GaWx

Thanks for posting this. For the first time I examined past versions of this because I just realized that the excellent website has these graphs going back to 4/11/2022. So, there’s ~1,000 of these to compare to. Now I know why the relative global AAM standardized anomaly graphs have been at extreme levels (>+4). This latitudinal based graph for 9/21/26 is an absolute beast overall and thus exhibits the strength of this Nino very well! Why? 1. Back to 4/11/22, I couldn’t find even one other day’s graph with a more positive avg std. AAM anomaly for the area from -40 to +30! The straight avg std. dev above the mean for that huge area on 9/21/26 is clearly >2. Keep in mind that 2 std dev above the mean at any one latitude is way up at the 98th percentile and thus nearly 1/3 of the globe (20N to 17S) is >>+2 std dev/>>98th percentile! 2. The # of degrees of latitude with a negative AAM anomaly of the 180 total degrees on this 9/21/26 graph adds to only 29 degrees of the 180 total (16%). I couldn’t find even one other with that small a % of latitudes <0! 3. As one can, most of the latitudes with a -AAM anomaly on 9/21/26 are only down to the 0 to -0.3 std dev range. The lowest is only down to -1.1, which is quite high for the lowest vs the other ~1,000 graphs. The highest lowest I could find is only barely higher (~-0.9) and it overall still had a much larger area below the 0 line than 9/21/26: So, it appears that 9/21/26 has the smallest area <0 of any back to 2022! Here again is this 9/21/26 beast of a +AAM anom by latitude. Keep in mind that just 1 std dev above the mean is up at the 84th percentile: Below is the corresponding relative global AAM, which since 0Z of 9/19/26 has been at an extreme level of >+4 std dev! This is probably the peak day as it should get back to <+4 by 9/23/26. How rare is 4 std dev above the mean? Per stat. formulas, only 1 in 31,574 days (that’s once per 87 years) gets to that level! (For 3 std dev, it’s ~once/year):

WB CFSv2 temp maps: 1. Have algo issues that lead to bad artifacts. I’ve posted on this here in the past. Here are some examples -note the cold spot over S Lake Michigan and warm spot only 250 miles to the north over N Lake Michigan that are almost always on these WB maps; the real CFSv2 output essentially never has these -note the E coast down to FL usually looks like a thin sharp warm barrier to any colder anomalies to the west; again, this essentially never is on the real output 2. WB CFSv2 maps are strongly cold biased for much of the E US. And this is the case even with WB using the colder 1981-2010 climo instead of 1991-2010, which the CFSv2 site uses! All one has to do to know this is to compare to the maps at the CFSv2 site. All 4 (DJFM) are notably too cold in almost all of the area away from the E coast. But I’ll focus on the Dec map to illustrate the strong cold bias: -Chicago’s +0.5C/1F vs 1981-2010 on WB but the CFS site’s relevant 10 day mean maps (see below) are all significantly warmer and they average +2.25C/4F vs the warmer 1991-2020 meaning WB is ~4F too cold. -SL’s 0.0 C/F vs 1981-2010 on WB but CFS actual maps are all significantly warmer and average +1.5C/3F vs 1991-2020 meaning WB is ~4F too cold -Knoxville’s -1C/-2F vs 1981-2010 on WB but CFS actual maps have it at +0.6C/+1F vs 1991-2020 meaning WB’s ~4F too cold -Despite the CFS site’s maps averaging AN from TN-NC north, WB in inland areas doesn’t have AN start til 500 miles to the N in N IL to N OH to NW PA and that’s even vs a colder climo base! That’s ignoring the immediate E coast, which has its own problems as already noted and is actually a bit too warm overall despite it being 4F too cold for much of the inland areas! Relevant 10 day mean Dec maps at CFSv2 site: 9/10-19: 8/31-9/9: 8/21-30: 8/12-21: @Stormchaserchuck1@donsutherland1

I’ve already compared the current Nino to the prior one for both GLAAM and for OHC (like night and day). 1. Here’s a similar comparison of PDO: Current: -0.71 Sept of 2023: ~-2.4 on Sept 19th though these are admittedly volatile as one can see it rose steeply Oct 1-20th: 2. SOI: 90 day mean is now way down at ~-22. The mean was down to only -8 for the same 90 day period in 2023. It was -18 in 2015 and -14 in 1997. A rough estimate using monthlies (due to no dailies that far back) for 1982 is -19 to -20. The only other 90 day mean as of Sept 20th back to 1876 that realistically could be about as low as 2026’s -22 based on monthlies is 1896 (incoming strong Nino) though that chance is <50% as it is more likely in the -20 to -21 range. So in summary: 2026: -22 2023: -8 2015: -18 1997: -14 1982: -19 to -20 1896: -20 to -21 The other super-Ninos’ rough estimates: none even close to 2026, 1896, 1982, 2015: 1972: -13 1965: -14 1957: -5 1888: -11 1877: -10

OHC of 180-100W appears to currently be at its highest yet (~+3.3): To compare, look how low was the same point in Sept of 2023 (only ~+1): Highest 2023-4 ever got was only just over +1.5:

Further to my global relative AAM post above showing it >+4 since yesterday and to show the huge contrast to the prior El Nino, here are some of the graphs I saved from 2023-4, which all show a much lower amplitude (keep in mind that CFSv2 progs looking out especially past 2 weeks have been biased positive as comparing these graphs to each other clearly show): 11/6/23: +0.15 12/16/23: +0.7: 12/30/23: +1.15 (highest of any I saved from 2023-4) 1/8/24: +0.15 2/3/24: +0.5 2/21/24: +0.2 3/1/24: -0.35

The global relative AAM remains in historic territory of +4+, which should last into tomorrow. This is probably its highest since Feb of 1998. For comparison, 2023-4 was no more than in the low +1 region on any of the graphs I saved with most of them <+0.5:

The AAM went over +4 as expected yesterday (9/19) per the image below, an extremely rare/historic event. The last time it was this high may very well have been in early Feb of 1998. It appears that it will likely stay >+4 into Sept 21st:

https://www.cpc.ncep.noaa.gov/products/CFSv2/CFSv2seasonal.shtml

The main CFS site actually has 40 run averages (10 days x 4 runs/day).

Also, the TT maps use colder 1984-2009 climo instead of the CFS site’s 1991-2020. The TT 3 day (12 run) avg does jump around more than the CFS site 10 day (40 run) avg. Regardless, I haven’t seen any 3 day avg Feb recently nearly as cold as those record cold Febs that were on some Aug 3 day means in the SE to S Plains. Then again, that degree of cold was hardly believable especially that far out. They were dreamable for me but not believable. I’ll keep following the trends of both the TT 3 day means and the CFS site 10 day means.

Tropical Weather Outlook NWS National Hurricane Center Miami FL 800 AM EDT Sat Sep 19 2026 For the North Atlantic...Caribbean Sea and the Gulf of America: 1. Eastern Subtropical Atlantic (AL99): Showers and thunderstorms have become more concentrated near the center of a low pressure system located several hundred miles southwest of the western Azores since yesterday. A short-lived tropical depression or tropical storm is expected to form later today or tonight while moving north-northeastward at about 5 mph, but remaining well to the southwest of the Azores. Upper-level winds are forecast to become less conducive for development beginning late Sunday. * Formation chance through 48 hours...high...90 percent. * Formation chance through 7 days...high...90 percent. Forecaster Berg

Recent 3 day mean CFS’ have cut way back on the degree of cold in Feb: 9/16-18 Compare that to this from 8/7-9:

There indeed have been UK runs with false TCs though not false canes. But in this case, all models have a low off the NE coast as opposed to it being just on the UK. But would it be T, ST, or ET and would it stay offshore?

Watch off the E coast next week. Here’s a hint of that from 0Z Uk: NEW TROPICAL CYCLONE FORECAST TO DEVELOP AFTER 150 HOURS FORECAST POSITION AT T+150 : 36.1N 70.8W LEAD CENTRAL MAXIMUM WIND VERIFYING TIME TIME POSITION PRESSURE (MB) SPEED (KNOTS) -------------- ---- -------- ------------- ------------- 1200UTC 25.09.2026 156 36.9N 69.7W 1000 47 0000UTC 26.09.2026 168 39.1N 67.9W 997 41

Google Deep Mind