2026-2027 Record-breaking El Nino

[Typhoon Tip]

Meanwhile the 3.4 sector is at historic levels relative to September +ENSO typology.  And I mean somewhere between gaspingly and laughably so ...

Kind of hard to think CANSIPs is onto much of anything other than introducing error when 3.4 is the opposite direction of eastern orientation.  huh

It is what it is I suppose but frankly - and I realize I've excoriated over this in the past... - anything with the letter "C" in the acronym of a model's title rolls my eyes.   Unless the C does not stand for Climatology, the given model's machinery has to have some sort of a updated "RONI" analog methodology or it's less useful to me.  Obviously ... because we'd be comparing a present day moving goal post on a fixed game field - the two have to change geometry together to be valid... 

No offense to CANSIPs.  LOL   if the model's adaptive than fine.   But even if so ... the orientation as of the Sep 9 update at CPC just looks like a massive pan-dimension heat bomb from eastern 4 clear to the masses of fish adorning the Peruvian shores.  Not sure any sector out there is really crucially 'bigger' in the holistic comparison when the entire thing is nuked like that

 

[40/70 Benchmark]

John, CANSIPS is really only a hair west of the EURO with regard to the orientation of El Nino, at this point.

[40/70 Benchmark]

CANSIPS is def. underselling 1.2 and 3, though, so I see what you are saying. JAMSTEC probably is, too.

[40/70 Benchmark]

Probably at least in part why it's overcooking the Modoki theme.

[Typhoon Tip]

9 minutes ago, 40/70 Benchmark said:

John, CANSIPS is really only a hair west of the EURO with regard to the orientation of El Nino, at this point.

Just op-ed'ing about the use of 'climate'-based tools in general.  No idea what and/or who is performing.

Like I said, if the model's adaptive than the criticism's less necessary there.   But that's something that seldom - in fact never as far as I'm aware ? - goes on in this 'enthusiast's sophistication', which is reporting about the construct of tools. Comparing it's components ( unit test) against empirically what is actually taking place, whereby confidences have foundations... etc. 

yeah right... 

[40/70 Benchmark]

1 minute ago, Typhoon Tip said:

Just op-ed'ing about the use of 'climate'-based tools in general.  No idea what and/or who is performing.

Like I said, if the model's adaptive than the criticism's less necessary there.   But that's something that seldom - in fact never as far as I'm aware ? - goes on in this 'enthusiast's sophistication', and that reporting about the construct of tools, then comparing it's components ( unit test) against empirically what is actually taking place, whereby confidences have foundations.  

yeah right... 

No argument there, it's initialization of the Pacific, both tropical and mid latitudes, has left a great deal to be desired. 

[bluewave]

The C3S model blend had the general temperature spatial pattern correct for the CONUS this summer.

But as usual it was too cool especially where the all-time warmth occurred out West. Plus the ridge axis verified further south which has become a reoccurring theme in recent years.

Many outlets that do seasonal forecasting use this blend as the basis for their forecasts which you see online. That’s why it’s a useful piece of guidance even if it often misses the magnitude of the warmth under the ridges. 

https://climate.copernicus.eu/charts/packages/c3s_seasonal/products/c3s_seasonal_spatial_mm_2mtm_3m?area=area06&base_time=202605010000&type=ensm&valid_time=202606010000

 

 

[Typhoon Tip]

12 minutes ago, bluewave said:

The model blend had the general temperature spatial pattern correct for the CONUS this summer. But as usual it was too cool especially where the all-time warmth occurred out West. Plus the ridge axis verified further south which has become a reoccurring theme in recent years. Many outlets that do seasonal forecasting use this blend as the basis for their forecasts which you see online. That’s why it a useful piece of guidance even if it often misses the magnitudes of the warmth under the ridges.

https://climate.copernicus.eu/charts/packages/c3s_seasonal/products/c3s_seasonal_spatial_mm_2mtm_3m?area=area06&base_time=202605010000&type=ensm&valid_time=202606010000

 

 

stating the obvious...  any "blend" technique is always going to miss the extreme, unless every member contributing to the mean are unilaterally nailing the result/coincidentally.   Since this is never going to be the case, the interquartile members that are middle road and less so, are always offsetting - basic stats.   Hence, blend missing the extremes in the west.

But part of that is also not really attainable anyway?  I don't believe so.  Because the heat in the W was very synergistic in nature.  It's basically combing a handful of inputs that are going to likely gestalt some gain that cannot really be preascertained.  That's how synergistic emergence is... it can't be mathematically predicted in excessively complex systems - at least not yet. Maybe QC will have a say in that some day.  But we have the perennial N/A western bulge in the westerlies, on top of a retrograde IA CGT hemisphere, on top of a terrain that promotes heat and then a subsequent bi-product of that is that the heat in turn bulges heights... on top of then even a kind of non-Markovian feedback when the quantum state of the whole system acts like it "wants" to be hot due to memory from aggregate occurrences...  Those are some of the parts that are creating an outcome > then just their sum.

I guess what I'm asking is, taking out the more extreme/warmest members contributing to any blend, I suspect even those were shy of the scalar ?

[LakePaste25]

9 minutes ago, bluewave said:

The C3S model blend had the general temperature spatial pattern correct for the CONUS this summer.

But as usual it was too cool especially where the all-time warmth occurred out West. Plus the ridge axis verified further south which has become a reoccurring theme in recent years.

Many outlets that do seasonal forecasting use this blend as the basis for their forecasts which you see online. That’s why it’s a useful piece of guidance even if it often misses the magnitude of the warmth under the ridges. 

https://climate.copernicus.eu/charts/packages/c3s_seasonal/products/c3s_seasonal_spatial_mm_2mtm_3m?area=area06&base_time=202605010000&type=ensm&valid_time=202606010000

 

 

This was essentially just a slightly east-shifted version of the long term trends of summer in CONUS. The strongest El Niño on record was unable to shake it. 

 

[bluewave]

12 minutes ago, Typhoon Tip said:

stating the obvious...  any "blend" technique is always going to miss the extreme, unless every member contributing to the mean are both nailing the result.   Since this is never going to be the case, the interquartile members that are middle road are always offsetting -

Even single model forecasts have been missing the magnitudes of the ridges and the warmth underneath. But this is to be expected due to the much lower resolution and nature of long range forecasts beyond 15 days.

In our rapidly warming climate there have been many more opportunities for the models to be biased too cold than too warm.

So it’s beyond technology of present day weather forecast models to have been able to see the historic 600 dm ridge this week from beyond 5-10 days out.

 

 

[Typhoon Tip]

8 minutes ago, bluewave said:

Even single model forecasts have been missing the magnitudes of the ridges and the warmth underneath. But this is to be expected due to the much lower resolution and nature of long range forecasts beyond 15 days.

In our rapidly warming climate there have been many more opportunities for the models to be biased too cold than too warm.

So it’s beyond technology of present day weather forecast models to have been able to see the historic 600 dm ridge this week from beyond 5-10 days out.

 

 

Yeah Chris ...I agree ...I added that exact aspect to that post 

[GaWx]

JAMSTEC Sept ‘26 Nino 3.4 forecast (this is NOT relative): mean is stronger than recent CFSv2 (though CFSv2 updates more recent since their updates are much more frequent) and mean cools more slowly, too, especially into Jan:

[GaWx]

Latest (0Z of Sep 15th) CFSv2 AAM forecast, which is already at a very high >+3, is like yesterday progging an off the charts extreme/historic? +4+ starting by Sep 19th (earliest in the forecast I’ve ever seen) and lasting at least through Sep 24th:

[mitchnick]

1 hour ago, 40/70 Benchmark said:

CANSIPS is def. underselling 1.2 and 3, though, so I see what you are saying. JAMSTEC probably is, too.

We're going to need the Icon to settle this. Lemme' take a close look with my good eye. 

[OSUmetstud]

1 hour ago, 40/70 Benchmark said:

John, CANSIPS is really only a hair west of the EURO with regard to the orientation of El Nino, at this point.

It's about 35°W with the warmest anomalies. 110°W on the Euro v. 145°W on the CANSIPS. Not trivial. 

[OSUmetstud]

[40/70 Benchmark]

4 minutes ago, OSUmetstud said:

It's about 35°W with the warmest anomalies. 110°W on the Euro v. 145°W on the CANSIPS. Not trivial. 

If you read on, I noted that. It's come into much better agreement, though.

[40/70 Benchmark]

The December output def. seems further off than the DJF mean, though, Nick.

[OSUmetstud]

10 minutes ago, 40/70 Benchmark said:

The December output def. seems further off than the DJF mean, though, Nick.

Right. Im not sure what is more relevant. Peak or post-peak during the matching winter months. Cansips is obviously wrong with its anomaly position either way and has been correcting each month for the past 6 or so. 

[GaWx]

 

38 minutes ago, OSUmetstud said:

Compare these to Sep ‘26 warmest JAMSTEC peak period (OND) SSTa longitude of 120-125W, which is between Euro’s 110W and CANSIPS’ 145W. It will be interesting to see which verifies the closest. Do you have a prediction for the warmest longitude at the progged OND peak? Between these 3 model progs, Euro 110W, JAMSTEC 120-125W, and CANSIPS 145W, which do you think will end up the closest?

[40/70 Benchmark]

20 minutes ago, OSUmetstud said:

Right. Im not sure what is more relevant. Peak or post-peak during the matching winter months. Cansips is obviously wrong with its anomaly position either way and has been correcting each month for the past 6 or so. 

I usually go with orientation at peak as most representative of the winter season, so December should be pretty representative. I agree, CANSIPS is off.

[40/70 Benchmark]

@GaWxI would say around 120W.

[LakePaste25]

2 hours ago, 40/70 Benchmark said:

The JAMSTEC pulls of the unicorn scenario that I was discussing with @Typhoon Tip in that it's an east-based event that essentially acts as a Modoki due to the residual cool ENSO influence/Pacific basin warmth biasing the forcing west of where one would expect based on a traditional synthesis of the basin. Very similar to what happened in 2023 when it looked like a Modoki at the surface, but ended up heavily MC oriented....forcing pulled westward. Do I want to bet on that actually happening in the seasonal mean? About as much as I want to be locked in thread alone with @forkyfork, BUT it does confirm my early hedge to advise caution to those forecasting a season largely bereft of winter on the east coast once we get to about mid January and beyond. 

This season appears bipolar to me.

I’m a little more skeptical of the “back loaded Modoki” idea than I was a month ago given how this keeps evolving. What you’re saying is potentially we see an 82-83 outcome which was exactly that. Only issue is we are east of 82 and even 97 on tao/tritton so it’s difficult to just say go with “basin waters are warmer, so it’s further west” when you also have to account for the actual position of the warmest waters being further east. I think i don’t disagree yet but I need another month to see where OLR’s land vs 82 and 97.