2026-2027 Record-breaking El Nino

[OSUmetstud]

If you want it, I was able to do some area averages in QGIS myself for 80 °N relative to ERA5 1991–2020 Normals

May -0.2 °C

June -0.4 °C

July -0.2 °C

August 0.9 °C

Another piece of evidence—Arctic Sea ice extent is near the 2011–2020 average. It was not a crazy melt summer but it was not particularly cold, either. 

 

[LakePaste25]

1 hour ago, bluewave said:

The Euro issues an extended forecast every two months. These really long range forecasts aren’t always the most accurate.

But Nino 1.2 has reached a new all-rime high of +5.345. This is even beyond what the Euro was forecasting. So the warm pool in the EPAC is stronger than anything that we have seen.

This makes it plausible that this event could linger at least into next June. Even models like the CFS which had this peaking in +1.2 at only +4.5 are showing it going into at least next June in 1+2.  

The record breaking WWBs just keep pushing the warm pool further east than model forecasts. So such an extreme heat build up in the EPAC will take to wind down.

 

As expected, the CFSv2 is going to be way off with Nino 1+2, forecasting a drop that never came. 

 

[bluewave]

9 minutes ago, LakePaste25 said:

As expected, the CFSv2 is going to be way off with Nino 1+2, forecasting a drop that never came. 

 

While the Euro is doing better with +1.2, it’s still too low also. Probably related to the models underestimating the strength of the WWBs in the longer range. The next Euro update on the 5th should be interesting when the warmer current conditions are incorporated into the new forecast. The EPS weeklies have been closer.

 

 

 

[Stormchaserchuck1]

1 hour ago, mitchnick said:

Did you notice that the 3 ensembles as of 0z runs continue showing a decently +NAO thru mid October? 

@Stormchaserchuck1

Yeah, October is the only month of the year that has an opposite correlation to DJFM NAO. I don't know if that's true given 250+ years of examples though. 

[GaWx]

4 hours ago, OSUmetstud said:

This plot is not consistent with 2 m temperature being 2 C below the much older 1958 to 2002 climatology all summer. This area is all ocean/ice. 

 Because the plot he posted was for May-Aug, I’ll analyze that period. Looking at the image below, May-Aug averaged ~1.3C, not ~2C, below the much older 1958-2002 climo: May was ~-2C, June was ~-2C, July was ~-1C, and Aug was ~-0.1C. This means ~-1.3C for the avg of May-Aug rather than -2C. However, I do realize that -1.3C vs 1958-2002 would be a good bit more negative vs 1991-2020….perhaps as much as ~-3C?? So, that would be colder than what that 925 mb anomaly map suggests for 80N+, which I roughly estimate to be -1.25C:

 

 However, that image is for 925 mb, not 2m. Isn’t it possible that the 2m anom was colder than the 925 mb anom?

[GaWx]

3 hours ago, OSUmetstud said:

The ecmwf was upgraded on May 12, 2026. What do you notice?

IFS users will also see the introduction of the new NEMO4-SI³ ocean and sea-ice model, bringing substantial improvements in forecasts and analysis over the marine areas, as well as over 40 new ocean and sea ice-related variables. Key improvements include a more accurate representation of how sea ice affects the power of waves and the interactions between waves and ocean currents, often at play in rougher seas. Other improvements for IFS include improved tropical upper-air temperature and wind forecasts, and improvements to temperature and humidity forecasts around the tropopause.

 

https://www.ecmwf.int/en/about/media-centre/news/2026/ifs-cycle-50r1-aifsv2-live

What's likely happening here is actually the new ecmwf 2 m temperatures over the Arctic are more accurate as the near surface inversion is more properly coupled to the ice and the cold ocean. The old 2 m ERA40 climatology is what's too warm. 

 Due to the timing of the upgrade, I’m also quite suspicious that this Euro 5/12/26 upgrade MAY have contributed to the colder 2026 vs previous summers. So, I do see what you’re saying. Thats why I went to that other forum and spent a couple of hours to see if anyone pointed to that. But nobody did in those hundreds of posts from very knowledgeable folks, including very experienced pro-mets. If this Euro upgrade is really a contributor or possibly even the main cause, don’t you think at least one person there would have mentioned that?

 That’s why I’d prefer to keep an open mind until there’s decent agreement among the met. community. Do you see anything wrong with that? You may be right and you’d then deserve major kudos, but I’d first Iike to see some consensus. But even if you’re right, can’t 2026 still be a cold outlier but to a lesser extent?

[TheClimateChanger]

11 hours ago, GaWx said:

Thanks for posting about this.

 After reading your post, I decided it would be worthwhile to check out another forum (which includes other pro-mets):

https://forum.arctic-sea-ice.net/index.php

 I read hundreds of posts in their “The 2026 melting season” thread. The first post I saw mentioning the unusually cold DMI 80+ N temperatures was on 5/25/26 (post #268) on the following page (p. 6) by very experienced pro-met Niall Dollard:

https://forum.arctic-sea-ice.net/index.php/topic,4507.250.html

 Niall said, “I suspect we will see a rapid increase on the DMI N80 graph in the next day or two, as that deep low advects warmth into the general DMI N80 area near the pole. 

Unusually, the daily mean has been below the 1958-2002 mean for much of the month of May.”

————————
  Then go to p. 9 and look at post #474 on 7/16/26 from pro-met Glen Koehler in here:

https://forum.arctic-sea-ice.net/index.php/topic,4507.450.html

Glen said: “The DMI 80N values are unprecedented.  I looked at every annual chart for 1958-2026, and no other year has such a large and prolonged period of below 2C temperature during June and July.”  
 
and

”The 2026 DMI 80N chart is so anomalous that it makes me wonder if a sensor or data processing change is affecting the 80N chart in summer 2026.  Anoother possibility is that the slushification of the ice near the North pole is creating a different relationship between ice and 2M air temperature.”

————————

Then @TheClimateChanger, who you may recall used to go by “Frivolous” here, posted in #490 (also on p.9) as “Frivolousz21” on 7/18/26 as a reply to Glen:

”It's not from an error. It's just that there has been a relentless PV anomaly and essentially next to no unimpeded solar insolation.  

 It's been pretty crazy.  This might end up the best year for ice retention since 2006.”

————————

 I went through the rest of the thread and there never was agreement on the cause. So, it remains a mystery there.

 I respect your knowledge and appreciate your post on this. But I plan to continue posting these DMI graphs at least until if/when there’s strong agreement in the overall pro-met community with you that 2026 really isn’t as cold relative to past years as the graphs show.

——————

 Today’s daily update had another sharp cooling (~1C) from yesterday keeping it colder than the blue line’s climo:

 I just looked at the model consensus and it is, indeed, solidly BN N of 80N. However, those same models are showing significant warming to AN very quickly. So, I’m expecting it to go back above the blue line within just a couple of days at most.

Not me. Friv was from Saint Louis. I think he did use a handle with global warming or something, so I can see the confusion.

[GaWx]

 As I said, I’ve been following model initializations of 2m at 80+ N. They were colder than climo most of late May-early Aug.

 Check out those of the Euro for yesterday: clearly solidly BN vs 1981-2010 for 80N+ on Sep 29th (associated with a storm)(they were also quite cold going back to Sep 23rd by the way), which may actually be cold enough to support the ~2C colder than the 1958-2002 mean shown on the newest DMI graph:

 

 Here’s the Euro forecast for Oct 2nd, which shows notable warming as the storm moves S into Siberia. This is why I’m expecting warming on the DMI graphs by then:

@so_whats_happening@OSUmetstud

[OSUmetstud]

15 minutes ago, GaWx said:

 Due to the timing of the upgrade, I’m also quite suspicious that this Euro 5/12/26 upgrade MAY have contributed to the colder 2026 vs previous summers. So, I do see what you’re saying. Thats why I went to that other forum and spent a couple of hours to see if anyone pointed to that. But nobody did in those hundreds of posts from very knowledgeable folks, including very experienced pro-mets. If this Euro upgrade is really a contributor or possibly even the main cause, don’t you think at least one person there would have mentioned that?

 That’s why I’d prefer to keep an open mind until there’s decent agreement among the met. community. Do you see anything wrong with that? You may be right and you’d then deserve major kudos, but I’d first Iike to see some consensus. But even if you’re right, can’t 2026 still be a cold outlier but to a lesser extent?

It's bad faith to make strong conclusions from the original because a) it's not comparing the reanalysis to the reanalysis climatology and b) there's an obvious change after the ocean and sea-ice upgrade. 

My analysis should speak for itself and I am confident in my conclusions. I don't need some other posters in another forum to back me up. 

 

[mitchnick]

1 hour ago, Stormchaserchuck1 said:

Yeah, October is the only month of the year that has an opposite correlation to DJFM NAO. I don't know if that's true given 250+ years of examples though. 

Don't break my heart like that.

[40/70 Benchmark]

1 hour ago, Stormchaserchuck1 said:

Yeah, October is the only month of the year that has an opposite correlation to DJFM NAO. I don't know if that's true given 250+ years of examples though. 

Yea, I saw that post concerning October and viewed it as good news. I still remember when you schooled me on that as a weenie back in the fall of 2006....we had this big block in October and I got all excited, and you went on about how it meant +NAO later and how it would be an epically warm winter.....rest is history.

[40/70 Benchmark]

1 hour ago, bluewave said:

While the Euro is doing better with +1.2, it’s still too low also. Probably related to the models underestimating the strength of the WWBs in the longer range. The next Euro update on the 5th should be interesting when the warmer current conditions are incorporated into the new forecast. The EPS weeklies have been closer.

 

 

 

I bet we see the PDO ascend to near neutral between now and the end of November.

[Stormchaserchuck1]

It looks like May-Aug NAO was -0.75/month and if October ends up +.. that is a formula for a more -NAO Winter. 

Plus since 2012 90N May-Sept SLP has reverse for the Winter every time, and it was strong neg anomaly this Summer. 

[40/70 Benchmark]

6 minutes ago, Stormchaserchuck1 said:

It looks like May-Aug NAO was -0.75/month and if October ends up +.. that is a formula for a more -NAO Winter. 

Plus since 2012 90N May-Sept SLP has reverse for the Winter every time, and it was strong neg anomaly this Summer. 

Geez...that is convincing...I hate to go against the descending solar/W QBO/multidecadal trend, but that makes it tough.

What you mean by -.75 PER MONTH.....isn't -.75 the DJFM aggregate mean per the formula?

[Stormchaserchuck1]

1 hour ago, 40/70 Benchmark said:

I bet we see the PDO ascend to near neutral between now and the end of November.

Not with this. -PDO is cold water in the GOA. This is also anti-El Nino October

[Stormchaserchuck1]

4 minutes ago, 40/70 Benchmark said:

Geez...that is convincing...I hate to go against the descending solar/W QBO/multidecadal trend, but that makes it tough.

What you mean by -.75 PER MONTH.....isn't -.75 the DJFM aggregate mean per the formula?

May, June, Aug, Sept NAO total was -3.00.. -0.75/mo. 3/4 months negative. 

I think the overall correlation from -0.75 MJJA to DJFM is -0.07/8 lol. 

[40/70 Benchmark]

9 minutes ago, Stormchaserchuck1 said:

May, June, Aug, Sept NAO total was -3.00.. -0.75/mo. 3/4 months negative. 

I think the overall correlation from -0.75 MJJA to DJFM is -0.07/8 lol. 

Okay....so DJFM mean NAO of -.75...so 50% chance of having a mean DJFM NAO between -.25 and -1.25.

[40/70 Benchmark]

We have been dead-on past few years, but I may have to stray a bit this year lol

[Stormchaserchuck1]

4 minutes ago, 40/70 Benchmark said:

Okay....so DJFM mean NAO of -.75...so 50% chance of having a mean DJFM NAO between -.25 and -1.25.

No, May-June-July-Aug NAO was -0.75 per month, that comes out to -0.07/month for DJFM, or DJFM total of -0.28, based on 75 years. Weak correlation, but generally those 4 months negative then Oct pos is an interesting point for more winter -nao tendency. 

[LakePaste25]

19 minutes ago, Stormchaserchuck1 said:

Not with this. -PDO is cold water in the GOA. This is also anti-El Nino October

 Close to Oct 1997 and 2015. The only difference is those years had more ridging north of the Bering sea. These were +PDO Ninos. 

 

[Stormchaserchuck1]

In the whole picture I'm not seeing H5 similarities to 1997 and 2015. 

[40/70 Benchmark]

4 minutes ago, Stormchaserchuck1 said:

No, May-June-July-Aug NAO was -0.75 per month, that comes out to -0.07/month for DJFM, or DJFM total of -0.28, based on 75 years. Weak correlation, but generally those 4 months negative then Oct pos is an interesting point for more winter -nao tendency. 

Ohhhhh...that makes more sense...so 50% of between -.78 and +.32...that is more palatable for me, but I would still hedge higher.

[40/70 Benchmark]

Just now, Stormchaserchuck1 said:

In the whole picture I'm not seeing H5 similarities to 1997 and 2015. 

I hate the 1997 analog for everything but ENSO....2015 is good from a polar standpoint.

This is the summer analog H5 composite I came up with.