GaWx

10 hours ago, GaWx said:

 

 Amazingly, the very next CFSv2 run of AAM (0Z on 8/2/26) is even more extreme overall than the posted prior run making it the most extreme I’ve ever seen although this one never quite gets all 4 at 4+ simultaneously (so the first week and last few days are only parts not more extreme): this one already has two at 4+ at day 20! @snowman19

The amazing runs keep coming! The 12Z 8/2/26 CFSv2 has the strongest AAM for forecast days ~12-23 that I’ve seen/saved! The mean hits 3 on day 12, 3.5 on day 16, and 4 on day 17 or 18! There are 3 of 4 members at 4+ on day 19 or 20 through day 22! Afterward, it isn’t quite as impressive as the last run and the far right isn’t as impressive as the last 2 although it’s still very impressive:@snowman19

It appears per radar that I’m about to get clobbered with heavy rain. Rain started ~5:50PM.

 Edit: The heaviest rain fell mainly 5:50-6:20PM on Aug 2nd. The total was ~1.75”. Thus I get a break from having to water!

12Z UKMET has no TC but it does have this vigorous TW on Thu evening moving NW and then NNW from there bringing convection into the NFL/GA/SC coast on Fri 8/7. I may have to watch this. Hopefully it ends up nothing more than this or a low end TS at the very worst:

The latest WCS daily PDO has fallen to -2.19, which is the lowest since way back on 8/29/25! This plunge was not predicted AFAIK:

1 hour ago, bluewave said:

NASA GISS uses 1951-1980 for its departures with Canada going with 1961-1990. Berkeley Earth uses the 1850-1900 period along with Copernicus and the Met Office.

1951-1980 was the last relatively stable temperature regime before emissions and temperatures really took off.

So using warmer 1991-2020 climate normals masks the warming relative to the past. 
 

https://science.nasa.gov/earth/earth-observatory/world-of-change/global-temperatures/

The NASA GISS team chose the period of 1951-1980 as its baseline largely because the U.S. National Weather Service uses a three-decade period to define “normal” or average temperature. The GISS temperature analysis effort also began around 1980, so the most recent 30-year period was 1951-1980. Their objective is to provide an estimate of temperature change that could be compared with predictions of global climate change in response to atmospheric carbon dioxide, aerosols, and changes in solar activity.

https://www.canada.ca/en/environment-climate-change/services/climate-change/science-research-data/climate-trends-variability/trends-variations.html

Much of Canadian economic and social activity is climate dependent. Understanding how Canada’s climate is changing, in the context of global climate change, is important for developing adaptive responses. The Climate Trends and Variations Bulletin (CTVB) helps communicate to Canadians how Canada's climate has changed  over the recent past and longer, over the period for which Environment Canada has climate observations.

The CTVB describes climate variability through maps of current and past departures from the mean (the 30-year average 1961-1990) temperature and precipitation conditions, both seasonally and annually. This approach provides visualization of the extent to which temperature and precipitation vary above and below longer term average conditions. Maps are presented for each season and year. The data are then analyzed to see if there are detectable trends of longer terms against this backdrop of climate variability for Canada as a whole, and for 11 climate regions. This information is presented as graphs with trend lines to assess climate change over the period of record, starting in 1948 when nation-wide records became available.

 

Winter National Temperature Departures and Long-term Trend, 1948–2025

 

 

 This isn’t about measuring GW. This is about using the most appropriate 30 year period to determine what normals are. Due to GW, normals have increased a couple of degrees over the last few decades as you know. Thus, 1951-80 has been outdated for decades to use to determine normals, which are always changing.

 The longer range model consensus is suggesting the possibility of a very long MJO phase 6 streak later this month, the 2nd this year possibly reaching 14+ days, and with a likely max magnitude of 2+.

Longest MJO ph. 6: # days; max mag

7/28-8/18/1984: 22; 2.5;

7/15-8/4/1996: 21; 2.0; 

10/13-31/2022: 19; 2.2

6/13-30/1984: 18; 2.0;

1/5-21/2026: 17; 2.7

11/16-12/1/1998: 16; 2.5

10/6-20/1988: 15; 3.0

7/19-8/1/2018: 14; 2.5

2/25-3/10/1997: 14; 3.6

On 7/31/2026 at 9:24 AM, GaWx said:

 2. This brings us to 82 straight days of -SOI. Going back to when records are readily available (since 1991), this is second only to the 100 days of 1998! It’s too early to tell whether or not the current streak will reach 100 days.

  Today’s SOI rose to ~-8, the highest in a month, from ~-22 yesterday due to a sharp increase in pressure at Tahiti. However, Tahiti has most likely about peaked and it will be falling back noticeably within the next few days. Meanwhile, Darwin’s pressure looks to be higher tomorrow than today. Thus, I’m expecting another -SOI tomorrow, which would make it 85 straight -SOI days vs the record back to 1991 of 100 (set in 1998).
 

 Darwin should peak tomorrow followed by a drop for 2-3 days. But with Tahiti’s concurrent fall during that period, I’m thinking the -SOI streak will likely continue for at least the next few days and likely make it to at least 90 days.

On 7/31/2026 at 11:59 AM, GaWx said:

 Darwin’s pressure has averaged 1015+ mb, a high pressure for there, for 22 straight days! That’s a new record (back to 1991) that breaks the old record of 20 days that was set during another developing super Nino, July of 2015. I’m expecting that streak to finally end tomorrow.

I was wrong as the Darwin 1015+ mb streak is still going! It’s now up to 24 days and tomorrow looks to also be 1015+. That would make it 25 straight days vs the prior longest of 20 days from July of 2015.

5 minutes ago, bluewave said:

They probably should just have left the departures fixed at 1951-1980 which was our last stable temperature period before the warming accelerated.

I strongly disagree. 1951-80 has not been relevant base climo for anomalies for a very long time. Normals are supposed to reflect the most up to date available climo. We’re obviously in a notably warmer climate than that of 1951-80. That earlier colder climate is no longer relevant for determining normals.

23 minutes ago, bluewave said:

The climatology period for the 1997-1998 super El Nino was 1961-1990. They don’t have 1941-1970 climatology maps on that site for 1972-1973. There also isn’t an 1951-1980 option for 1982-1983. So I used the closest time period that they offered at the site.

 I fully realize that the 30 year climo period available during 1997-8, just like at the time for any winter, always has to end prior to that winter for obvious reasons. There’s no choice.
 
 But once one gets further into the future, a more appropriate 30 year period to use as a basis for the analysis of anomalies for any winter is imho the 30 years most centered on that winter. Otherwise, in a warming climate using a 30 year period that’s prior to that winter is always going to bias that winter to look warmer than it really was vs the best centered 30 year climo.

 So, I strongly disagree with you on this and I’ll leave it at that. I don’t expect for you to agree with my thinking here. That’s fine as there’s nothing wrong with that. These are solid debates, which I enjoy, rather than them being arguments/throwing cheap shots.

28 minutes ago, bluewave said:

Yeah, there weren’t any cold winters in the East with super El Niños relative to the climatology at the time. 
 

 

 Chris,

 The following is a better way to look at how 1997-8 did:

using the more appropriate/well centered climo base of 1981-2010 instead of 1961-90, which is earlier than 1997-8. Using the earlier, colder 1961-90 makes 1997-8 anomalies ~1F warmer than they actually were in many areas as you know vs the available best centered climo of 1981-2010: still mild in NE but not as mild and SE NN:

9 minutes ago, snowman19 said:

Which super El Niños were cold winters (Dec-Mar)? 72-73: warm,  82-83: warm, 97-98: warm, 15-16: warm

I assume you’re referring specifically to the NE US as the first 3 winters weren’t warm in the SE. They were pretty close to normal there.

11 hours ago, GaWx said:

 I know that the right sides of these are typically overdone and that’s nearly certainly the case here, of course. But the last 2 weeks of the 12Z 8/1/26 CFSv2 AAM prog is one for the books with the earliest 4+ amp. in a run from a member that I’ve ever seen (day 20) as well as the first run I’ve seen with more than two members at 4+ at the same time! This one has ALL 4 members at 4+ on the far right allowing them to actually have room to put the 4 member color key on the far right instead of on the far left like it always has been: 

@snowman19

 Amazingly, the very next CFSv2 run of AAM (0Z on 8/2/26) is even more extreme overall than the posted prior run making it the most extreme I’ve ever seen although this one never quite gets all 4 at 4+ simultaneously (so the first week and last few days are only parts not more extreme): this one already has two at 4+ at day 20! @snowman19

 I know that the right sides of these are typically overdone and that’s nearly certainly the case here, of course. But the last 2 weeks of the 12Z 8/1/26 CFSv2 AAM prog is one for the books with the earliest 4+ amp. in a run from a member that I’ve ever seen (day 20) as well as the first run I’ve seen with more than two members at 4+ at the same time! This one has ALL 4 members at 4+ on the far right allowing them to actually have room to put the 4 member color key on the far right instead of on the far left like it always has been: 

@snowman19

CFS is at warmest yet!

Nov peak +3.75 vs record +2.69! Also, still has a record crazy rapid plunge Dec-Mar!

OND peak +3.5 vs record +2.52!

1 hour ago, donsutherland1 said:

The map is dated. Nevertheless, winter fire activity in South America has so far been largely suppressed by a wetter to much wetter than normal start to winter. The U.S. has seen a fairly sizable pickup in fire activity in late July. 

Yeah, Don, I just looked at the following, which is as of Aug 1. It tells me that whereas Aug 1st is NN for NA, SA is MBN. Projecting from Aug 1 to EOY gets NA to ~2012-25 avg but SA is headed for MBN when projected.

 July added only ~4 mill ha in Americas vs avg of ~8 due to wet SA thus putting Aug 1st even further behind than Jul 1st:

Opinions about this, which shows lowest cumulative YTD burnt area as of July 1st in Americas since at least 2012?@donsutherland1@WolfStock1

27 minutes ago, bluewave said:

No red herrings involved. My first post showing the warmth this month was based on the warmest 1991-2020 averages. So that is why I kept it for continuity between posts.

But your cooler base period doesn’t show any material difference from the main point how much warmer this summer has been than past developing El Niños due to the record ridging from north of Hawaii across the CONUS.

So it just seems like you are inventing ways to be argumentative without addressing the main point. 

 I never said this summer wasn’t hot and never said 1987 was as hot. But 1987 was still hot and not normal for Nino. That’s why I pointed it out as another hot summer exception for Nino. That’s all.

Edit: I agree with your original idea that 2026 hot summer to date is atypical of pre-super-Nino. Then I brought up ‘91 being hot because it was the other exception to pre-super Nino hot summers. Then you correctly said that ‘91 was still just weak during summer. Then I brought up 1987 because it was mod to strong Nino in summer and hot.

4 minutes ago, GaWx said:

Chris,

 1. Just “A little warmer in the East” relative to the rest of CONUS”? For summer, which has relatively lower variance, this is not just “a little”.

2. This discussion started based on you posting cool Nino summer US maps. Why are you bringing up “Pacific to the Atlantic”? Red herring again.

3. I never said 1987 summer was as warm as 2026 summer to date. But it was still warm in the E US, which is not normal for Nino.

@bluewave

6 minutes ago, bluewave said:

It matches the El Niño composite in most areas from the Pacific to the Atlantic. So a little warmer in the East relative to the rest of the CONUS isn’t really a dealbreaker. This summer so far is much more of an outlier than 1987 was.

Chris,

 1. Just “A little warmer in the East” relative to the rest of CONUS”? For summer, which has relatively lower variance, this is not just “a little”.

2. This discussion started based on you posting cool Nino summer US maps. Why are you bringing up “Pacific to the Atlantic”? Red herring again.

2 hours ago, snowman19 said:

@Gawx Over/under that the MEI is already at or above +2 given the current obs of the parameters it measures? If so, given that those parameters are projected to only get stronger the next few months and the record-breaking rise since April, I think we set a new MEI record (old record: +2.9)…..

Over

3 hours ago, bluewave said:

2026 is near the warmest July on record in my first post using 1991-2020 climate normals at +2.5 since that is the base period Brian used. The departure would be even higher if an older base period was used.

My 2nd post uses 1971-2000 climo and even that cooler one shows that 1987 wasn’t a warm summer for the CONUs even during the 1980s. Notice the warmest summer of that decade was 1988 which came in the summer after the El Niño and not before. So this El Niño is getting off to a much warmer start than usual. 

It’s El Ninoish since it matches most of the pattern for an El Niño summer from the Pacific across to the Atlantic. Every spot isn’t going to be a 100% match. But most of the area from the Pacific to the Atlantic is a reasonable 500 mb match whatever the base period is used for an El Niño summer.

A hallmark of a developing very strong El Niño pattern during the summer is higher 500mb heights further north and lower heights underneath from the Pacific to the Atlantic. This summer the pattern is reversed which lead to all the record warmth across the CONUS.

The record ridge north of Hawaii this month is the reason that the PDO had the steep drop in July. This is not the 500 mb pattern that we have typically had during any El Niños which were on their way to super status. So this event is in a league of its own in numerous ways. 
 

 

 Here’s 1980-89 summer mean vs 1971-2000 climo: 1980s were NN

Here’s 1987 summer vs same climo: notably warmer than the NN 1980s mean E 1/2 US and it did that during Nino: that was not Ninoish at all; comparing it to other hot summers like 1980, 1983, or 1988 is a red herring as they weren’t Nino

22 minutes ago, 40/70 Benchmark said:

No one thought it was right with a straight Modoki in the seasonal mean.

The maps I was basing my post on were from weathermodels.com and were using 1991-2020 for base climo.

 But since I posted that, the WxBell Aug 1st version came out. It uses 1981-2010 for climo. WB had the warmest anomalies centered in Nov near 180 for the April-June 1st runs and significantly further E (near 160W) for the July 1st run. WB (using 1981-2010 base) has its Aug 1st Nov warmest centered near 150W, which is C based/basinwide and not E based like weathermodels.com’s (1991-2020 base) 130W. So, the WB 1981-2010 base version also moved E, but not nearly as far as the weathermodels 1991-2020 base version, which has it all the way E to 130W (pretty E based):

WB 7/1/26 SSTa run for Nov (1981-2010 base) centered ~160W

WB 8/1/26 SSTa run for Nov (1981-2010 base) centered ~150W

19 minutes ago, bluewave said:

Deflect from what? 1987 had a summer pattern that was a reasonable match for El Nino expectations.This summer is completely different. 
 

 

 

1. Why are you using 1991-2020 for climo for a 1987 summer? That diminishes the above normal H5 in the E US on your 1st map.

2. The highest height anomalies/warmest temp annomalies are in the E US, regardless. How is that El Ninoish?

6 minutes ago, bluewave said:

That was still a cooler summer by recent standards. 

 

 

June-August 1987

71.77°F

-0.54°F

 

But the summer of 1987 was hot based on the climo centered on the period it was within as I posted. Why are you comparing it to more recent (hotter) climo that makes it not seem as hot? Also, the hottest was in the E 1/2 of the US regardless of what climo you use.