I’ve moved to WordPress. This post can now be found at Sea Surface Temperature Anomalies – East Pacific Versus The Rest Of The World
################I will be adding two volcano-adjusted SST anomaly subsets to the monthly updates starting with the February 2011 update. They are the East Pacific, which mimics NINO3.4 SST anomalies, and the Rest of the World, which rises in very clear steps during significant ENSO events.
You may want to bookmark the LINKS TO SST ANOMALY UPDATES for future reference.
INTRODUCTION
I’ve written numerous posts about the upward steps in the Sea Surface Temperature (SST) anomalies of the East Indian-West Pacific Oceans (60S-65, 80E-180). Many of them are linked at the end if this post under the heading of “Further Discussions”. In an upcoming post, which is a continuation of my Comments On Tamino’s AMO Post, I’ll also be illustrating and discussing similar upward steps in the Atlantic Multidecadal Oscillation (AMO) data, which are detrended North Atlantic SST data (0-70N, 80W-0). Refer to Figure 1, which is a .gif animation. As illustrated, ENSO- and Volcano-adjusted AMO data rise in steps during the transitions from El Niño to La Niña, but between the upward steps, they mimic the inverted NINO3.4 SST anomalies.
http://i56.tinypic.com/2repfo0.jpgFigure 1
Similar upward steps can also be illustrated if we divide the global oceans into two subsets so that one of them contains both the North Atlantic and the East Indian-West Pacific datasets. The two subsets, Figure 2, are the East Pacific Ocean (90S-90N, 180W-80W) and the Rest Of The World (90S-90N, 80W-180E), the latter of the two containing the North Atlantic and East Indian-West Pacific datasets. As we shall see, the East Pacific SST anomalies mimic the variations of ENSO proxies, with little trend, while the SST anomalies of the Rest of the World rise in very clear steps that coincide with the 1986/87/88 and 1997/98 El Niño events.
http://i53.tinypic.com/141qcyd.jpgFigure 2
ACCOUNTING FOR THE IMPACTS OF VOLCANIC ERUPTIONS
I’ll be removing the linear effects of the two major volcanic eruptions, El Chichon and Mount Pinatubo, from the two SST datasets. To determine the scaling factor for the volcanic aerosol proxy, I used a linear regression software tool (Analyse-it for Excel) with global SST anomalies as the dependent variable and GISS Stratospheric Aerosol Optical Thickness data (ASCII data) as the independent variable. The scaling factor determined was 1.431. This equals a global SST anomaly impact of approximately 0.2 deg C for the 1991 Mount Pinatubo eruption. Refer to Figure 3. I’ll use that scaling factor for the East Pacific and Rest of the World datasets.
http://i55.tinypic.com/33acpwp.jpgFigure 3
EAST PACIFIC
As shown in Figure 4, the Volcano-adjusted East Pacific SST anomalies vary in concert with the scaled NINO3.4 SST anomalies. There are periodic divergences, but the variations in the East Pacific SST anomalies mimic the commonly used ENSO proxy.
http://i53.tinypic.com/2zi4ig2.jpgFigure 4
The linear trend for the East Pacific SST anomalies since November 1981 is basically flat, at only 0.08 deg C per Century. Refer to Figure 5.
http://i56.tinypic.com/1565ldz.jpgFigure 5
This obviously means the rise in the global SST anomalies since the start if this satellite-based SST dataset must occur outside of the East Pacific.
REST OF THE WORLD
Figure 6 compares the SST anomalies and the linear trends of the East Pacific and the Rest Of the World. Since November 1981, the SST anomalies of the Rest of the World (90S-90N, 80W-180) have risen at a rate of approximately 1.01 deg C per Century, while the trend of the East Pacific SST anomalies is only 0.08 deg C per Century.
http://i55.tinypic.com/2mxh9pf.jpgFigure 6
The East Pacific data used in this post represent approximately 33% of the global ocean surface area. (The percentage is based on the NCEP/DOE Reanalysis-2 “Land Mask” data available through the KNMI Climate Explorer.) So we can place the two datasets in perspective by scaling the East Pacific data by a factor of 0.5. Refer to Figure 7. Notice how flat the Rest of the World SST anomalies are after the 1997/98 upward shift. There are some minor fluctuations, but the Rest of the World SST anomalies are essentially flat until 2009/10. Backing up in time, the same could be said for the period from 1987/88 and 1997/98; the volcano-adjusted Rest of the World data also have not risen over that period.
http://i55.tinypic.com/15hgy0j.jpgFigure 7
Adding period-average values to the Rest Of The World SST anomalies, Figure 8, makes the upward steps stand out even more. It will be interesting to see where the “July 2009 to Present” SST anomaly average settles out, if it does before the next significant El Niño drives them higher.
http://i51.tinypic.com/1zmf3ev.jpgFigure 8
CLOSING NOTE
In a recent discussion at another blog (I believe it was a discussion of the adjusted AMO data in Figure 1.), an AGW proponent noted that upward shifts in SST anomalies did not disprove the hypothesis of anthropogenic global warming.
As illustrated in this post, the SST anomalies of the East Pacific Ocean, or approximately 33% of the surface area of the global oceans, have risen very little since 1982. And between upward shifts, the trends of the SST anomalies for the rest of the world (67% of the global ocean surface area) remain flat. What processes could cause anthropogenic global warming to work only during the significant El Niño events of 1986/87/88, 1997/98 and 2009/10?
SOURCES
SST anomaly data is available through the NOAA NOMADS website:
http://nomad1.ncep.noaa.gov/cgi-bin/pdisp_sst.sh
or:
http://nomad3.ncep.noaa.gov/cgi-bin/pdisp_sst.sh?lite
The GISS Global Stratospheric Aerosol Optical Thickness data is available here:
http://data.giss.nasa.gov/modelforce/strataer/tau_line.txt
FURTHER INFORMATION
My first detailed posts on the multiyear aftereffects of ENSO events are:
Can El Nino Events Explain All of the Global Warming Since 1976? – Part 1
And:
Can El Nino Events Explain All of the Global Warming Since 1976? – Part 2
And:
Supplement To “Can El Nino Events Explain All Of The Warming Since 1976?”
And:
Supplement 2 To “Can El Nino Events Explain All Of The Warming Since 1976?”
And for those who like visual aids, refer to the two videos included in:
La Niña Is Not The Opposite Of El Niño – The Videos.
The impacts of these El Nino events on the North Atlantic are discussed in:
There Are Also El Nino-Induced Step Changes In The North Atlantic
And:
Atlantic Meridional Overturning Circulation Data
As noted earlier, I’ve also written a rebuttal post to Tamino’s AMO Post. I hope to have a new post on the North Atlantic posted in a few weeks.
The posts related to the effects of ENSO on Ocean Heat Content are here:
ENSO Dominates NODC Ocean Heat Content (0-700 Meters) Data
And:
North Atlantic Ocean Heat Content (0-700 Meters) Is Governed By Natural Variables
More detailed technical discussions can be found here:
More Detail On The Multiyear Aftereffects Of ENSO – Part 1 – El Nino Events Warm The Oceans
And:
More Detail On The Multiyear Aftereffects Of ENSO - Part 2 – La Nina Events Recharge The Heat Released By El Nino Events AND...During Major Traditional ENSO Events, Warm Water Is Redistributed Via Ocean Currents.
And:
More Detail On The Multiyear Aftereffects Of ENSO - Part 3 – East Indian & West Pacific Oceans Can Warm In Response To Both El Nino & La Nina Events
26 comments:
"What processes could cause anthropogenic global warming to work only during the significant El Niño events of 1986/87/88, 1997/98 and 2009/10?"
I've always speculated (arm waved!) that the so-called AGW heat in the pipeline gets kicked out (from I don't know where) during El Nino. The Step Up in the rest of the world is delayed by 9-months while the East Pac stands pat. It is possible that these kicks that result in step changes are the result of warming that has occurred several decades in the past. In any event, I have no doubt that the average global air temperature is governed by SST.
The next ten years or so should show another step change upward if the standard IPCC prediction of AGW is true.
Howard: The following animation disagrees with your thought that the "step changes are the result of warming that has occurred several decades in the past."
http://i51.tinypic.com/2qiwscz.jpg
The upward steps are caused by the respective El Nino events.
Hi Bob:
I have a few questions connected with some work I am doing.
1. Do El Niños always cause a step-function increase in SST? As I recollect there was a large El Niño in 1941 or thereabouts that was followed by a temperature decrease, at least in SAT.
2. Do the El Niños/La Niñas have to exceed some kind of a threshold value before the effect occurs?
3. How might the effect work in the opposite sense, i.e. cause cooling?
Apologies if you've already dealt with these questions and I missed it.
Roger
Roger: Sorry I can’t provide more specific answers to your questions:
1)Every El Niño event does not cause the upward steps, as far as I can tell. The lesser El Niño events that are not followed by a La Niña event do not appear to have the same effect.
The 1939-1942 El Niño lasted a long time but it wasn’t too strong:
http://i56.tinypic.com/i3850k.jpg
2)If there is a threshold I haven’t identified it. The 1972/73 El Niño was stronger than the 2009/10 El Niño, yet the 1972/73 El Niño did not cause an upward step while the 2009/10 El Niño appears to be doing so.
And while we’re discussing this, both El Niño Modoki and “traditional” El Niño events can cause them.
3)I have been hoping the AMO would “make a turn”, or that there would be a Pacific Climate shift in the other direction, so that we could see if they’re providing feedback of some kind.
Hi Bob
Thanks, I get that. It was clear from your most excellent post.
However, where does the warmth come from that causes the El Nino events resulting in the step up in world-wide SST and Air temps?
Conceptually, for the step model to work, heat needs to build up over time somewhere not evident in SST or Air temps, then be released to SST via El Nino to cause the step up.
I'm probably all wet, thanks for you interesting blog.
Bob
Thanks for your answers.
Another question is what causes the Niño and Niña events. FWIW I've been messing around superimposing sine curves with the dominant ENSO frequency (+/-7 years) on the sunspot record. Adding the two together gives a plot that doesn't of course match the ENSO Index, but it doesn't look all that different either.
Howard: The fuel (warm water) for an El Nino comes from below the surface (and the surface) of the Western tropical Pacific, an area known as the Pacific Warm Pool.
There doesn't need to be a build-up over years. The warm water for the 1997/98 El Nino built up during the 1995/96 La Nina, resulting from unusually high trade winds in the tropical Pacific. Higher trade wind strength yields less cloud cover, which yields more downward shortwave radiation (visible light) warming the ocean.
Roger: A relaxation of the trade winds starts an El Nino. The next question is, what initiates the relaxation of the trade winds? And the answer to that is unknown. There are many hypotheses, but none stand up to scrutiny. In other words, there are probably many factors that can initiate an El Nino.
The El Nino is the truly anomalous phase, while a La Nina is an exaggeration of "normal" conditions. In fact, there are hypotheses that there only two phases of ENSO (El Nino and other) not the three phases that are usually discussed (El Nino, La Nina, ENSO neutral).
Some researchers consider the formation of a Rossby wave in the far eastern tropical Pacific (at about 10N) as the start of the La Nina. Refer to the westward moving band of elevated sea level at about 10N in the following:
http://www.youtube.com/watch?v=MF5vZErQ6HM
That's the video that started me looking at the SST anomalies of the East Indian and West Pacific Oceans, and at the processes that could cause the upward steps there after an El Nino.
Regards
Bob
"A relaxation of the trade winds starts an El Nino." You've probably dealt with this question somewhere else, but is there in fact a detectable time lapse between the two?
Roger: The trade winds are in effect holding the "pile" of warm water in place in the Pacific Warm Pool. As soon as the trade winds relax, gravity dictates the response.
Reference the relaxation in Trade Wind "pressure" that holds the 'pile' of warm water in place in the Pacific Warm Pool, would imagine that a corresponding "pressure" cell (or two?) at (an)other position(s) on the globe is/are responsible for this hold/release phenom. Since, next to ENSO, the Monsoons are the next biggest observable weather pattern, a close connection would be found in their pattern shifts. (-;Still of a mind that variation in temp/salinity within the global conveyor are #3 in the big picture.) Your work is so interesting. Thank you again!
Hi Bob:
Sorry about the delay in replying but I've been running some checks.
I cross-correlated the monthly Niño 3.4 Index with the monthly NOAA 850mb Trade Wind Indices (http://www.cpc.ncep.noaa.gov/data/indices/) for the West, Central and East Pacific. The correlations were all negative (i.e. less wind = higher Niño 3.4), which fits your hypothesis. I also got the following lags in Niño 3.4 relative to changes in the wind indices:
W. Pacific: 2 months (R=0.61)
C. Pacific: 0-1 month (R=0.77)
E. Pacific: no lag (R=0.48)
These results suggest that El Niños are in fact preceded by a reduction in trade wind strength that begins in the W Pacific and works its way eastwards.
Regards
Roger
Roger: Your question was whether there was a detectable lag between the trade winds relaxing and an El Niño, so I’m not sure why you ran correlations at different lags of the full terms of the datasets. The additional data that has nothing to do with the evolution of the El Niño would skew your results. Comparing the two…
http://i52.tinypic.com/3129zk3.jpg
…they appear to be in lockstep during El Niño development.
Bob:
"Roger: Your question was whether there was a detectable lag between the trade winds relaxing and an El Niño, so I’m not sure why you ran correlations at different lags of the full terms of the datasets. The additional data that has nothing to do with the evolution of the El Niño would skew your results."
It doesn't skew the results in the slightest. The lag between West Pacific winds and the Niño 3.4 index stays reasonably constant at about 2 months whether there's an El Niño or not. The implication is that West Pacific winds aren't correlated just with the evolution of El Niños. They're correlated with the entire Niño Index.
The reason I ran the correlations in the first place was simply that if one is going to postulate that A causes B then it's a good idea to make sure that A starts before B, or at least no later. I have now satisfied myself that this is the case.
In previous posts, you have shown how the running total of the montlhy ENSO since say 1950 produces a curve that closely correlates with SST, or in fact with aggregate world temperature. So my question is not exactly what forces initiate an episode of El Nino, but what is it that accounts for the fact that in the period since the late 1970s, the ENSO episodes have been predominantly El Ninos whose strength and duration seems to "outweigh" the LaNinas during the same period? If this question cannot be answered, then don't AGW modelers simply "answer" it with the steady upslope of CO2 increase?
Bill: You asked, “…what is it that accounts for the fact that in the period since the late 1970s, the ENSO episodes have been predominantly El Ninos whose strength and duration seems to 'outweigh' the LaNinas during the same period?”
There is a multidecadal component to ENSO. This can be seen if you smooth NINO3.4 SST anomalies with a 121-month running-average filter:
http://oi43.tinypic.com/33agh3c.jpg
When I first used the running total of NINO3.4 SST anomalies, I was trying to determine if there were periods when El Niño events outweighed La Niña events, and vice versa.
BTW, the running total of scaled NINO3.4 SST anomalies reproduces the basic shape of global land plus sea surface temperature anomaly curve not only since the 1950s. It does a good job of it since the early 1900s, including the two warming periods and the 1940s to 1970s cooling period:
http://i42.tinypic.com/2zqufzp.jpg
The above graph is from this post:
http://bobtisdale.blogspot.com/2009/01/reproducing-global-temperature.html
Or if you like, you could use global sea surface temperatures for the comparison:
http://i53.tinypic.com/29fcjl2.jpg
And that graph is from this post:
http://bobtisdale.blogspot.com/2010/11/multidecadal-changes-in-sea-surface_17.html
Back to the frequency and magnitude of El Niño events: as illustrated above, they outweighed those of La Niña events during the early warming period also. We can use period average NINO3.4 SST anomalies to show the same thing:
http://i56.tinypic.com/zxmsg8.jpg
Personally, I like the looks of a graph that compares period average NINO3.4 SST anomalies (from the above graph) to global SST anomalies:
http://i55.tinypic.com/33cwt4j.jpg
Thanks, Bob, I recall these charts from your previous posts, and it’s good to have them all together in one file.
But my basic question remains, I can’t shake it. Let’s take your last chart, the period average NINO3.4 SST anomalies from 1910 to 2010. There are three multidecadal components, the plus 0.15, the minus 0.06, and the plus 0.20 degree C. These three steps fit the SST curve.
But how do we know that this entire 10-decade pattern is independent of the smooth increase of CO2 during the same period of time? Is ENSO perhaps some kind of “chaotic rhythm” in the fluid dynamics of the ocean-atmospheric system? Does ENSO simply do its own cosmic thing? Is an unfettered ENSO the underlying causal “drum beat” for multi-decadal SST, with no feedback loop from changes to the atmosphere—either solar-induced or anthropogenic?
To me, if we cannot establish ENSO firmly as a mulitdecadal independent variable, then modeling arithmetic almost automatically makes it a dependent noise. In other words, despite your clearly charted “step structure,” the trend over 100 years is up.
"What processes could cause anthropogenic global warming to work only during the significant El Niño events of 1986/87/88, 1997/98 and 2009/10?"
I work with large data myself, and typically one needs to normalize the data -- and normalization is a wide-open back door for introducing experimenter-driven bias, because one builds one's assumptions into the grand average -- as you are doing in your volcano adjustment. I recommend you halve your adjustment in this and everything -- else your models will dominate the data -- exactly what you don't want.
Bill Hudson asked, "But how do we know that this entire 10-decade pattern is independent of the smooth increase of CO2 during the same period of time?"
No one has proven anything in either direction, but what I can show you is that the linear trend of NINO3.4 SST anomalies is flat since 1900:
http://i56.tinypic.com/2ag0u2u.jpg
data grinder: Thanks for the thoughts, but the coefficients I've presented are toward the low end of the ranges of past research studies. Cutting them in half would put them well outside the range.
Also, with respect to normalization, right or wrong, it is very common for the data to remain in its raw (not normalized) form for basic research like this. Example:
http://www.atmos.colostate.edu/ao/ThompsonPapers/ThompsonWallaceJonesKennedy_JClimate2009.pdf
Regards
Bill Hudson:
PS: Let me attempt to address your concerns in another way. If there was a known anthropogenic component to ENSO, climate studies wouldn't attempt to treat it simply noise and remove it using linear regression.
"What processes could cause anthropogenic global warming to work only during the significant El Niño events of 1986/87/88, 1997/98 and 2009/10?"
In my point about normalization, I was intending to address the above quote. As normalization is a back door to introducting researcher bias, and normalization is needed to handle large events like El Niño, thus the "process" that "could cause" the AGW to ratchet up during the El Niños could be nothing more than the researchers' methodology. This is a general point hailing from what I've seen elsewhere, where researchers publish graphs of their results, not conscious that their graphs are actually showing only their own misapprehension of their source data.
data grinder: I can remove the volcanic adjustment. The upward steps remain. So your argument doesn’t ring true.
http://i56.tinypic.com/icl6jp.jpg
By adjusting for the volcanic aerosols the graph doesn’t require the note about the periods impacted by volcanic eruptions. In other words, it’s a cleaner graph in the post.
Bob, the ElNino triggered "step up" signal in the global SST's is very interesting. I took a look to the areas, where we can identify this signal very clear. It can't be found in the southern hemisphere: http://climexp.knmi.nl/data/ihadisst1_-80-180E_-90-0N_n_su_1980:2011_6month_low-pass_box_6month_low-pass_boxa.png , much better in the northern: http://climexp.knmi.nl/data/ihadisst1_-80-180E_0-90N_n_su_1980:2011_6month_low-pass_boxa.png.
It seems 2 b very clear in the SST's of northern extratropics: http://climexp.knmi.nl/data/ihadisst1_-80-180E_25-60N_n_su_1980:2011_6month_low-pass_boxa.png.
So we should look for the reason of this behaviour in the northern extratropic oceans 80 W...180 0, the western northpacific and atlanic ocean?
dh7fb: In the Pacific, the leftover warm water from a significant El Nino is returned to the Western Pacific during the La Nina. Does the shape of the landmass in the Western tropical Pacific cause more of the leftover warm water to be spun up into the KOE, than into the South Pacific Convergence Zone?
In the Atlantic, we might come to that conclusion, but creating animations that actually show this has proven elusive for me. The tropical Atlantic warms during an El Niño due to the weakening of the trade winds. During the La Niña, the trade winds resume. Where does the warm water go that was created by the El Niño? The South Equatorial Current in the Atlantic splits at the Brazilian coast, with some heading into the Northern Hemisphere. Does the El Niño-warmed water from the Southern Hemisphere merge with the El Niño-warmed water in the Northern Hemisphere and get swept up into the Gulf Stream, to appear higher latitudes at what I’ll call the Gulf Stream Extension?
And to answer your question, “So we should look for the reason of this behaviour in the northern extratropic oceans 80 W...180 0, the western northpacific and atlanic ocean?,” you should find the behavior is most prevalent in the Kuroshio-Oyashio Extension…
http://bobtisdale.blogspot.com/2010/12/enso-related-variations-in-kuroshio.html
…and in the Gulf Stream Extension.
Regards
Bob, thanks for reminding me of your chart showing that a property of the NINO3.4 SST Anomalies since 1900 is that their linear trend is flat.
But the property of the NINO3.4 SST Anomalies of most interest to me is that their running total bears a striking correlation to global temperature changes for the same 110-year period.
Another way of displaying the correlation which you have so well highlighted is the step-structure in the sequence of ENSO events, where we observe three main multi-decadal periods (approx 1910-1944, 1945-1975, and 1976-present).
So an important question is whether the sequential structure of ENSO is itself an independent, natural cause of the observed extent and pattern of global warming over the past century?
Or, because ENSO events originate in the complex fluid dynamics of the ocean-atmosphere system, is their sequential structure a co-dependent result of other long-running forces, perhaps not as yet well understood? Maybe even of the smooth increase of CO2 in the atmosphere?
Why doesn’t the academic climate research community address the sequential structure of ENSO and its correlation to climate change, which you have so aptly demonstrated?
Post a Comment