic hurricane season of 2005. In the abstract of that paper they wrote:

However, previous AMO indices are conflated with linear trends and a revised AMO index accounts for between 0 and 0.1 C of the 2005 SST anomaly.

And Michael Mann and Kerry Emmanuel wrote about the same thing in EOS in 2006:

Defining  the  AMO as  the  residual  pattern  after  linear  detrending  assumes that  any internal  oscillation  is superimposed  on  a  linear  background  trend. Yet  neither  observed nor  modeled  twentieth-century  global  mean temperature  series  exhibit  a  linear  trend. It is more  plausible  [see  Trenberth and  Shea,  2006]  to  assume  that  North  Atlantic SST trends  result  from  a combination  of  back­ground  large-scale warming,  believed  to  be largely radiatively  forced [e.g., Crowley, 2000], and an internal AMO signal that projects onto regional SST trends.

So this idea that linear detrending is not the way to go isn't new. It's been around for at least nine years, and possibly longer.

What about the purported drop of 0.5°C?

The other thing is that nowhere in the McCarthy paper that I could find, is there any suggestion that there will be a drop of 0.5°C in global surface temperature. If the Atlantic didn't warm as much or even if it cooled, it wouldn't mean that the earth as a whole will cool. There's the great big Pacific ocean out there and it's getting mighty hot right now - especially around the equator - but not only there. Here's a recent snapshot of the sea surface temperature anomaly in the Pacific.

Source: NOAA

It seems to me that this drop of 0.5°C was because they reckon that the AMO is on its way down and is going to drop the sea surface temperature in the North Atlantic by 0.5°C. However, as discussed, other research shows that linear detrending is not measuring internal variability. Therefore even were the AMO about to drop from a warm phase to a cool phase, that would probably at most lower the mean temperature in that region only, by about 0.1°C or a bit more. And would barely register on the global mean surface temperature, particularly if the Pacific continues to get hotter. (Note that by other measures, a drop from a warm phase to a cool phase would mean that the AMO would have had an fairly short warm phase - around 20 years.)

Steinman15 supports that notion. That paper concluded with:

We find that internal multidecadal variability in Northern Hemisphere temperatures (the NMO), rather than having contributed to recent warming, likely offset anthropogenic warming over the past decade. This natural cooling trend appears to reflect a combination of a relatively flat, modestly positive AMO and a sharply negative-trending PMO. Given the pattern of past historical variation, this trend will likely reverse with internal variability instead, adding to anthropogenic warming in the coming decades.

The implication being that when the PMO (Pacific Multidecadal Oscillation) turns positive, then warming will increase quite noticeably.

So I don't put a great deal of stock in this new paper by McCarthy et al. I do take the press release with a pinch of salt.

Deniers are all over this paper

At WUWT, Bob Tisdale is touting this paper as being the ants pants, even though it relies to some extent on models. He wrote:

As could be expected, the alarmist mainstream media have so far chosen to ignore a paper that discusses an upcoming multidecadal natural suppression of global warming…probably because indicates the slowdown in global surface warming should continue and it implies the natural variability of the North Atlantic contributed to the global warming we have seen since the mid-1970s.

Bob posted the two charts below, showing how the AMO would look if it were merely calculated by linear detrending of the sea surface temperature of the north Atlantic:

Source: WUWT

That WUWT chart shows the problem with linear detrending. It assumes that if you just take away the trend what's left is internal variability. But that's a huge assumption. It's much more likely that a lot of the external forcing has been left in the AMO chart (the bottom one).

What Steinman15 did was much more nuanced. Those researchers took away the estimated forced component and what was left was the internal variability. The internal variability using this latter method was considerably less than the simple linear detrended approach.

Bob also suspected what "alarmist" would do. He wrote:

HEADING THE CO2-OBSESSED OFF AT THE PASS

I suspect the true blue believers in catastrophic human-induced global warming will attempt to downplay the role of the AMO by citing the curious paper Steinman et al. (2015), which clearly illustrated model failings, even though they were attempting (and failing) to make other points.

Bob is a greenhouse effect denier. He used the Steinman paper as an excuse to complain that climate models don't model weather in the medium to longer term. Of course they don't - they are climate models not weather forecasts. Leaving that aside, Bob doesn't complain about McCarthy15 using climate models. What Bob doesn't know is that other scientists have pointed out that linear detrending is not the right way to determine internal variability such as the AMO.


Bob might have tried to head the CO2-obsessed off at the pass, but Michael Mann got to the pass before him. He tweeted this a few hours ago:

Note: Articles defining "AMO" via linear detrending of SST data are fatally flawed. see: http://t.co/mPAQOahtIl and: http://t.co/HX6uj24Wpy

— Michael E. Mann (@MichaelEMann) May 28, 2015

From the WUWT comments

Bloke down the pub

May 29, 2015 at 3:46 am
So my prediction is that when it turns negative and the rains in the Sahel dry up, cagw will get the blame. Don’t suppose I’d get any takers in a bet against that happening.

While this is the hottest ever start to a year after the hottest year on record, RH writes of what he thinks is the ice age about to cometh, any day now:

May 29, 2015 at 4:01 am
Within a couple of years we’ll be treated to stories of “unprecedented” cold, “historic” snowfalls, and “extreme” frost causing “catastrophic” crop losses and famine. Of course it will be attributed to human-released carbon, and the solution will be a continued goose-step to Marxism. 

While all RexAlan can think about is money - and that all the science of the last two centuries must be wrong:

May 29, 2015 at 5:20 am
But wouldn’t that mean that CO2 is not a greenhouse gas after all and they were wrong.
Silly me what am I thinking. How could they be wrong, they know everything…just send more money!

Greg wants to blame global warming on internal variability. He doesn't explain how the earth could suddenly get so hot without any forcing.

May 29, 2015 at 5:12 am
If anyone read the whole paper, did they ever even mention the herd of elephants in the room, that the AMO being in a warm phase previously is responsible for most of the prior warming?

Resourceguy used the WUWT notice board to write a random thought that popped into his head that had nothing at all to do with the topic at hand.

May 29, 2015 at 6:58 am
Either climate science does not know how to deal with irregular multi-decade cycles along with their political partners, or they don’t want to know it. The former case is pathetic and latter case is not science inquiry. 

References and further reading

Gerard D. McCarthy, Ivan D. Haigh, Joël J.-M. Hirschi, Jeremy P. Grist, David A. Smeed. "Ocean impact on decadal Atlantic climate variability revealed by sea-level observations." Nature, 2015; 521 (7553): 508 DOI: 10.1038/nature14491 (subs req'd)

Global climate on verge of multi-decadal change - press release from the University of Southampton

Byron A. Steinman, Michael E. Mann, Sonya K. Miller. "Atlantic and Pacific multidecadal oscillations and Northern Hemisphere temperatures". Science, 2015 DOI: 10.1126/science.1257856 (pdf here)

Mann, Michael E., Byron A. Steinman, and Sonya K. Miller. "On forced temperature changes, internal variability, and the AMO." Geophysical Research Letters 41, no. 9 (2014): 3211-3219. DOI: 10.1002/2014GL059233 (pdf here)

Trenberth, Kevin E., and Dennis J. Shea. "Atlantic hurricanes and natural variability in 2005." Geophysical Research Letters 33, no. 12 (2006). doi:10.1029/2006GL026894 (open access)

Mann, Michael E., and Kerry A. Emanuel. "Atlantic hurricane trends linked to climate change." Eos, Transactions American Geophysical Union 87, no. 24 (2006): 233-241. DOI: 10.1029/2006EO240001 (open access)

From HotWhopper

A short pause: Bob Tisdale thinks climate models are weather forecasts of the Northern Hemisphere - an article about Steinman15 (as above)