18 October 2009

Question Place

About time again for one of these. Questions, suggestions, ideas, and so forth.


Jim said...

With a moderate-to-strong El Niño predicted to for this winter, this nagging question returns to my mind: Does an El Niño event "delay" the global warming trend?

During an El Niño event, ocean heat outcrops along the West Coast of South America. This is stored heat that's been accumulating for a long time, and if I understand correctly, much of this stored energy gets dumped into the atmosphere over a short period (several months). From there, it radiates/convects into space.

My first thought is that El Niño is an internal oscillation that doesn't affect the total energy budget, and so it can't "delay" warming. But then I thought that the ocean-atmosphere system will respond to this pulse by ringing is some nonlinear way, and during this transient response, the warming trend would be "delayed."

Maybe the question is really: what is the characteristic time constant of the ocean-atmosphere system?

Robert Grumbine said...

Some interesting things involved here.

To back up a little, the El Nino is a sort-of periodic event. It peaks every 3-8 years, give or take. The trough (cold period) is called La Nina.

One way of looking at the cycle is that there is a sloshing across the Equatorial Pacific Ocean. In the Western Pacific, the waters are always quite warm (the Pacific Warm Pool). As we reach the peak of an El Nino, that pool of warm water sloshes to the east -- burying the colder waters that are normally alone the Eastern Equatorial Pacific. At this point, as you suggest, the greatly increased area of warm water gets to interact with the atmosphere. We see some mean global warming in El Nino periods as compared to La Nina. For this part of the story, the time scale (I wouldn't call it a 'constant') is the 3-8 years.

But that expansion of the warm waters to the east also affects weather in some other parts of the world. This comes out because of the changes to the cloud patterns and atmospheric circulation. The time scale for this response is a relatively rapid few weeks to months. Namely, this year's El Nino will affect weather for this year and a bit in to next year.

Between the two, for El Nino-related weather, we can say that the time scales are weeks to months to a few years.

In terms of trends, El Nino (at least by itself) doesn't change the trends. It's an oscillation, so what it does is re-distribute the energy. After sinking heat in to the warm pool, El Nino releases it. But if there were no trend in the energy being buried in the warm pool, there would be no trend elsewhere related to El Nino.

Reminds me that I have a short note to be finishing on ENSO and Global Mean Temperature. (ENSO is the full cycle -- El Nino/Southern Oscillation).

wag said...

Apparently a WUWT post, about Mann's hockey stick being "upside down," was the 2nd most viewed piece on Wordpress a couple days ago:


I know anything Watts says is wrong, but felt that if this one had been viewed so many times, it needs a proper response. Any help?

Jesús R. said...

Skeptics seem to think that there is a contradiction between the accelerated decrease in mass balance in Antarctica and the decrease in summer snowmelt.

In a blog post dealing with the first paper, a commenter asked if declining recorded snow melt could be due to increased precipitation due to warmer waters, and the blog owner suggested that you were the right person to ask.

I would also like some perspective on how both papers are mutually related.


Robert Grumbine said...

(by the way, how do I get the accent over the u for your name? and a tilda over n for in El Nino?)

I wouldn't call such people 'skeptics'. Skeptic is an honorable, very hard work, position. It involves, for instance, being skeptical of all information sources. And being even more skeptical of sources that reach conclusions you'd like. Most places that claim to be 'skeptical' are exceedingly credulous -- any source that reaches a conclusion they like must be right, and any source that reaches a conclusion they don't like must be wrong.

Arrgh. Sorry, not the point of your question, but a long-standing peeve of mine.

The real question you raise is interesting, and not trivial. So I'm addressing it in a post to appear tomorrow morning (21 October). Might be a lot more than you were looking for. But, if it's less, followup questions are welcome!

Jim said...

Jesus, I don't know if this helps, but here's the elevation/ice thickness data for Antarctica:

Graph of the Day: Rate of Change of Surface Elevation for Antarctica and Greenland, 2003-2007.

You can compare to Figure 2 in Tedesco.

Jim said...

Thanks for the thoughts. Looking forward to your ENSO post!

thingsbreak said...


Hold the Alt key and type 164 for the ñ, Alt and 0176 for the degree symbol °, and Alt and 0250 for the ú.

Robert Grumbine said...

Thanks thingsbreak. I was hoping to avoid learning character codes. Oh well. Can't always get what I want.

Jesús R. said...

To avoid ASCII codes (163, 164 / 0241, 0250) I guess that copy & paste should work (e.g. from my own comment for Jesús, or from a bookmarked page like this one). As for me, there’s no problem if you write my name without the accent ;-) It’s very similar and I understand that you don’t have the accent at hand in an English keyboard :) Thanks for your consideration!

I share your view regarding skepticism. When the skepticism is blindly biased only one-way, that’s not skepticism, that’s prejudice, that’s denialism.

Thanks, Jim! It's curious, because Tedesco shows negative melting anomalies in almost all the perimeter while Pritchard shows negative elevation trends. However, Tedesco shows 2009 anomalies, while Pritchard shows a (4-year) trend. If they are comparable, I suppose that that might possible point at losing mass through basal melting in the ice sleves (see the new post).

Anonymous said...


Can you suggest a good source for a narrative of the Vostok Ice core chart? Something that might show possible events for the carbon rise and cooling periods would be great. I am responding to a denier.

Jesús R. said...

Anonymous / Robert:

Ice Ages were triggered by orbital variations known as the Milankovitch cycles (see here, here or here). The initial warming melts ice, thus reducing the Earth's albedo and thus causing further warming (positive feedback). Ice melting also releases greenhouse gases that were trapped in the ice, which, in turn, cause more warming (positive feedback). Positive feedbacks have a gain factor (<1); given that they cause further warming, they cause a bit more positive feedback. See this for more information or this for a different event).

The "peak" in the Milankovitch cycles was also the cause of the warming peak about 6,000 years ago (known as the Alti-thermal or Climatic Optimum) and of the long-term cooling trend since then (stopped we began to emit greenhouse gases).

Milankovitch cycles are predictable and the most recent work says that (without the human influence) the next ice age would take place in about 50.000 years.


carrot eater said...

I would encourage a post about ENSO, if you're already working on one.

In it, could you expand on your note above on how unforced variability can cause changes in mean surface global temp through redistribution of heat below the surface, with graphics of the heat flow? Also, could there be effects on clouds, which may actually change the radiation balance?

If you have time, of course. I just think it'd be good to clearly show that ENSO can cause a warm year like 1998, but that ocean cycles couldn't have caused the observed 30 years of warming.

Anonymous said...


I'm in a rebuttal/counter-rebuttal debate online with an AGW Denialist who's engaging in the usual routine of throwing out reams of links to denialist papers and papers that can be interpreted to back-up the denialist view.

One particular paper splurge came from Ipso's 'CO2 Science' blog. Ipso claims his referenced papers show that the Medieval Warm Period was a global event triggered by some common cause.

I'm not au fait enough with paleoclimatology to really nail these claims to to wall - but I suspect that 'Medieval Warm Period' is being used as a catch-all phrase to cover any warm period over a certain historical time range. My own understanding of the MWP was that it was essentially a European thing.

Does that gel with your understanding of the MWP Robert? Would you have any good references/posts out there that would enable me to rebut this denialist line?

Jesús R. said...


AFAIK, the MWP was a northern hemisphere phenomenon, mainly located in the North Atlantic region.

You can start here:

If they are quoting the paper by Loehle, you can read this:

That paper was published in the infamous Energy and Environment:

You can also read these comments, where Kate was having a similar problem:

crandles said...

Probably a silly question but why are ENSO indexes done as anomaly from a 30 year average rather than anomaly from a 30 year trend?

(Ok it won't matter in many cases eg if you are doing multiple regression and one of your other factors is date.)

Would the El Nino and La Nina periods be less likely to change when you change the base period if done against a 30 year trend than if done against a 30 year average? Would/Could such an index be more reliable at explaining various relationships?

Of course the particular 30 year trend to use might make differences eg should it be a linear trend or a nonlinear trend that accounts for expected effects from other known forcings.

Robert Grumbine said...

Catching up some:

I haven't gotten the coding to work for me. Is this a windows-specific trick? I'm on Mac and Linux.

Jesús mentioned some good sources. A more general book, whose author I like, is The Two Mile Time Machine by Richard Alley. Richard is one of the scientists involved in using ice cores to reconstruct climate. This book is his writing for non-professionals on the subject.

I'll be getting to the El Nino note sooner, I guess. That requires some homework on my part as the equatorial Pacific hasn't been a major concern of mine.

I think you mean Idso, not Ipso. A link to what paper you mean would be a plus. It's possible that they have a some real science.

My impression of how Medieval Warm Period is used at places like the Idsos' is that any warm weather, anywhere, in certain time range (and a flexible one at that) is called proof of Medieval Warm Period. Unfortunately, such places seem to ignore the cooler weather that happens in some parts of the world within the same span. An early serious look at the question is the Lamb Climate: Past, Present, and Future I mentioned earlier. It was published in 1972. The data were much less extensive then, so I excuse Lamb reaching the rather strong conclusions he did -- there was reason to think the other data would, as it was found, be in line with what he had found. We now know the MWP was, at minimum, a complex time.

For back and forths, let me suggest something. For the sort of people you're talking to the science on the MWP is irrelevant to their actual point. The real point they are typically trying to make is something in the line of 'climate changed before without human activity, therefore climate change now cannot be because of human activity'.

And this is simply logically false, as well as scientifically false. There's the well known reality, for either science or logic, that more than 1 thing can cause a result. For instance, usually when I can't find my keys, it's because I forgot where I put them. According to the sort of argument I'm referencing, I would have to conclude that if I can't find my keys today, that I must, again, have forgotten where I put them. That's obviously false. It could be that, as has happened, someone else moved my keys this time. More than one (logically or scientifically) possible explanation for a given observation.

Jesús gives some good sources. But his 'if they mean this' comments show why it is I'd like to see the exact source your correspondent is using. I don't like guessing.

Robert Grumbine said...

I think you have the method wrong for SOI (Southern Oscillation Index) computation. But there are several, and I'd like to check out exactly how they are computed.

One key, for the ones I'm familiar with, is that they don't use an absolute number. The original (which goes back to the early 20th century) was the atmospheric pressure difference between Darwin and Tahiti. If there is a general trend in pressure, that cancels out since both places feel the trend.

But, as I said, there are several methods. Could you provide a link to the one(s) you mean?

Robert Grumbine said...

Interesting paper. Thank you for the link. It's also good to see the review comments openly. Standard sorts of comment, so maybe something I can take up for a note.

I think you have much more room for writing than you took advantage of :-) There's an article titled something like 'what if surgery were done like wikipedia'.

Anon - GaryB:
I haven't pursued all articles, or many, on alternate uses of the sedimentary data. I'd be surprised if Mann, et al., ran with the data set inverted. Why would they? Either they used the data correctly, so there's no need. Or they didn't, in which case there's more to be done than just invert that one data set. As far as I did see, none of the critics of Mann and others have run the Mann analysis with the inversion.

Mann and others considered removing not just this questionable (for their purpose) data set but also removing some others. The results, I'll suggest you check out the graphs for. To my eyes, the result is that change in the reconstructions between how they used the data to not using the data was too small to notice. Going to inverting the data usage would be twice as large as too small to notice. So it's either also too small to notice, or only barely noticeable. Your concern is actually already addressed.

Yes, I'm the Grumbine from talk.origins. I'm not, however, the Grumbine from misc.invest.* !

My time in talk.origins has given me a very low tolerance for what you described as 'simple overstatement to make a point'. The overstated conclusion (what you agree is overstated) is one that many WUWT readers seem to believe is entirely correct. The 'overstatement' contributes to people having a false understanding. If the interest were science, using the sediment cores wrongly is more than serious enough a charge. It's still likely overstated to be going to plots and 'purposely selective', but as least there might be evidence presentable for that case.

crandles said...

OK, I should have been more specific. The index I was looking at was NOAA's ONI:

"Oceanic Niño Index (ONI)
•The ONI is based on SST departures from average in the Niño 3.4 region, and is a principal measure for monitoring, assessing, and predicting ENSO.•Defined as the three-month running-mean SST departures in the Niño 3.4 region. Departures are based on a set of improved homogeneous historical SST analyses (Extended Reconstructed SST –ERSST.v3b). The SST reconstruction methodology is described in Smith et al., 2008, J. Climate, vol. 21, 2283-2296.)•Used to place current events into a historical perspective•NOAA’s operational definitions of El Niño and La Niña are keyed to the ONI index."

as per page 20 of http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/lanina/enso_evolution-status-fcsts-web.pdf

There appears to be a trend in those figures over 50 years and I suppose if you want to detrend the data that is fairly easy to do.

(Note to self - don't assume other indexes do similar things just because the one I am looking at does something.)

jyyh said...

I'd be interested of an article that would track the possible relation of the speed and volume of the North Atlantic drift (Gulf stream) to the extent of the Arctic sea ice, since I'm presuming the downwelling vortices in the Arctic are mostly present under the ice. Does the Gulf stream 'kickstart' when enough ice has formed over there? In northern Europe there has been abnormally cold months right before winters in recent years but the cold period has ended once the winter should set in, and the pattern of arctic amplification prevails over much of the winter (giving relatively high temps), the same has sometimes happened to the opposite direction in the spring, a cold period after the onset of the arctic melt which could be linked to the fresh water melting and subsequent potential weakening of the Gulf stream.

Anonymous said...

Hi Robert,

The sources are papers like Zhang De'Er's "Evidence for the existence of the medieval warm period in China"


"Winter half-year temperature reconstruction for the middle and lower reaches of the Yellow River and Yangtze River, China, during the past 2000 years" By Ge et al


"Climatic changes during the past 1300 years as deduced from the sediments of Lake Nakatsuna, central Japan" by Adhikari and Kumon


There's more of the same at the bottom of this page, where I was trying to engage some denialists (rather unsuccessfully sadly):


The papers I referenced above either make a direct reference to a MWP in Asia, or have data in the right range. The papers themselves are hidden behind paywalls so any subtleties in the researchers' interpretations are hidden from me.

If you want me to post more, let me know - but I think those three are probably a good representation of the types of papers that my denialist correspondent posted.

Eamon (AKA Yea-mon)

(Sorry for the anon posting - OpenID reported an error when I tried to post using my LJ Identity.)

Anonymous said...

Thanks for the links Jesús. Most of them would be dismissed out of hand by the person I'm having the discussion with. He's one of those "The Hockey Stick is broken!" guys.

However, there was a gem on the NOAA Page showing how out-of-phase regional temperatures were historically:


Eamon (AKA Yea-mon)

Jesús R. said...

Eamon (AKA Yea-mon),

My suggestion would be moving the debate to the global scale.

That figure in NOAA's sote comes from IPCC 2007:

"In medieval times, as now, climate was unlikely to have changed in the same direction, or by the same magnitude, everywhere (Box 6.4, Figure 1). At some times, some regions may have experienced even warmer conditions than those that prevailed throughout the 20th century (e.g., see Bradley et al., 2003a). Regionally restricted evidence by itself, especially when the dating is imprecise, is of little practical relevance to the question of whether climate in medieval times was globally as warm or warmer than today. Local climate variations can be dominated by internal climate variability, often the result of the redistribution of heat by regional climate processes. Only very large-scale climate averages can be expected to refl ect global forcings over recent millennia (Mann and Jones, 2003; Goosse et al., 2005a)."
http://www.ipcc.ch/pdf/assessment-report/ar4/wg1/ar4-wg1-chapter6.pdf (p. 468)

IPCC 2001 said:
"Regional evidence for medieval warmth elsewhere in the Northern Hemisphere is so variable that eastern, yet not western, China appears to have been warm by 20th century standards from the 9th to 13th centuries. The 12th and 14th centuries appear to have been mainly cold in China (Wang et al., 1998a,b; Wang and Gong, 2000)."

It's not easy to discuss when you are outnumbered in the enemy's battlefield! Too many statements to check! And, generally, when proved wrong, they just jump from one argument to another. Good luck!

Polaari said...

Dr. Grumbine,

I wish to bring to your attention this fascinating paper by a compatriot:

Geomagnetic activity and polar surface air temperature variability

And the related dissertation:

Observations of production and transport of NOx formed by energetic particle precipitation in the polar night atmosphere

I have hard time grasping the essence in all of the above.

Robert Grumbine said...

Folks: an ENSO post is in progress. I originally planned it for today, but want to do more fact checking before putting it up.

Interesting papers. To the extent they're in english (the one is, the other seems only to have its abstract in english), I'll be taking a look.

The issue they address is an interesting one of long standing. The obvious thing about the atmosphere is that the troposphere is what matters for weather. 90% or more of the mass of the atmosphere is in the troposphere, and everything that rains on us, etc., is in the troposphere. Nevertheless, there have been observations back at least 30 years that -- somehow -- stratospheric processes get translated in to tropospheric effects. There have been many suggestions as to how it happens. This isn't my area, so I may have missed something important. But it seems to me that, so far, none of the explanations has collected general support.

Hank Roberts said...

CO2science is at it again:

Climateshifts is a reliably interesting science blog that keeps getting better.