2025-2026 ENSO

[A-L-E-K]

ready 4 spring

[PhiEaglesfan712]

41 minutes ago, GaWx said:

 Tweets are often exaggerations as we know. The new Euro for March is still mainly warmer than normal but no torch as it’s mainly only slightly AN since it’s cooler than the forecast from a month ago:

Last month’s Euro for March:

 

Today’s Euro for March: not as mild with only ~1F AN NE, 2 AN DC, Chicago, Detroit, and RDU, and 2-2.5 AN ATL/MEM for example:

 

April also came in cooler than it had last month.

 I am wary, however, about WxBell Euro 2m temp anomalies having a cold bias due to their algos. So, you probably have to add at least 1F to what I said to adjust for that bias. So, more like at least 2F AN NE and at least 3-3.5 AN in the other cities I listed. So, again not as warm as last month’s forecast and not really a torch, but still fairly mild especially after the bias correction. I can’t find March by itself at the ecmwf site, itself, unfortunately.

 

Here's the April 2015 temperature anomoly: 

This map doesn't look much different than those March 2026 Euro maps.

[MJO812]

Weeklies for last week of February into March look stormy and cold 

[LakePaste25]

16 minutes ago, MJO812 said:

Weeklies for last week of February into March look stormy and cold 

 

8 minutes ago, snowman19 said:

 

 

I highlighted a colder risk for the last third of february this morning and gave reasons…low-medium confidence for now. I’m also a little too regionally biased/focused with this because even if the blocking pattern breaks down and a cutter pattern develops, i tend to be in a better spot being 300-400 miles west of I-95 in the eastern great lakes. 

[GaWx]

4 hours ago, MJO812 said:

Weeklies for last week of February into March look stormy and cold 

Anthony,

 It’s important to be aware that the WxBell Euro EW 2m maps have a notable 2m temp cold bias as I’ve posted about repeatedly. I have an easy way to show this by posting the in-house Euro weeklies maps covering the same period and comparing to this WB map.

Examples:

 -The Rockies and west coast are a very cold -3F to -10F for 2/21-3/22 on your map. What?!?
 
 -Chicago/OH Valley are -2F, Asheville is -3F, Detroit is -4F, DC/Richmond are -3F, ATL/CLT are -1F.

 -Now check out ECMWF’s in-house maps for the truth: no BN on any maps covering the same period for the locations I listed

2/23-3/1:

 

3/2-8:

 

3/9-15:

 

3/16-22:

 So based on the truth, the Rockies and much of the West should be a bit AN instead of well BN. So, that means they’re mainly 5F to 10F too cold on WB!

 The E cities I listed should be NN in Chic, Det, Ohio Valley, DC, Richmond, and +1F in ATL/CLT/ASH. So, that means these locations are 2-4F too cold on WB.

 WxBell has NYC at -4F. The truth is that they should be -1F. So, NYC is 3F too cold on WB.

@donsutherland1this remains a problem

[michsnowfreak]

6 hours ago, PositiveEPOEnjoyer said:

They really haven't though. 2024 featured the 6th warmest April on record at Detroit, and your last "cold" April was in 2022. Even then, it was a mere 2 degrees below average, so nothing to write home about. I haven't said anything about how April will shape up this year as I have no idea what will happen (way too early), but if we're basing forecasts on how our recent weather months shaped up (which isn't entirely correct to do), I'd def lean warmer than average. Also still waiting to see how this supposed PV split will shake up the pattern. Cold may favor the West instead (assuming it even does happen).

April is one of my least favorite months (Apr and Aug). I was recalling recent April snowfalls and frigid Opening Days moreso than the monthly temp departure. A closer look shows that in the last 10 Aprils, 4 were colder than avg, 1 avg, 5 warmer than avg. 

But regardless my stance remains. Calling for a torch month 2 months out with no basis other than a gut feeling is automatically wrong. I wouldn't have even responded if someone said "id lean warm for April". But torch? Ill wait and see.

[LakePaste25]

41 minutes ago, GaWx said:

Anthony,

 It’s important to be aware that the WxBell Euro EW 2m maps have a notable 2m temp cold bias as I’ve posted about repeatedly. I have an easy way to show this by posting the in-house ECMWF weeklies covering the same period and comparing to this WxBell map.
 

Examples:

 -The Rockies and west coast are a very cold -3F to -10F for 2/21-3/22 on your map. What?!?
 
 -Chicago/OH Valley are -2F, Asheville is -3F, Detroit is -4F, DC/Richmond are -3F, ATL/CLT are -1F.

 - Now check out the ECMWF’s in-house maps for the truth: no BN on any of the maps covering the same period for the locations I listed

2/23-3/1:

 

3/2-8:

 

3/9-15:

 

3/16-22:

 So based on the truth, the Rockies and much of the West should be a bit AN instead of well BN. So, that means they’re mainly 5F to 10F too cold on WxBell!

 The E cities I listed should be NN in Chic, Det, Ohio Valley, DC, Richmond,  and ~+1F in ATL/CLT/ASH. So, that means that these locations are 2-4F too cold on WxBell.

 WxBell has NYC at -4F. The truth is that they should be ~-1F. So, NYC is 3F too cold on WxBell.

@donsutherland1this remains a problem

I believe the ECMWF native site is still using 1981-2010 climate normals, while the wxbell site is using the warmer 1991-2020 normals. The actual 2m temps should be the same, but the newer 1991-2020 normals will appear colder because the baseline is warmer, which is what we should expect in a warmer climate. 

Raw 2m temps are produced by the models themselves. It’s up to what dataset you want to use for the anomalies. 

[TheClimateChanger]

52 minutes ago, LakePaste25 said:

I believe the ECMWF native site is still using 1981-2010 climate normals, while the wxbell site is using the warmer 1991-2020 normals. The actual 2m temps should be the same, but the newer 1991-2020 normals will appear colder because the baseline is warmer, which is what we should expect in a warmer climate. 

Raw 2m temps are produced by the models themselves. It’s up to what dataset you want to use for the anomalies. 

Shouldn't be that dramatic of a discrepancy though... the 1991-2020 normals are generally ~0.5F, maybe 1F warmer, than the 1981-2010 normals for most places. After all, they share 20 years.

[TheClimateChanger]

5 hours ago, GaWx said:

It’s the WxBell CFS, not their Euro, that has those weird issues. WxBell Euro just has a straight 2m cold bias. And you must have meant a S. Lake Michigan - departure (not Superior, which actually typically has the opposite, a relative + departure in the E portion) just as this CFS has:

Yeah, now that I think about it... I think it's there CDAS or whatever that shows the past anomalies that always seems to show an anomalously cooler area over Lake Superior. But that might be a problem with the lake temperature dataset.

[MJO812]

[LakePaste25]

9 minutes ago, MJO812 said:

Who said there was going to be a torch in Feb? Seems like a straw man. 

[LakePaste25]

34 minutes ago, TheClimateChanger said:

Shouldn't be that dramatic of a discrepancy though... the 1991-2020 normals are generally ~0.5F, maybe 1F warmer, than the 1981-2010 normals for most places. After all, they share 20 years.

Happy to dig deeper into this. I always thought actual 2m temps were universal vs. the differing anomaly datasets. 

[GaWx]

1 hour ago, LakePaste25 said:

I believe the ECMWF native site is still using 1981-2010 climate normals, while the wxbell site is using the warmer 1991-2020 normals. The actual 2m temps should be the same, but the newer 1991-2020 normals will appear colder because the baseline is warmer, which is what we should expect in a warmer climate. 

Raw 2m temps are produced by the models themselves. It’s up to what dataset you want to use for the anomalies. 

 There’s no indication that the Euro native site is using 1981-2010 for its climo base. Is it possible you’re mixing up Tropical Tidbits with the Euro native base? TT does use 1981-2010 as that’s what’s shown on many of their maps. Regarding the Euro native site itself, it refers to the last 20 years for its climo base. So, that means a base of 2006-2025 for the in-house Euro maps. Compared to the 1991-2020 base of WxBell, that means the climo base for the native Euro maps is actually warmer than rather than colder than WB. So, if the only difference between WB and native Euro were the climo base and WB didn’t have algorithm issues, then the WB anomaly maps would actually be a little warmer rather than colder than the native Euro maps. So, that means that the WB cold bias is actually even worse rather than better when considering climo base differences!

 So, there really is a big problem with WB Euro maps. Thus, if anything, WB Euro Weekly NYC 2m anomalies on that map are >3F too cold…perhaps closer to 4F too cold! And don’t forget that out west, that WB map is 5-10F too cold before considering climo base differences based on what I just showed! @MJO812@donsutherland1

[LakePaste25]

33 minutes ago, GaWx said:

 There’s no indication that the Euro native site is using 1981-2010 for its climo base. Is it possible you’re mixing up Tropical Tidbits with the Euro native base? TT does use 1981-2010 as that’s what’s shown on many of their maps. Regarding the Euro native site itself, it refers to the last 20 years for its climo base. So, that means a base of 2006-2025 for the in-house Euro maps. Compared to the 1991-2020 base of WxBell, that means the climo base for the native Euro maps is actually warmer than rather than colder than WB. So, if the only difference between WB and native Euro were the climo base and WB didn’t have algorithm issues, then the WB anomaly maps would actually be a little warmer rather than colder than the native Euro maps. So, that means that the WB cold bias is actually even worse rather than better when considering climo base differences!

 So, there really is a big problem with WB Euro maps. Thus, if anything, WB Euro Weekly NYC 2m anomalies on that map are >3F too cold…perhaps closer to 4F too cold! And don’t forget that out west, that WB map is 5-10F too cold before considering climo base differences based on what I just showed! @MJO812@donsutherland1

All 2m temp forecasts are the same though. It’s just how the anomalies are interpreted. 

Would agree that EPS are running slightly cold of GEFS/GEPS suites. 

[so_whats_happening]

12 hours ago, PhiEaglesfan712 said:

The law of averages says a warm month has to happen at some point. The Eastern US hasn't had an above average month since September, and a near normal month since November.

With the exception of 2013-14 (and even that had a warm May and June), all of the big cold years produced torch months at some point in the spring. 2010 (from March-September), 2011 (from April-July), and 2015 (in May). Even last year, March produced a Top 10 warmest month, and 2018 had the anomolously warm February, as well as a Top 10 warmest May.

By using that same logic the law of averages would assume we would have extensive cooler than average year(s) coming up too since we have been an endless year after year warm regime.

[GaWx]

33 minutes ago, LakePaste25 said:

All 2m temp forecasts are the same though. It’s just how the anomalies are interpreted. 

Would agree that EPS are running slightly cold of GEFS/GEPS suites. 

 I hate to see such large discrepancies in anomalies when comparing sources. It’s not comforting to say the least.

 Maybe they are and I have no evidence to the contrary at least right now, but how do you know for sure that the 2m temp forecasts themselves between Euro native and WB are the same?

[so_whats_happening]

On 1/23/2026 at 10:35 AM, so_whats_happening said:

Phase 8 in February starts to get ok but phase 1 in February in Nina is a large warm up. If we have more of a neutral state rather than La Nina by then with a passage through phase 1 in February it is a bit more muted of a warm up. Looks fun though coming up for at least the next maybe 10 days.

https://www.meteonetwork.it/models/mjo/

Unfortunately those 10 or so days were not all that fun for here. Eastern NC though I'm sure was just in love with this pattern lol.

From the looks of the Nina phase 1 passage is taking place this weekend. Large cold dump with warming through much of the plains showing up. Going forward it does show cold building back up into Canada to be unleashed again at some point maybe the last week or so of February into early March as we move into Phase 2/3?

[nycwinter]

dt says next couple of weeks going to be bad if you want snow in the east

 

[snowman19]

[donsutherland1]

9 hours ago, GaWx said:

 There’s no indication that the Euro native site is using 1981-2010 for its climo base. Is it possible you’re mixing up Tropical Tidbits with the Euro native base? TT does use 1981-2010 as that’s what’s shown on many of their maps. Regarding the Euro native site itself, it refers to the last 20 years for its climo base. So, that means a base of 2006-2025 for the in-house Euro maps. Compared to the 1991-2020 base of WxBell, that means the climo base for the native Euro maps is actually warmer than rather than colder than WB. So, if the only difference between WB and native Euro were the climo base and WB didn’t have algorithm issues, then the WB anomaly maps would actually be a little warmer rather than colder than the native Euro maps. So, that means that the WB cold bias is actually even worse rather than better when considering climo base differences!

 So, there really is a big problem with WB Euro maps. Thus, if anything, WB Euro Weekly NYC 2m anomalies on that map are >3F too cold…perhaps closer to 4F too cold! And don’t forget that out west, that WB map is 5-10F too cold before considering climo base differences based on what I just showed! @MJO812@donsutherland1

Yes. That's correct. The ECMWF is actually using a potentially somewhat warmer period than 1991-2020 for calculating its subseasonal anomalies. The site mentions:

The mean anomalies are derived from the ECMWF Sub-seasonal range ensemble consisting of 100 ensemble members plus a control member and averaged over a seven day period. Anomalies are calculated with respect to the Model Climate, which is derived from re-running an 11-member ensemble over the last 20 years, giving 220 realisations in total.

See bottom of: https://charts.ecmwf.int/products/opencharts_extended_meteogram?base_time=202602050000&lat=51.4333&lon=-1.0&station_name=Reading

A fuller discussion, which also notes 20 years of values, is here: https://confluence.ecmwf.int/display/FUG/Section+5.3.2+SUBS-M-climate%2C+the+sub-seasonal+model+climatehttps://confluence.ecmwf.int/display/FUG/Section+5.3.2+SUBS-M-climate%2C+the+sub-seasonal+model+climate

Below is a quick comparison of the 1991-2020 period with the last 20 years for 10 select cities to provide an idea of how things have changed.

 

[jconsor]

Top 10 cold period past two and a half weeks for eastern US:

https://x.com/wxjerdman/status/2019729810914992195

[michsnowfreak]

24 minutes ago, jconsor said:

Top 10 cold period past two and a half weeks for eastern US:

https://x.com/wxjerdman/status/2019729810914992195

Wow thanks for posting this. I knew it was cold but didn't realize the impressive nature of it.

@bluewave frequently mentions that cold anamolies are short in duration. This is the opposite. January 15 - February 4 is the coldest 3 weeks of the year on avg to begin with. 2026 now ranks 3rd coldest for that 3 week stretch at Detroit in 153 years of record. Only 1918 & 1963 were colder.

Pittsburgh also ranks 3rd in 152 years of record. Cleveland 4th in 156 years. Toledo 6th in 153 years. And even NYC 17th in 158 years.

 @FPizz @mitchnick @MJO812

[EastonSN+]

2 hours ago, snowman19 said:

 

Great - another cold and rainy spring

[EastonSN+]

12 hours ago, GaWx said:

Anthony,

 It’s important to be aware that the WxBell Euro EW 2m maps have a notable 2m temp cold bias as I’ve posted about repeatedly. I have an easy way to show this by posting the in-house Euro weeklies maps covering the same period and comparing to this WB map.

Examples:

 -The Rockies and west coast are a very cold -3F to -10F for 2/21-3/22 on your map. What?!?
 
 -Chicago/OH Valley are -2F, Asheville is -3F, Detroit is -4F, DC/Richmond are -3F, ATL/CLT are -1F.

 -Now check out ECMWF’s in-house maps for the truth: no BN on any maps covering the same period for the locations I listed

2/23-3/1:

 

3/2-8:

 

3/9-15:

 

3/16-22:

 So based on the truth, the Rockies and much of the West should be a bit AN instead of well BN. So, that means they’re mainly 5F to 10F too cold on WB!

 The E cities I listed should be NN in Chic, Det, Ohio Valley, DC, Richmond, and +1F in ATL/CLT/ASH. So, that means these locations are 2-4F too cold on WB.

 WxBell has NYC at -4F. The truth is that they should be -1F. So, NYC is 3F too cold on WB.

@donsutherland1this remains a problem

Just for the northern MA and Northeast, thats a decent look for BN temps, albeit slightly.

[EastonSN+]

6 hours ago, so_whats_happening said:

Unfortunately those 10 or so days were not all that fun for here. Eastern NC though I'm sure was just in love with this pattern lol.

From the looks of the Nina phase 1 passage is taking place this weekend. Large cold dump with warming through much of the plains showing up. Going forward it does show cold building back up into Canada to be unleashed again at some point maybe the last week or so of February into early March as we move into Phase 2/3?

Hopefully phase 3 occurrs by 1st week of March (considering lag).

[Violentweatherfan]

6 hours ago, nycwinter said:

dt says next couple of weeks going to be bad if you want snow in the east

 

He also pointed out how off the forecast was for the MJO. It’s definitely going in the wrong direction so here is to hoping if is off about that.

[George001]

22 hours ago, MJO812 said:

February doesnt look warm at all for the east. Same thing has been happening all winter. Looks warm in the east in the long range only to correct colder.

I agree it will end up being a cold and possibly snowy month overall, but I do agree with the idea of a mid month relaxation period.

[TheClimateChanger]

3 hours ago, donsutherland1 said:

Yes. That's correct. The ECMWF is actually using a potentially somewhat warmer period than 1991-2020 for calculating its subseasonal anomalies. The site mentions:

The mean anomalies are derived from the ECMWF Sub-seasonal range ensemble consisting of 100 ensemble members plus a control member and averaged over a seven day period. Anomalies are calculated with respect to the Model Climate, which is derived from re-running an 11-member ensemble over the last 20 years, giving 220 realisations in total.

See bottom of: https://charts.ecmwf.int/products/opencharts_extended_meteogram?base_time=202602050000&lat=51.4333&lon=-1.0&station_name=Reading

A fuller discussion, which also notes 20 years of values, is here: https://confluence.ecmwf.int/display/FUG/Section+5.3.2+SUBS-M-climate%2C+the+sub-seasonal+model+climatehttps://confluence.ecmwf.int/display/FUG/Section+5.3.2+SUBS-M-climate%2C+the+sub-seasonal+model+climate

Below is a quick comparison of the 1991-2020 period with the last 20 years for 10 select cities to provide an idea of how things have changed.

 

Wow, that's quite impressive IMO, considering 1991-2020 was, by far, the warmest climatological normal period in recorded history. And the last 20 years are outpacing them by 0.6-1F across the board. One would think the warmer norms would tilt the scale towards more cooler than normal years, but, in fact, one finds the opposite phenomenon. Very intriguing. 

[GaWx]

2 hours ago, jconsor said:

Top 10 cold period past two and a half weeks for eastern US:

https://x.com/wxjerdman/status/2019729810914992195

Dominant indices Jan 15-Feb 4:
- moderate to strong +PNA
- strong to very strong -AO
- moderate to very strong -EPO
- moderate to strong -WPO
- so one of the greatest 3 week long combos of +PNA, -AO, -EPO, -WPO on record
- strong to moderate MJO phases 6-7 and weak phase 8
- neutral NAO averaged out with no strong anomalies 

[donsutherland1]

Uncertainty about the February 15-28 period has increased. What is fairly certain are the following:

1) Substantial warmup in the Plains States
2) Brief cool shot to the Southwest and then a return of warmer conditions. The cool shot likely won't be sufficient to preclude Phoenix from experiencing its warmest winter on record.

Moderation is likely to occur across the Great Lakes Region, Northeast, Mid-Atlantic, and Southeast during and after February 10th. What happens beyond 7-10 days is where things become increasingly uncertain. The CFSv2 favors a continuation of above normal temperatures through the end of February. The ECMWF weeklies bring colder conditions for the February 23-28 period.

CFSv2:

ECMWF Weeklies:

Even as wave lengths are shortening, the Arctic Oscillation (AO) could hold the key. The EPS and GEFS agree on the development of a WPO-, EPO+, PNA- pattern. During such patterns, conditions in the East tend to be much colder when the AO is negative than when it is positive. 

New York City provides a good example. Since 1980, WPO-, EPO+, PNA-, AO- patterns have seen a mean February 15-28 temperature of 33.9°. In contrast, WPO-, EPO+, PNA-, AO+ patterns have seen a mean temperature of 44.5°. So, if the AO goes positive, the warmth from the Plains will likely come eastward.

WPO-, EPO+, PNA-, AO- Patterns:

WPO-, EPO+, PNA-, AO+ Patterns:

The latest AO forecast is below. 

A near neutral AO would favor near seasonable to somewhat above seasonable temperatures in the East. Historic experience following the breakdown of long-duration (25-day or longer) AO- regimes that began in February favors AO variability following the breakdown of such regimes. Only 1986 saw a prolonged and strong AO+ regime develop.

As noted previously, due to shortening wave lengths, the consensus outcome of a resumption of above normal temperatures in the Southwest to close out February is the base case.