The following figure shows the AMO+PDO (black line above changed to red below) superimposed on the Arctic average annual temperature shown at the beginning of this document.
Note: I tried to provide an excerpt for readers of this website Appinsys in the post there about Arctic cycles, but gave up. The website is written in MS-Word HTML export which is quite frankly the worst possible way to publish a website. The amount of garbage code it creates that makes it impractical for sharing and pretty much ruins the effectiveness of the website for others that want to reference it with excerpts. Trying to paste even short excerpts into WUWT’s WordPress publisher caused massive visual entropy. So, all I can manage is this sentence and image above.
I hope he’ll take a cue from this and use a real publishing platform (WordPress, Blogger, Typepad, anything but MS-Word) designed for the web so we can help spread the word more often. Good works shouldn’t be saddled by bad web publishing systems.
Here’s the link – http://www.appinsys.com/GlobalWarming/ArcticCycles.htm
UPDATE: Well, maybe not so good after all. Bob Tisdale writes in comments:
The AMO+PDO graph strikes again. It is a prime example of the adage “correlation does not mean causation”, because the AMO+PDO graph is meaningless. You can’t add the AMO and the PDO. I’ll cut and paste a comment I made on an earlier thread here at WUWT to save myself some time.
Unfortunately, the PDO and AMO are not similar datasets and cannot be added or averaged. The AMO is created by detrending North Atlantic SST anomalies, while the PDO is the product of a principal component analysis North Pacific SST anomalies, north of 20N. Basically, the PDO represents the pattern of the North Pacific SST anomalies that are similar to those created by El Niño and La Niña events. If one were to detrend the SST anomalies of the North Pacific, north of 20N, and compare it to the PDO, the two curves (smoothed with a 121-month filter) appear to be inversely related:
http://i52.tinypic.com/fvi92b.jpg
Thanks Bob for teaching us all something. Not being an ocean data specialist, I wasn’t aware in the difference in datasets. Hopefully this exposure of this issue here will prompt wider understanding that while they seem similar, you can’t appropriately combine the two datasets – Anthony
Feet2theFire says: “In other words, the heat energy that goes into the El Niño apparently comes by taking that heat from the northern Pacific (by what mechanism I do not speculate), and vice versa for La Niña.”
The Pacific Warm Pool furnishes the warm water for an El Nino. Some of that supply of warm water in the Pacific Warm Pool is leftover subsurface warm water from the preceding El Nino that has been returned to the west via Rossby Waves at about 10N and 10S during the La Nina. Add to that, some of the warm water supply comes from the preceding El Nino when the leftover warm surface waters are pushed back to the west by the strong trade winds of the La Nina. And some of the warm water supply is created during the La Nina by the stronger-than-normal trade winds reducing cloud cover and allowing Downward Shortwave Radiation to warm the waters.
You said, “Maybe I am doing what Bob says we shouldn’t do – but I am pretty sure there is SOME link between PDO and ENSO, being in the same ocean and all.”
There is, but the PDO pattern is an aftereffect of ENSO. The warm water in the northeastern Pacific (an indicator of a positive PDO) is created by changes in atmospheric circulation caused by an El Nino and through coastally trapped Kelvin Waves. As described above, that warm water is returned to the west during the La Nina. Some of the leftover warm water gets spun up into Kuroshio-Oyashio Extension east of Japan. That creates the warming of the western and central North Pacific that’s associated with the negative PDO pattern. And of course, the northeast Pacific cools during the La Nina. The PDO pattern is also affected by the Sea Level Pressure of the North Pacific, which alters how ENSO events impact the Sea Surface Temperatures there.
Does the leftover warm water from the El Nino that was spun up into the northwest and central North Pacific during the La Nina eventually feed back to the tropics? Yup. The North Pacific gyre would assure that, and there’s also mid-depth meridional overturning circulation in the North Pacific that would return some of the warm water to the tropics via subsurface routes. But those are not functions of the PDO.
You wrote, “That has always been my big question: El Niño is not a CAUSE, it is an effect, so where does all that energy derive from? I haven’t seen anyone address this question yet.”
ENSO is the cause. I’ve addressed it in posts at my blog for a couple of years. Many have been cross posted here at WUWT.
Personally, I think we’d be better off presenting the PDO the same way as the AMO, and that is by detrending North Pacific SST anomalies north of 20N. We should be more interested in the temperature of the North Pacific than in the pattern (appearance) of the temperature anomalies there. The PDO is useful, but not as an indicator of the sea surface temperature of the North Pacific.
I think we’re dealing with this all wrong.
Bob points to an almost inverse relationship between the heat distribution in the North Pacific and the North Atlantic, this I believe is the essence of what’s going on with the cycles that are roughly 70 years in length (35 up, 35 down give and take).
A fairly simple mechanism can explain at least some of this behaviour:
Land heat distrubution is fairly seasonal (because it loses heat fast), so is the ocean, but the ocean stores a lot of energy, which gives us cycles of heat distrubution, which are not primarily seasonal.
I suggest that the distribution of heat in the ocean determines polar cell meandering especially in winter. I also suggest that lower and higher pressures are pushed across the Arctic in response to this ocean heat distrubution, because the massive lows over the ocean expands the atmosphere pushing surrounding air masses away. Such shifts in atmospheric behavior determined by ocean heat distrubution works back on the ocean via the atmosphere to again change the heat distribution.
In essence, what we should be contemplating is a giant shuffling mach in which low pressure systems expand and moves the polar cell and prevents the vortex from transporting high level air to the surface in one region, and as a consequence forcing it to strike down in another region. The arctic dicotomy, I call it.
For instance: if there is massive cyclonic activity in the North Pacific and these cyclones wander into Northern Canada all the way to Hudson bay and the disco bay, then if cyclone activity is weak in the Barents sea and the norwegian sea, the vortex will more readilly touch ground in these areas: Iceland and Scandinavia and Siberia producing blocking highs. This will change conditions and on average change the oceanic flow. More blocking over the Scandinavian Icelandic region will push more warm water from the Golf Stream into the Labrador Sea, Disco Bay, Hudson bay and the Canadian Archipilago.
I think the rhythm of the sun has much to do with these changes, and it is even possible, that these flow changes in the ocean can agrevate vulcanic eruptions and earth quakes. Might even modulate to some small extent the magnetic field of the earth itself.
Sorry for the poor quality of this comment, but I’m in a hurry. :-/
Per
Art Ford says: “Possibly Joe can add some explanation here as to what he actually combined. ”
I asked Joe D’Aleo about this a couple of years ago and he was kind enough to email me a copy of his spreadsheet. He standarized the AMO data from the ESRL website, and added it to the PDO data from JISAO, which is already standardized. But the AMO data is North Atlantic SST anomaly data that has been detrended, and the PDO data is the first principal component of detrended North Pacific SST anomalies. One represents Sea Surface Temperature anomalies and the other represents the pattern (appearance) of the Sea Surface Temperature anomalies.
Alan S. Blue says: “IOW: Try AMO minus scaled-North-Pacific-via-SST.”
Since the intent is to illustrate a naturally caused, sea surface temperature-based explanation for Arctic land surface temperature anomalies, why not simply start with the actual SST anomalies of the North Pacific and North Atlantic, north of a specific latitude? SST data is readily available through the KNMI Climate Explorer, and so is Land Surface Temperature data. The KNMI Climate Explorer even creates correlation maps, so anyone could start with the Arctic Land Surface Temperatures and see which areas of the North Pacific and North Atlantic actually correlate with them.
Per says: “Bob points to an almost inverse relationship between the heat distribution in the North Pacific and the North Atlantic…”
I didn’t write that. I showed the PDO is inversely related to the North Pacific SST anomalies, north of 20N. Detrended North Pacific SST anomalies for the area north of 20N run in and out of synch with the AMO:
http://i56.tinypic.com/t9zhua.jpg
That graph is the last one in this post:
http://bobtisdale.wordpress.com/2010/09/03/an-introduction-to-enso-amo-and-pdo-part-3/
Juraj V. says: “There is one simple method, how to identify the greenhouse fingerprint in the Arctic; winters should warm much more than summers.
“http://i56.tinypic.com/vfv70g.jpg
“It is not the case.”
Juraj V., your graph appears to contradict you. The Dec-Feb anomalies appear to have risen more than Jun-Aug anomalies. Maybe a graph that compares the two would help confirm or contradict your statements.
For Bob Tisdale:
“Detrending” is being used a lot in this post. To “detrend” a monthly observation dataset, does one do the following?
1. Add up all the monthly observations.
2. Determine the average from the sum in #1.
3. Subtract the monthly observations from the average determined in #2.
4. Plot the differences determined in #3 and that provides a “detrended” curve.
If the above is incorrect, Bob, could you provide a simple explanation of how one detrends a monthly dataset? Anyone?
Is “detrending” another way of saying “normalizing” or “standardizing”? Anyone or Bob.
December anomalies and trend since 1928.
Francis White says: May 22, 2011 at 7:05 pm
…………..
Thanks for the nomenclature note. For the detrending I subtract linear trend line from the data sequence, while for rate of change over fixed period of time (delta T) I use term ‘first differential’. Two produce somewhat different result, depending on the delta T chosen. Here is an example:
http://www.vukcevic.talktalk.net/PDO-ENSO.htm
(see the lower graph)
Bob Tisdale wrote (May 23, 2011 at 2:37 am)
“The Pacific Warm Pool furnishes the warm water for an El Nino.”
“And of course, the northeast Pacific cools during the La Nina.”
There are times when generalities get 180 degrees out of phase. (Linear methods such as cross-correlation, of course, overlook these.)
–
Bob Tisdale wrote (May 23, 2011 at 2:37 am)
“Personally, I think we’d be better off presenting the PDO the same way as the AMO, and that is by detrending North Pacific SST anomalies north of 20N. We should be more interested in the temperature of the North Pacific than in the pattern (appearance) of the temperature anomalies there. The PDO is useful, but not as an indicator of the sea surface temperature of the North Pacific.”
Basically I support Bob here with 2 notable exceptions:
1) In general (I can think of a few exceptions such as reaching a lay audience with easy-to-follow illustrations), nevermind linear detrending.
2) Measures of spatial pattern have utility, so instead of suggesting what we should be “more interested in” (e.g. PMO) I would rather caution that we should be “more careful about how we interpret summaries” (PDO vs. PMO or whatever).
Nandie: How about an illustration to answer your questions about detrending? It’s relatively simple:
http://i37.tinypic.com/14ln95l.jpg
The graph is from the introduction to the AMO:
http://bobtisdale.wordpress.com/2010/08/16/an-introduction-to-enso-amo-and-pdo-part-2/
Bob Tisdale@at 3:09 am: “Since the intent is to illustrate a naturally caused, sea surface temperature-based explanation for Arctic land surface temperature anomalies”
That wouldn’t be my goal. My goal would be to predict GMST a year in advance with available data. (Yes, if I’m limited to the data presented, I might restrict myself to the Arctic temperatures.)
This is the core difference between attempting to “find a correlation/explanation” and trying to “find an empirical model”.
R. Gates says:
May 22, 2011 at 12:21 pm
If it doesn’t, and we see 2010-2019 as warmer than 2000-2009, and then 2020-2029 as warmer on avergage than 2010-2019, and the arctic sea ice continues its long-term downward trend, what then my skeptical friends?
If all this is observed, we still need some actual evidence that CO2 is the cause. We need evidence from people that are open to FOI requests and aren’t trying to push an agenda. We need science journals that aren’t taken over for political purposes. We need science journals that require release of data. We need honest science rather than fraudulant piecing together of a graph from disparate sources to defend sketchy proxies. We need a lot more evidence that includes better understanding of clouds, oceans, and a whole lot more. We need climate models that get back casts of rainfall correct as well as temperatures. If CO2 is the cause, then show it and don’t hide behind anything. That’s what we need.
There’s nothing wrong with mixing PDO & AMO in an empirical model – someone with a good handle on spatiotemporal dynamics might even successfully defend such a model on physical grounds – but there is something FUNDAMENTALLY WRONG with perceiving PDO as analogous to AMO (which far too many erroneously do).
Solar max interrupts the semi-annual heat pump. The frequency of pump outages controls multidecadal oscillations (via hydrology). Interannual spatiotemporal chaos makes this difficult or impossible to see using LINEAR methods.
Bob Tisdale says:
May 23, 2011 at 2:37 am
Personally, I think we’d be better off presenting the PDO the same way as the AMO, and that is by detrending North Pacific SST anomalies north of 20N. We should be more interested in the temperature of the North Pacific than in the pattern (appearance) of the temperature anomalies there. The PDO is useful, but not as an indicator of the sea surface temperature of the North Pacific.
====================
That makes so much common sense it hurts. That way you have apples to apples.
Chris
Norfolk, VA, USA
Isn’t the issue then how you CAN correlate relevant inter-decadal temperature indices (which both PDO and AMO are) with actual temperature graphs?
Because presumably the point of doing so is saying ‘there is some component of the future in the past data’ (which presumably is an indication of oceanic heat content absorbed in the past feeding into future temperature)?
I’ll leave the technical decisions to the experts, but if my logic is wrong, will someone please explain why and what logic would be better?
@Rhys Jaggar May 23, 2011 at 11:06 pm
Start by asking:
What is the difference between PDO & North Pacific SST?
To R.Gates,
I notice you are putting a heck of a lot of store in the Artic ice levels as they stand today. In fact you go so far as to say that such low levels have not been seen for ‘a long time’.
But…thats just not true is it, be honest.
The Artic ice levels in the 20s/30s was just as low as it is today……this is evidenced by the many photographs and reports at the time. You cannot have missed the ‘disapearing island’ stories or the ice free shipping routes from that era…for sure it was hardly the Caribbean, and it was only a few ships for a short time of the year, but overall, a 20-30yr period where ice cover was most likely less than it is today…because today we cannot even achieve those shipping routes, and that inconveniant little island is still today joined by an ice bridge.
If you can agree on that, then your arguement ref CO2 stands on its head….increasing CO2 levels have failed (all other inputs being static…and after all we are ‘told’ it cannot be Enso/GCR/TSI by you guys..heh!) to melt the ice any further than it was 70yrs ago.
Same arguement goes for Manns hockey stick….if we agree, as seems likely now (do you accept this?) that the MWP was at least as warm as it is today AND globaly, then increases of CO2 by 30% since that time have failed totally to melt the ice or warm the earth at all.
Which means…….in ain’t CO2 mate…
Or….there is certainly no postive feedback from CO2.
I find your arguements on the artic rather self defeating. But then…I guess every other part of the theory has not played out – no tropho hotspot/kinda kills the reasoning for the strato cooling/no real sea rise increase rate. Al there is left is the Artic to hang the hopes on.
We could of course have several hundred years of very very low, ice free waters in the Artic from now on, just like during the MWP (which..to labour the point…had nothing to do with CO2..) and it would prove absolutely zilch as far as causation is concerned.
Do you not think you are hanging too much on sea ice?