• Hello, please take a minute to check out our awesome content, contributed by the wonderful members of our community. We hope you'll add your own thoughts and opinions by making a free account!

Pattern ENSO Updates

Today’s 2 Euro runs suggest that after the next two days that there will be a dominating -SOI for most of the subsequent 8+ days (I think tomorrow and Saturday will actually be positive as Darwin comes down). Much of this is due to relatively high Darwin SLPs most days. As a matter of fact, 7/8 and 7/12 could be close to 1016 there. Also, Tahiti gets low late in the run. So, the most -SOIs of the next 10 days may actually end up toward the end.
Considering what lies ahead SOI-wise, El Niño could return within a few weeks though there often is a couple of week lag.

It’s been a strange year with the SOI like it has and the Niño fading. I think the Niño will continue a slow fade, it’s pretty close to neutral in all sectors right now and barely hanging on.
 
It’s been a strange year with the SOI like it has and the Niño fading. I think the Niño will continue a slow fade, it’s pretty close to neutral in all sectors right now and barely hanging on.

There's nothing "strange" about what's happening, the SOI is literally the worst index out there to measure ENSO. Essentially, the SOI is much noisier compared w/ virtually every other ENSO index because it's calculated at single grid pts & the spectra of variability associated w/ ENSO's atmosphere component is blue-shifted, i.e. comprised of some high-freq "white noise" (weather) that's not representative of ENSO. It's made worse by the fact that the MSLP variable being measured is itself noisy and in northern summer, the SOI is even less useful because more variability in the southern hem MSLP (where the geographic poles of the SOI are located) is consumed by mid-lat planetary waves rather than convectively coupled equatorial waves which ultimately determine the near-equatorial zonal wind anomalies which implicate the timing, amplitude, and excitation of oceanic kelvin waves.
 
There's nothing "strange" about what's happening, the SOI is literally the worst index out there to measure ENSO. Essentially, the SOI is much noisier compared w/ virtually every other ENSO index because it's calculated at single grid pts & the spectra of variability associated w/ ENSO's atmosphere component is blue-shifted, i.e. comprised of some high-freq "white noise" (weather) that's not representative of ENSO. It's made worse by the fact that the MSLP variable being measured is itself noisy and in northern summer, the SOI is even less useful because more variability in the southern hem MSLP (where the geographic poles of the SOI are located) is consumed by mid-lat planetary waves rather than convectively coupled equatorial waves which ultimately determine the near-equatorial zonal wind anomalies which implicate the timing, amplitude, and excitation of oceanic kelvin waves.

I'm aware of what the SOI is and why it's not the best index. My point was that you still typically expect to see the SOI behave like it has in a Nino regime and the WWB's this year have been pretty far apart and not helping the Nino much at all. The overall OHC has dramatically declined and things haven't unfolded quite like many thought it would in the spring when the claims on Twitter by many mets were that Nino would be slowly or steadily strengthening throughout the summer.

The temporary warming/cooling spikes are not unusual for borderline ENSO events but the underlying long-term tendency is for El Nino when the warming is being driven by low frequency variability while the subsequent cooling is not, that's really the point at I'm getting at here. The whole system is still clearly going towards El Nino and the warming spike, while temporary in space & time was attributable to changes that were shifting the entire system towards +ENSO, the cooling that's followed is occurring due to a modest easterly trade wind burst as most of the subeasonal forcing has shifted into the Maritime Continent (temporarily)

My thoughts posted in April were as follows, "Overall I still remain quite skeptical of a full blown mod to strong Nino developing this year. In fact I myself am leaning towards the idea of a transition into a warmish neutral state by June or July, something in the .2 to .4C range. Obviously we are still in the spring barrier where things can change quickly but what I'm seeing so far leads me to believe we might have a neutral or weak Nina this upcoming winter. Many hurricane season outlooks are basing their predictions off an El Nino lasting through the peak of the season. If it actually weakens to neutral before then that could change things somewhat."

CDAS on Tidbits has all 4 sectors now either in neutral or Nina territory. Although it may be a bit too cool it gives a good idea of the overall picture that Nino is dying and neutral conditions look quite likely just in time for hurricane season.

1562678430546.png
 
" My point was that you still typically expect to see the SOI behave like it has in a Nino regime and the WWB's this year have been pretty far apart"

That's not actually completely true though which was the gist of my initial response to you & as usual you completely overlooked it & blew me off. a) the SOI isn't actually measuring equatorial trade winds, it's a proxy for them yes, but the poles of the SOI are entirely in the southern hemisphere. b) in the northern summer, the SOI's correspondence to ENSO is weakest due to seasonal changes in the ITCZ and monsoon circulations. This means that more variability is contained within planetary-scale rossby waves in the northern summer and not CC equatorial waves that ultimately have a direct impact on the trade winds which effectively says that the SOI is less indicative of equatorial trade wind variability at this time of the year. This is also supported by the seasonal cycle in EOF loading patterns and correlation structure which show higher loadings and correlation in the northern hemisphere.

I.e. so is this behavior for the SOI as unusual as you're claiming/asserting it to be?

No, it's really not.
 
Last edited:
"My thoughts posted in April were as follows, "Overall I still remain quite skeptical of a full blown mod to strong Nino developing this year. In fact I myself am leaning towards the idea of a transition into a warmish neutral state by June or July, something in the .2 to .4C range. Obviously we are still in the spring barrier where things can change quickly but what I'm seeing so far leads me to believe we might have a neutral or weak Nina this upcoming winter. Many hurricane season outlooks are basing their predictions off an El Nino lasting through the peak of the season. If it actually weakens to neutral before then that could change things somewhat.""

Cool, cool I made a prediction in the previous spring, backed up w/ reason and science and adjusted it thereafter when the observations changed and didn't support it. I also made several statements on this to back that claim up on my different opinion as well as released data in real-time to complement it (like the ENS ONI and NCEP MEI indices)

You made yours earlier largely based on pure speculation and guessing with an underlying confirmation bias that you want to observe an active atlantic hurricane season and now are vehemently beating your chest out to me because of it. Must be nice to be an armchair meteorologist.
 
Last edited:
"My thoughts posted in April were as follows, "Overall I still remain quite skeptical of a full blown mod to strong Nino developing this year. In fact I myself am leaning towards the idea of a transition into a warmish neutral state by June or July, something in the .2 to .4C range. Obviously we are still in the spring barrier where things can change quickly but what I'm seeing so far leads me to believe we might have a neutral or weak Nina this upcoming winter. Many hurricane season outlooks are basing their predictions off an El Nino lasting through the peak of the season. If it actually weakens to neutral before then that could change things somewhat.""

Cool, cool I made a prediction in the previous spring, backed up w/ reason and science and adjusted it thereafter when the observations changed and didn't support it. I also made several statements on this to back that claim up on my different opinion as well as released data in real-time to complement it (like the ENS ONI and NCEP MEI indices)

You made yours earlier largely based on pure speculation and guessing with an underlying confirmation bias that you want to observe an active atlantic hurricane season and now are vehemently beating your chest out to me because of it. Must be nice to be an armchair meteorologist.

My prediction was based off observations and science as well. There was a lot of hype on Twitter in general from various mets about the Nino strengthening in March/April and I understand the science behind why they were making their predictions. At the same time, the computer models were consistently showing otherwise, the subsurface cold pool was expanding and the WWB's were pretty far apart. These were red flags that I pointed out why I was skeptical of it verifying in light of the conflicting signals and why I favored a slow fade this summer into the fall.
 
" My point was that you still typically expect to see the SOI behave like it has in a Nino regime and the WWB's this year have been pretty far apart"

That's not actually completely true though which was the gist of my initial response to you & as usual you completely overlooked it & blew me off. a) the SOI isn't actually measuring equatorial trade winds, it's a proxy for them yes, but the poles of the SOI are entirely in the southern hemisphere. b) in the northern summer, the SOI's correspondence to ENSO is weakest due to seasonal changes in the ITCZ and monsoon circulations. This means that more variability is contained within planetary-scale rossby waves in the northern summer and not CC equatorial waves that ultimately have a direct impact on the trade winds which effectively says that the SOI is less indicative of equatorial trade wind variability at this time of the year. This is also supported by the seasonal cycle in EOF loading patterns and correlation structure which show higher loadings and correlation in the northern hemisphere.

I.e. so is this behavior for the SOI as unusual as you're claiming/asserting it to be?

No, it's really not.

My point is that, in general, the SOI is typically correlated with the El Nino changes especially when you filter out the noise and look at running 3 month averages. I base that off the CPC and other agencies which broadly link it to changes in ocean temperatures across the eastern tropical Pacific.

The CPC says "The Southern Oscillation Index (SOI) is one measure of the large-scale fluctuations in air pressure occurring between the western and eastern tropical Pacific (i.e., the state of the Southern Oscillation) during El Niño and La Niña episodes. Traditionally, this index has been calculated based on the differences in air pressure anomaly between Tahiti and Darwin, Australia. In general, smoothed time series of the SOI correspond very well with changes in ocean temperatures across the eastern tropical Pacific. The negative phase of the SOI represents below-normal air pressure at Tahiti and above-normal air pressure at Darwin. Prolonged periods of negative SOI values coincide with abnormally warm ocean waters across the eastern tropical Pacific typical of El Niño episodes. Prolonged periods of positive SOI values coincide with abnormally cold ocean waters across the eastern tropical Pacific typical of La Niña episodes."

1562697421778.png

The Australian Bureau of Meteorology defines it in this general sense, "The SOI measures the difference in surface air pressure between Tahiti and Darwin. The index is best represented by monthly (or longer) averages as daily or weekly SOI values can fluctuate markedly due to short-lived, day-to-day weather patterns, particularly if a tropical cyclone is present.

Sustained positive SOI values above about +8 indicate a La Niña event while sustained negative values below about –8 indicate an El Niño."

The current 90 day average is -8.51 according to this page. While the SOI index isn't a metric I follow my original intent in my post was discussing what Larry had posted and not to get into a debate on it's strengths and weaknesses. In the context of how it's defined by various agencies and what Larry was discussing, yes I would call the fading El Nino (by the definitions above) a bit unusual since the SOI has been -8.51 over the past 90 day average.
 
My prediction was based off observations and science as well. There was a lot of hype on Twitter in general from various mets about the Nino strengthening in March/April and I understand the science behind why they were making their predictions. At the same time, the computer models were consistently showing otherwise, the subsurface cold pool was expanding and the WWB's were pretty far apart. These were red flags that I pointed out why I was skeptical of it verifying in light of the conflicting signals and why I favored a slow fade this summer into the fall.

There's a lot more to ENSO evolution even during weak events than the appearance of a subsurface cold pool, extrapolating that out in time even w/ some regard to WWB activity doesn't do it enough justice. As for the models, while showing a "weakening" El Nino (it's actually within the realm of noise I wouldn't call a 0.2C drop in NINO 3.4 SSTs "weakening") it was still an El Nino nonetheless thru the summer.

figure4.png
 
Last edited:
My point is that, in general, the SOI is typically correlated with the El Nino changes especially when you filter out the noise and look at running 3 month averages. I base that off the CPC and other agencies which broadly link it to changes in ocean temperatures across the eastern tropical Pacific.

The CPC says "The Southern Oscillation Index (SOI) is one measure of the large-scale fluctuations in air pressure occurring between the western and eastern tropical Pacific (i.e., the state of the Southern Oscillation) during El Niño and La Niña episodes. Traditionally, this index has been calculated based on the differences in air pressure anomaly between Tahiti and Darwin, Australia. In general, smoothed time series of the SOI correspond very well with changes in ocean temperatures across the eastern tropical Pacific. The negative phase of the SOI represents below-normal air pressure at Tahiti and above-normal air pressure at Darwin. Prolonged periods of negative SOI values coincide with abnormally warm ocean waters across the eastern tropical Pacific typical of El Niño episodes. Prolonged periods of positive SOI values coincide with abnormally cold ocean waters across the eastern tropical Pacific typical of La Niña episodes."

View attachment 20942

The Australian Bureau of Meteorology defines it in this general sense, "The SOI measures the difference in surface air pressure between Tahiti and Darwin. The index is best represented by monthly (or longer) averages as daily or weekly SOI values can fluctuate markedly due to short-lived, day-to-day weather patterns, particularly if a tropical cyclone is present.

Sustained positive SOI values above about +8 indicate a La Niña event while sustained negative values below about –8 indicate an El Niño."

The current 90 day average is -8.51 according to this page. While the SOI index isn't a metric I follow my original intent in my post was discussing what Larry had posted and not to get into a debate on it's strengths and weaknesses. In the context of how it's defined by various agencies and what Larry was discussing, yes I would call the fading El Nino (by the definitions above) a bit unusual since the SOI has been -8.51 over the past 90 day average.

Wow, copy & pasted information from a website on an index that I've already known about for at least a decade and am using in my masters thesis, breaking new ground for me there bud, I simply had no idea that the SOI corresponded to changes in ocean temperatures in the eastern Pacific.

These values are arbitrarily chosen by the Australian BOM and depending on the dataset and amount of smoothing used you'll find that in some cases some aren't supporting El Nino in terms of the SLP/atmospheric component of ENSO. Equatorial SOI &/or SLP EOF are arguably much better proxies of the trade winds that the SOI is trying to get at.


" While the SOI index isn't a metric I follow my original intent in my post was discussing what Larry had posted and not to get into a debate on it's strengths and weaknesses."

Smh, my point is which you're still continuing to ignore is that it's not that unusual at all for the SOI to disconnect with the trade winds especially at this time of the year. That's why actually understanding its strengths and weaknesses are important, you can't just simply ignore that.

When you correlate every ENSO index over the reference period of 1950-2005 with everything else, you'll quickly find that the SOI in both the smoothed and 20CR reconstructed versions explains the least amount of variance in virtually every other available index, again a testament to its weakness and futility. Thus one would expect it to be among the least representative of ENSO behavior and thus SOI values of a specific threshold (-8 in the BOM, others define this differently) would tend to correspond the least with ENSO phase, which also suggests it is not, i repeat, not that unusual to have a disconnect between even tri-monthly SOI and ENSO. If you've actually took a hard look at the data you'd see this is indeed true.

Anyways, I've been doing this for several years and have extensively studied every ENSO index known to man so I guess I don't know what I'm talking about then...

Screen Shot 2019-07-09 at 3.17.03 PM.png
 
There's a lot more to ENSO evolution even during weak events than the appearance of a subsurface cold pool, extrapolating that out in time even w/ some regard to WWB activity doesn't do it enough justice. As for the models, while showing a "weakening" El Nino (it's actually within the realm of noise I wouldn't call a 0.2C drop in NINO 3.4 SSTs "weakening") it was still an El Nino nonetheless thru the summer.

View attachment 20943

Check out my original post on page 19 to see the models I cited as reference to the decline. Regions 1+2 & 3 are in Nina territory per CDAS (although a bit cold biased probably) and sector 3 is in neutral. Sectors 3.4 and 4 are neutral and have declined significantly over the past few weeks as well. I believe the latest update had things officially at .6C but I'd expect that number to come down into neutral territory by the next update with the continued cooling ongoing.
 
Check out my original post on page 19 to see the models I cited as reference to the decline. Regions 1+2 & 3 are in Nina territory per CDAS (although a bit cold biased probably) and sector 3 is in neutral. Sectors 3.4 and 4 are neutral and have declined significantly over the past few weeks as well. I believe the latest update had things officially at .6C but I'd expect that number to come down into neutral territory by the next update with the continued cooling ongoing.

CDAS & OISSTv2 are noisy satellite-based datasets that are susceptible to subseasonal variability significantly impacting the SSTAs, their decline has only happened very recently and wasn't forecast by a majority of models to occur at least anywhere to this extent. The appearance of tropical instability waves in the eastern pacific throughout the NINO 1-2 & NINO 3 regions are indicative of a transition to & representative of a colder ENSO base state. The prevalence of a +NPMM/-SPMM in northern summer doesn't favor NINO growth, the same regime persisting into the fall & subsequent winter may be a different story however.

As for the weekly OISSTv2 CPC data that uses a 1971-2000 base period (which ironically is computed outside the range of OISSTv2 itself which begins in Sep 1981), a more modern, warmer base period will push that anomaly closer to or even slightly below 0.5C. I don't actually know why the CPC hasn't reverted to a 1982-2018 base period or something of the like where they actually have data for it in OISSTv2. It's likely easier to construct the 1971-2000 real-time anomalies since the climatology netcdf files are already created.
 
Wow, copy & pasted information from a website on an index that I've already known about for at least a decade and am using in my masters thesis, breaking new ground for me there bud, I simply had no idea that the SOI corresponded to changes in ocean temperatures in the eastern Pacific.

These values are arbitrarily chosen by the Australian BOM and depending on the dataset and amount of smoothing used you'll find that in some cases some aren't supporting El Nino in terms of the SLP/atmospheric component of ENSO. Equatorial SOI &/or SLP EOF are arguably much better proxies of the trade winds that the SOI is trying to get at.


" While the SOI index isn't a metric I follow my original intent in my post was discussing what Larry had posted and not to get into a debate on it's strengths and weaknesses."

Smh, my point is which you're still continuing to ignore is that it's not that unusual at all for the SOI to disconnect with the trade winds especially at this time of the year. That's why actually understanding its strengths and weaknesses are important, you can't just simply ignore that.

When you correlate every ENSO index over the reference period of 1950-2005 with everything else, you'll quickly find that the SOI in both the smoothed and 20CR reconstructed versions explains the least amount of variance in virtually every other available index, again a testament to its weakness and futility. Thus one would expect it to be among the least representative of ENSO behavior and thus SOI values of a specific threshold (-8 in the BOM, others define this differently) would tend to correspond the least with ENSO phase, which also suggests it is not, i repeat, not that unusual to have a disconnect between even tri-monthly SOI and ENSO. If you've actually took a hard look at the data you'd see this is indeed true.

Anyways, I've been doing this for several years and have extensively studied every ENSO index known to man so I guess I don't know what I'm talking about then...

View attachment 20944

My main point was that "In the context of how it's defined by various agencies and what Larry was discussing, yes I would call the fading El Nino (by the definitions above) a bit unusual since the SOI has been -8.51 over the past 90 day average." My original statement "It’s been a strange year with the SOI like it has and the Niño fading" was in a general/broad sense based on how the agencies I cited define it and how it hasn't correlated much for awhile (both winter, spring and now). In addition, it has been strange in the sense that many mets were forecasting a strengthening of the Nino, some even calling for a strong one, yet the opposite has happened in the past few months.

I think part of the disagreement here is that I was referring to the broad sense of how the index is used/correlated and the results for awhile have not matched up very well (making it unusual in that broad sense) while your statements are asserting that the SOI isn't a very useful index to begin with, especially in summer (which I agree with). Hopefully that clears up my intent and views on things but if not then that's ok to agree to disagree :) I'm not a met and there is always much out there that I'm willing to learn and study and you've definitely studied things far more in depth than I have in all sorts of various fields.
 
My main point was that "In the context of how it's defined by various agencies and what Larry was discussing, yes I would call the fading El Nino (by the definitions above) a bit unusual since the SOI has been -8.51 over the past 90 day average." My original statement "It’s been a strange year with the SOI like it has and the Niño fading" was in a general/broad sense based on how the agencies I cited define it and how it hasn't correlated much for awhile (both winter, spring and now). In addition, it has been strange in the sense that many mets were forecasting a strengthening of the Nino, some even calling for a strong one, yet the opposite has happened in the past few months.

I think part of the disagreement here is that I was referring to the broad sense of how the index is used/correlated and the results for awhile have not matched up very well (making it unusual in that broad sense) while your statements are asserting that the SOI isn't a very useful index to begin with, especially in summer (which I agree with). Hopefully that clears up my intent and views on things but if not then that's ok to agree to disagree :) I'm not a met and there is always much out there that I'm willing to learn and study and you've definitely studied things far more in depth than I have in all sorts of various fields.

Even when you use a 90-day sliding average for just the AMJ period, in no way is a value < -8 during non-NINo conditions somehow "unusual" as you're assuming to be. It's happened as recently as 2005, 2003, & most notably 1993 in this tri-monthly and 1993 actually ended up being a complete flop of a NINO although the 3-month SOI was a whopping -15 in AMJ 1993. There are other years when you extend the range of tri-monthlies around AMJ
 
Even when you use a 90-day sliding average for just the AMJ period, in no way is a value < -8 during non-NINo conditions somehow "unusual" as you're assuming to be. It's happened as recently as 2005, 2003, & most notably 1993 in this tri-monthly and 1993 actually ended up being a complete flop of a NINO although the 3-month SOI was a whopping -15 in AMJ 1993. There are other years when you extend the range of tri-monthlies around AMJ

Your definition of unusual is probably different than mine ;) We just have to agree to disagree on things here and it's best to end the debate here.
 
Your definition of unusual is probably different than mine ;) We just have to agree to disagree on things here and it's best to end the debate here.

"your definition"?

Umm correct me if I'm wrong but you didn't come up w/ an ENSO-SOI definition to begin with. Instead you just copy & pasted what the Australian BOM showed & agreed w/ their definition of what SOI values define specific ENSO conditions of +/-8 and I used that very definition and tri-monthly values and got answers that didn't exactly support your claims of how unusual this behavior is with at least 3 recent examples in the past few decades and one that was far more extreme than what we're observing now. Not surprisingly, when you were actually confronted with real data & facts that hurt your argument you immediately deflected & tried to throw it under the rug. Sigh.
 
"your definition"?

Umm correct me if I'm wrong but you didn't come up w/ an ENSO-SOI definition to begin with. Instead you just copy & pasted what the Australian BOM showed & agreed w/ their definition of what SOI values define specific ENSO conditions of +/-8 and I used that very definition and tri-monthly values and got answers that didn't exactly support your claims of how unusual this behavior is with at least 3 recent examples in the past few decades and one that was far more extreme than what we're observing now. Not surprisingly, when you were actually confronted with real data & facts that hurt your argument you deflected & tried to throw it under the rug. Sigh.

No, 3 times out of 30 years is the exception, not the rule, which I would define as "unusual." It doesn't mean it can't happen just that it doesn't all that often.
 
No, 3 times out of 30 years is the exception, not the rule, which I would define as "unusual." It doesn't mean it can't happen just that it doesn't all that often.

30 years? Umm did it ever occur to you that El Ninos roughly occur one-third of the time right? So only 10 of those years roughly are NINOs, and 3 of those 10 years are actually false alarms, that's 30% of them, not something you can just immediately dismiss and sweep under the rug & dismiss.

Nice try tho.
 
Back
Top