Sunday, May 20, 2012

Natural Variability Is Not the Lukewarmers Friend

There have been a number of posts and comments about James Hansen's op ed in the NYTimes, featured here at Rabett Run, and of course, many other places.  As Eli predicted, the key graph for those in disagreement was

The global warming signal is now louder than the noise of random weather, as I predicted would happen by now in the journal Science in 1981. Extremely hot summers have increased noticeably. We can say with high confidence that the recent heat waves in Texas and Russia, and the one in Europe in 2003, which killed tens of thousands, were not natural events — they were caused by human-induced climate change. 
As he said then Eli's POV on this is rather simple.  If you have natural variability on top of a rising base, then the hotter extremes are going to get hotter and you don't need a degree in meteorology or statistics to figure that out, so the Bunny is with Hansen.  If you want to argue that it is going to cool like crazy Roy in the next decade, well, good luck with that, but Eli has a slightly used Sky Dragon to sell you. 

Andy Revkin, in particular was keen to go out and find disagreeers, in particular Martin Hoerling from NOAA, and Marty did disagree
This is patently false. Take temperature over the U.S. as an example. The variability of daily temperature over the U.S. is much larger than the anthropogenic warming signal at the time scales of local weather. Depending on season and location, the disparity is at least a factor of 5 to 10.
I think that a more scientifically justifiable statement, at least for the U.S. and extratropical land areas is that daily weather noise continues to drum out the siren call of climate change on local, weather scales.
Dot Earth arranged a further, more formal exchange of position papers between Hoerling and Dan Miller who worked with Hansen on the op ed.  As usual, Eli encourages all to RTFRs, but here he wants to call attention to Hansen's brief comment
I have several papers well along in the publication process that make clear your characterizations are far off the mark. The editors prefer, indeed are insistent, that I not discuss these in blogs. Some scientists may be able to spend their time blogging and e-mailing without a significant impact on their scientific productivity — I’m not one of them — but I do make an effort to make my papers understandable to a wide audience.
Dan Miller's paper at Dot Earth and Kate, over at Climate Sight, drew Eli's attention to Public Perception of Climate Change and the New Climate Dice by Hansen, Sato and Ruedy which was posted to arXiv April 5, so perhaps this is not one of the papers Hansen was specifically referring to, but phrases in the abstract show that it clearly bears.
"Climate dice", describing the chance of unusually warm or cool seasons relative to climatology, have become progressively "loaded" in the past 30 years, coincident with rapid global warming. The distribution of seasonal mean temperature anomalies has shifted toward higher temperatures and the range of anomalies has increased. An important change is the emergence of a category of summertime extremely hot outliers, more than three standard deviations (3{\sigma}) warmer than climatology. This hot extreme, which covered much less than 1% of Earth's surface in the period of climatology, now typically covers about 10% of the land area. It follows that we can state, with a high degree of confidence, that extreme anomalies such as those in Texas and Oklahoma in 2011 and Moscow in 2010 were a consequence of global warming, because their likelihood in the absence of global warming was exceedingly small. We discuss practical implications of this substantial, growing, climate change.
 The manuscript is there to be read, off you go little bunnies, but Eli wants to look at a key figure.

The authors examined the distribution of local anomalies using the 1951-1980 base period using local standard distributions.  This was also done for the Northern Hemisphere alone, and for other base periods.  The areas under each curve are normalized.  The black curve is a Gaussian and it should be noted that the 1951-61and 1961-71 distributions are slightly more peaked, with slightly less area in the cold and hot wings, in other words there are fewer extremes.

On the other hand (kenw that was coming didn't ja) with time after 1971, the distributions shift to hotter temperatures, and the shift increases with time, good additional evidence that global warming is continuing, but more importantly, the distributions noticeably broaden and the wings grow significantly.  Whereas before 1971 the probability of a 3 σ event was zilch, today it is well above the base line.

But wait, there is more.  Eli as an old curve eyeballer thought there were something funny about the purple curve and tried to fit it to a gaussian.
and sure enough, the fit is fine up to the peak, but on the hot side the blue Gassian is much lower than the observed frequency of hot anomalies beyond about 1 σ  This shift from symmetric Gaussian to skewed hot anomalies is as, if not more important for addressing the issue of whether extreme hot spells are meteorology or climatology.

Although Hansen, Sato and Ruedy have not discussed the skew, they have looked simply at the implications of the anomaly distributions and this explains their confidence about the source of the extreme summer anomalies
Seasonal-mean temperatures in the category defined as "cold" in 1951-1980 climatology (mean temperature below -0.43σ), which occurred about one-third of the time in 1951-1980, still occur with a probability about 10% over land areas. Thus an occasional unusually cool winter is not evidence against global warming. Temperature is less "noisy" in the summer than winter. The chance of summer falling in the "hot" category of 1951-1980 is now about 80% (Fig. 7). The climate dice are now loaded to a degree that the perceptive person (old enough to remember the climate of 1951-1980) should recognize the existence of climate change. 
The most important change of the climate dice is the appearance of a new category of extremely hot summer anomalies, with mean temperature at least three standard deviations greater than climatology. These extreme temperatures were practically absent in the period of climatology, covering only a few tenths of one percent of the land area, but they have occurred over about 10% of land area in recent years. The increased frequency of these extreme anomalies, by more than an order of magnitude, implies that we can say with a high degree of confidence that events such as the extreme summer heat in the Moscow region in 2010 and Texas in 2011 were a consequence of global warming. Rahmstorf and Coumou (23), using a more elegant mathematical analysis, reached a similar conclusion for the Moscow anomaly. 
It is not uncommon for meteorologists to reject global warming as a cause of these extreme events, offering instead a meteorological explanation. For example, it is said that the Moscow heat wave was caused by an atmospheric "blocking" situation, or the Texas heat wave was caused by La Nina ocean temperature patterns. Certainly the locations of the extreme anomalies in any given case are related to specific weather patterns. However, blocking patterns and La Ninas have always been common, yet the large areas of extreme warming have come into existence only with large global warming. Today's extreme anomalies occur because of simultaneous contributions of specific weather patterns and global warming.
Martin Hoerling has a hard walk back.


William Connolley said...

Hansen's refusal to discuss with Hoerling on the grounds of I-can't-talk-on-blogs is bollocks.

EliRabett said...

You should get out more

Anonymous said...

This one is rich: "It is not uncommon for meteorologists to reject global warming as a cause of these extreme events, offering instead a meteorological explanation."

Imagine that - using meteorology to describe meteorological events.

Heatwaves are a function of subsidence associated with anticyclonic air masses - not much controversial there.

That such episodes, based on the pile up of high pressure areas,
occur chaotically around the world is not particularly surprising
( rather like bumper cars, one knows a pile up will occur eventually, just not when or where ).

How exactly additional warming from GHGs would effect this is not clear to me.
High pressure systems are a function of polar cooling, after all.

But lacking a meteorological theory much less evidence, claims about heatwaves and AGW
don't surpass idle speculation.


Anonymous said...

Waldo-- weren't you paying attention to the argument here? The point is pretty obvious-- change the statistical distribution of temperatures (i.e. the climate) and the probability of some kinds of events (i.e. extreme hot events) can (and has) increased substantially. Does repeating the point make it clear to you? Or are you lost in the Dunning-Krueger fog?

Bryson Brown

Scrooge said...

Sometimes its hard to see the forest because of all the trees. GW is shifting the jet stream pattern. So climate patterns change. Another thing that may be happening is with the arctic warming faster the amplitude of the pattern may be increasing. So the easy way to explain heat waves is to say blocking high. What should be asked is why so intense. The rotation of earth and geography pretty much stay constant. Over the long run patterns should be fairly stable. The one thing we are changing is the temperature.

jyyh said...

I've been sporting to use the (monthly/seasonal) temperature anomaly maps from noaa:
( and
to track down where 'the global warming resides' currently, whenever someone has mentioned a local cool spell (ok we had a cold spell two winters ago too). The method is simply to take the upmost quarter of the scale, and look, f.e. now 21/5/2012 (well the map says 18/5/2012, but it's monday so it's just waking up) looks like GW is in Barents sea, Midwest plains and Antarctica (90days) and Antarctica, Wyoming,Russian far east and north of Black Sea (30days). It seems that this 'global warming' has got no pattern in its occurrences, but I maybe wrong, as it seems to reside much in the Arctic or in the Antarctic.

Anonymous said...

Not just temperatures, floods as well.

"A record La Niña event coupled with tropical cyclone Tasha generated most of the record deluge of rain that devastated much of Queensland in December 2010, but a new study has found that another big culprit was also in play - record high sea-surface temperatures off northern Australia.

It was thought that the twin impacts of the La Niña and the cyclone alone could explain why Queensland’s December rainfall was an all-time high at 154% above normal.

But the new calculations by climate researchers have revealed that evaporation from the warmer seas to the north and north-west of Australia probably contributed about a quarter of the total.

Sea-surface temperatures off northern Australia in the Indian Ocean, Arafura Sea and Coral Sea were unusually warm at the time, in places as much as 2 degrees C, the study notes: analysing 30 years of historic measurements, the study identified a general warming trend there of at least 0.2 degrees C per decade.

“If the observed warming trend in the sea-surface temperatures continues, this result suggests that future La Niña events are more likely to produce extreme precipitation and flooding than is present in the historical record,” says Dr Jason Evans, of the UNSW Climate Change Research Centre. Dr Evans led the study, to be published in the journal Geophysical Research Letters, with a French co-author, Dr Irène Boyer-Souchet.

“If the sea-surface temperature increases can be attributed to global warming, then the probability of La Niña events producing extreme precipitation responses similar to December 2010 will increase in the future.”

Global warming on top of the already extreme conditions = a disaster. The Queensland floods were unprecedented, and now large areas of the state are uninsurable.

Unknown said...

But lacking a meteorological theory much less evidence, claims about heatwaves and AGW
don't surpass idle speculation.

Lizard reflex; article cited in post did not reach neocortex. Tsk.

Hank Roberts said...

> not discussed on blogs

Editors do embargo papers until publication.

> chart
That chart needs labels on the X and Y axis to explain the obvious:

Total number of events/observations is the same for each curve (?)

X axis is temperature, 0 = normal, + is hotter, - is cooler.

Y axis is the number -- how many events/observations at any particular temperature.

Did I get that right?

Brian Dodge said...

Denialist refusal to present a coherent argument against AGW in peer reviewed publications on the grounds of there-is-an-international-scientific-conspiracy is bollocks.

Anonymous said...

This is such a simple common-sense argument, I really don't understand why NOAA seems desperate to deny it.

Timothy Chase said...

"Our results provide to our knowledge the first formal identification of a human contribution to the observed intensification of extreme precipitation. We used probability-based indices of precipitation extremes that facilitate the comparison of observations with models. results also show that the global climate models we used may have underestimated the observed trend,which implies that extreme precipitation events may strengthen more quickly in the future than projected and that they may have more severe impacts than estimated. There are,however, uncertainties related to observational limitations,missing or uncertain external forcings and model performance. "

Seung-Ki Min,Xuebin Zhang,Francis W. Zwiers,Gabriele C. Hegerl(2011 Feb 17) Human contribution to more-intense precipitation extremes, Nature,470(7334):378-81

Timothy Chase said...

My previous somewhat cryptic comment (written from my cell phone) was in response to Waldo, providing a theoretical basis for why we expect more extreme weather under global warming, albeit with respect to precipitation, but given the focus on temperatures in this discussion the following may be more pertinent.

There is good reason to think that the ice melt in the Arctic is already having a big effect on the mid-latitudes, resulting in more extreme weather. According to recent research, Arctic amplification decreases the temperature differential between the Arctic and the mid-latitudes, slowing the jet stream, causing extreme weather, whether flooding or drought, to stall out over a region. For example, high temperatures will increase evaporation, drying out a location until there is no more moisture, no moist air convection and temperatures climb further. Similarly, Rossby waves, which one sees in the swings north and south in the storm tracks swing further north and south, causing cold air to make further incursions to the south and warm air to make further incursions to the north. Thus Arctic amplification results in more extreme weather in the mid-latitudes.

Please see:

Francis, J. A. and S. J. Vavrus (2012), Evidence linking Arctic amplification to extreme weather in mid-latitudes, Geophys. Res. Lett., 39, L06801, doi:10.1029/2012GL051000

susan said...

Trying to "locate" where global warming is seems a false argument to me, though I sympathize with the frustration that leads to the effort. It's a question of perspective - taking a very narrow view and blocking out any information that does not fit the preconceptions doesn't work for the long haul, so allowing the narrowing premise is kind of giving in to the false argument.

A normal person with a habit of observation can see that things are changing over time, but if the brain is unable to get that there are highs and lows over time with an overall rising trend, not much can be done.

As to meteorologists not comprehending climate change due to global warming (caused by small increases in small amounts of heat-trapping greenhouse gases) that's not really true for all of them by a long shot, particularly the more educated and younger ones, who can see it all around them. Try Jeff Masters at Wunderground, for instance.

Anonymous said...


"The point is pretty obvious-- change the statistical distribution of temperatures (i.e. the climate) and the probability of some kinds of events (i.e. extreme hot events) can (and has) increased substantially."

You and I would agree that, lacking any other information, a shift in the mean would likely mean a shift in the distribution.

Great But that's only one point of the presentation. The other point is an increase in the variance.

Statistics do not cause heatwaves, physical phenomenon cause heatwaves.
What variance of the physical phenomena that give rise to heatwaves will change?
And if you have a theoretical basis for change, why would that be toward INCREASED variance and not DECREASED variance?

Heatwaves are interesting because the units of action - 'air masses' - become elongated, merged, non-distinct, and difficult to quantify, leading to a lot of superstition.

But I would like to observe:

1. Like snow events, they are relatively infrequent, episodic, mult-factoral, and variable.
As such, they are subject to more uncertainty with respect to 'global warming' than are more continuous variables such as temperature or heat content.

2. During the Altithermal, Northern heatwaves were probably much more intense than today's because summer sunshine was so much more intense. When large areas of subsidence occurred, the change in solar
input was probably greater.

3. Urban effects are much stronger during heatwaves than they are during 'normal' weather because heatwaves are drier and calmer. When the weather is wetter and windier, the gradient between urban areas and the surroundings is reduced - during heatwaves, it is increased.

4. As noted in 1, heatwaves are episodic, and not a good category for analysis,
but it is worth reflecting on the most extreme heatwaves of the past:

90 years ago: Africa 136F El Azizia, Libya, 13 Sep 1922
99 years ago: North America 134F Death Valley, CA (Greenland Ranch), 10 Jul 1913
70 years ago: Asia 129F Tirat Tsvi, Israel, 22 Jun 1942
123 years ago: Australia 128F Cloncurry, Queensland, 16 Jan 1889
131 years ago: Europe 122F Seville, Spain, 4 Aug 1881
107 years ago: South America 120F Rivadavia, Argentina, 11 Dec 1905
100 years ago: Oceania 108F Tuguegarao, Philippines, 29 Apr 1912

and with a brief record:
38 years ago: Antarctica 59F Vanda Station, Scott Coast, 5 Jan 1974

These are part of the natural range.

But ask yourself, if they were all tied this year, would you say
"I expect that, it's part of the natural range", or would you say
"Oh my god, we're frying the planet because of global warming"?

Question to yourself: if these records were all tied


EliRabett said...

Putting more energy into a system increases the excursions from the mean, something you learn in Stat Mech I. Eli is not saying that the weather is so simple, but it is a useful first cut at the issue.

Brian Dodge said...

In addition to putting more energy into the system, we have decreased the gradient to the poles. In the lower Missippi, the low gradient results in large meanders; when there is large flow(more energy), the meanders grow faster. When the meanders reach a critical size, and the flood reaches a critical rate, the flow cuts across the neck - creating an oxbow lake. IMHO, The physics of meanders in the jetstream follow similar physics.

The Hadley cells have expanded because of global warming, bringing relatively warmer air aloft further north. A northbound meander in the jetstream entrained some of this warmer(and less humid, therefore relatively denser) air, and delivered it to the sinking high pressure air to the east, where it sank and warmed adiabatically. the westward component of the outflow of the sinking air at the surface was converted by coriolis force into stronger northward flow, positive feedbacklocally enhancing that vector of the jetstream, while opposing the weakened force from the decreased polar/tropical gradient because of global warming that tends to drive weather eastward. This created an "oxbow lake" of hot cut off high pressure - and tens of thousands died prematurelybecause of global warming.

Anonymous said...


"Putting more energy into a system increases the excursions from the mean,"

Though not a perfect analog, for climatological temperature variation on the seasonal level, precisely the opposite is true.

Temperature variation peaks in winter and troughs during summer.

Here is a good example for Chicago:

Your locale probably shows something similar.

It is important to recognize that the significant "energy" of variation in the atmosphere is not the total thermal energy.
Rather it is the gradients that arise. The thermal wind is 'driven' not by the temperature, but by the gradient of temperature.

Imagine a uniform earth with the same temperature around the globe. The wind would not blow and there would be no weather or climate.
Now, raise the earth's temperature but maintain uniformity. Still, the wind would not blow and there would be no weather or climate.

The earth is not uniform, but an oscillating, radiatively differentiated spheroid.
The 'energy' of the atmosphere arises from the pole to equator gradient, making polar air relatively denser and thus pressing turbulently
upon the tropical regions.

During winter the gradient is greater, during summer the gradient is lower ( because of seasonal solar load ).

This goes a long way toward explaining the reduced temperature variation during summer.

Now consider what 'global warming' is alleged to do.

One aspect is 'Arctic amplification' - greater warming at the pole (North Pole, anyway) than the tropics.
To the extent this does occur, the meteorological ramification is clear - the pole to equator gradient is reduced.
The thermal wind is reduced, and like the summer analog, variability decreases.

I am not convinced that variability will decrease.

But I fail to see any compelling physical basis for it changing at all.


Scrooge said...

Seems to me with things like "lazy" jets that when an omega or blocking high forms they would hang around and cook in the frying pan a bit longer. The jet plays a role in breaking them down or moving them out. I don't now how that would effect the other side of the ridge.

Anonymous said...

Hmm, Waldo does not seem to understand what a heat wave is, if he thinks individual station record temperatures are relevant. Heat waves have long periods of abnormally high temperature over large areas. If they *were* rare, but now are not, they are perfect for showing the effects of global warming. Also, decreased latitudinal gradient is not a plausible argument for why a temporal range at one position should decrease instead of the observed increase in mean, variance, and skew.

As far as definition of a heat wave being vague or not, here is Hansen:"A more important change is the emergence of a subset of the hot category, extremely hot
outliers, defined as anomalies exceeding +3σ. The frequency of these extreme anomalies is
about 0.13% in the normal distribution, and thus a typical summer in the period of climatology
would have only about 0.1-0.2% of the globe covered by such hot extremes. We show that
during the past several years the portion of global land area covered by summer temperature
anomalies exceeding +3σ has averaged about 10%, thus an increase by about a factor of 50
compared to the period of climatology."
Contrast to Waldo's contributions.

Rib Smokin' Bunny

Anonymous said...

Assigned reading for Waldo:
Letters to Nature

Nature 427, 332-336 (22 January 2004) | doi:10.1038/nature02300; Received 15 September 2003; Accepted 17 December 2003; Published online 11 January 2004

The role of increasing temperature variability in European summer heatwaves

Christoph Schär1, Pier Luigi Vidale1, Daniel Lüthi1, Christoph Frei1, Christian Häberli2, Mark A. Liniger2 & Christof Appenzeller2

Atmospheric and Climate Science, ETH Zürich, Winterthurerstrasse 190, 8057 Zürich, Switzerland
MeteoSwiss, Krähbühlstrasse 58, 8044 Zürich, Switzerland

Correspondence to: Christoph Schär1 Email:

Top of page

Instrumental observations1, 2 and reconstructions3, 4 of global and hemispheric temperature evolution reveal a pronounced warming during the past approx150 years. One expression of this warming is the observed increase in the occurrence of heatwaves5, 6. Conceptually this increase is understood as a shift of the statistical distribution towards warmer temperatures, while changes in the width of the distribution are often considered small7. Here we show that this framework fails to explain the record-breaking central European summer temperatures in 2003, although it is consistent with observations from previous years. We find that an event like that of summer 2003 is statistically extremely unlikely, even when the observed warming is taken into account. We propose that a regime with an increased variability of temperatures (in addition to increases in mean temperature) may be able to account for summer 2003. To test this proposal, we simulate possible future European climate with a regional climate model in a scenario with increased atmospheric greenhouse-gas concentrations, and find that temperature variability increases by up to 100%, with maximum changes in central and eastern Europe.

Rib Smokin' Bunny

Anonymous said...


what physical basis do you have for believing that variability will increase?


Anonymous said...

Waldo asks:"what physical basis do you have for believing that variability will increase?"

The paper I cited used a model in addition to observations, and climate models are based on physics.

As far as simple explanations go, one factor is that intensification of the hydrological cycle gives more frequent, intense droughts such as those seen in Texas last year. This increases variability and introduces a skew in the distribution on the high end.

I missed your explanation for why the mean temperature is increasing, and why there should be a skew to the high side. I would prefer a cite to the scientific literature.

Rib Smokin' Bunny