GaWx

I’m not aware of any that are free. I can tell you that 11/17’s +233 (the latest daily available), which was the highest daily since way back on June 4, will probably end up being the highest for quite awhile. Models are unanimous in dropping it very sharply to at or near the lowest daily of 2024 on 11/21 or 11/22 (~-250)! The consensus then has it remaining negative the rest of Nov fwiw. When combining the low point and duration of this upcoming -EPO, it could end up the most impressive since way back in Oct of 2023! https://downloads.psl.noaa.gov/Public/map/teleconnections/epo.reanalysis.t10trunc.1948-present.txt

1. 12Z GEFS/EE appear to my eyeballs to probably have little change in total US pop weighted HDDs vs their respective prior runs. 2. OMG, there’s finally now an OMG reaction! Wow! This was much needed.

How are the autumn leaf colors currently looking in NC and other areas in relation to peak? Down here the first leaves have only just started to turn and hardly any have fallen. Thus a peak here in early to possibly as late as mid Dec would be my guess. The late week cooldown with coldest lows getting to down ~40 should really wake them up. The lowest in much of this region to date has been down only to the mid to upper 40s, with upper 40s the climo avg for today.

Through the 0Z/6Z EPS/GEFS the colder late solutions are continuing. NG is up near multi-session highs as a result. *Edit for 12Z GFS fwiw showing hard freezes much of E/SE US early Dec.

The left side of the image below shows the cooling of the 12Z EPS (purple) vs 0Z (yellow) by 6 HDD along with the coldest days being at the end, which helped NG to close in the higher portion of its session range (its lowest of session was before the 12Z models were released thanks largely to the then 10 HDD cooler Mon 0Z EPS vs Sun’s 12Z run). Again, it will be interesting to see as said by Huffman above whether this next progged cold period gets muted like was the case with late 11/24-7 as a new SE ridge appeared seemingly out of nowhere. This underplaying of the SE ridge has been a forecast problem for years coincident with the very warm W Pac. To reiterate: the 11/14 0Z EPS had 11/23-7 with 89 HDD. This 11/18 12Z run despite being cold at the end had only 68 HDD during 11/23-7, a whopping 4 degrees warmer per day (AN) vs the 11/14 0Z run’s BN! Will this later period eventually have the same fate? I hope not. Stay tuned!

The 12Z EPS came in significantly colder than the 0Z EPS, which itself had been significantly warmer than yesterday’s 12Z.

Despite the dominance of +EPO so far this autumn, the rest of Nov is favored to have -EPO on most days. That would likely mean Nov ends up pretty neutral. So, we have solid +EPO Sep and Oct followed likely by (pretty) neutral Nov. Regarding NP-EP, what other autumns since 1948 had similar assuming Nov ends up pretty neutral (a big if, of course)? -2021 excellent analog and also Nina -2007 good analog and also Niña -2005 excellent analog and also Nina -2004 good analog though was Nino -No other good analogs back to 1948! So, these are the only four potentially good to excellent autumn NP-EP (keeping in mind Nov still TBD). And with 2004 being Nino, it should probably be thrown out. So, I’ll focus on especially 2021 and 2005 and secondarily on 2007. Unfortunately it is a very small sample. So, fwiw, how did EPO (based on dailies) end up in winter for these three? -2021-2 averaged neutral EPO: -10 Dec mixed/neutral: -20 Jan mixed/neutral: 0 Feb mixed/neutral: -10 -2005-6 averaged neutral EPO: +20 Dec mixed but more -EPO: -50 Jan strong +EPO: +140 Feb mixed but more -EPO: -30 -2007-8 averaged moderate +EPO: +60 Dec mixed but more +EPO: +40 Jan mixed but more +EPO: +50 Feb +EPO: +90

Thanks, Mitch. From that: “This index replaces the old EP index which is no longer maintained by CPC.” I don’t understand why that is the case when it is still maintained on a daily basis here: https://downloads.psl.noaa.gov/Public/map/teleconnections/epo.reanalysis.t10trunc.1948-present.txt Meanwhile, a whopping 46 of the last 55 days have had a +50+ EPO. But the correlation of autumn EPO and winter EPO has been only weakly positive from what I can tell looking at actual winter vs autumn.

Yes, I agree that taking those NP-EP monthlies and flipping the sign seems to give a good (though not perfect by any means as Aug of 24 shows, for example) indication of what the classical/original EPO would calculate to be based on the dailies. Is the EP portion of that the actual classical EPO? If so, that would explain why it isn’t a perfect negative correlation since “NP” is also in the calc. Why are we having so much trouble finding a monthly EPO table based on the pure original calc? I’m going to continue looking at the daily EPO table as my go to for EPO back to 1948.

So that means these aren’t at all the monthlies for the original EPO. Actually, these monthlies usually have the opposite sign the dailies calculate to. So, a strong negative monthly on your link’s table actually correlates rather well to +EPO (I mean the original and only version I have ever followed). But even that negative correlation is far from 100% as Aug of 2024 shows. Your link’s table has a strong negative for Aug of 2024 vs a very slight negative (essentially neutral) based on the avg of the dailies of the original EPO.

What exactly is that? It isn’t EPO, itself, as one can see just looking at Sept and Oct of 2024. This monthly table has a near record strong negative for Sep of -2.75 (2nd lowest to April of 1968’s -2.88) even though the dailies show that they were solidly positive. Also, this has Aug of 2024 way down at -2.62 despite the dailies averaging neutral.

Ray, Unfortunately I don’t have a source for monthly EPO. To get any monthly I literally have to calculate them based on these dailies. Because that’s very tedious/time consuming, I rarely have done that. More often I’ll instead roughly estimate them or better yet just count days over or under a certain level because it is quicker and still telling.

I agree that E US cold is still clearly coming for 11/21 through early 11/24. But late 11/24-11/27 has warmed considerably in the E US from how it looked just 4 days ago as I’ve been showing via the HDD graphs. So, the upcoming cold period is now looking to be shorter vs how it appeared 4 days ago. Now we look ahead to a second potential cold period in very early Dec. Will it also diminish or shorten like often occurs in the E US and did occur for 11/24-7? Or will it fully verify this time? We’ll see. By the way, NG has turned out somewhat lower than it was as I was guessing would occur at about this time (due to the models warming on and near 11/26) since I noted it late last night fwiw. We’ll see what occurs the rest of today.

This is illustrative of how I like to describe the tendency of model/ensemble means to often (though of course not nearly always as I keep emphasizing as these are general tendencies (cold bias) rather than absolutes) show cold being delayed: it’s often like a mirage. Instead of fake water, it is fake blue/cold. When this occurs it will look cold at two weeks out and as that period gets closer the cold often wanes and may never materialize for that same period/could even turn warm/red like Nov 26th has dramatically already done. In this case (Dec 3rd), is it the common fake blue like Nov 26th in retrospect appears to have been, will it end up real this time, or something in between? The cold bias says better chance cold wanes rather than intensifies and of course it could remain about the same.

Looking at TT EPS at H5 and 2M: one can see the huge difference between the 0Z 11/14 run and the 12Z 11/17 run on 11/26: 1) a) 0Z 11/14 EPS for 11/26 at H5: conducive H5 flow allowing CDN air to dominate E US 1) b) 0Z 11/14 EPS for 11/26 at 2m: BN dominates east resulting in 18 US HDD Now compare those to the 11/17 12Z EPS 2) a) 12Z 11/17 EPS for 11/26 @H5: sig. SE ridge, which has often been underplayed late in week 2 and beyond for 7-8 years and then appears as get closer in: 2) b) 12Z 11/17 EPS for 11/26 @2m: AN instead of BN dominates east resulting in a mere 12 instead of 18 HDD with that near the normal for way back on 11/10:

AAM looks solidly Ninoish starting in ~1 week as of this CFS run of 2 days ago fwiw. Will this verify? Warning: these graphs are from the very unreliable 4 member CFS ensemble (you know, the model with the very unrealistic record obliterating plunge in Nino 3.4 by next month) and they are very jumpy/fickle! Look how much they changed from just 4 days prior (2nd image). So, be careful not to read too much into this: Prior run wasn’t as solidly Ninoish:

Here’s a better illustration of a main reason why NG has been up this evening, which is interesting: it is because the 12Z 11/17 EPS (purple) is coldest at the end, Dec 1st, with 20.4 HDD and rising. However, keep in mind that 11/23-7 total HDD are down to only 71 vs 89.5 on the Thu 11/14 0Z run, when NG was about where it has been this evening. Check out 11/26: it’s only at 12 on the 11/17 12Z, which is a whopping 6 BN and the normal for 11/10, 16 days earlier. On the 11/14 0Z, it was way up at 18.1 or near normal! This is the nature of these cold biases. They’ll often be too cold late in week 2 and then get back to reality and warm up as those days get closer. Back on 11/14, 11/23-27 were out 9-13 days. Now that these days are only out 6-10, they’ve warmed substantially. They don’t always do that, of course. But they tend to do that more than the opposite (nature of cold bias). If I were a betting man, I’d bet based on the cold bias that Dec 1st will end up verifying lower than 20 HDD. Of course it may not but the odds favor that. Also, even though I’m not trading commodities, I’m guessing that NG will be lower than it is now by morning. That’s partially because 11/23-7 are so much warmer than they were as of 3 days ago. But that’s just a guess. If the overnight GEFS/EPS runs turn out cold enough especially late, NG probably wouldn’t turn down in the morning and may even rise further.

The last few advisories have been saying isolated total amounts in N Honduras up to 40”!! RAINFALL: Additional rainfall amounts of 1 to 3 inches are expected over northern Honduras, with storm total amounts locally as high as 40 inches. The risk of catastrophic and life-threatening flooding impacts will continue, especially along and near the Sierra La Esperanza.

Dan, why? All mkts are manipulated and NG is no exception. However, despite this, I’ve found NG current price trends to be a good barometer of days 10-15 of ensemble model E US temperature trends.

NG price this evening is up 4% likely due to the end of the 12Z GEFS and EPS being colder than how the late in the run 12Z runs looked Fri/BN in the E US, where NG is most heavily used in cold season. This negates the price drop of Thursday that was largely due to warmer models that day.

1. Neither should be considered anywhere near gospel out at hour 360 of even the ensemble means as they are typically of pretty low confidence though they’re much better than their respective operationals. 2. Due to cold bias of ensemble means that has existed for years, especially that far out, quite possibly due to the very warm W Pacific, warm solutions have a somewhat better chance to verify closely than cold solutions. 3. As I assume you and many of the members know, cold bias doesn’t at all mean always too cold or anywhere near always. Sometimes they can be either too warm or about right. But overall, too cold is typically more common than too warm (definition of cold bias). 4. Be that as it may, I’ve also noticed the model trend for the start of Dec has turned colder. But again, it remains to be seen whether or not these colder solutions will actually verify closely. Keep in mind that much earlier in the forecast period (11/23-7) that the EPS warmed considerably from an avg of +0.5 AN HDD per US pop weights to 2.5 HDD BN.

Thanks for posting. I hope your hunches are wrong and that we get a much quieter season. Interestingly, I saw this exact post at another BB but I don’t see it anymore for some reason.

For the period Sept 1st through Nov 14th back to 1948, 2024 is tied with 2021 for the fewest -EPO (sub -50) days with just 3: Lowest # of sub -50 EPO days 9/1-11/14 back to 1948 (7 or fewer): 2024: 3 2021: 3 2011: 4 1994: 4 1958: 4 1948: 4 1968: 5 I didn’t see any notable correlation to # of sub -50 days one way or the other in the subsequent respective winters, however. https://downloads.psl.noaa.gov/Public/map/teleconnections/epo.reanalysis.t10trunc.1948-present.txt

How are the autumn leaves looking in N GA and the W Carolinas? Are they running way behind? Have they hit peak yet?

So based on my prior post: IF the CFS cooling in Nino 3.4 from Nov to Dec were to somehow verify closely, it would more than double the old record cooling from Nov to Dec back to 1950 and more than triple the strongest forecasted cooling of any other model (JMA’s 0.3)! For some reason I feel CFS will bust very badly but we’ll see. In light of this likely upcoming huge bust in Nino 3.4, how much weight should be put on the CFS current winter forecast for the E US?