It is time to stop quivering in our boots in pointless fear of the future and just roll up our sleeves and build it.
- Ray Pierrehumbert

Friday, August 20, 2010

Fair is Fair

I'm not a great fan of Pat Michaels. But whether you like a person or not doesn't affect whether they are right or wrong.

In the context of whether recent events in eastern Europe and western Asia have a precedent, Michaels comes up with this.

Surface temperature anomalies (°C) for July, 1936
(figure from the Goddard Institute for Space Studies).

It certainly looks like the same pattern we saw this year over the eastern hemisphere, perhaps slightly less intense, but not much so. North America is different, but ENSO is neutral to cool and the Atlantic is hot, similar to this summer. This calms me down a bit, since I don't actually want an unprecedented climate disaster.

Further, let's admit this gives ammunition to RP Jr., who says disasters are MORE about human habitation and management practices than about weather extremes. In fact that is true, and Roger's tendency to overemphasize it at the expense of other factors shouldn't cause us to forget it.

While this may take a bit of the edge off of recent arguments, neither of these items should encourage us to let down our guard too much. First of all, the Pakistan human disaster is BECAUSE of water management, not despite it: the population has greatly increased because the Indus river has been channelled and diverted all across the otherwise barren land. So now there is a much enhanced population in a much expanded flood plain. (I think Russian mismanagement of their environment both pre and post communist is well-known and demonstrated here.)

Secondly, even if the pattern is not entirely unprecedented, it is not a good pattern. So while the attribution to climate change may be weakened (and again, let's not jump the gun in either direction here, let's let the weather guys think about it for a while) it still isn't good news that we are seeing wild oscillations, and we still have reasons to suspect these are becoming more frequent.

Finally, we still have an open question. How will we know when we've "rolled a 13"?


Chip said...


I would contend that, at least according to the analysis from the NOAA folks at PSD:

that the types of wild oscillations that led to this summer’s heat wave in Moscow don’t seem to be becoming more frequent—at least in that part of the world.


Steve Bloom said...


Michael, never, ever, take anything from Pat at face value. One of his standard tricks is to mix apples and oranges and not say so, as with his figures 2 (base period 1971-2000, as it turns out) and 4 (base period 1951-1980). The latter is a serve-yourself graphic with an adjustable base period, so the lack of a match isn't a mistake.

Also, a calendar-month comparison for blocking events that have roughly that scale of duration has an inherent potential to misrepresent the situation.

So, without looking into the details further (and this sort of crap from Pat bores me so much I'm not going to bother, but I think you ought to given that you fell for it), Pat has demonstrated that Russia has experienced at least one mid-summer blocking event-induced heat wave in the past. The question is the relative intensity/duration, and Pat's post sheds no light on that whatsoever.

Hi Chip! Don't forget to say hi to Don Blankenship for me.

WV puns, suitably: cateso

Chip said...

Hi Steve,

When we post the WCR article, the GISS July 2010 analysis wasn't yet available, so I just grabbed one from CPC. Perhaps Michael can post the GISS temps here for direct comparison. As Michael observed, it is a bit less intense, but the patterns are similar.

Also, if you are not happy with our analysis, I suggest that you spend time with NOAA's.


Steve Bloom said...

Chip's link is interesting, notwithstanding. Basically they make the case that even though this is was 4-sigma temperature event, the lack of an apparent trend toward such events in that region means that it's likely not a consequence of climate change. That seems reasonable enough, but I'd like to see them extend the analysis to include the monsoon event and the associated record-high North Atlantic SSTs.

Steve Bloom said...
This comment has been removed by a blog administrator.
Michael Tobis said...

Steve just a hair too confrontational; please tone it down.

As for "Basically they make the case that even though this is was 4-sigma temperature event, the lack of an apparent trend toward such events in that region means that it's likely not a consequence of climate change." I consider that argument precisely backwards.

On the other hand, clearly there is precedent for the spatial pattern and the season, which was one of the questions that I was asking. That is somewhat reassuring.

Hank Roberts said...

"The interannual variation of blocking signature frequency over the Northern Hemisphere showed substantial fluctuations, particularly with regard to the more persistent episodes. However, because of the small size of the sample (33 years) relative to the time scale of the fluctuations (order of one decade) it was not possible to draw statistical conclusions."

Blocking signatures in the northern hemisphere: Frequency distribution and interpretation

International Journal of Climatology
Volume 5, Issue 1, pages 1–16, Jan/Feb 1985 John L. Knox, John E. Hay
(first published online only recently: 28 NOV 2006)

Scholar's "cited by"

finds more that a statistician might find useful, e.g.

"The results show that the frequency of blocking formations in the North Atlantic region is sensitive to the phase of the NAO. Sixty-seven percent more winter blocking days are observed during the negative than the positive phase of the NAO. The lifetime of blocking episodes is also sensitive to the phase of the NAO. When the NAO is in the negative phase, the distribution of the length of blocking varies considerably. The average length of blocking during the negative phase is about 11 days, which is nearly twice as long as the 6-day length observed during the positive phase of the NAO. The NAO accounts for about 30% of the variation in the wintertime North Atlantic blocking episodes."

Shabbar, A., Huang, J. and Higuchi, K. (2001), The relationship between the wintertime north Atlantic oscillation and blocking episodes in the north Atlantic. International Journal of Climatology, 21: 355–369. doi: 10.1002/joc.612

Much else out there. I trust the people familiar with the field are getting their papers written and into the journals before the next IPCC deadline.

Steve Scolnik said...

And cherry-picking is unfair. Given that Michaels' books are a veritable feast of cherries jubilee, I'd expect a little more, what is it called, oh yeah, skepticism here. A coupla things stand out at first glance:
- Extremes are important, but ultimately it's about the means. The NCDC data show that this July was at least twice as anomalously warm as anything in the 1930's, and for the Northern Hemisphere, the difference is even more pronounced.
- Yeah, there's a correlation between the locations of the North American and western Asian ridges, but the magnitudes are not comparable. The Moscow extremes completely blew up the record book, as Jeff Masters has shown. With only 11 days to go, we're also about to smash the summer heat record right here in Michaels' Spin City by as much as half a standard deviation beyond the old record. (If the mean to date were to hold, it would be 90% of a standard deviation beyond.)
- Even if they were comparable, the fact that some event happened at some point in the past is irrelevant if the recurrence period is different.
- The Atlantic is not hot in that 1936 chart. The vast majority is near or below average, but this year it has been record-breaking hot.

Steve Bloom said...

Sorry, Michael, but the whole employed-by-the-fossil-fuel-industry-to-produce-propaganda thing bothers the hell out of me. In a reasonable world it would be criminal behavior since, to the exctent that it succeeds, people will inevitably die. I'm quite surprised it doesn't bother you more.

Feel free to axe this one as well.

Steve Bloom said...

I should have added "as far as it goes" to that. It seems to me that regardless of what analytical approach one takes (rarity of event vs. rarity of atmospheric pattern) it's preferable to look at the other pieces of the pattern. Taking the dice analogy, the Russian bit might be a seven, but of course one can obtain a seven with other than a seven die. It's the thirteen combo that seals the deal IMHO.

Speaking of which, the cartoon is fabulous, right up there with Easter Bunny Island but with more mass appeal.

Michael Tobis said...

It bothers me a lot, but insofar as I can tell many people believe what they are saying when I find it unbelievable. So I try to give anyone this side of Singer the benefit of the doubt. Michaels may or may not be this side of Singer, but it appears to be a real map.

It's one thing to say it's unprecedented in magnitude, anotehr to say it's unprecedented in configuration. I have been saying I suspect the latter, and insofar as I suggested that it appears I was wrong.

I'll have more to say on this.

Steve Bloom said...

Also note that the analysis Chip linked doesn't even identify 1936 as a notably warm year in Russia, although it sure was in the U.S.

Also, Michael, re blocking events generally, my impression was that they're quite common and tend to recur in the same locations, so the bare fact that they've occurred previously in mid-summer in Russis doesn't seem notable, so long as one is comparing them using unjuiced graphics.

Michael Tobis said...

They're not rare in some places. I thought I had read somewhere that the jet was in an unheard of configuration. But maybe I misunderstood.

Gareth said...

Michael, if you haven't seen Stu Ostro's thoughts on this, they're worth reading. He has been writing and commenting about a marked increase in 500mb heights over recent years - and as the interesting NOAA link indicates, a large 500mb anomaly drove the heatwave.

Weather delivers climate change: as an intensification of "normal" variability (ie existing patterns, but with changes in intensity and perhaps frequency of extremes), and also through the emergence of new patterns. The main driver for that, IMHO, is declining Arctic sea ice, but I would expect that to have its major impacts in autumn and early winter (ie during freeze-up).

Francis said...

I come to this issue as a California land use lawyer. I would not underestimate the importance of the decisions already made in placing infrastructure. As seen from our most recent jaunt into Mesopotamia, there's much more to building reliable infrastructure than handing billions of dollars to American contractors. Down that path lies waste, corruption and infrastructure that is worse than useless, because it interferes with developing alternatives.

Here in the US, major infrastructure tends to take 20 years or more to go from conception, through planning, initial approval, litigation, design, (more litigation), funding (and more litigation), construction (and yet more litigation). China does better on the timeline, but is making billions of dollars of investments that will be hard to undo.

In other words, the planning window for 25-50 years out is RIGHT NOW. Yet people like Pielke and Manzi appear to think that we can wait until impacts become more clear before responding. That, I think, is an extremely risky approach.

Steve Bloom said...

The overall configuration of the jet does sound highly unusual if not unique for that region, speaking if I have it right of the extent of the southerly dip. But that just goes back to my point about the bigger picture.

Note that PSD says: "Blocks are not an uncommon occurrence over Eurasia in summer, with a episodes of July blocking in the region between 0-60°E evident during the past half century. This region is vulnerable to episodes of blocking owing to physical factors related to the region's location downstream of the Atlantic westerly jet."

William T said...

You should put up the 2010 comparison - - the Russian event is a fair bit bigger than in 1936 (and more extreme).

omniclimate said...

wouldn't it make more sense to finally abandon the rather unpleasant rushing after the latest tragedies in the hope of being able to blame them on (anthropogenic) global warming? Rather than behaving like "climate looters", it would be far more effective to figure out where in the world a "climate signal" might be materialising (eg where trends in disasters are present or on the edge of being detectable), in order to concentrate minds on forecasting what it anything might happen in those specific places also with the goal of pushing adaptation projects forward.

this is not all too different from what vulcanologists already do.

EliRabett said...

1936 was no walk in the North American Park. If anything it makes a stronger case for avoiding climate change

"The 1936 North American heat wave was the most severe heat wave in the modern history of North America. It took place in the middle of the Great Depression of the 1930s, and caused catastrophic human suffering and an enormous economic toll. The death toll exceeded 5,000, and huge numbers of crops were destroyed by the heat and lack of moisture. Many state and city record high temperatures set during the 1936 heat wave still stand to this day. The heat wave followed one of the coldest winters on record."

Aaron said...

My feeling is that MT is absolutely correct, and the blocking of Eurasia in the summer of 2010, is a new configuration.

I see analysis where they say that a “4-sigma” higher block is more of the same-old, same-old. However they do not seem to want to talk about how frequently “4-sigma higher blocks” occur.

Moreover, in the last decade, Russia has seen more July warmth without a proportional increase in blocking. With all the talk about heat in Russia and floods in Pakistan, folks did not talk about Nuuk, Greenland being warmer than San Francisco, CA. Add this unusual warmth to an unusual blocking intensity and the combination is a new configuration.

The question is: How common is this new configuration? And will we see it again? I think it is now a stable weather pattern that will show up every so often, as long as there is both significant open water and significant ice in the Arctic. Russia, Pakistan, and China, should expect more summers like this. As can Nuuk, SF, and Tennessee.

There is more moisture in the atmosphere and any analysis that does not account for that increase in the latent heat in the atmosphere is going to get it wrong.

Alastair said...

We seem to be saying that throwing a 13 is not proof of catastrophic global warming so lets wait until we throw a 14, or even a 15 'cause then we can be sure.

But when that happens it will be too late!

Perhaps we should point out that, although they may not be caused by global warming, events like the Russia heat wave, Pakistan floods, Chinese floods, Katrina, Californian fires, Paris heat wave etc. are all examples of what will happen when global warming takes off. As the late Stephen Schneider said, we have to present some scary scenarios.

David B. Benson said...

Well, there is a recent book with subtitle Mathematics of Extreme Events. Comes with fine reiews and is accessible to the intelligent layperson.

Hank Roberts said...

> how will we know ...?

Hmmmm. Anyone following up with public discussion/blogging/reports from the meetings of the IPCC this year?
This might have been relevant:

6-7 July
IPCC Cross-Working Group Meeting on Consistent Evaluation of Uncertainties and Risks
Jasper Ridge, CA, USA


Aaron said...

Managing the Risk of Extreme Weather Events in a
Changing Climate
First Wednesday Seminar
February 3, 2010

Also consider the video at

Hank Roberts said...

> The 1936 North American heat wave

Hmmmmmmm. Anyone done numbers on emission of and forcing by sulfate aerosols (from coal) from the boom during the Roaring '20s through the global economic Great Depression years? Pretty minor change compared to current fuel use, but I wonder if anyone's looked for a change in forcing and precipitation associated with a drop in coal use in that time span.

Steve Bloom said...

Record extreme rain in Indonesia in the dry season:

Indonesia has been experiencing its most extreme weather conditions in recorded history, meteorologists warned on Wednesday as torrential rains continued to pound the capital.

All regions across the archipelago have been experiencing abnormal and often catastrophic weather, an official from the Meteorology, Climatology and Geophysics Agency (BMKG) said.

“We have reached a super-extreme level of weather this year, the first time in our history, and this is much worse than what we experienced back in 1998, when the La Nina caused extreme weather in the country,” Edvin Aldrian warned.

Edvin, who leads the climate change and air quality division at the agency, told the Jakarta Globe that a combination of a heating planet and the La Nina climate cycle were behind the unseasonable downpours.

“The combination of global warming and the La Nina phenomenon makes everything exceed normalcy,” he said, adding that global warming causes higher temperature in sea waters, and La Nina boosts humidity and the likeliness of rains.

Sea temperatures, Edvin said, were also at a level considered normal for Indonesia’s rainy season, not for the dry season. “It is about 28 to 29 [degrees] Celsius now. Normally, for August it should have been around 24 to 26 degrees.”

Generally at this time of year, Indonesia is supposed to be in the midst of the dry season and entering the transition to wetter months.

Indonesia is pretty far from Pakistan, so I assume there's no direct connection, although the duration makes it sound like the same sort of blocking event.

Rattus Norvegicus said...

I guess the question that needs to be asked is "how many thirteens do you need to roll in one year?".

Nashville was a 13, Russia is a 13, Pakistan is probably a 13, China might be a 13. Something odd is happening.

Steve Scolnik said...

Nashville/Middle Tennessee has been hit twice in the same year. That makes it at least a 13.

Alexander Ac said...

Hi Michael, have you seen this paper?

Drought-Induced Reduction in Global Terrestrial Net Primary Production from 2000 Through 2009

regarding this PM picture: it is for July - but we know the heatwave continued well into august..

and I agree with Eli, that pointing to high natural variability in the past is not very good argument - since there is a suggestion warming is and will exacerbate natural wariability

mybach62 said...

ghd straighteners
ghd hair straighteners
ghd australia
ghd uk
pink hair straighteners
pink hair straightener
pink straighteners

Vinny Burgoo said...

Does anyone know whether it's appropriate to multiply the Wiedenmann index of block intensity by the duration of blocking events, then sum the annual products? Am I double-counting in some (easily avoidable) way? I'm looking for a rough measure of trends in the severity of summer Eurasian blocking events.

NewYork said...

Steve Bloom beat me to it. Never take anything Pat Michaels says at face value. His graph does not match the actual data.

July 1936 vs 1951-1980 baseline

compare to:

July 2010 vs 1951-1980 baseline

2010 vs 1936 diff:

July 2010 vs July 1936

Create your own graphs here

Using the correct maps, Michaels actually makes the point for climate science - that under a similar configuration, global warming is making such extreme events considerably more intense. What was a dice value of 12 in July 1936 is now a 13 or a 14.

Now whether or not global warming is leading to more blocking events and more extreme configurations is an open question.

Steve Bloom said...

Thanks, NY. Also note that upthread Chip Knappenberger (Wormtongue to PM's Saruman, or should that be Eyegor to Frahnkensteen?) took credit for the post.