GaWx

Though it may be subtle, the 12Z CMC especially and the GFS to a lesser extent had changes vs the 6Z/0Z that went in a good direction toward possibly reviving the 12/26 threat. The CMC has a storm form offshore the SE that actually gives some sleet to coastal NC while the GFS has a W GOM low (though weak) form for the first time in 5 runs.

For a wx history lover like me, the highest Montana/N Wyoming SLP next Thursday will be fun to track. That's because the 12Z ICON forecasted a max of 1067 mb on 12/22 and both the 12Z GFS and Euro forecasted 1065 mb on 12/22 in that area, which would be a new record high for the lower 48! The highest on record back to the late 1800s is 1064, set in Miles City, MT, on 12/24/1983. The 2nd highest is 1063.3 mb, which was at Helena, MT, on 1/10/1962. But the other three main models are lower. On 12/22 the 12Z JMA is at 1061, CMC is 1060, and the UKMET is only near 1050. So, we'll see whether or not these ICON/Euro/GFS runs end up too high. The odds are that they're too high but that remains to be seen. On 12/24/1983, the PNA was -1.1, the AO was +2.8, and the NAO was 0.0. On 1/10/1962, the PNA was +0.5, the AO was -+1.1, and the NAO was +0.9. Per today's GEFS forecasts for 12/22: 1. PNA neutral and rising 2. AO still very strongly negative 3. NAO neutral and rising What might this very strong high portend for the SE US? To get an idea of the potential, let's look at what the 1983 and 1962 Arctic highs lead to at ATL and RDU: At Atlanta: - The 12/24/1983 high lead to 3 days in a row that didn't exceed 32 (12/24-26) with a coldest of 0 on 12/25. They had some ZR on 12/27 as they warmed. They had a followup cold blast that got them down to 10 on 12/30. - The 1/10/1962 high also lead to three days in a row that didn't get above 32 (1/10-12). Atl got down to 8 on 1/11. It also had 3.5" of snow preceding it on 1/9. At RDU: - The 12/24/1983 high lead to two days in a row that didn't rise above 32 (12/25-6) and a coldest of 4 on 12/25. They then had some wintry precip on 12/28 as they warmed followed by another cold blast that got them down to 11 on 12/31. - The 1/10/1962 high lead to two days in a row that didn't get above 32, 1/11-12. Their coldest was 5 on 1/11. This was preceded by 5" of snow on 1/10. US pressure records from here: https://www.wunderground.com/blog/weatherhistorian/world-and-us-anticyclonic-high-barometric-pressure-records.html

Whereas the 0Z GEFS backed down on a coastal 12/26 threat, the 0Z Euro did the opposite. It had at 240 in the SE Gulf a weak low that appeared ready to make a turn NE across FL and strengthen. This would be due to the 500 mb flow becoming a more moist SW in response to strong upper level energy diving SE from Canada. Cold enough air at the time was still along the coast. Had this run gone just a little bit further, I think it would have shown a coastal SE winter storm. More importantly, the 0Z EPS supports the Euro with a good number of members as well as the mean looking threatening on 12/26. In addition, 8 of 20 (40%) of 0Z GEPS members have wintry precip, mainly ZR, to the coast on 12/25. Edit: 6Z GFS is MUCH warmer on 12/26 with no winter storm anywhere closeby, following the lead of the 0Z GEFS.

Whereas the 18Z GFS backed off for 12/26, the 0Z brought it right back as a sleet/snow mix to the coast and inland 25-50 miles. See map below, which counts it all as snow and thus inflates the amounts. Further inland is all snow. This makes the 0Z GFS a highly unusual 5th of the last 9 and 3rd of the last 4 with significant wintry precip all the way to the SE coast on 12/26 fwiw (still not much value other than entertainment that far out). OTOH and more importantly since ensemble means have a good bit less inaccuracy than operational runs, the 0Z GEFS mean has the 500 mb trough a good bit west of earlier GEFS runs and is thus warmer/too warm for coastal wintry precip. on 12/26. Since having a winter storm on the coast of this magnitude is rare (only once every 20 years in our area), I normally view modeled winter storms in our area with a lot of doubt since most don't materialize. Let's see whether or not this GEFS run is a harbinger of GFS future runs.

Good point though since we're talking 11 days out (well out in almost unforecastable fantasyland) a lot could and probably will change as you know. But fwiw the 12Z EPS mean (and some other recent ensembles to an extent) has a textbook setup for a rare SE winter storm of some form near 12/26 that includes the coast because it is so cold. The mean 500 mb trough is centered to the west allowing for a moist WSW flow on top of the very cold air forecasted to be already in place. The mean qpf is most concentrated from the FL panhandle through the eastern sections of GA, SC, and GA. So, still just a far out fantasy, but rather interesting.

For the record/entertainment since it is still way out in fantasyland (12/26), a large portion of that 2nd storm is sleet/ice during the majority of it within 75-100 miles of the GA-NC coast as 850 mb temperatures are or go above 0C. The 6Z GFS was colder at 850 with more snow but it had a lighter event. Regardless, the 12Z GFS is the 4th with significant wintry precip to the SE coast on 12/26 fwiw (still not much value other than entertainment that far out).

And this 0Z GFS storm is in reality actually almost all snow since 850s are all below 0C. The southern part of this goes all of the way down to the FL/GA border! This is still way out in fantasyland and is thus mainly for entertainment as of now. However, this is about the third GFS run over the last few days showing SE coastal wintry precip. The 0Z 12/14 GFS also had a winter storm for the same timeframe that included the SE coast though it was mainly IP/ZR there (see 2nd image). This one, like that one, actually starts on 12/25 and thus would mean a white Christmas for many! From 0Z 12/14 GFS run: This was the first one showing significant wintry precip for the coast (mainly IP/ZR with this from the 18Z 12/12 run for 12/27):

Followup as I just compiled the effects on RDU and ATL from the strongest -AOs averaged out 12/16-31 since 1950. 2022's 12/16-31 is per today's GEFS flirting with ending up with an average of near -3. Here are the strongest: 2009: -4.5 1950: -3.8 2010: -3.7 1976: -3.5 2000: -3.4 1995: -2.8 So, 2022 is as of now aiming for 6th or 7th strongest 12/16-31 -AO since 1950. That would mean 92-93 percentile. How were these six last half of Decembers at RDU and ATL in terms of temperature anomaly, lowest temperature, and wintry precip? 1) RDU: 2009: 6 BN, 23, 0.2" SN/IP 1950: 9 BN, 13, 0.3" SN/IP, 0.41" ZR 2010: 8 BN, 19, 7.5" SN/IP, 0.47" ZR 1976: 6 BN, 9, 0.0" SN/IP 2000: 11 BN, 12, 0.1" SN/IP 1995: 8 BN, 18, 0.0" SN/IP AVG 8 BN, 16, 1.4" SN/IP, 0.15" ZR 2) ATL: 2009: 5 BN, 27, T of SN/IP 1950: 6 BN, 18, 0.0" SN/IP, 0.77" ZR 2010: 5 BN, 23, 1.4" SN/IP 1976: 7 BN, 13, 0.0" SN/IP 2000: 13 BN, 13, 3.0", SN/IP 1995: 7 BN, 21, 0.0" SN/IP AVG 7 BN, 19, 0.7" SN/IP, 0.13" ZR What does this tell me about the last half of this month in the SE? 1. Cold to very cold most days very likely. 2. Good chance for RDU coldest to be teens and ATL coldest to be teens or low 20s. 3. Assume non-cold dominated operational runs not supported by ensembles are likely wrong. 4. Significant wintry precip hard to predict as is normal for the SE. One third of the analogs at RDU and half at ATL had significant. That's a larger % than the overall climo % for just the last half of December, which is intuitive since it is so cold. Thus the average wintry precip for the six analogs is well above normal for that period even though there could still easily be nothing significant or possibly none at all.

This afternoon was more like normal here for winter with temps only in the 50s along with a nice easterly breeze and lots of clouds.

Ensemble means as you must know are much more credible that far out. The 12Z EPS, whereas it isn't quite as cold as the 0Z EPS, it doesn't have the SW trough/SER and is thus much colder than the operational.

The latest GEFS forecast for AO implies a good chance for a ~-3 AO for 12/16-31. Since 1950 (assuming I didn't miss any), the only lower AOs for 12/16-31 were in 1950, 1976, 2000, 2009, and 2010. So, 2022 is aiming to have the 6th lowest AO for 12/16-31. 1995 was just above -3 and would be 7th lowest if this forecast holds. Since 2010, there have been none that were -2 or lower. What were the SE US temps and wintry precip for those 12/16-31s? I haven't checked that yet. If I get time, I'll look them up.

Several (~5) 18Z GEFS members have a significant wintry event in this area, something not seen often on model runs. Significant wintry events (say ~1"+ of snow/sleet and/or ~0.25"+ of ZR) have occurred here only ~once every 10 years. So, when these show up on models, they get my interest and they're fun to see but I take with a grain as they more often than not don't materialize. The last two significant wintry events here were biggies: Jan of 2018 and Dec of 1989. The models did a pretty good job of forecasting Jan of 2018 with the ICON doing best here. 18Z GEFS mean (treat these as 10 times wintry qpf as some of this is sleet or ZR rather than snow):

For the record/entertainment way out in the fantasyland period, the 18Z GFS has for here snow flurries on 12/24 from one coastal followed by a whopping ~1" of qpf in the form of sleet/freezing rain on 12/27 from the next coastal. (Some clown maps have this as snow, which it obviously isn't since 850s are clearly several degrees over 0C.) That 1" of qpf of IP/ZR would be the worst here since the devastating icestorm of 1/25-26/1922, the predecessor to the deadly Knickerbocker Storm that hit DC.

18Z GFS is about the coldest GFS run yet for the days just before Christmas.

It still looks like cold is coming, but I want to clarify my post regarding the NG market's 7% drop back since early this morning after the huge (10%) rise last evening. On a national basis (not specifically SE), the three major 12Z ensemble means were all somewhat less cold overall than their 0Z counterparts on the full 2 week runs. So, in summary, the ensemble means are still significantly colder than they were on Friday, but they're all not quite as cold at 12Z vs 0Z today for the full 2 weeks in the US as a whole.

The trend of the 6Z and now 12Z model consensus is clearly warmer vs the 0Z/18Z models. Look no further than the NG price, which went back down 7% from its highs early this morning. Keep in mind that it opened up a whopping 10% Sun evening on much colder models vs Fri PM. Now a good portion of that has been given back.

The TT 12Z GFS map posted below has Alexandria, LA, getting down to -4 F as of 12Z on 12/25. Their coldest on record in December is +7 F (1989) and their coldest on record for any month is +3 F (1899). There's no way this can be anywhere close to real considering that history as well as the run on TT having no colder than ~+13F at Tulsa, OK, for example, from this airmass (on 12/24). Tulsa's all-time coldest in Dec is way down at -8 F (1989), which compares to Alex.'s coldest in 1989 of +7 F. So, in 1989, Tulsa's coldest was 15 colder than Alex, which makes sense. But the 12Z GFS has Alex. ~17 colder than Tulsa's coldest for this airmass. In other words, the 12Z GFS, itself, isn't progging anything near a historic airmass well to its NW but it then gives Alex. an alltime coldest Dec by 11!! I call BS:

Looking at the climo of the other eight 3rd year cold ENSO winters since the mid 1800s: 1) Winter of 1894-5: mainly Feb of 1895 (but also the last week of Dec/first few days of Jan) 2) Winter of 1917-8: Dec-Jan of 1917-18 These were two of the most historic lengthy periods on record in the SE US in terms of a combination of multiple wintry events and persistent extreme cold. For example, Dec of 1917/Feb of 1895 are RDU's coldest Dec/Feb and ATL's 2nd coldest Dec/coldest Feb on record. Also, Feb of 1895 is the snowiest ATL month of any on record due to not just one but two major snowstorms along with a third one! In addition, ATL had two major winter storms in Dec-Jan 1917-8. But I'm not just talking about well inland/northerly locations like RDU and ATL. I'm also talking about locations not at all accustomed to this that are far to the south, including right on the Gulf and SE coasts. N.O, had a historically very heavy snow in Feb of 1895. Right here Feb of 1895 was both the coldest Feb on record and it gave a very rare ~2" of snow (that much accumulation has occurred on average only ~once every 20 years). Dec-Jan of 1917-8 had not one but two accumulations. Both were 1/2" or smaller but only ~1/8 of winters have gotten ANY accumulation. Dec of 1917/Feb of 1895 were the 4th coldest Dec/coldest Feb on record here.

To give you an idea of how much colder the model consensus for especially the E 1/3 of the US is now vs what it was Friday at 12Z, the natural gas market this evening opened up a whopping 10% over the last price traded of Friday!! That's almost, if not totally, unheard of in the 30 years this market has traded!

Some interesting climo for major Atlanta ice and sleet storms based on my own research done several years ago: 1. Every one since Feb of 1979 has occurred when the MJO was within the COD with one exception: the Feb 1979 sleet, which started in phase 3 just outside. But then even it moved inside for most of it. 2. Every one since 1940 had predominantly east winds, not NE winds. Any major winter storm that had mainly NE, N, NW, or west winds (~18 of them) was mainly or all snow as opposed to mainly sleet or ZR.

12Z GFS fwiw has not just one, not just two, not just three, but FOUR different short waves that produce wintry precip over LA. Details that far out (1-2 weeks) aren't important due to very low credibility of the very jumpy GFS. But this shows the potential threat for wintry precip in parts of the SE US at some point(s) in the second half of December. Looking at ensemble consensus for then, practically the entire US is a cold island sticking out like a wintry thumb, which is rare. If enough of that cold were to reach the SE US, it could easily mean significant wintry precip in parts of the SE and quite possibly more than once, especially as usual in the more favorable northern parts. Down my way, where wintry precip is rare (only about once every 4-5 years longterm average and more often trace or light than not), I'm mainly just wanting a cold pattern. Down this far SE, I don't live or die based on getting wintry precip as I'd be dead almost all of the time lol though I never give up hope for a miracle when it is cold. The indices per today's GEFS forecast for that period are rather favorable for cold overall with a continued strong -AO (-2 to -3), weak to borderline moderate -NAO (-0.25 to -0.50), and a rising PNA from the prior three weeks of -PNA to a neutral to possibly weak +PNA during this period. Also, the MJO is pretty conducive for then per forecasts for allowing cold with it mainly low amped inside the bottom part of the circle although the EPS has it going outside right side in about 10 days, which itself would favor mild in the SE.

This month the AO plummeted to just below -4. The NAO plunged to -1.95 while the PNA fell to -1.86. For all months since 1/1950, there's only one other month that had a sub -4 AO day that also had BOTH a daily PNA at least this low AND a daily NAO at least this low: December of 2010. Its AO dipped to -5.3, NAO to -2.0, and PNA to -2.1. So, a more extreme for all 3, especially the AO.

Here's something one will see only very rarely, a 360 hour EPS map with BN essentially in all of the lower 48! It is especially hard for all of Florida to be included in these situations, where usually some ridging nearby keeps it at least normal especially in S FL. Note also that this is for Christmas morning:

We're pretty much on course for a cold most of 12/16-20 in most of the SE. Shortly after, we warm to normal to above for ~2 days at least. So, I'd like to look further: Take it with a grain considering recent models being too cold in the 11-15 day period, but fwiw the 12Z EPS is decidedly colder at 850 mb and the 500 mb pattern is also more conducive for it to not warm up as much or for as long after the aforementioned brief warmup. It suggests cold likely returns to much of the SE by 12/23. Again, fwiw! But I prefer models showing this as opposed to not because I would strongly prefer the return of cold over the warmth of the first half of this month!

The 6Z GEFS was about the coldest GEFS run yet in the SE in general for 12/16-20 and the 12Z is almost as cold. Looking ahead to after that period (11-15), here's some food for thought regarding models recently being too cold to keep in mind when looking ahead from a non-hyping pro met company (Maxar) (unlike the hyping Bastardi): "These are larger swings and much is due to poor model performances of late. As an example, here is how the past five days verified:" This is much warmer than the model consensus had when this period was 11-15 days out. "Looking into these errors, we found that the models underestimated the warmer influence from tropical circulations. Our method was to find MJO cases which also featured the already established –AO. These cases decently matched the upper air pattern that was observed and provides proof that the MJO had a role in what was a warmer period for the Eastern Half. " "We can’t guarantee that the models are mishandling this forcing again, as one case does not necessarily represent a bias that will continue. That said, the MJO looks to return to those warmer phases within the 11-15 Day period, and this contributes to our forecast leaning warmer than the models for the South and East during that period." "The alternative is based on model projections at 500mb. The upcoming cold pattern is due to a strong polar ridge, with signals like the –AO and –EPO. As of now, the models show colder than normal air reaching the South and East, at least for a few days, but are more anomalous from the Northwest to the North-Central:" "When we look at the –AO and –EPO correlations, they imply a colder risk for the South and East as well:" "So mixed signals is the message here, between the warmer forcing from the tropics to the colder pattern signals being projected. Confidence is lower than usual as a result."