The Tao of El Nino

Guest Post by Willis Eschenbach

I was wandering through the graphics section of the TAO buoy data this evening. I noted that they have an outstanding animation of the most recent sixty months of tropical sea temperatures and surface heights. Go to their graphics page, click on “Animation”. Then click on “Animate”. When the new window comes up, click on “60 months”, and then click on the play button. Figure 1 shows a couple of screen captures of different points in time, an El Nino in early 2010 and the succeeding La Nina half a year later.

nino nina tao triton temp and dynamic heightFigure 1. 3D section of the Pacific Ocean looking westward alone the equator. Each 3D section covers the area eight degrees north and south of the equator, from 137° East (far end) to 95° West (near end), and down to 500 metres depth. Click on image for larger size.

Let me go get a cup of coffee so y’all can go look at the animation, it is most informative. When you come back, allow me to point out a few things that are happening, and make a testable prediction

The main issue that the animation brings to mind for me is, what happens to all the warm water? In March of 2010, shown in the left hand 3D section, there is a thick layer of warm water covering the entire surface of the tropical Pacific. This is the El Nino condition. Half a year later, it’s gone, and its departure has exposed the cooler water which was previously below the surface.

What has happened is that when the Pacific gets to a certain threshold warmth (other conditions being equal), the rising air from the heated surface waters of the El Nino reinforces and strengthens the eastern trade winds. And these strengthened winds simply blow the warm surface water mass to the west, where it divides and goes towards both poles. This exposes the atmosphere to the cooler waters from below. At some point in this process, the exposure of the cool subsurface waters reduces the rising air over the Pacific. This reduces the trade winds, a neutral condition prevails, and the surface once again begins to warm.

Contrary to some students of the El Nino phenomenon, I do not see these two as separate phenomena. You could think of it as a two-stroke pump. In the first part, the tropical Pacific gets warmer and warmer from the sun. At some point, this starts reinforcing the easterly trades, leading to a high trade wind La Nina situation. This pumps the warm tropical surface water poleward. Finally, the Pacific returns to a neutral condition, and sets the stage for the next cycle.

There have been differences of opinion as to whether warmer conditions lead to more El Ninos, or to more La Ninas. Within the constraint that this is a natural system and that half-cycles are possible, I hold that warmer or colder conditions do not strongly favor either one. I say that the reason for the differences of opinion among students of the phenomena is that the data doesn’t clearly favor one position or the other.

Finally, I say the reason neither position is favored is that by and large, the El Nino and La Nina operate as a pair, a two-stroke pump that kicks in to move the hot water to the poles and expose the cooler water underneath. When the heating of the ocean reaches a certain point, it triggers the winds that power the pump. After the winds of the discharge stroke of the pump have moved the warm water first west and then poleward, the trade winds die down, a neutral condition prevails, the heating resumes, and the cycle starts over.

However, if I am correct this implies that there will be a difference between warmer and cooler times—in warmer times, the pump should cycle more frequently. That is to say, there should be more of both El Ninos and La Ninas during warmer times. In cooler times the Pacific should take longer to warm up, and the pump should cycle less frequently, so there should be less of both.

Now, I have made this prediction without recourse to any detailed information about how the relative frequencies of the El Ninos and La Ninas vary with temperature. If we have enough data to calculate it, this should make a good test of how good a student of the El Nino I actually am …

Testable predictions … don’t leave the lab without them. I’ll report back with my further findings on this question at a future date, unless someone beats me to it, because it’s 1 AM Monday morning and I’m back to pounding nails tomorrow … and what is the night doing?

Outside, a crisp night, the moon being at the zenith at 1 AM means it’s about a day past full. It has a large ring around it signaling the distant approach of a front, the smaller the ring the nearer the storm. A cool wind is sighing in from the ocean and backing to the west, it carries the sound of the Bodega Bay buoy six miles (10 km) inshore to tickle my sailor’s earbones, the sound nags and pulls at me like an outgoing tide. Big waves tonight, the buoy says ten foot swell at 13 seconds. I can just make out the distant susurration of the breakers ceaselessly striving to claw the cliffs and beaches back down to the ocean bottom … aaah, do what you love to do, my friends, skate on the underside of the ice, life is far too short …


PS—My explanation of the El Nino pump is a functional explanation, by which I mean I am focusing on what the El Nino does, what its function is in the larger climate system. The explanation leaves out a host of practical details of the mechanism of how it does it, not discussing Ekman transport and Kelvin waves and the effects of sloshing water in the Pacific basin and the influence of the Madden-Julian Oscillation and the like … so please don’t bust me for leaving them out. I am not ignorant of them by any means, I’m just ignoring them to focus on what I want to understand, which is what the El Nino is doing, and not the exact details of how it does it.


newest oldest most voted
Notify of

Wind from the East moves the water west, not East.

Vince Causey

The question is, how long does it take the waters to flow to the east, and thence poleward? I know these are surface waters, but I heard on some tv doc, that the deep ocean currents take 1,000 years to make a complete circuit of the globe.

Nick Stokes

I’ve been making some animations of recent ENSO region SST AVHRR data here.

Doug Proctor

If the warming threshold is reached by solar heating, it is possible that the threshold will be reached mid-year, initiating an El Nino at a time when it cannot be sustained due to the cooling portion of the year. This would mean that there is a natural buffer or counter to new El Ninos, locking the rate to a minimum period, i.e. a maximum frequency rate.
On this basis, I would look to a stabilization of El Nino-La Nina periodicity as well (perhaps) as an increased frequency.


Third paragraph after the diagram (which I’ve not yet properly deciphered) I was puzzled by the easterly winds driving the warm water /east/. Should this be West? Later you explain further with the western drift turning towards the pole, so I guess that’s OK.
Just wondering!


I had the same problem as Ed_B. Also, aren’t the Easterlies somewhat north and south of the equatorial doldrums? All the same, an extraordinarily interesting post with a movingly lyrical ending.

Willis, love your sailors guide to weather explanations with regard to rings around the moon, thanks for the great post, joe.


What you describe, Willis, is what I have seen during my decades of studying the commodity markets. What I marvel at is how chaotic systems behave similarly. I have modeled these cyclic movements in the markets using the psychology of the participants. George Soros used the word “reflexivity” to describe this, but well before I read his book on this, I called it “knowledge of the fundamentals changes the fundamentals.” We trade that changing *perception* of the fundamentals. There is no psychology to the behavior of what you describe here, but it is just so interesting that things like markets and El Nino/La Nina behave so much the same.
Thanks for the link.

steve salter

The effect of wind on surface water temperature can be seen on a smaller scale. I live on a spit…. large bodies of water on either side. When it is hot and a prevailing wind blows during the day, the water temperature on either side of the spit can vary several degrees.
My reading is that on the windward side, the prevailing wind pushes the very top layer of warm water towards the spit, where it piles up and creates several feet of warm water. On the lee side, that same wind drags the top layer away from the spit, and since water must come from somewhere to replace that ‘dragged’ top layer, it has to come from the bottom…. it circles from the colder bottom.


Willis: As noted by Ed_B at 9:11 AM, I think the “east” in this sentence should be “west.” Of course, things get confusing because, crossing the dateline, things are now “east” again…
“And these strengthened winds simply blow the warm surface water mass to the east, where it divides and goes towards both poles.”


Willis: let me hasten to say that I think this is another superb post. WUWT (and us readers) are lucky to have such talent. And your style is transparent in several good ways. First, it’s clear and the ideas are developed and stuck together so we can follow along. Second, it’s got your personal stamp on it. We can watch you think. Which you do very well indeed. Thanks. …As for your prediction about faster cycling of Nino-Nina with greater heating, that makes intuitive sense. But instead of faster cycling, could the “engine” move the greater heat in some other way, e.g. a larger stroke volume? Do the trades blow more strongly (total air mass moved over time) or do the changes in surface height become greater, if there is more heat?

“In warmer times the pump should cycle more frequently”
The other day I graphed the AMO and the individual months of the AMO and strangley it looks like the warm phase of the AMO is almost over (and damned short compared the previous cycle), but just for the winter months (Nov-Apr). The other months are still going strong.
For example, January looks to be in the positive phase for only 15 years while in the previous warm AMO phase January’s warm phase lasted for 30-35 years.

Willis, thanks once again for the stimulation of thought . I rarely post here but read this site daily . I suspect there are thousands like me . Hey , your explanation explains the dodrums quite well . The oldtimers suffered them but couldnot explain them .


Vince, there is no one time frame for the deep ocean currents. There are different deep water circulations based on the source of the water. These are traced, as well as one can do, by their T-S characteristics. The main types are North Atlantic Deep Water, Antarctic Deep Water, and Mediterranean Deep Water. There is no direct measurement of the time for water to return to the surface for any of these. For example the longest circulation time is thought to be for NA Deep Water (forming the deepest “layer” in the ocean). This is created at various locations in the N. Atlantic and is thought to return to the surface over a broad and nebulous region of the Pacific. Med. Water (which spills over the sill in the Straights of Gibraltar and then finds it’s density level above the NA Deep Water) is thought to have a circulation time of several hundred years. A varied range of times are possible for different waters and different portions of the circulation.
It’s a mistake to think of the deep water circulation as a rigid “conveyor belt”. Just as the atmosphere exists in 3-D, so does the ocean. If less dense water were to sink in the N. Atlantic, it wouldn’t just replace denser water already existing in the deep ocean. It would ride over the denser water and modify the circulation pattern. If my memory of things I read years ago is still valid, the Pacific has had times in it’s history where the deepest waters have become anoxic (i.e. the deepest waters were essentially not replaced until all oxygen had been depleted – representing greatly reduced circulation). There is much to be learned in this area, just as there is still much to be learned about the atmosphere.

The Tao that can be spoken is not the true Tao.
The Climate Science that can be quantified is not the true Climate Science.
The sacred Models alone lead to a realisation of the mystery that we call Climate Science.

Nick Stokes

“And these strengthened winds simply blow the warm surface water mass to the east, where it divides and goes towards both poles.”
What I found remarkable, looking at the animations I linked above, is the persistence of the jet that comes from the East Pacific. It goes a long way before dividing.
What I also found amazing was the vortex street patterns. I don’t think wind driven motion could do that. There has to be deep water dynamics.
There’s a large collection of daily global SST data pictured here. It’s where the animations came from.

Willis Eschenbach

Thanks to those that pointed out my east-vs-west error, fixed now, gotta run, back in a bit.


I think the frequency of the El Nino / La Nina events is what requires more attention rather than the magnitude, especially when looking for trends in the historical data.


Willis the Bodgea Bay area is one of my favorite places in California. Love the Sailor’s lore on the
moon used such in my years in aviation years too..
Like knowing that fire season’s about to end and we go home for the winter, looking off the starboard side and see a nice, tight ring, around the moon,,,
Great post and food for thought..


“The Tao that can be spoken is not the true Tao.
The Climate Science that can be quantified is not the true Climate Science.
The sacred Models alone lead to a realisation of the mystery that we call Climate Science.”
Love this, MaTC


The interplay of the Hadley cell and the Polar cell and the Ocean and the resulting heat transfers are the key to understanding climate. I am reading Marcel Leroux right now on this subject.

Matthew R Marler

Thanks for the update.

Bill Illis

The Trade Winds and the ENSO are inversely related. The stronger the winds, the more likely there will be a La Nina. The weaker the winds, and especially if they blow backwards to the East, the more likely there will be El Nino.
Charted here.
Data source.
Now this relationship exists all the way back to the beginning of the record. This is Nino 3.4 versus the Trade Wind Index from the ESRL Reanalysis2 dataset which was just released recently.
In this case, I have inversed the Trade Wind numbers so that it is easier to see the relationship.
But I think it is really the temperature of the ocean itself which causes these winds to change or that they move together. The ENSO versus the Equatorial Upper Ocean Temperature Anomaly.
Equatorial Upper Ocean temps here.

Willis says: “When the heating of the ocean reaches a certain point, it triggers the winds that power the pump.”
The trade winds and sea surface temperatures are coupled with positive feedback—Bjerknes feedback. The triggers for initiating an El Niño and for initiating its demise (and the start of a La Niña) are Kelvin and Rossby waves—see the delayed oscillator theory:
Willis says: “However, if I am correct this implies that there will be a difference between warmer and cooler times—in warmer times, the pump should cycle more frequently. That is to say, there should be more of both El Ninos and La Ninas during warmer times.”
See Ray and Giese (2012) “Changes in El Niño and La Niña characteristics in an ocean reanalysis and reconstructions from 1871-2008”. Their abstracts ends with, “Overall, there is no evidence that there are changes in the strength, frequency, duration, location or direction of propagation of El Niño and La Niña anomalies caused by global warming during the period from 1871 to 2008.”
And, the Giese et al (2009) paper “The 1918/19 El Niño” argued that the 1918/19 portion of the 1918/19/20 El Niño was underestimated in the NINO3.4 sea surface temperature reconstructions, and that it was likely comparable in strength to the 1982/83 and 1997/98 El Niño events. Giese et al (2009) also suggested that the 1912/13 and 1939/40/41/42 El Niño events were also under-rated.
Also for informational purposes, most SST reconstructions show no warming of the eastern equatorial Pacific since 1900. That is, the oceans warm around the eastern equatorial Pacific.
And the trends of the Pacific sea surface temperature anomalies on a zonal-means basis since 1976 show the sea surface temperatures for the equatorial Pacific have cooled:


In a previous ‘incarnation’ (1980’s) I worked a laboratory chemist, and had occasion to use a marvelous gadget called a Soxhlet Extractor:
In operation, fluid in the flask at the bottom is heated, vapors rise and are condensed in the tube at the top. The recondensed fluid collects in the middle section until enough has accumulated to start the siphon (the inverted U-shaped tube on the side). The siphon returns the cooled liquid to the flask at the bottom, chilling it somewhat. The cycle restarts as more heat is added to the flask at the bottom.
The whole system is a two-stroke heat engine and the duration of the cycle depends on the rate at which heat is put in (at the bottom) and removed (at the top), as well as the capacity of the middle reservoir.
This suggests that there are 3 key variables in the El Nino/La Nina cycle: the ocean insolation (radiation), the atmospheric heat removal rate (convection), and the heat capacity of the atmosphere. This last variable is only very slightly dependent on the concentration of CO2, which can hold a little more heat than nitrogen, oxygen, or water (constant volume, molar heat capacity).

Duke C.

In figure 1, the amount of ocean water at 7C is hugely disproportionate to the higher temps closer to the surface. A neophyte (like me) would think that La Nina is always lurking 100 meters or so below the surface and needs little disturbance to activate an El Nino. I wonder if the TAO data shows any correlation between recent tsunami events and a change from La Nina to El Nino conditions…

Graeme M

Thats a great animation Nick Stokes. Thanks.

tom in indy

Clouds – Could variation in cloud cover between cooler and warmer times negate the impact of the hypothesized effect of temperature change on the frequency El Niño and La Niña?

Greg Goodman

Willis, maybe your testable should include magnitude. Will warmer times induce more frequent or more powerful ENSOs?
The two are linked , but the asymmetry comes from how they act in terms of energy balance. The factors determining how a warmer surface outputs heat to the atmosphere and how a cooler surface absorbs more incoming solar mean that the process is not symmetrical.
I think this makes magnitude even more important.

John Campbell

Dear mr Eschenbach,
I love the way you write, and I admire the good sense you exhibit. Keep it up – please.

Ulric Lyons

@Bill Illis
January 28, 2013 at 10:39 am
The trade wind changes often precede the change in ENSO status:
which what I would expect if changes in solar plasma speeds are responsible for the trade wind variations.
Your internally driven pump model cannot account for extended La Nina periods such as 1973-76, 1983-85 and 1998-2001. While all three of these periods had elevated solar plasma speeds:


I was about to say “fluidyne Stirling engine” – your thingimy looks a bit similar.
Ulric Lyons
Where did he say “solar plasma speeds” ? – I thought he was talking about heat

Hey Willis, I hate to put a spanner in your works but there was a certain Doctor, who’s name can’t be mentioned here but he is now deceased and it was a German name and he lived in Canada, who could predict El Ninos and La Ninas and their intensity using the spin and orbit of the sun. Here in Australia he was revered by government and farmers alike since his predictions were years in advance and farmers could plan stuff. So with that in mind that means there is an external periodic driver whereas your story can be rather random.
Just a thought.

Glad someone who actually knows about these things came in to comment. Thanks, Bob Tisdale.

Ulric Lyons

Gareth says:
“Where did he say “solar plasma speeds” ? – I thought he was talking about heat.”
Clearly Willis made no mention of external forcing. The plasma is very hot and causes Joule heating of the upper atmosphere and produces strong circulation changes that transfer right down to the troposphere.

“So with that in mind that means there is an external periodic driver whereas your story can be rather random.”” Should be
So with that in mind that means there is an external periodic driver whereas IN your story, EL NINOS and LA NINAS can be rather random.

Gary Pearse

I think there is a turbocharger in this pump in the form of ocean currents also moving the water westward and bringing in some cooler water (California Current from the N and its partner the Peruvian Current from the S heading for the equator). What do you think?

Integrated Atmospheric Angular Momentum (height of the HL stress fields and Hadley Cell correspondence) precedes variations in tropical SSTs..ENSO is driven by atmospheric circulatory processes. There are of course feedbacks (imbalance/re-perturbation via the MJO, etc).
The phenomena mentioned above are influenced by many factors, including:
-The Quasi Biennial Oscillation (QBO)
-The annular modes, and their connection to stratospheric O^3 content hence the solar wind
-The strength of the Earth’s magnetic field (through the same mechanism).
We’ve seen a long term weakening of the magnetic field, which has both increased the photodissociation rate of O^3 at lower altitudes and decreased H2O content in the lower levels of the stratosphere. We’ve thus seen a warming of the Hadley Cell caps as well as a gradual poleward migration, stronger polar vortexes, and decreased mobile polar high production. Also manifesting is a decrease in tropical/subtropical cloud cover.

I like the analogy of a pump and have used that myself but you have to understand that the pump has a variable stroke length and duration. Sometimes you can go from La Nina to neutral to La Nina again without an El Nino in between. It isn’t EXACTLY like a pump because a pump cycles from one state to the other. This particular case doesn’t have to. I have seen some research literature that showed evidence of a persistent La Nina condition toward the end of the last glacial around the LGM. Persistent La Nina during glacial conditions might make sense if you consider that it would mean persistently strong trade winds. One might expect that condition when you have decreased polar insolation and equatorial insolation is relatively unchanged. The difference in temperature between the poles and the equator is greater in that situation and the winds may want to carry more heat away from the equator to try to equalize that situation. In a period such as at the Holocene Climate Optimum, there is increased polar insolation and so the possibility of less of a temperature difference. I would be willing to bet the trade winds were more erratic during the HCO,

William Larson

Mr. Eschenbach–
I know that you don’t read all these comments (well, I THINK I know that). But in case you are reading this one, my-oh-my I like your poetic writing there at the end (Bodega Bay et al.). I did not know that you had that in you. And yes, testable predictions–now we are talking real science! Signing off for now–thanks.

Bill Illis

If you have a few minutes, watch this animation of equatorial temperatures in a cross-section down to 1000 metres. Note what happens in the 1982-83 Super-El Nino, the 1988-89 largest La Nina to-date, the 1997-98 Super-El Nino and the more recent La Ninas.
Other info and animations can be seen throughout this directory.
And here as well.


So, if warmer means more ENSO, this means more warm water sent to the poles. If this happens, the biggest effect will be at higher latitudes, since the whole ENSO thing keeps the middle lattitudes at an even temperature. Thus, the planet will not warm up, only the higher latutudes, where they actually want to warm up, will.
Conversly, if it gets colder, mostly it will get colder at high latutudes, ice ckating in Holland again
If ENSO stops altogether, such as when isolation gets too low from the tilt of the planet, then no warm water will get to the high latutudes and the ice will build up (starting with the summer the snow stays). Then the ice will reflect sunlight and it will get colder rapidly.

Ulric Lyons

crosspatch says:
“I would be willing to bet the trade winds were more erratic during the HCO,”
Going even further back in time, in a study of a recently revised New England varve chronology derived from proglacial lakes formed during the recession of the Laurentide ice sheet some 17,500 to 13,500 years ago, Rittenour et al. (2000) determined that “the chronology shows a distinct interannual band of enhanced variability suggestive of El Niño-Southern Oscillation (ENSO) teleconnections into North America during the late Pleistocene, when the Laurentide ice sheet was near its maximum extent … during near-peak glacial conditions.” But during the middle of the Holocene, when it was considerably warmer, even than it is today, Overpeck and Webb (2000) report that data from corals suggest that “interannual ENSO variability, as we now know it, was substantially reduced, or perhaps even absent.:

El Nino/La Nina historical graphs look to me like a random walk between upper and lower bounds. Which suggests a cooling mechanism triggered when SSTs warm and a warming mechanism triggered when SSTs cool, The warmer SSTs get the greater the probability the cooling mechanism is triggered and ditto cooler SSTs.
So, rather than a cycle, it’s random (or chaotic if you prefer) variation constrained in a range. The warming and cooling mechanisms may well be those described by Willis.

BTW, weekly NINO3.4 sea surface temperature anomalies have swung back up toward zero.
I published the very preliminary January 2013 sea surface temperature update today:
The results won’t be official for 2 weeks.


I agree with you Gary. During one direction of the PDO, the El Nino’s are much stronger and La Nina’s much weaker. When the PDO reverses, the El Nino’s weaken and La Nina’s strengthen, depending on which side of the Pacific you are .
A long jump from a lab to an ocean tadchem but the same principles apply, might help indicate why variation in the strength of the El Nino and La Nina, allowing for complexity of ocean currents.

Chuck Nolan

That warm area disappears from 31C down to 19C real quick.
I wonder how much of the surface heat is lost due to evaporation?

Keith Minto

For entertainment purposes, I view this TAO/Triton link, that overlays wind speed/direction with SST’s.
The ‘means’ make more sense to me, I can’t quite get a handle on a wind anomaly.


When that warm tropical water gets to the poles, displacing the water already there, where do you suppose that cold arctic water goes, and is this anything but a net heat conveyor to move ocean heat to the arctic where it loses that heat to space? I dunno who invented this pump but the way it works is a bit of a miracle as it has been doing this for billions of years.

Bill H

dp says:
January 28, 2013 at 6:33 pm
When that warm tropical water gets to the poles, displacing the water already there, where do you suppose that cold arctic water goes, and is this anything but a net heat conveyor to move ocean heat to the arctic where it loses that heat to space? I dunno who invented this pump but the way it works is a bit of a miracle as it has been doing this for billions of years.
Deep Water Cold Return..
This is a known oscillation. One which if disrupted would cause massive cooling of the involved hemisphere.