Study: El Niño forecasting reliability during the period 2002-2011 declined

From the “bigger computers and newer models don’t necessarily produce better predictions” department and the INSTITUTE OF ATMOSPHERIC PHYSICS, CHINESE ACADEMY OF SCIENCES comes this curious observation.

Why has ENSO been more difficult to predict since 2000?

El Nino-Southern Oscillation (ENSO), which is one of the most striking interannual variability in the tropical Pacific, has been extensively studied for several decades. Understanding the changes in its characteristics is still an important issue for worldwide environmental and socioeconomic interests. Clear decadal variations exist in the ENSO’s predictability, with the most recent decade having the lowest ENSO predictability in the past six decades.

Linking decadal variations in ENSO predictability and in BF intensity. Standardized decadal variations in the averaged 1-3 leading months persistence (black solid line), averaged 4-6 leading months persistence (black dashed line), Niño3.4 SD (blue line) and BF intensity (red line, defined as the regression between dSST/dx and the wind index). CREDIT Fei Zheng

Linking decadal variations in ENSO predictability and in BF intensity. Standardized decadal variations in the averaged 1-3 leading months persistence (black solid line), averaged 4-6 leading months persistence (black dashed line), Niño3.4 SD (blue line) and BF intensity (red line, defined as the regression between dSST/dx and the wind index). CREDIT Fei Zheng

“The forecasting reliability during the period 2002-2011 was relatively lower than that in the 1980s and 1990s.” Observed Dr. Fei Zheng from Institute of Atmospheric Physics after he and his team examined the ENSO prediction skills of 20 state-of-the art models. They explored the possible reasons of modulating the decadal variations in ENSO’s predictability, and recommended the Bjerknes Feedback (BF) intensity, which dominates the development of ENSO, to determine ENSO predictability.

They found that decadal variations in BF intensity are largely a result of the sensitivity of the zonal winds to the zonal sea level pressure (SLP) gradient in the equatorial Pacific. Furthermore, the results show that during low-ENSO predictability decades, zonal wind anomalies over the equatorial Pacific are more linked to SLP variations in the off-equatorial Pacific, which can then transfer this information into surface temperature and precipitation fields through the BF, suggesting a weakening in the ocean-atmosphere coupling in the tropical Pacific.

For future research and forecasting practice, Zheng suggested, “More attention should be paid to off-equatorial processes in the prediction of ENSO during the low-ENSO predictability period.”

Their findings were recently published in Geophysical Research Letters.


The paper:


65 thoughts on “Study: El Niño forecasting reliability during the period 2002-2011 declined

  1. ENSO is quasiperiodic and involves a complex mechanism dependent on interaction of winds, humidity, temperatures, lunar & solar tidal forces, sunlight, and who knows what else. Plankton? ENSO periodicity is like a grandfather clock with a rather large mouse running up and down the pendulum.

    • I suspect Antarctic sea ice formation to be instrumental in ENSO as well, as sinking highly saline water provides the push for the cold bottom currents that originate there and upwell along the West coast of South America.

    • “More attention should be paid to off-equatorial processes in the prediction of ENSO during the low-ENSO predictability period.”

      If your prediction method sometimes works and sometimes doesn’t it’s a pretty good sign that you have it fundamentally. wrong or are missing a key driver.

      It’s like those who claim surface temperature correlated to SSN as a proxy of solar activity …. except for periods when it doesn’t. Back to the drawing board.

    • Why would anybody decline the el Nino forecasting reliability ??

      Seems like one would graciously accept any help they can get.


    • Are these the same folks who conclude that if a dog has four broken legs and you tell him to run, and he doesn’t, the conclusion has to be that the dog can’t hear…

      • The Old Man.

        Yeah. Ever tried herding cats?
        About as successful are King Canute and his demonstration of his impotence over tides.
        Which was very well-judged, however.

        Auto – aware of the limits of human ‘outreach’.

    • Or drawing up the predicted movements of the presumed dancing angles on the proposed point of the assumed needle – and debating the size and weight of those same angels.

    • That would be when the giant mouse has been awakened by the clock striking. Hickory-dickory-dock. The mouse ran up the clock. The clock struck one; the mouse became irregular.

      • jorgekafkazar: or as comedian Jackie Vernon once said: “Three mice ran up the clock. The clock struck one, the other two escaped without injury.

      • I think it goes:

        Hickory Dickery Dock.
        Three mice ran up the clock.
        The clock struck one.
        The other two escaped with minor injuries.

      • Hickory, dickory dock
        3 Mexican mice ran up the clock
        The clock struck Juan
        Jose and Antonio ran down safely

  2. A tally of all the unpredictability of the major parts of the system might say something about the whole, except in IPCC circles. Aloofness of the total from the parts is what arm waver predictions do.

  3. It’s a chaotic system. It is therefore unpredictable. Even though it is deterministic, the slightest change in initial conditions leads to a completely different outcome. Same thing as 3 celestial bodies orbiting about one another. Same thing as the Earth’s climate.

  4. The balance between El Nino and La Nina within the underlying 60 year cycle is shifting back towards La Nina after many years of El Nino dominance.

    The ENSO predictions were based on continuing El Nino dominance and so are failing.

    The cause is increased global cloudiness from solar induced more meridional jet stream tracks. That first became evident to me around 2000 and I have been pointing it out continually since 2007.

    In due course, if the sun stays quiet, La Nina will establish dominance and global atmospheric temperatures will begin to fall.

      • The source of the energy in the oceans is solar shortwave radiation and NOT energy transferred from the atmosphere. Thus global atmospheric temperatures do not induce ENSO phenomena such as La Nina or El Nino.
        Separately, winds will contribute to the location, intensity and timing of ENSO phenomena but that is another issue.

    • The balance between El Nino and La Nina within the underlying 60 year cycle is shifting back towards La Nina after many years of El Nino dominance.

      The ENSO predictions were based on continuing El Nino dominance and so are failing..”

      I am mighty curious, …… just why is it that so many learned people keep referring to an El Nino and/or a La Nina ……. as being the “cause” of a weather, temperature or whatever “event” …. when in fact they are nothing more than the measured “effect” of an event that was “caused” by one or more physical “acts” occurring in the biosphere?

      • Simply because the rate at which energy is released from the oceans provides a substantial bottom up thermal effect on the atmosphere.
        The oceans receive varying amounts of solar energy depending on solar induced cloudiness changes and then internal ocean cycles of which ENSO is the largest will vary the rate of energy release back to the atmosphere.
        Not really a lot to do with the biosphere at all.

  5. What is the driver of the ENSO?

    If you know what the driver is, can you predict what it is going to do over the next year?

    Do you know all the other factors that have to be taken into account and what they will do over the next year?

    The driver is really the ocean circulation system in the equatorial Pacific. East to West at the surface driven by the Trade Winds, and then backwards West to East in the undercurrent at 200 metres dept. It flows back to the East because all the water flowing West at the surface has pushed it down and it has to go somewhere, After it flows to far eastern side of the Pacific, it finds somewhere to go as it gets to surface and replace all that water moving to the West because of the Trade Winds.

    What it is going to do over the next year is push some warmer than average water in the undercurrent toward the eastern side over the next six months (this is really the warm water left over the last El Nino, it got pushed down). When it flows all the way back to the Eastern side, there is just mainly neutral normal temperature water here so it will get mostly neutralized.

    So normal, neutral for 6 months and that is all one can forecast in this scenario. 50:50 crap shoot afterward.

    • Some back-up for what I said above.

      First, the Trade Winds Drive the ENSO (without saying what drives the Trade Winds).

      Pretty tight relationship here.

      And if you go way, way back, this relationship still holds in a very eerie manner. The SOI which also lends its name in part to the ENSO is also a reflection of the Trade Wind Index in the (eastern) Pacific.

      I wrote a post about this long ago. It is interesting to see how things continue to work (well, 8 years later).

      What the Trades really do, however, is they drag or push or pull the surface waters along with them to the West. The surface waters at the equator flow East to West because of the Trade Winds. (The Indian Ocean sometimes exhibits a reversal of this for short periods of time and the currents go backwards, but for the most part the Indian Ocean follows this pattern as well as the Atlantic ocean does – which is actually how the Gulf Stream starts. The remnants of Flight MH370 which crashed somewhere in the Indian Ocean were found at Madagascar. Why? Because the currents and the Winds go East to West).

      What this pattern does in the Pacific is that it sets up a scenario where all this water is piling up against the Indonesian continental shelf. The water cannot keep rising working against gravity (although it does actually do this) but it has to go somewhere. All the water piling in behind forces it to go somewhere. A small amount gets squeezed through the Indonesian Flowthrough but this is too small to hold a big current. Some of gets moved northwest (and down first so it looks like it disappears) but it eventually becomes part of Kuroshio Current next to Japan.

      But most of it simply gets forced downward. It becomes the equatorial under-current and flows back to the east now. When it get all the way to the Galapagos Island continental shelf, again it has to go somewhere. Fortunately, the Trade Winds are evacuated water westward at the surface and now the water has somewhere to go. It can just replace the water flowing West now.

      It surfaces and itself becomes the next “wave” of ENSO waters at the surface some months later. One big self-reinforcing current that also becomes a plus or minus 3.0C at the surface just based on how warm or cold the water in the under-current originally was. It makes a big circle like this (copied from one Bob Tisdale did long ago).

      And it is this temperature anomaly itself which causes the Trade Winds to pick up or slow down. It is the temperature of the water itself which forces the thunderstorms to form in the middle of the Pacific or over Indonesia/Australia. It is so much energy considering it can fluctuate so much and it is a very big area.

      The Circulation Pattern is the DRIVER at the end of all of it. It is such a big area that it actually influences how fast the Earth rotates (believe it or don’t) and the even the amount of Out-going Longwave Radiation to space.

      This is the best predictor of what the ENSO is going to do. It gives at least 2 months notice and one can predict what it is going to do for maybe 6 months out. The East Equatorial Upper Ocean Temperature Anomaly in the upper 300 metres of the ocean (which primarily reflects the temperature of the water in the under-current which is getting ready to surface).

      Most of the data in this post can be found here:

  6. he and his team examined the ENSO prediction skills of 20 state-of-the art models.

    If there are 20 (or more) models then the state of the art is akin to kindergarten finger-painting, so in a very poor state indeed.

    In what other area of human endeavour (other than horse racing or stock market predictions) would one find a proponent with supposed skill being able to offer 20 different models of a process, all of which are supposed to be believed to function correctly and on which you should spend your money?

    • They should be required to put some money on their models. In no time we’ed have fewer models and most of our grant money back.

  7. Maybe they’re actually doing quite well. It sounds like their predicting accuracy is at least as good as weather forecasting.

  8. well, I figure in just another decade or two the self declared experts on climate, who must never be questioned or doubted, will discover what many have been saying right here for years now. That the 60 year cycle was in it’s warm phase through the 70’s, 80’s and 90’s, and is now in it’s cool phase. and will stay there at least another dozen years.

    Unless they just plan to keep ‘Adjusting’ the cooling trend away. ~_~

    • So, if temps go down that will be accurate; that will prove you right and them wrong, but if they don’t; it’s fake.
      But if they aren’t fake, well, it can’t be us doing it? And if it is, it’s a good thing afterall.

      That seems to be where it always leads.

  9. Oh, you silly rabbits! It’s climate change that obviously make the climate behave in an unpredictable manner. The brilliance of this observation not only arises from the fact that it’s a tautology, and therefore utterly truthful when read literally, but that it can also be (mis)used by alarmists to dismiss the fact that climate models have no predictive value while still clinging desperately to the output of those same computer models.

  10. “Why has ENSO been more difficult to predict since 2000?”

    Well duh global warming and CO2. It can’t be the scientists.

  11. Why are you using such a short time period? El Nino’s basic long-term period is about 5 years. The entire length of every global temperature graph is a concatenation of El Nino peaks alternating with La Nina valleys. You don’t see this because usually the curves are computer-smoothed to eliminate the peaks. You are now interpreting a temperature stretch that barely takes in two El Ninos. And why are you so cheap with data points? Put in enough data points so that the sharp angles in the image disappear. Your chosen data period is difficult to interpret because of interference with ENSO warm deposits by some left-over material from the 1998 super El Nino. It starts in 1999, immediately after the departure of that super El Nino. The leftover deposit inherits its warmth from the super El Nino and no greenhouse effect is needed, thank you. Its warmth raises the global temperature by 0.6 degrees Celsius in only three years. It partially buries a La Nina in its middle, the traces of which are still visible around 2003. The warmth is left over from the super El Nino and after it left I expected the temperature of this mass to start cooling. Early views of it at the beginning of the century did not show this, however, and it looked more like a hiatus than a cooling. But if you now look at the UAH satellite view from December, you will note that a distinct cooling is evident from the year 2002 on. It is hard to understand how it all fits together because the La Nina of 2008 and El Nino of 2010 together block any view of the real background. But starting in 2012 the outer parts of the El Nino of 2015/16 begin to rise and you realize that there is the other end of the background curve that starts in 2002. Just draw a straight line from 2002 to 2012 and join it with the beginning of the rise that will form the next El Nino. Consider it the best approximation to the background you cannot see. I also suggest that you ought to extrapolate the straight line section between 2002 and 2012 to the base line that is beyond 2016. It will tell you approximately when to expect the cooling to reach the background level that existed before 1997. And incidentally, the eighties and the nineties, from 1979 to 1997, were a flat hiatus region. You don’t see this because a cabal consisting of GISS, NCDC, and the Met Office has substituted a fake warming for that region, and they control what is shown to the public. Their act is actually worse than a hockey stick because Mann does admit that he added data to the end of his curve.

  12. The Bjerknes feedback? Peruvian upwelling? What could either of them POSSIBLY have to do with ENSO?? I thought they were banned words already here on WUWT? (/sarc)

  13. There should be more of this – objective testing of the effectiveness of research spending. Ironic that it should come from Chinese scientists, doing what should be the “western” governments’ work for them.

    This suggestive transition around 2002 – just after the last major genuinely Bjerknes-based el Nino event – would be called by some authors a change in the “ground state” of the Pacific. It’s also when global warming stopped.

  14. Bjerknes feedback is just that: a feedback. What climatologists have failed to look at is what the trigger is.

    IMO they should be looking at tidal forcing causing movements at the top of the thermocline. There is a density difference between the warm “mixed layer” and deeper colder waters. This density difference will lead to tidal waves in a similar way to surface waves.

    The density difference is about 1/1000 of the difference between air and water, so the main resonances will be 1000 times longer. The dominant response at the surface is the twice daily tides, so 12h * 1000. That’s of the order of 1.4 years.

    Tidal waves at the bottom of the mixed layer will be of this order, ie multi-annual time-scales.

    There is too much focus on equatorial sections of temp profiles. The way temperature anomalies appear and disappear in such sections indicates a N/S movement of heat energy in and out of the equatorial zone. this needs to be studies as 3D maps not 2D sections.

    The BJ f/b is a +ve f/b which reinforces and locks in changes started by tidal forcing, it is NOT the cause. That is why models relying on it are not reliable. They are missing the underlying cause.

  15. Dr. George Philander, the father or at least uncle of ENSO, told me over a decade ago that his research, published in one of his books, indicated that in an increasingly CO2 warmed world El Nino would dominate over La Nina, I believe in a positive feedback mechanism.

  16. Isn’t the ultimate driver of all this climate silliness the Sun? Hence the name “Solar System” ? Just asking, I’m new here.

    • Phone may be powered by a battery but that does not explain all the things a phone does, Kivy. Of course a flat battery is one thing to consider if the phone behaves unusually, but not the only thing. Is that what you were asking? I think we all ask that at first.

  17. It’s about fricken time someone published this. That it took this long to look back to see if the models were any good seems to me to be evidence they were not altogether enthralled with performance and only published when they observed a “badder” string of predictions. Talk about making lemonade out of lemons!

  18. Notice that their ability to reliably forecast El Nino started in the 1970s with the advent of the renewed natural warming trend. To my mind that shows that part of any warming trend will include a dominant positive/El Nino pattern in the ENSO regions.

    Their failure to reliably forecast conditions after the mid 2000s coincides with the switch back to a cool trend. The ENSO regions should favor negative/La Nina conditions until around the mid 2030s, imo.

Comments are closed.