Extending El Niño event predictions to a year

Posted: December 31, 2019 by oldbrew in Analysis, ENSO, predictions, research, Temperature

The El Niño of 1997-8

Let’s see how this theory works out in practice.

A group of researchers from the Potsdam Institute for Climate Impact Research, Beijing Normal University and Justus-Liebig-Universität Giessen has found a way to predict El Niño events up to a year before they occur, says Phys.org.

In their paper published in the Proceedings of the National Academy of Sciences, the group describes their complexity-based approach to better predicting the seemingly random weather events.

El Niño is a weather event in which the water surface temperatures in some western parts of the Pacific grow warmer than normal and then seep eastward.

Scientists are eager to learn more about such events because they can contribute to excess rainfall in some parts of the world and drought conditions in others.

Knowing when they are going to occur would give people living in both areas time to plan for them. Until now, the best weather experts could do was offer predictions six months in advance. In this new effort, the researchers have found what they believe to be reasonably accurate predictions a year in advance.

The new method involves a novel analysis of sea surface temperature data based on complexity theory. It involves dividing the eastern part of the Pacific Ocean into square cells and comparing the temperature of each cell with others near it and others that are more distant.

Such comparisons allow for measuring the degree of disorder in the temperature data. Then, by comparing the degree of disorder year over year, they were able to spot a pattern—years of high disorder tended to be followed by El Niño events the next year.

Once they found that pattern, the researchers then analyzed yearly surface temperature data from 1984 to 2018 to make predictions about El Niño events in the past.

They report that their method correctly predicted nine out of 10 El Niño events (and had three false positives.) Additionally, they found that the higher the disorder the previous year, the stronger the following El Niño event.

The researchers conclude that it is now possible to predict El Niño events up to a year in advance with reasonable accuracy.

Source here.

  1. oldbrew says:

    Then there’s La Niña…

    A warning from ancient tree rings: The Americas are prone to catastrophic, simultaneous droughts
    By Paul Voosen Dec. 31, 2019


  2. Adam Gallon says:

    Ah, I see they’ve got to kowtow, to the Great God Global Warming.

  3. oldbrew says:

    Yes, the Potsdam Institute is in the front line of climate alarm promotion…er, research 😎


  4. Kip Hansen says:

    Maybe…..they had four misses and 9 hits.

    If that was a baseball batting average it would be pretty good. If it was success in “bridge building”, it would be pretty dismal.

  5. They have looked at a relatively short period to justify their number fiddling. They should look at the the extended wet and dry period poster here https://www.longpaddock.qld.gov.au/rainfall-poster/ and note the graph of SOI and IPO SOI has a long record of something like 160 years and has not been fiddled. Daily SOI can be found here https://www.longpaddock.qld.gov.au/soi/. Please do not trust BOM. They have been hiding raw data and certainly fiddling with data for charts and tables (particularly with respect to temperature) Wet and dry periods are periodical with length within certain bounds but it is difficult to predict the extact start and finish. It would appear that daily SOI have some relation to the phases of the moon. I would suggest that planetary alignment also has an influence.

  6. oldbrew says:

    From Ian Wilson’s 2014 blog post – ‘Evidence that strong El Nino events are triggered by the Moon’ (see Final Comments):

    The first prediction is that because we are currently in a 31 year Full Moon Epoch for El Nino events. Hence, there should be heightened probability of experiencing a strong El Nino in the following years:

    2015-2016 (see figure 1)
    2019-2020 and
    as these are the years where the lunar line-of-apse aligns with the Sun at the times of the Equinoxes.

    – – –
    We’ve had the first one, but 2019 is over so we’re looking at 2020. Is it too soon for another strong one after 2016? Of course ‘heightened probability’ is a qualified prediction, and solar cycle 25 hasn’t got going yet.

  7. Phoenix44 says:

    Nine out of ten and three false positives is pretty useless. The actual rate was 10 in 34 so you would not be miles out flipping a coin each year or rolling a dice.

    Their model was wrong four times and right nine.

  8. oldbrew says:

    Re the possible 2020 La Niña…

    Solar minimum and ENSO prediction
    Andy May / July 5, 2018
    By Javier

    Two solar physicists, Robert Leamon from NASA Goddard Space Flight Center, and Scott McIntosh from the High Altitude Observatory at Boulder, CO, have made an interesting observation that links changes in solar activity with changes in the El Niño Southern Oscillation (ENSO).

    As they reported at the AGU 2017 Fall Meeting, the termination of the solar magnetic activity bands at the solar equator that mark the end of the Hale cycle coincides since the 1960’s with a shift from El Niño to La Niña conditions in the Pacific.


    They think they see a pattern…

    More info in the link. But if it’s a 60~ year cycle it might be about to change 🤔

  9. Christine Nore says:

    Some geologists believe that El Nino is caused by sea water seeping into a magma chamber between New Guinea and The Solomon Islands, where it is under extreme pressure and heated to100’s of degrees C. The superheated water is forced out in waves, this has been tracked by NASA as pulses of warm water rather than a stream moving across the Pacific. As it does not fit the paradigm of atmospheric activity and CO2 warming it is largely ignored.

  10. tallbloke says:

    OB, I’ve been pointing out the onset of big El Ninos coinciding with solar minimum for years.
    e.g. https://tallbloke.wordpress.com/2010/02/06/el-nino-and-the-solar-cycle/
    and https://tallbloke.wordpress.com/2014/06/23/are-the-sun-and-moon-both-acting-on-el-nino-events/

    Ian’s research with Nikolay Sidorenkov is now reaching a level of maturity where it’s making accurate hindcasts and potentially useful predictions. Given that we are now also entering solar minimum, I think a 2020 El Nino is pretty likely.

  11. oldbrew says:

    That account is restricted 🙁

  12. tallbloke says:

    Ah. OK, here’s the link from the second tweet

  13. tallbloke says:

    And the image from the first

  14. oldbrew says:

    From the paper:
    It was found that El-Niño tended to occur in the 1st, 10th, and 13th years after the maximum diurnal tide in the 18.6-yr cycle, and La-Niña tended to occur in the 3rd, 12th, and 16th years.

    The 10th year i.e. the half-period (minimum) after the 2006 lunar standstill (maximum) was 2015/2016 = the most recent El-Niño.

    OTOH no La-Niña in 2017/2018. Could the solar cycle minimum have a role there?

  15. oldbrew says:

    12 December 2019
    ENSO Alert System Status: Not Active

    Synopsis: ENSO-neutral is favored during the Northern Hemisphere winter 2019-20 (70% chance), continuing through spring 2020 (~65% chance).


    The next ENSO Diagnostics Discussion is scheduled for 9 January 2020.

  16. Bob Weber says:

    Determining an El Nino a year ahead based on ‘disorder’ is disordered thinking, imo. In my work I’ve identified two main types of El Nino, those following solar max, and a ‘solar cycle onset El Nino’.

    In mid-2014 I successfully predicted the return of El Nino as exclusively contingent upon a return to higher solar activity, which did happen, driving TSI higher, and subsequent deeper tropical warming.

    I’ve predicted the next ‘solar cycle onset El Nino’ to start summer 2020, then a La Nina in 2021.

    The conditions for tropical warming are enduring positive Central Pacific OLR, indicating fewer clouds and higher insolation warming. Low TSI sets those conditions, which are in effect now.

    We are now in a 2009/10 solar-ENSO analog period. The cycle onset El Nino is a short-term impulse response to the usual new cycle TSI burst, the highest year-year change of a solar cycle.

    The ocean responds quickly to the new cycle TSI rise and fall, and upon a reduction in TSI, the tropics cool and the warm-water pulse dissipates as in ’97/98 and ’09/10, and the ocean re-aligns with the longer-term absorbed solar energy deficit (into La Nina), awaiting the usual energy climb to the solar max (which drives a temperature ‘step’ when cycle max SN>95).

    If SC25 is similar to SC24 we should see a near repeat of the pattern, subject to sunspots, with the random emergence and strength of new cycle sunspot activity as the only ‘disorder’ involved.

    If in 2014 solar activity hadn’t returned to high levels in the year’s latter six months, the Nino would’ve fizzled or been substantially weaker and delayed, responding in half the time of whatever this study’s ‘disorder’ is a measure of. That means my system is at least 2X as predictive because it’s based on the actual mechanisms involved.

    I used to say a lot more about lunar-driven atmospheric tides that I thought could drive clouds away from the tropics, allowing more warming from fewer clouds/higher insolation, but after seeing how long CP OLR can stay positive (fewer clouds), there doesn’t appear at first glance to be an evident direct connection to lunar perigees.

    PS- my 5-minute solar-geo app product is now live

  17. oldbrew says:

    Thanks Bob W – testable predictions are good.
    – – –
    astroclimatelink says: January 2, 2020 at 1:08 pm
    You may have proof of a 31.0-year Perigean New/Full Moon lunar tidal cycle + 11.2-year Solar Cycle origin for the ENSO cycle.

    Re 31 years: worth repeating that 3 * 31y = 93y = ~5 lunar nodal cycles (93.066y).

    I also like to say (‘/’ = divided by):
    Apsidal cycle / Full moon cycle = 7.84753
    93.066 (5 nodal) / Jupiter orbit = 7.84532
    (34/13) * 3 = 7.8461538
    [3,13 and 34 are Fibonacci numbers]
    Nodal cycle / QBO = 7.85473 (where nodal cycle / 36.3 trop. months = 6.85473 i.e. one less) – see note below.
    Phi² * 3 = 7.854102

    See: https://tallbloke.wordpress.com/2017/01/08/why-phi-a-lunar-ratios-model/
    (doesn’t mention Jupiter )
    – – –
    Note: I may not have explained this in a blog post to date.

  18. oldbrew says:

    Surprising only to folk who think a change to 0.01% of the atmosphere is a big deal.

  19. Paul Vaughan says:

    Golden Tether Four Won Rabbit Hale

    “Perhaps even more remarkably […] we can replace the base 2 by any real number bigger than 1 […]”

    BRItest numbrrrr con spy R ace C 4 shore sum west turn politica11y elite wi11 just knot bet 2 base kin D-Light.

    Try base = See V-E ‘n’ to phi-nail e say Halo to JEV J-S.

    Really IT’s ~7.001315:
    1. Replace the base.
    2. Remark ably:
    We R floor ring the 4th power of phi O V-E R Σun sing 2 C just ⌊(φ/Σ)^4⌋.

    “Once in a lifetime
    there is water at the bottom of the ocean
    REMOVE THE WATER” — Talk King Heads “Water UNderground….”

    Audio logic a11y fathoming numeric depth trackground hare gurgles O sun drain.

    “UNder the rocks and stones there is….”

    “Same as it ever was” – 1. Reversing:
    “Saw reve ti sa” – 2. Cleansing:
    “Saw river to sea.”

    “1s in a lifetime”
    Note Knott strings rules guaranTEEing polarIT swITch = polariTEE CHange.

    Once in a rabbIT’s lifetime 4 cast goal done string “same as IT ever was”.

    “in 2 the blue again in 2 silent water”:

    Dearest Moderators: Shed O wing cens’or’ ship on the ‘eye’V-E seize sum we11-hidden Suggestions-41. Please reV-E’AI IT 4 US sew EU’11 C R 5-star red’or’ can know.

  20. oldbrew says:

    JANUARY 3, 2020

    Atlantic and Pacific oscillations lost in the noise
    by Pennsylvania State University

    According to the researchers, if the Atlantic Multidecadal or Pacific Decadal oscillations existed, there would be evidence for their existence across the suite of current state-of-the-art climate model simulations.

    “Given the current sophistication of climate models as seen in their ability to capture the El Niño/Southern Oscillation, we would expect to see consistent evidence for oscillations across a suite of climate models,” said Michael E. Mann, distinguished professor of atmospheric science at Penn State. “We found no such evidence.”

    – – –
    So climate models can’t ‘see’ them – that’s it then 😛

  21. oldbrew says:

    5 lunar nodal cycles = 93.0662 tropical years = 98.0662 draconic years
    but also
    98 draconic years = 93 sidereal years