El Nino – same but different


There Is Nothing New About The El Nino Modoki

Guest post by Bob Tisdale

The press release…


…for the Hye-Mi Kim, et al (2009) paper “Impact of Shifting Patterns of Pacific Ocean Warming on North AtlanticTropical Cyclones”…


…includes the quote from Peter Webster, one of the authors of the paper, “’Normally, El Nino results in diminished hurricanes in the Atlantic, but this NEW TYPE is resulting in a greater number of hurricanes with greater frequency and more potential to make landfall,’ said Peter Webster, professor at Georgia Tech’s School of Earth and Atmospheric Sciences.” (Caps added for emphasis.)

The press release continues, “That’s because this NEW TYPE of El Nino, known as EL NINO MODOKI (from the Japanese meaning “similar, but different”), forms in the Central Pacific, rather than the Eastern Pacific as the typical El Nino event does. Warming in the Central Pacific is associated with a higher storm frequency and a greater potential for making landfall along the Gulf coast and the coast of Central America.” (Caps added for emphasis.)

Hye-Mi Kim et al (2009) references (Their Reference 10) the Ashok et al (2007) paper “El Nino Modoki and its Possible Teleconnection.”


Ashok et al introduce a new index to document the occurrence of an El Nino Modoki event. They write, “Based on the EOF2 pattern presented in Figure 2b and the PC2 time series shown in Figure 3, we derive an El Niño Modoki index (EMI). Because of the unique tripolar nature of the SSTA, the index is defined as follows:

“EMI= [SSTA]A-0.5*[SSTA]B-0.5*[SSTA]C (1)

“The square bracket in Equation (1) represents the area-averaged SSTA over each of the regions A (165E-140W, 10S-10N), B (110W-70W, 15S-5N), and C (125E-145E, 10S-20N), respectively.”

The areas of the western, central, and eastern tropical Pacific included in the equation are illustrated in Figure 1.


Figure 1

Figure 2 is a short-term (1979 to 2008) times-series graph of the El Nino Modoki Index, using the calculation from Ashok et al. I’ve also standardized the data. Ashok et al define a typical El Nino Modoki event as, “We call an El Nino Modoki event ‘typical’ when its amplitude of the index is equal to or greater than 0.7α, where α is the seasonal standard deviation.” I’ve also highlighted 0.7 on Figure 2.


Figure 2

Figure 3 is a longer-term (1900 to 2008) times-series graph of the El Nino Modoki Index. As you will note, El Nino Modoki events as determined by the El Nino Modoki Index have occurred over the entire term of the data. There is nothing NEW about El Nino Modoki events.


Figure 3


HADISST data is available through the KNMI Climate Explorer website:


0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
Pieter F
July 7, 2009 8:26 am

If there is nothing different about Modoki ENSOs and the record is showing an overall decline in total annual cyclonic energy, how can Dr. Webster say, “this NEW TYPE is resulting in a greater number of hurricanes with greater frequency and more potential to make landfall”? Where is the observed cyclonic energy data to support his claim? As Bob points out, there appears to be nothing happening here.

July 7, 2009 8:30 am

The “new” type is an old type (well, at least, older than 1979). The “new” type is El Nino juxtaposed on a negative PDO. I doubt that anyone really knows for sure what impacts on Atlantic Hurricanes will be. During the period 1940 – 1979, when we last had a negative PDO, there was barely any comprehension of ENSO, PDO was unknown, and satellite tracking of hurricanes was in its infancy during the second half of the period.

gary gulrud
July 7, 2009 8:34 am

Hurricanes making landfall? Are these NH hurricanes striking the West Coast, Japan, etc?

July 7, 2009 8:45 am

gary gulrud: You asked, “Are these NH hurricanes striking the West Coast, Japan, etc?”
The Hye-Mi Kim et al (2009) paper “Impact of Shifting Patterns of Pacific Ocean Warming on North AtlanticTropical Cyclones” deals with the statistics of different types of El Nino events (where they form–central or eastern equatorial Pacific) on the patterns of North Atlantic Hurricanes.

Dan Lee
July 7, 2009 8:52 am

So, what does moving that standard deviation line to the other side of zero in that third graph tell us? That there were much stronger “La Nina Modoki” events recently than el nino ones? I’m not getting the point of that press release, unless its their way of saying “we’re just making stuff up now.”

July 7, 2009 9:00 am

That´s the “canonical nino”.

Ron de Haan
July 7, 2009 9:23 am

Some of us are waiting for an El Nino Okidoki.

July 7, 2009 10:12 am

Let me make sure I understand Figure 2. You say you “standardized” the data, which I take to mean “relative to standard deviation.” If that’s the case, shouldn’t the y-axis be dimensionless? In other words, I would think that what you are measuring on the y-axis is “temperature (anomaly) divided by standard deviation,” and since “Deg C” is in both numerator and denominator, it cancels out, leaving the metric dimensionless. At least that is what I think of when I read that you “standardized” the data. Is that what you did, or did you do something else?
I concur with Dan Lee’s head scratching. Do we also here have a new kind of La Nina, when the index falls below -0.7α?
I haven’t read the paper, but did they bother to plot their “new” El Nino index along side of an “old” El Nino index? Did they look at how the standard deviations of the old and new compare? I’d bet that if you standardized an “old” El Nino index, and plotted it with the standardized “new” El Nino index, you’ll find it hard to tell the difference between the two. Are you in a position to do that (i.e., have the data handy)? It would be interesting to see.

Retired Engineer
July 7, 2009 10:16 am

MODOKI (from the Japanese meaning “similar, but different”)
Isn’t that the definition of climate?
It’s always the same, except when it isn’t.

July 7, 2009 10:46 am

Arigato 🙂

July 7, 2009 10:52 am

I took “new type” to mean that El Nino Modokis haven’t been previously documented in the literature, not that they have only recently began to exist.
In the paper they show that from 1950-1990 1 of 9 El Nino events were classified as “Modoki” while since 1990 4 of 5 El Nino events were “Modoki”, so that’s where they get the increasing relative frequency from.
I guess I don’t really understand all the harsh criticisms… I thought the paper was interesting, especially when they showed the clear differences between tropical cyclone tracks and seasonal predictability associated with the different El Nino types.

John F. Hultquist
July 7, 2009 10:53 am

Our young puppy will sometimes roust out of slumber and bark while running to the sliding glass door. The other dogs seem as surprised at this as we are – none of us having heard anything. We say “She heard a ghost.” and go let her out to chase it.
This report gives me the same feeling: These folks are hearing ghosts.
Where are the “…greater number of hurricanes with greater frequency and more potential to make landfall…” ? I know AGW-folk have been saying things such as this but the data seem not to support it. We need to see proof of the problem before we are served an answer.
What does “more potential to make landfall” mean? That “potential” is used rather than “probable” seems to imply, to me, that strength or power is involved in the same sense that a large truck rather than a small car will take out more guard rails if run off the road. These two words are not synonyms.
And there is this “…making landfall along the Gulf coast and the coast of Central America.”
Is there a part of the report (or is there anywhere) a study of landfalls by place that supports this statement? Even in a general sense – does frequency of land fall by latitude vary over time?
Ron de Haan (09:23:38) : Some of us are waiting for an El Nino Okidoki.
That’s a gem!

July 7, 2009 10:56 am

Peter F (08:26:48) : “… the record is showing an overall decline in total annual cyclonic energy…”
I assume you are referring to the GLOBAL decline in ACE. This particular study focuses on the Atlantic basin where there has been an increase in overall activity since around 1995.

July 7, 2009 11:00 am

Basil and Dan Lee: I provided the red line simply as an indicator of the threshold of a El Nino Modoki. The El Nino Modoki Index is only for El Nino events, not La Nina events.
Basil: Thanks for noting the units on the y-axis on the graph of standardized data. I apologize. I added the Deg C out of habit. And yes, Ashok et al did compare the El Nino Modoki Index to NINO3.4 in their Figure 4, cell b. There are significant differences. In fact, you can see that in my Figure 2 above. The El Nino Modoki index is negative during 1997/98 when the NINO3.4 would be elevated. The PDO would also be elevated in 1997/98 so the comparisons to the El Nino Modoki Index are unfounded.

jack morrow
July 7, 2009 11:33 am

Ron de Haan–okidoki
Lol–very funny-wish I had thought of that!
I’m waiting with bated or baited breath(We’ve already discussed the spelling here)

George DeBusk
July 7, 2009 1:03 pm

Forgive me for what might be a rather basic observation, but I often check the “ENSO Wrap-Up” of the Australian Bureau of Meteorology just to look at the SOI graph that is updated daily. I notice that the SOI has become slightly positive now despite consensus predictions of an El Nino this year. Shouldn’t the SOI remain negative if an El Nino is forming? Can there be an El Nino with a positive SOI? Is a positive SOI one of the different things about an “El Nino Modoki?.
Perhaps the positive SOI is merely temporary and the index will plunge directly, but it certainly seems a rather glaring inconsistency with El Nino forecasts.

July 7, 2009 1:05 pm

Would this indicate we “might” beginning seeing more hurricane/tropical storms starting over Central (and western South America) and continuing (harmlessly!) on their way out over the uninhabited lands of equatorial Pacific?
Those storms I have seen lately – last two years or so – but there has been no sign of ANY increase in the “classic” Caribbean-Gulf Coast storm pattern.

July 7, 2009 1:35 pm

Adam: You wrote, “In the paper they show that from 1950-1990 1 of 9 El Nino events were classified as “Modoki” while since 1990 4 of 5 El Nino events were “Modoki”, so that’s where they get the increasing relative frequency from.”
My complaint was with the press release, which has found its way to a number of websites and blogs. The term El Nino Modoki does not appear in the Hye-Mi Kim et al (2009) paper. They use CPW (Central Pacific Warming) and EPW (Eastern Pacific Warming) as identifiers. To clarify what I wrote in the post, one of the authors used new, and the press release expanded on it to include the term El Nino Modoki.
And another point of clarification, it’s not every El Nino Modoki event that has been statistically found to effect Atlantic Hurricane storm track. Hye-Mi Kim et al (2009) only consider the months of August-September-October, months that are El Nino formation months, in their supplement graph S-2. (Makes sense because when El Ninos reach their peak, hurricane season is over,) And there are special criteria that have to be met. For a CPW classification, it’s not a simple comparison of SST anomalies for the NINO3 and NINO4 regions that qualifies the El Ninos as having a statistical impact on Atlantic Hurricanes, because there were more than one El Nino events before 1990 with higher Central Pacific (NINO4) SST anomalies than Eastern Pacific (NINO3) SST anomalies.
BTW, the four CPW events they identified after 1990 occurred after the significant El Nino events of 1986/87/88 and 1997/98. Those two periods are very similar in terms of NINO3.4 SST anomalies. I’ve got a post in the works on those similarities.

Milwaukee Bob
July 7, 2009 1:55 pm

Ron de Haan (09:23:38) :
Some of us are waiting for an El Nino Okidoki.
Thank you! I needed that! Bob Hope is smiling down at you sir.
You are one funny guy.

July 7, 2009 1:58 pm

George DeBusk: You asked, “Shouldn’t the SOI remain negative if an El Nino is forming? Can there be an El Nino with a positive SOI? Is a positive SOI one of the different things about an ‘El Nino Modoki?'”
George, last month or the month before (sorry that I can’t tie it down better than that), Bill Illis ran through a list of variables that impact and are impacted directly by ENSO events and very few were “in line” with what one would consider normal for an El Nino event. I wish I could remember which thread, because the detail was remarkable.
In the traditional El Nino, warmer water from the western equatorial Pacific sloshes east. With the El Nino Modoki, there can be any number of things that “feed” the central Pacific (NINO3.4) warming. These wouldn’t necessarily be reflected in the SOI. But there are times when the central equatorial Pacific warming precedes the traditional El Nino. Does the Central Pacific Warming trigger the traditional El Nino on these occasions? Dunno. I’m working on a video that shows the traditional 1997/98 El Nino versus the El Nino Modoki events of 1994/95 and 2002/03 to illustrate the different “sources”.

Milwaukee Bob
July 7, 2009 2:06 pm

And a question – is the El Nino Modoki the cause/reason the High that is normally resident off the East coast of Florida, has been and is over the Gulf of Mexico? and why the equatorial trade-winds have been abnormally blowing west to east until just the last 2 weeks or so?
As long as that High is where it is and those TWs are weak or stationary – no hurricanes. Meanwhile the gulf SST is higher and rising faster than normal for this time of year and Katy-bar-the-door if and when it moves back over the Atlantic. we’ll have Gulf formed Hs early AND – – – hmm, maybe it’s time to move again…

Richard M
July 7, 2009 3:12 pm

Yet another unknown in the climate puzzle. However, I’m sure this is all covered in the models, right?

July 7, 2009 5:13 pm

It always better to claim “new” when rediscovering the old and threadbare. That way you get published.

July 7, 2009 5:30 pm

I noticed that Peter J. Webster, who is quoted in the “Media Newswire”, and who is the corresponding co-author of the July 3 Science Mag report, “Impact of Shifting Patterns of Pacific Ocean Warming…,” is part of “Carbon Management” — a division of Georgia Tech Strategic Energy Research. Webster’s specialties are “Increasing Efficiency/Mitigating Environmental Impact” in the unit that “is addressing growing concern over carbon cycle issues….investigating whether greenhouse warming is causing a global increase in hurricane intensity….[and concerned with] carbon capture and sequestration related to power generation. The unit’s researchers are “working with major industry partners to discover and refine new processes for eliminating the threat of climate change from carbon dioxide emissions.”
My guess that El Nino Modoki is an “okidoki” (Ron de Han) to find something “new” on which to hang global warming for December’s meeting. With a negative PDO and cooler temperatures we are likely to have fewer El Ninos and larger La Ninas, but, there must be some other warming disaster that rising CO2 levels can cause in the oceans.
Bob Tisdale, you have shown that there is nothing NEW in the heating of the Central Pacific Ocean during the entire 20th Century up to the present. So now climate realists — e.g., Bob, WUWT — have to turn their attention and research efforts toward outing this NEW okidoki.
A second item appeared In the same issue June 3rd of Science Mag in “Perspectives” by Greg J. Holland that appeared to innocently be about “Predicting El Nino’s Impacts”. Holland is from NCAR in Boulder CO where he is director (since 2005) of “Mesocale and Microscale Meteorology Division” He asserts that it has been shown, citing Webster’s co-authored article, Kim, ea, that El Nino should be broken into two independent modes: eastern Pacific warming (EPW) and a central Pacific warming (CPW). The latter has a high level of predictability and functions like a La Nina in this respect — North Atlantic tropical cyclone activity rises during CPW.
The detective story is getting hot. Webster and Holland co-authored an article published online in 2007 by the Philosophical Transactions of the Royal Society of London. They proposed that “climate change” is a major factor in the increasing number of Atlantic hurricanes”. They showed step changes over the century of periods of annual averages of 6, 10, and now 15 tropical storms correlated closely with a 1.3 degree F rise in SSTs in the last 100 years. “The new study indicates that natural cycles are probably not the entire cause because the increase has happened across the last century rather than oscillating in tandem with a natural cycle.” Finally, “the proportion of major hurricanes to lesser hurricanes…has increased significantly in recent years.”
Webster and Holland appear like Al Gore’s storm troopers in the battle to purify our environment of the pollution and immenent disaster of CO2.

July 7, 2009 6:24 pm

Bob T., could you recreate your Figure 3 with just the averages of August-September-October plotted? I’d like to see how that compares with Kim Figure S2. Thanks.

July 7, 2009 7:18 pm

davidsmith1: El Nino Modoki Index averaged for August through October:
Looks like there was a run around 1900, too, but the tropical Pacific SST data is questionable before 1914. Then there were another pair around 1920. What I enjoy is that the 1986/87 El Nino shows up, but the second part (1987/88) of it doesn’t. Apparently, the warm spot moved east, which they have a tendency to do as the El Nino progresses. (IMO, the 1986/87/88 El Nino was the second of the Super El Ninos in the 20th century, not 1982/83, since El Chichon suppressed most of the heat transport of the earlier one.)
You’ll soon discover there are a few differences between the Hye-Mi Kim et al Figure S-2 and this one.
I’ll also work up a couple of NINO3.4 and El Nino Modoki comparison graphs tomorrow morning.

July 7, 2009 7:55 pm

Bob Tisdale
“(IMO, the 1986/87/88 El Nino was the second of the Super El Ninos in the 20th century, not 1982/83, since El Chichon suppressed most of the heat transport of the earlier one.)”
Are you talking impacts or relative strength? The ’82-’83 El Nino was huge in terms of anomalies in the Equatorial Pacific. Even the ’72-’73 event had quite a peak: http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/quick_look/sst34.gif

July 7, 2009 7:57 pm

“IMO, the 1986/87/88 El Nino was the second of the Super El Ninos in the 20th century, not 1982/83, since El Chichon suppressed most of the heat transport of the earlier one.”
Are you saying the impacts from the ’86-87 El Nino were greater than the ’82-83 event? Looking at 3.4 anomalies, the latter was a huge event. The ’72-’73 El Nino was also quite a show: http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/quick_look/sst34.gif

David Smith
July 7, 2009 8:20 pm

Thanks, Bob.
It looks like ’65, ’66, ’67, ’77,’ 86, ’90, ’91, ’94, ’02, and ’04 were Modoki seasons.
For comparison, Kim lists ’69, ’91, ’94, ’02 and ’04 as CPW seasons. Pretty big mismatch.
Related to Kim’s Figure S2, I easily replicated the Nino 3 values in the Figure but was way off on the Nino 4 values. I still haven’t identified the reasons for the differences.

July 7, 2009 9:04 pm

Come back to first principles then – to the physics and chemistry involved (supposedly!) in GISS’s work.
What size the “cubes” How are they set around the world by latitude, elevation, and longitude (?) and by ocean and ground? Are the “cubes” the same size in x, y, and z? _ if Not, how are changes vertically handled in temperature, pressure, winds, water and humidity and clouds? What are modeled for mountains, seas, and high “flat” elevations? What about slowly changing elevations – such as Rockies CO to Western Mississippi? How are peninsula’s like FL and Thailand (?) and islands like New Zealand and Greenland and New Guinea handled? How is Antarctica handled? The Arctic?
Are jet streams produced (internal to the model) by the math as a function of the math itself, or are they artificially simulated by the equations or are they ignored altogether?
We understand “clouds” are a single, fixed, simple factor: A single assumed value. If this is correct, does that value change with time, humidity, “rainfall” predictions output or “drought” or higher temperatures or do the “climate modelers” simply keep one cloud value forever across 100 years or 200 years of processing?
Are clouds changing with elevation? With the different time steps, do “clouds” get changed by the output of previous steps, and what exactly IS fed back from step n to step n+1, n+2 , n+3 etc?
What is the physics in each “cube” – there ARE hundreds (if not thousands) or engineers and doctors and even scientists capable of understanding each equation, and easily understanding the “physics and chemistry” of each interaction inside each cube. WHAT do the models assume iis going on in the “cube”? and who has checked or audited or peer-reviewed (by name!) these steps.
Who has checked the computer coding in each “cube” to verify that what the equations are supposed to be is what what the computer is calculating?

July 7, 2009 11:29 pm

First I thought of making a snide comment about some super massive el nino but then I thought about it. Given that the ocean current model is failing it must be a very exciting time to be in this particular field of expertise. With the addition of new tech toys and models to be made well… i guess it depends on what spin you want to put on the data. Sadly, I bet we know about as much about our oceans secret life then we do about the surface of Mars. 🙂

July 8, 2009 2:49 am

Karl: You asked about my 1986/87/88 El Nino comment, “Are you talking impacts or relative strength?” and Are you saying the impacts from the ‘86-87 El Nino were greater than the ‘82-83 event?”
The peak NINO3.4 SST anomalies for the 1982/83 and 1997/98 El Ninos were similar, but the El Chichon eruption impacted the heat transport of the 1982/83 El Nino, so it was a non-Nino. That is, the 1982/83 El Nino had little to no measurable impact on global temperature. The peak NINO3.4 SST anomalies of the 1982/83 and 1997/98 El Nino events were substantially higher than the 1986/87/88 El Nino, but the 1986/87/88 El Nino lasted through the entire year of 1987, plus the ramp up in 1986 and transition to La Nina in 1988. The 1986/87/88 El Nino was so strong it caused a noticeable upward step change in global temperature, like the 1997/98 El Nino. The heat transport of the 1986/87/88 and 1997/98 El Nino events can be seen in the RSS MSU TLT Time-Latitude Plots and Time-Series graphs I provided here:
So to answer the first of your questions above, I was talking about global impacts and about relative strength and duration.

July 8, 2009 3:38 am

David Smith: You wrote, “Related to Kim’s Figure S2, I easily replicated the Nino 3 values in the Figure but was way off on the Nino 4 values. I still haven’t identified the reasons for the differences.”
I also had some problems trying to duplicate the Hye-Mi Kim et al Figure S2. Standardizing the data helped. Keep in mind that they described how they determined EPW and CPW events as, “EPW, CPW, and Eastern Pacific cooling (EPC) events are defined as follows: Niño 3 (fig. S1) warming greater than 1 standard deviation (SD), for EPW; Niño 3 or Niño 3.4 cooler than 1 SD, for EPC; and for CPW, Niño 4 warming exceeding 1 SD, while Niño 3 stays below this range. The results are shown in fig. S2.” So there is another qualifier other than NINO4 having a higher SST anomaly than NINO3 for the CPW event.

July 8, 2009 7:04 am

For those interested in short-term (~30 year) comparisons of scaled monthly NINO3.4 SST anomalies and El Nino Modoki Index data, I’ve just posted them.

July 8, 2009 7:59 am

I remember reading about 2 years ago (I don’t know where) that El Ninos in negative PDOs are more centered in the Pacific and during positive PDOs they are centered right on the South America coast. This does cause a different pattern in North America; positive PDO El Ninos are very warm and wet in the eastern US, and negative PDO El Ninos are cooler but still wet(think snowy winters in the late 70s). So it makes sense that it might affect where hurricanes go. But this is all natural and has nothing to do with AGW.

July 8, 2009 8:42 am

Karl: Thanks for bringing up the 1972/73 El Nino. The post I’m working on would not have been complete without it. The post is about the similarities between significant El Nino events and the lesser ones that follow (aftereffects?). Here’s a comparison of the 1972/73, 1986/87/88, and 1997/98 El Nino events:
And here’s the same data smoothed with 25-month filters:
The shapes of the curves of the NINO3.4 SST anomalies that followed those big three events are strikingly similar.
Thanks again for the reminder.

Pamela Gray
July 8, 2009 2:31 pm

It is only a matter of time before the AGW crowd discovers that oceanic oscillations dictate our weather pattern variations, therefore it is incumbent on us to reduce the amount of pollution draining into our oceans. That means that anything that makes it into our oceans should be regulated. CO2, animal waste, irrigation drainage, etc. Protect the oceans. Can you imagine the list of banned substances? This is a case easier to make than greenhouse gasses. Actually, they would have done a better job of this AGW to have focused on ocean pollution sources. Instead of monitoring the skies for greenhouse gasses, they should be monitoring the oceans for pollution.

Bill Illis
July 8, 2009 5:34 pm

This Georgia Tech group is constantly torturing data “in creative ways” trying to prove the case that global warming is increasing hurricane numbers and increasing hurricane intensity.
Is the Modoki index values available somewhere?
I plotted the Nino 4 and Nino 3 region values against US landfalling hurricanes in various different ways and there is just no “different type of El Nino” relationship to landfalling hurricanes whatsoever. I haven’t tried this novel Modoki method however.

July 8, 2009 6:48 pm

Bill Illis: You asked, “Is the Modoki index values available somewhere?”
I had to create my own using the SST anomalies for the areas listed in Ashok et al and the simple formula they included.

July 8, 2009 7:46 pm

You can take “modoki” as an equivalent of a postfix “-oid”.
Modoki not only means “similar but different”, but also conveys a nuance of “somewhat inferior to the original.”

July 11, 2009 2:25 am

They need to be looking in the Atlantic too. The stored ocean heat from the run of high solar cycles will come out into the lower tropospheric temperatures. I predict north Atlantic Ocean Heat Content will plummet following this el nino.
Josh Willis will be wondering if his Atlantic Argo floats have thrown another wobbly, and it’ll probably get ‘adjusted’ away in the processing of the raw ARGO data.

July 11, 2009 2:55 am

“El Nino Modoki and its Possible Teleconnection.”
Please, water is heavy. It doesnt get teleported from the Pacific to the Atlantic. The heat coming out of the surface of the Atlantic is the stored heat from below the surface of the Atlantic.

%d bloggers like this:
Verified by MonsterInsights