2026-2027 Record-breaking El Nino

[Terpeast]

35 minutes ago, mitchnick said:

Average for last 5 days ending 384 hours off the 0z Geps fwiw.

At least its showing up on modeling, now lets see if it actually verifies. 

[GaWx]

OHC monthly rose to a new record high that last month had set (note these aren’t adjusted for GW in a “relative” sense, but they’re still very impressive):

2025   11    -0.02     -0.33      -0.46
2025   12     0.26      0.01      -0.03
2026    1     0.64      0.45       0.49
2026    2     1.08      1.02       1.15
2026    3     1.38      1.39       1.36
2026    4     1.85      2.07       2.24
2026    5     1.67      1.97       1.99
2026    6     1.79      2.14       2.26
2026    7     2.03      2.50       2.96
2026    8     2.19      2.75       3.20

 The following image is for the same longitudes as that for the 3rd column and it jibes fairly well:

Prior to 2026, this had been the record high for the last 2 columns:

1997   10     1.35      2.05       2.56

[GaWx]

 The new relative OISST (not ERSST) has just been released and the change from July jibes well for Nino 3.4 with the slope of the graphs we’ve been following (though as I’ve also been saying the weeklies/last 2 months are notably cooler than what the cyclonicwx graphs have been suggesting).

2026    7      2.84     1.73     0.34      1.38
2026    8      3.35     2.52     0.11      1.78

https://www.cpc.ncep.noaa.gov/data/indices/rel_mthsst9120.txt
 
  This 1.78 is a new record high (back to 1981) for August Nino 3.4 as it barely beats the 1.74 of 1997. Here are the warmest Augusts since 1981:

2026 1.78
1997 1.74
2015 1.30
1987 1.24
1982 1.19
1991 0.82
2002 0.80
1994 0.72
2004 0.71
2023 0.63

 Note that Nino 3 rose twice as much and Nino 4 fell 0.23! Also, Nino 1+2 rose the 2nd most at 0.51. All of this indicates that this Nino has been becoming increasingly E based and that the W portion of Nino 3.4 is what is holding it back vs what model projections had been.

 Also, note that Nino 3.4 rose only 0.4C from the July avg to the Aug avg. I say “only” because several models were suggesting (as I’ve posted about several times) a record rapid rise for 3.4 of ~1C. It remains to be seen what the rise was per ERSST, but it’s clear that August didn’t warm nearly as much in 3.4 as model consensus had suggested, which has caused some at least near-term cooling of progs in some models’ recent runs.

 Here’s the latest OISST relative 3.4 chart through Sept 1st: STILL no sign of resumption of warming with it practically flat the last 31 days!

 

The CDAS (not relative) agrees with the continuing flatness:

[GaWx]

To add to the above, here are Nino 3 and Nino 4 (note that these aren’t relative, but that’s ok since my point is to show the slopes):

Nino 3: still a slow rise through Aug instead of nearly flat like 3.4:

Nino 4: had been falling through Aug 17th though a bumpy net slow rise since:

[MJO812]

7 hours ago, qg_omega said:

Record warmth summer and winter 

Screenshotted

[Stormchaserchuck1]

6 hours ago, 40/70 Benchmark said:

@Stormchaserchuck1

Thanks! They don't have SST so I will still have to find something for May-Sept N. Atlantic SSTA index. 

Here's that arctic -H5 for the warm season this year

90N.. just where my reverse correlation for the Winter is the highest. Expect 90N H5 ridging for the DJFM mean. 

[Stormchaserchuck1]

2 hours ago, mitchnick said:

Average for last 5 days ending 384 hours off the 0z Geps fwiw.

Yeah.. better models don't really show it. 30 days until the correlation increases by 3-4x!

[LakePaste25]

1 hour ago, GaWx said:

 The new relative OISST (not ERSST) has just been released and the change from July jibes well for Nino 3.4 with the slope of the graphs we’ve been following (though as I’ve also been saying the weeklies/last 2 months are notably cooler than what the cyclonicwx graphs have been suggesting).

2026    7      2.84     1.73     0.34      1.38
2026    8      3.35     2.52     0.11      1.78

https://www.cpc.ncep.noaa.gov/data/indices/rel_mthsst9120.txt
 
  This 1.78 is a new record high (back to 1981) for August Nino 3.4 as it barely beats the 1.74 of 1997. Here are the warmest Augusts since 1981:

2026 1.78
1997 1.74
2015 1.30
1987 1.24
1982 1.19
1991 0.82
2002 0.80
1994 0.72
2004 0.71
2023 0.63

 Note that Nino 3 rose twice as much and Nino 4 fell 0.23! Also, Nino 1+2 rose the 2nd most at 0.51. All of this indicates that this Nino has been becoming increasingly E based and that the W portion of Nino 3.4 is what is holding it back vs what model projections had been.

 Also, note that Nino 3.4 rose only 0.4C from the July avg to the Aug avg. I say “only” because several models were suggesting (as I’ve posted about several times) a record rapid rise for 3.4 of ~1C. It remains to be seen what the rise was per ERSST, but it’s clear that August didn’t warm nearly as much in 3.4 as model consensus had suggested, which has caused some at least near-term cooling of progs in some models’ recent runs.

 Here’s the latest OISST relative 3.4 chart through Sept 1st: STILL no sign of resumption of warming with it practically flat the last 31 days!

 

The CDAS (not relative) agrees with the continuing flatness:

1+2 has been marching back upward and the 20C isotherm in the subsurface itself has still been pressing deeper in the 1+2 region (i can’t show the GIF here due to file size limits), so if anything it’s been getting even more east-based despite Nino 3.4 flattening. I am keeping an eye out for a sudden jump in 3.4 because of this. 

 

 

[mitchnick]

31 minutes ago, Stormchaserchuck1 said:

Yeah.. better models don't really show it. 30 days until the correlation increases by 3-4x!

If you're referring to the Eps, I agree. Except you can see a dip in the height lines in the GOA not too far from the one on the Geps. Suggesting maybe it might be showing some blue in the GOA in future runs....OK, look real hard then! Lol

[Terpeast]

1 hour ago, LakePaste25 said:

1+2 has been marching back upward and the 20C isotherm in the subsurface itself has still been pressing deeper in the 1+2 region (i can’t show the GIF here due to file size limits), so if anything it’s been getting even more east-based despite Nino 3.4 flattening. I am keeping an eye out for a sudden jump in 3.4 because of this. 

 

 

It will be interesting to see how the negative PDO interacts with the east based nino. Historically east based isn’t good for us, especially if the PDO is positive because too strong of a GOA low would flood the continent with mild air cutting off any polar flow. How will the negative PDO interacts with that is the big question. GOA low still amplified with a stronger jet further north? Or weaker and further west low that allows polar air to move south? I think we’re in uncharted territory and the analogs can only do so much here. 

[Stormchaserchuck1]

28 minutes ago, Terpeast said:

It will be interesting to see how the negative PDO interacts with the east based nino. Historically east based isn’t good for us, especially if the PDO is positive because too strong of a GOA low would flood the continent with mild air cutting off any polar flow. How will the negative PDO interacts with that is the big question. GOA low still amplified with a stronger jet further north? Or weaker and further west low that allows polar air to move south? I think we’re in uncharted territory and the analogs can only do so much here. 

I understand that has happened in the past, but I would still prefer a +PDO in east-based Nino. Remember that correlates with warm water in the GOA and off the West Coast. 

PDO is further west than east-based Nino position, and doesn't necessarily only fuel it. It can shift things a bit west if strong enough

Some of the east-based Nino composite is warm in the NE because of +NAO (which is a more random part - if anything warmer 10mb in +ENSO may slightly favor -nao over a large enough sample). 

Jan 1998 is an anomaly, but I'm willing to say if that upper latitude pattern happens 10 times in east-based Nino it's cold in the east 8/10. Some times things just don't line up like they usually do.  See Dec 2022 and March 2023 as 2 good examples of that. It's only because the correlation isn't 100%.