OMG Nature Article Reports models matching the pause in AGW

[zdas04]

In spite of frequent assertions on eng-tips.com that "there ain't no pause, damnit", Gerald Meehl and company found that dang, there really is a pause, and if you take 262 models, tweak them for the current state of the art in modeling, input 1990 data as initial conditions, then by god 10 of the models show the hiatus clearly. The big problem in modeling is "how do you know which 10 to select while rejecting the 96.2% of the models that didn't?" when you don't have the answer book in advance? Meehl posed the question

If we could be transported back to the 1990s with this new decadal prediction capability, a set of current models, and a modern-day supercomputer, could we simulate the hiatus?

. That is kind of like saying "if we had today's tools we could have predicted and prevented..." The stock market crash? The Challenger Disaster? The plane crashes on 9/11?" Where the heck does it stop?

You can get the full text for $32 at Climate model simulations of the observed early-2000s hiatus of global warming. The opening paragraph is pretty telling

The slowdown in the rate of global warming in the early 2000s is not evident in the multi-model ensemble average of traditional climate change projection simulations1. However, a number of individual ensemble members from that set of models successfully simulate the early-2000s hiatus when naturally-occurring climate variability involving the Interdecadal Pacific Oscillation (IPO) coincided, by chance, with the observed negative phase of the IPO that contributed to the early-2000s hiatus. If the recent methodology of initialized decadal climate prediction could have been applied in the mid-1990s using the Coupled Model Intercomparison Project Phase 5 multi-models, both the negative phase of the IPO in the early 2000s as well as the hiatus could have been simulated, with the multi-model average performing better than most of the individual models. The loss of predictive skill for six initial years before the mid-1990s points to the need for consistent hindcast skill to establish reliability of an operational decadal climate prediction system

David Simpson, PE
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. —Galileo Galilei, Italian Physicist

 

[GregLocock]

So the consensus of 97% was wrong and the 3% was right ? Oh sorry 96% and 4%. Hmmm...

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[hydroman247]

It amazes me how so many professionals can completely misunderstand statistics.

 

[rb1957]

i'd expect 262 models will give you 262 different results ... you just need to be careful in selecting which is right; in the 90's they wanted the results to say "it's getting hotter".

another day in paradise, or is paradise one day closer ?

 

[rconnor]

For the Nth time, no one has said that the recent short term slowdown of surface temperature rise doesn’t exist. What we have stated is that the “pause” does not entail a “pause” in energy accumulation as OHC has increased throughout this period, especially >700 m depths which is exactly what you’d expect during a La Nina dominated period. Furthermore, when you compare ENSO neutral years with ENSO neutral years, they continue to warm at similar rates to the 30 year trend and faster than the 60 year trend. We have stated, and I will continue to state, that the “pause” is not a valid argument against the ACC theory – this paper is one of many that explains why. (I sound like a broken record but when you folks keep regurgitating the same arguments without addressing the reasons why it’s a faulty argument, then the problem lies with you not us)

No one can forecast which ENSO events will occur when and with what magnitude, neither can models. However, what we do know is that they the short term storage and release of heat from the oceans to the atmosphere, with a roughly oscillating cycle between positive and negative PDO. While ENSO events can be the single most dominate factor in a single year’s temperature, they have no notable long-term impact. During El Nino years, heat in the oceans is released causing a hotter than average year. But during the next non-El Nino year, the temperatures tend to return to average. This is not “cooling” this is moving from an anomalously hot year to a year with less internal variability. During La Nina years, warm water is forced deeper in the oceans and less heat is released to the atmosphere causing a cooler than average year. But during the next non-La Nina year, the temperatures tend to return to average. This is not “warming” this is moving from an anomalously cool year to a year with less internal variability. ENSO events cannot effect long term climatic trends by itself because it only works to move heat around the system (oceans to atmosphere) and has no effect on energy imbalances. They can however act to hide trends over short time periods or allow people to cherry pick dates to suit a narrative as I will describe below.

If someone wanted to deceive people into thinking that the warming trend was greater than it actually is, they’d select a period starting with a strong La Nina (anomalously lower starting point) and ending with a strong El Nino (anomalously higher end point). For example, let’s take the period of 1992 (La Nina and effects of the 1991 Mount Pinatubo eruption) to 2010 (El Nino).
[image

http://s3.postimg.org/8gd0slbxf/ENSO_GISS_Trend_1992_to_2010.png

]
During this 18 year period, the warming trend would be 0.29 deg C/decade. This is close to double the trend since 1975 of 0.17 deg C/decade. But what did the scientific community say about this period? Did they deceptively use it to further their plot for world domination?

Well, Rahmstorf et al, 2007 was comparing climate observations to computer model projections and noticed that the warming was faster than expected leading up to 2006. Did they hit the blogosphere and start ranting about how “IPCC models failed! Earth expected to be ball of molten lava by 2050! Surrender your individual rights and freedoms to me!!!”? No, they looked at the situation in its entirety and attempted to actually and honestly understand the data in front of them. They concluded:

The first candidate reason is intrinsic variability within the climate system.

Conversely, if someone wanted to deceive people into thinking that the warming had stopped, they’d select a period starting with a strong El Nino and ending with a strong La Nina. For example, let’s take the period of 1998 (strongest El Nino on the modern record) and 2012 (double dip La Nina).

Well, you know how that story goes.

So while, yes, temperature trends are lower than they have been during this recent short term period, that in no way acts to weaken the core argument for the ACC theory. OHC is still rising, ENSO neutral years continue to warm at rates similar to the 30 year period, ice and snow continue to melt, sea level continues to increase. All while solar activity is dropping and aerosols are increasing.

Now that we’ve cleared up zdas04’s straw man, let’s revisit what the paper is actually saying. As previously stated, ENSO events are stochastic, short term and roughly oscillate between El Nino dominated periods and La Nina dominated periods. Different model runs estimate different ENSO events at different times. Due to this, looking at the “average” model run as a comparison against observed temperature trends can be misleading for short periods that are dominated by a specific type of internal variability (negative PDO in this case). A better comparison is to looks at model runs that estimated ENSO events more or less accurately and see how they compared to observed temperature trends. That’s exactly what Meehl et al 2014 does. And when you do that comparison, you find that those model runs represent the recent temperature changes quite well. This demonstrates that CO2 sensitivity, feedbacks, cloud parameterization and other such factors that skeptics are skeptical about, appear to be quite accurately represented in models.

 

[TGS4]

We have stated, and I will continue to state, that the “pause” is not a valid argument against the ACC theory – this paper is one of many that explains why. (I sound like a broken record but when you folks keep regurgitating the same arguments without addressing the reasons why it’s a faulty argument, then the problem lies with you not us)

Couple of questions:
1) Are you confused about whether or not you are an individual or a group - you vacillate between I, we and us?
2) Are these your words or are you regurgitating them from elsewhere? If so, where?
3) Please be so kind as to tell us EXACTLY what you (not the royal you, but you personally) mean by the ACC Theory. Exact numbers would be nice. Something falsifiable would also be good...

You also mention internal variability (PDO for example, as well as ENSO). How are these included in the CIMP5 models? How well do they actually simulate internal variability?

Of course you still have that the models, until they were told that a hiatus in atmospheric temperatures were happening, didn't predict the hiatus. How long of a hiatus is necessary to invalidate the model assumptions?

 

[IRstuff]

""there ain't no pause, damnit""

No, the "pause" is there, it's just not what you want to paint it as, which is a repudiation of climate change. The proof of the pudding will be when the "pause" ends, or when the oceans get so warm that all sea ice melts away.

"How long of a hiatus is necessary to invalidate the model assumptions?"

It would need to be infinite on our timescale, wouldn't it? The increasing OHC uptake rate over the same period could possibly account for a substantial portion the pause., but warming oceans aren't necessarily that much better for us than a warming atmosphere, and the warming ocean is telling us that there is still a net increase in the heat balance when all other environmental factors point to what should be a decrease in the heat balance. When the global temperature and OHC rise of the last 50 yrs that started this discussion get erased, that would something.


TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

Of course I can. I can do anything. I can do absolutely anything. I'm an expert!

 

[rconnor]

I’m going to assume your first three questions were a joke. Even if they weren’t intended on being, they are. Your last two questions are less inane and I will address them.

Currently, models don’t attempt to predict when a specific ENSO event will occur but produce a multitude of different ENSO scenarios which average out across the various runs. Some will, by chance, match what actually happened, most will not. The average model run will not match a period where one ENSO event dominates the other, as was demonstrated by the two cherry picked periods in my last post. However, the average model run should correspond with the trend of ENSO neutral years. This is consistent with what happened throughout the “pause”. Furthermore, long term trends should not be affected by this method as ENSO does not impact the long-term signal but merely dominates the short-term noise.

So while this method can cause models to stray temporarily from observations, the only way this could lead to long term deviations is if we enter a period where PDO stays negative indefinitely. Even at that, OHC will still continue to rise because ENSO/PDO does not dictate the energy imbalance, which drives long-term, significant global climate change. However, as this is highly unlikely, the more reasonable expectation is that once we enter into a positive PDO cycle, atmospheric temperatures will resume warming.

Your final question continues to miss the point (purposefully, I believe). The duration of “pause” doesn’t matter. What’s causing the “pause” does. If we stay in a La Nina dominated period for another 10 years, all while OHC continues to rise at rates consistent with the theory, then it STILL doesn’t discredit the ACC theory.

The key in all of this, and the point you continue to dodge, is that “skeptics” use the “pause” as “proof” that climate models are too sensitive to CO2 forcings or global warming has magically stopped. However, in reality, the proof continues to stream in that this is dead wrong. The discrepancy between models and observations can be boiled down to a “failure” to predict stochastic, temporary , roughly oscillating changes in ocean-atmosphere dynamics (ENSO/PDO). As this is rather unimportant to long-term climate change trends, nothing about the “pause” invalidates the ACC theory.

In fact, when you input actual ENSO index data, models produce temperature trends that match observations extremely well, which directly validates the ACC theory and adds confidence to model projections. If, conversely, when you input actual ENSO index data and the accuracy does not improve (or gets worse) THEN “skeptics” might have an argument because it would suggest that some other aspect (sensitivity, cloud parameterization, etc) is off. Too bad that isn’t reality…

 

[TGS4]

No, the first two questions weren't jokes - I seriously want an answer to them. Just like I seriously wanted an answer to my other questions about your peer review pedigree. You may dismiss them as silly/stupid, but that dismissal is rather telling to me...

What's interesting from your perspective is that just as ENSO and PDO can be used as excuses for why the models "miss" reality, isn't the reverse also potentially true - that the late 1900's ramp up in temperature was due (or at least partially due) to the reverse occurring - the oceans dumping thermal energy into the atmosphere while losing energy themselves? And remember, our data for OHC prior to 2007 is CRAP, so we don't know what the oceans were really doing...

And again, what is this so-called ACC theory that you speak of, again? Not an inane question - we seem to be pulling out a lot of strawman arguments here - let's put some meat to the discuss and define some terms here. What is it and what, exactly, are the numbers? Remember that I agree that CO2 is a greenhouse gas and that its sensitivity without feedbacks is about 1°C per doubling. And I also agree that the evidence seems to indicate that the increase in CO2 concentration has an anthropogenic signature to it.

 

[rb1957]

"You also mention internal variability (PDO for example, as well as ENSO). How are these included in the CIMP5 models? How well do they actually simulate internal variability?"

i'd ask this question differently ... why do the models include PDO and ENSO ? i mean we're repeatedly been told that they don't produce long term effects, and also that models are not tuned to past history.

another day in paradise, or is paradise one day closer ?

 

[rconnor]

TGS4 - what’s telling about dismissing questions that are so far away from the actual point that they become absurdly irrelevant? Regarding the peer review pedigree (and, yes, I’ve been published), what you ACTUALLY mean to say is “the peer review process isn’t perfect, and if you’ve published/reviewed you’d know”. And yes, I do know it’s not perfect but it is absolutely the very best system we have for assuring scientific accuracy. Do you disagree with this? Of course a paper here and there will slip through that shouldn’t. However, when we discuss the scientific support of the ACC theory, we aren’t talking about a handful of papers. We’re talking about thousands of papers from hundreds (if not thousands) of authors from dozens of institutions publishing into dozens of journals for decades! Are you saying that the failure in the peer-review process is this far wide spread? What system is better? Do you think some random blog holds articles to more scientific rigor than Nature, Science, PNAS? Of course you don’t! And that makes your original question extremely inane and is the reason why I never answered it.

ACC Theory – by and large, I agree with the culmination of scientific evidence that suggests that anthropogenic CO2 emissions, if allowed to continue unabated, will lead to future global temperature rises that will have adverse effects on the ecosystem and, thusly, humans. I take it your issue is with the ECS or TCR value? This is fair. There is still uncertainty in that value. However, as previously stated (hence my confusion on why you’re asking again) I’ve yet to read substantial, credible evidence that suggest that these values should be on the low end. So while maybe you’re not fully in agreement, what is your evidence that it is undoubtedly wrong? Doubting and having credible evidence that something is wrong are two different things. However, “skeptics” appear to blend the two together. And ironically, the only thing that “skeptics” don’t doubt is any and all opinions that attack the ACC theory.

“isn’t the reserve also potentially true” – did you read what my first post? See that big graph that shows a warming trend nearly double that of the long-term average? See the quote from the Rahmstorf et al 2007? But I’ll answer it again…
YES – the reverse is true and ENSO can cause artificially accelerated warming periods. It’s happened. It’s been noted.
BUT –climate scientists who are actually attempting to understand the data honestly, unlike “skeptic” bloggers, understood that it was likely just internal variability, unlike “skeptic” bloggers, and concluded that sensitivity wasn’t actually higher than previously estimated. That’s the difference between honest science and blogger “science”. That’s the difference between skepticism and “skepticism”.
AND – ENSO/PDO, which are purely about the redistribution of heat from oceans to atmosphere, cannot possibly explain long-term warming trends especially warming trends in both temperature and OHC. (and I know pre-2005 OHC data is a dirty word for you but you’ll have to provide evidence that pre-2005 is unusable because nothing in the post-2005 data suggests that the pre-2005 data was off base…and regardless, post-2005 data, DURING THE “PAUSE” shows steady increases)

rb1957 – That is actually a good question that I’ve wondered myself. What effect on models would you have if you removed their ability to estimate ENSO conditions and forced them to have ENSO neutral conditions all the time? Would the “average” model run be the same in both cases? I would think it would be close but I’m not sure.

One reason I could see for including it is that it provides a good test of how ocean-atmosphere interactions are handled. Put in historic ENSO index data and see how the model represents it compared to reality. Again, rb9157, you have to be very careful to differentiate between historic temperature data being used as a test versus being blindly tuned to. The former is true, the latter is false. In this example, if models aren’t able to reproduce temperatures for different ENSO events then there is an issue with ocean-atmosphere interactions that result from changes in trade winds. This specific area is refined, such that it’s more accurately represented, with the hopes that, consequently and secondarily, global temperatures are also more accurate.

However, neither you nor TGS4 nor anyone else that’s used the “pause” as an argument against the ACC theory have addressed the crux of the issue here :

“skeptics” use the “pause” as “proof” that climate models are too sensitive to CO2 forcings or global warming has magically stopped. However, in reality, the proof continues to stream in that this is dead wrong. The discrepancy between models and observations can be boiled down to a “failure” to predict stochastic, temporary , roughly oscillating changes in ocean-atmosphere dynamics (ENSO/PDO)... when you input actual ENSO index data, models produce temperature trends that match observations extremely well, which directly validates the ACC theory and adds confidence to model projections… This demonstrates that CO2 sensitivity, feedbacks, cloud parameterization and other such factors that skeptics are skeptical about, appear to be quite accurately represented in models

(I’m not sure if I’ve referenced Huber and Knutti 2014 but it’s very relevant (as of course is Schmidt et al 2014). Breakdown of the paper at ATTP)

 

[rb1957]

"anthropogenic CO2 emissions, if allowed to continue unabated, will lead to future global temperature rises that will have adverse effects on the ecosystem" ... that is the key word "future".

i think most of the disagreement is that most of the CC graphs (yours included) are saying, IMHO, that the CO2 we put into the air yesterday is having an effect today, ie that the effect is very close coupled. otherwise there'd be a delay in the curves that everyone matches so carefully (ie temperature "anormality" and CO2 with time). Of course the other disagreement is how much are AGHG affecting climate as opposed to other factors (including "unknown").

another day in paradise, or is paradise one day closer ?

 

[GregLocock]

I must admit I rather like this graph.

taken from

https://ec.europa.eu/jrc/en/news/climate-anomaly-causes-global-warming-slow-down.

However apparently the science isn't settled, they don't know what causes the cycles, or what causes the shape of the early part of the 'secular' curve, so they need our money. An odd choice of language given the widespread accusation that AGW is a religion-substitute.

Anyway by eye it looks as though we are on course for a temperature rise of about 0.8 deg by the end of the century, which is rather good news for the human race.

Taking more of a punt, the pause is likely to continue for another fifteen years, which may try the patience of the believers.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[IRstuff]

There's this

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

Of course I can. I can do anything. I can do absolutely anything. I'm an expert!

 

[TGS4]

(and I know pre-2005 OHC data is a dirty word for you but you’ll have to provide evidence that pre-2005 is unusable because nothing in the post-2005 data suggests that the pre-2005 data was off base…and regardless, post-2005 data, DURING THE “PAUSE” shows steady increases)

No I don't have to prove evidence that the pre-2005 data is unusable. It is so unbelievably sparse that any scientist (or engineer) with any experience would discard it as unrepresentative of the whole ocean.

Yes, indeed, the oceans in aggregate are heating up. However, it's not uniform as I have stated before. The Southern Ocean (predominantly the Indian Ocean) is heat up rather quickly, but the ALL other parts of the oceans are COOLING -

http://ceres.larc.nasa.gov/documents/STM/2013-10/14_Global_averages.pdf

Again, how does the ACC theory predict that?

 

[rconnor]

GregLocock, I’m unsure how that paper supports your position that the “pause” invalidates the ACC theory. Frankly, the paper is a little odd. It states something that is already well known, that there has been research into the causes, doesn’t bother going into the research in any depth and then concludes more is required. While I don’t disagree that more research is always a good thing, it certainly does not offer support that lower climate sensitivity is the most likely explanation for the “pause”. It offers no commentary on why the authors feel that the research up to this point is inconclusive. It almost seems to suggests that the various lines of research (ENSO, aerosols, etc.) mean that the scientific community is at odds with what the cause is (i.e. they are incompatible with each other), which is just false. This is explained in Schmidt et al 2014 and Huber and Knutti 2014 that they never reference (likely because they were published too recently) which demonstrate that, when taken together, the various lines of research explain the “pause” very well. Furthermore, the authors fail to reference the coverage bias inherent in HadCRUT (see Cowtan and Way 2013).

And your eye-balling analysis is, well, suspect.

TGS4, while very sparse and predominately coastal, the pre-ARGO data came from different areas around the globe which provides for a diverse, yet sparse, sampling. The thing that matters is long-term anomalies. So while specifics would be difficult to conclude with any accuracy, long-term trends could be determined from the pre-ARGO data. And those estimates of long-term trends do not seem out of place in light of the ARGO data. (If you haven’t already, you should read the Levitus papers (various years, various updates). They detail the process in calculating deep OHC in pre-ARGO conditions. It certainly does not shy away from the uncertainty in working with such a sparse sample set. While I don’t expect it to change your mind, it would be important for you to know, given your strong stance on the data.)

Regarding the non-uniform heating, the oceans are a dynamic place. No one expects uniform heating to occur. Just like ENSO events move heat between the atmosphere and oceans, the oceans move heat between themselves. However, if year after year, the total energy content continues to increase, we continue to have a problem. Also, I’d check the 2013 data.

 

[rconnor]

rb1957, there are two things that you need to dissociate when thinking about the rate in which CO2 affects temperature. The first is the direct effect of increased CO2 decreasing outgoing radiation (the greenhouse effect). The second is the indirect effect of positive feedbacks.

The direct greenhouse effect is quite fast and can be thought of as happening within the same year. In reality, it might not be that fast but given the rate of CO2 emissions, we can, for all intents and purposes, assume that each ppm increase in atmospheric CO2 concentration will increase the greenhouse effect that same year. However, this effect causes an energy imbalance not a direct increase in atmospheric temperatures. The majority of the energy imbalance is taken up by increased OHC. As we’ve discussed, this energy can be released to the atmosphere or stored, depending on oceanic cycles and trade winds (ENSO, PDO, AMO, etc.). So the effect on atmospheric temperatures may be delayed a month, a year or a decade. Details aside, to address the crux of your question – the recent warming is attributable to recent CO2 emissions according to the ACC theory

The indirect, positive feedback effects can have a fast response or a slow response, depending on the feedback. Positive feedbacks beget more positive feedbacks. However, in the absence of a primary driver (which is CO2 according to the ACC theory), the positive feedbacks weaken over time until they too settle to a new climate state. So, for example, if you were to double the atmospheric CO2 concentration and then hold the concentration there, the planet would continue to warm for some time (multiple decades) but slowly settle to a certain temperature. What I just described is referred to as the Equilibrium Climate Sensitivity (how much the planet will warm, in response to a doubling of atmospheric CO2 concentrations).

So, to conclude in a simplified sense, increases in atmospheric CO2 emissions today, have an effect on the energy balance tomorrow. However, due to a positive feedback loop, the full effect of increases in atmospheric CO2 emissions today, will not be fully realized for decades to come.

 

[rb1957]

"However, in the absence of a primary driver (which is CO2 according to the ACC theory), the positive feedbacks weaken over time until they too settle to a new climate state." ... that's going to take some thinking to appreciate.

a question, what causes the huge tidal changes in CO2, as recorded at Mauna Loa ? I assume that they are not globally synchronised, but globally phased.

another day in paradise, or is paradise one day closer ?