GaWx

@Awesomesauce81 @gtg947h FYI for the record, the KSAV (airport) automated temperature sensor suddenly became erroneously too warm around the start of this month and is thus supposed to be replaced within a few days per CHS. This is of extra importance now because the 1PM there is showing 85 F, which if real would mean 2 F warmer than the alltime December record of 83. In reality, it is very likely "only" in the lower 80s as it is too warm by ~3 F. The Hunter (KSVN) station's 82 is likely about what it really is. CHS NWS has human observers there. So, at some point later the official record will reflect what they record. As it is though, 82 would be only 1 F from tieing the alltime December record at KSAV. So, KSAV is already flirting with the 83 monthly record. Strong -PNA/La Niña ftw if you like warmth in December! This was posted starting on December 1st by the CHS office: "EQUIPMENT TEMPERATURE DATA AT THE KSAV ASOS ARE SUSPECT. PARTS ARE ON ORDER AND WILL ARRIVE THIS WEEK. USE DATA WITH CAUTION."

This isn't just due to La Niña. It is also likely due to heavy convection in and near the MC, something that has been a factor for at least several years due partially to near record SSTs there.

The 12Z EPS 11-15 is milder in the E US than recent runs with it only down to near normal in the SE vs BN on prior runs. This likely has lead to NG dropping now to session lows (-4% on the day).

1. Agreed. With a cold E US December, NG would likely now be much more expensive! There are clear benefits to a mild winter. 2. Fwiw, a stronger +PNA coming out now on the 12Z GEFS in the 11-15.

1. Yes, the jet crash was in Jan of 1982. During the second week of Jan of 1982, Atlanta was just about totally shut down. First, they had their first below zero F since 1966. Then they had a three day barrage of a heavy mix of wintry precip (front part of this was the infamous "Snow-jam 1982") due to back to back Gulf of Mexico storms followed by a second very cold airmass dropping temperatures back to zero! 2. Just before Christmas of 1989, the SE coast/N FL got a historic snow giving first on record white Christmas to many with a dusting as far south as Sarasota!: 3. Nothing for me on Veterans Day 1987.

I agree about the south to southern mid-Atlantic (pretty strong correlation to cold of +0.5 to +0.6) though it is still a pretty cold signal even up into the SNE with a respectable +0.4 to +0.5 correlation even up there (nearly +0.5 at NYC):

But for many of us down here in the SE US, the 80s was a very wintry decade with quite a few major and widespread winter storms including March of 1980, Jan of 1982, March of 1983, Jan of 1987, Jan of 1988, Feb of 1989, and Dec of 1989. In some parts of the SE, the 1980s had the heaviest liquid equivalent for wintry precip in the better part of 100 years. That extreme cold that deterred snow in the NE US enhanced it further south.

NG market down another 2-3% as of now (~25% down vs Thursday's high!): https://www.investing.com/commodities/natural-gas As previously mentioned, this market is often as good a barometer as any of temperature trends in the 14 day forecast in the E US as a whole. SE cold lovers generally want NG to rise on any day. The trend in the E half of the US and especially E 1/3 in the most recent EPS and GEFS run means often has the most effect though the operational GFS and especially Euro 6-10 can have significant short term effect. Thus, NG is telling me that milder trends since Thursday in the E US over the next two weeks are still not reversing as of this moment anyway. We'll see how the rest of the 12Z GEFS as well as the 12Z Euro suite affect NG over the next few hours. A deadcat bounce is due regardless of the wx outlook.

The 12Z EPS is still another EPS run establishing a +PNA around 12/16-17. So, it has been persistent with this for several days. We'll see whether or not the model will finally be right and the pattern actually change to cold dominated around then. The PNA is forecasted to rise to neutral to possibly a slight positive then while a -NAO per the GEFS at least is forecasted to then be getting reestablished along with a -AO. The MJO per GEFS is forecasted to possibly then be headed toward low amplitude/inside COD on left side, which in combo with a -AO/-NAO/neutral to low amp +PNA would typically be cold in the SE in mid to late December. We'll see.

If you want to know how the 14 day model consensus (mainly GEFS and EPS runs) forecast for the E US is trending from day to day, especially in late fall and winter, a great clue more often than not is the natural gas (NG) market. If warmer/colder than the previous trading day (M-F), it will often be down/up. Currently, it is down a whopping 10% vs Friday's close! Also, it is down an amazing 22% just since the high of the Thursday session, which is only two trading days ago! So, the NG market is telling us that the models have been adjusting much warmer in early to mid December in the E US vs what they showed just 4-5 days ago meaning they've been strongly cold biased. I believe that this cold bias (models being much too cold) is probably mainly the result of intense convection in and near the Maritime Continent and Indian Ocean leading to a warmer pattern vs where the official MJO is. It thus acts like a warm MJO phase for the SE US. The models don't handle this well and thus it has resulted in a cold bias on many occasions over recent years. The intense convection has often been the result of (near) record warmth in that area. For this reason and as long as this too cold trend in the models continues, it is best to be more wary about cold runs than warm runs of models. At some point, this trend will cease and cold will start verifying better. But when? Current NG prices: https://www.investing.com/commodities/natural-gas-historical-data

I think the following might be of interest as I happened to have recently looked at the December 1-15 temperature anomalies at Raleigh-Durham, NC, (RDU) for the seven years since 1950 with an -NAO of -1 or stronger averaged out for the period 12/1-15 (year: average NAO for 12/1-15, RDU temperature anomaly): 1963: -1.5/very cold -7 2009: -1.3/cool -2 2010: -1.2/very cold -12 1977: -1.1/cold -5 1989: -1.1/very cold -7 1995: -1.1/cool -2 2002: -1.0/very cold -9 *Average was a cold -6 *None were mild. All were colder than normal with 5 of 7 cold to very cold. *2010 with -12 was the coldest of these and yet it somehow did this with a PNA way down at -1.0, the 4th lowest 12/1-15 PNA since 1950! The lowest 12/1-15 PNA since 1950 isn't that much lower (-1.2 in 1972). *2022's 12/1-15 NAO/PNA combo is forecasted by the GEFS to end up by far the closest to 2010 vs any other year since 1950. The latest forecast is for an NAO of ~-0.8 to -0.9 and a PNA of ~-1.1. In 2010, the NAO was -1.2 (3rd lowest since 1950) and the PNA was -1.0 (4th lowest PNA since 1950). *But despite the similarity to the NAO/PNA of 2010, the latest forecast temperature anomaly for RDU for 12/1-15/2022 is for it to average significantly warmer than normal, the opposite of 2010's extreme cold! With it only December 5th, the actual 12/1-15/2022 is still somewhat up in the air especially considering that this is a difficult pattern to forecast. Regardless, the period as a whole clearly isn't currently looking cold even with the possibility of it ending cold and thus looks to be far different from 2010. RDU temperatures: https://www.weather.gov/wrh/Climate?wfo=rah Daily NAO since 1950: https://ftp.cpc.ncep.noaa.gov/cwlinks/norm.daily.nao.cdas.z500.19500101_current.csv Daily PNA since 1950: https://ftp.cpc.ncep.noaa.gov/cwlinks/norm.daily.pna.cdas.z500.19500101_current.csv

The end of the 0Z Euro fwiw is by a good margin the coldest in the SE US at 850 mb for any run yet this season. The 240 hour is absolutely frigid at 850 mb with 850 mb temperature anomalies a whopping 10-15 C (18-27 F) colder than normal! I don't have 2m temps yet. Again, fwiw that far out. Runs like this for day 10 don't tell me what's likely or anything close to likely for day 10. They just tell me the kind of thing that potentially can happen, I.e., a small chance for something similar in day 10, especially considering that models have been too cold. Edit: Although not as cold in the 11-15 in the SE as the coldest yet prior run because the 500 mb trough axis is a bit further west, the mean 0Z EPS is still chilly with a +PNA forming. Moreover, with the trough axis near the Mississippi River instead of the SE, this run is more moist and includes a few members with Miller A Gulf lows moving ENE at the bottom of the cold air. I don't have access to mean snowfall or individual members' snowfall, but I bet a couple of members have snow around the end of the run.

Further to the above, I just saw the 12Z EPS and it is I believe easily the coldest run in the SE of any recent run for any 11-15 day period. It is significantly colder than any run of the last few days for sure. At face value, it is hard to beat this run if you like cold. Linking this to my post above, this is the kind of thing that is quite possible with an MJO forecast like what the GEFS has. But the EPS and other models have been too cold during earlier runs for the 11-15 day periods. As I said, I suspect high levels of tropical convection in the MC/IO are at least partial culprits. If that convection were to diminish, then that might reduce the tendency of models to be too cold in the SE and thus allow for cold similar to the 12Z EPS' 11-15 to actually happen. The timing will be crucial to be able to take advantage of the progged cold favoring MJO should it really end up as the GEFS has.

Here is the PNA for the seven strongest -NAOs since 1950 for Dec 1-15 along with the RDU temperature anomaly for the respective Dec 1-15: 1963: +0.8/-7 1977: -0.4/-5 1989: +0.9/-7 1995: 0.0/-2 2002: +1.2/-9 2009: +0.1/-2 2010: -1.0/-12 Here in 2022, the GEFS is now forecasting a Dec 1-15 PNA of -1.1, which would be tied for second lowest since 1950, and an NAO of -0.9, which would be ~8th lowest NAO for Dec 1-15. So, 2022's -PNA/-NAO for Dec 1-15 is actually most similar to 2010, which despite a strong -PNA was the coldest Dec 1-15 at RDU of the strong -NAO group! It is the only Dec 1-15 similar to 2022. Thus as mentioned earlier, this makes for an unusually difficult forecast for the models as well as forecasters. Therefore, my confidence in model runs is lower than normal and I expect continued volatility in model runs. Currently, for the first half of this month, warmth is winning out over cold for the SE US on the recent model runs as runs have largely been turning warmer for the same period. This may be due to enhanced tropical convection in or near the Maritime Continent/Indian Ocean, which tends to act like a warm phased MJO regardless of what the official MJO phase is. This has often occurred in recent years due to record or near record warm SSTs there and has often resulted in model runs being cold biased. This often results in runs showing cold and then backing off in later runs for the same period as it gets closer. Also, La Niña in general favors a stronger SER (southeast ridge) than other ENSO. Whereas La Niña isn't just going to disappear, the enhanced tropical convection that I mentioned can change. I suspect that IF a -NAO/-AO regime were to hold, the best chance for SE cold to take hold for a several week period would be if that enhanced convection were to diminish and possibly allow a +PNA to become dominant. The GEFS is saying to look for a slow moving low amp/inside COD MJO on the left side of the circle to possibly occur starting around Dec 17th. Research has shown that especially during and near midwinter that a slow moving low amp MJO just outside to inside the left side of the COD (especially rotating from 7 to 8 to 1 to 2) has on average been the coldest MJO for the SE US. IF this occurs, IF the -NAO/-AO were to still be around, and IF that convection were to diminish, I'd be looking out for a +PNA to dominate along with a cold 2nd half of this month. Lots of IFs though. So, this isn't a prediction. (Sorry for the big font, but I didn't do that on purpose. I wasn't able to change it using my phone.)

It was a warm day in this area today with highs near 79. Strong -PNA in control.

Today's 0Z GEFS based forecast for 12/1-15 is for an NAO of ~-1.0, which would be about tied for 7th lowest NAO since 1950. The quoted post above shows an average at RDU of a cold 6 colder than normal for 12/1-15 for the 7 NAOs since 1950 of -1.0 or lower. OTOH, today's GEFS based forecast for 12/1-15 is for a PNA of ~-1.0. This would be for 12/1-15 the 4th lowest PNA since 1950 and the lowest since 1973! Here are the RDU temperature anomalies for the six 12/1-15 periods since 1950 with a PNA of -0.8 or lower: 1956 (PNA -1.1): quite mild +7 1961 (PNA -1.05): normal -1 1972 (PNA -1.2): mild +4 1975 (PNA -0.9): normal 0 2012 (PNA -0.8): quite mild +7 2021 (PNA -0.8): mild +4 *Average: mild +4 *None were cold and 4 of 6 were mild to quite mild. So, there is a forecasting conundrum for the first half of this month with the strong -NAO analogs averaging a cold -6 and the strong -PNA analogs averaging a mild +4. This illustrates the complexity of the pattern and thus the higher level of difficulty of the models sorting this out. This tells me to keep confidence in any one model run (including ensembles) lower than normal and to expect larger than normal changes from run to run. --------------------- *Edit on 12/4/22: Although my list of strongest -NAOs for Dec 1-15 is correct, I inadvertently left off 2010's -1.0 from my list of strong -PNAs. Thus, I need to update the -PNA analysis and will do so with today's GEFS based forecast, which is ~-1.1 and ~tied for the second strongest -PNA since 1950 for the first half of December. The GEFS NAO forecast has risen from ~-1.1 to ~-0.9 though that's still strong enough to be ~8th strongest. Here are the RDU temperature anomalies for the seven 12/1-15 periods since 1950 with a PNA of -0.8 or lower: 1956 (PNA -1.1): quite mild +7 1961 (PNA -1.05): normal -1 1972 (PNA -1.2): mild +4 1975 (PNA -0.9): normal 0 2012 (PNA -0.8): quite mild +7 2010 (PNA -1.0): very cold -12 2021 (PNA -0.8): mild +4 *Average: normal +1 So, adding 2010's extreme cold despite a strong -PNA changes the average considerably since it is a near normal +1 vs a mild +4 without 2010. Including 2010 is extra important because it is by far the closest strong -NAO/-PNA analog to 2022 with 2022 ~-0.9/-1.1 vs 2010's ~-1.2/-1.0.

So, are these models right? Or are they underestimating the cold potential of the forecasted -1.1 NAO for 12/1-15? Which is more likely? I don't know but consider this for RDU's 12/1-15 temperature anomalies for the seven years since 1950 with a -NAO of -1 or stronger: 1963: very cold -7 1977: cold -5 1989: very cold -7 1995: cool -2 2002: very cold -9 2009: cool -2 2010: very cold -12 *Average: cold -6 *None were mild. All were colder than normal with 5 of 7 cold to very cold. RDU temperatures: https://www.weather.gov/wrh/Climate?wfo=rah

Per the 0Z 11/30/22 GEFS run, the first half of December is forecasted to have an average NAO of near -1.1. That would by a good margin be the strongest -NAO in the first half of December since 2010. Looking back to 1950, here are the strongest -NAOs for the first half of December: 1. 1963: -1.5 2. 2009: -1.3 3. 2010: -1.2 4-6. 1977, 1989, 1995: -1.1 7. 2002: -1.0 The remainder since 1950 are -0.8 or weaker. So, based on today's GEFS, 2022 is looking to be at about tied for 4th with 1977, 1989, and 1995 for the strongest -NAO for the first half of December since 1950. Only 2010, 2009, and 1963 were stronger than this forecast and there's a slight chance that it could reach 2010's strength. Daily NAO data since 1950: https://ftp.cpc.ncep.noaa.gov/cwlinks/norm.daily.nao.cdas.z500.19500101_current.csv Edit: Since 1963 and 2009 were during El Niño, 2022 is aiming to be tied for 2nd for the strongest non-Nino -NAO in the first half of December since 1950 (behind only 2010).

This would be pretty darn good in a very general sense for cold SE prospects IF this were to verify since a very important +PNA would be joining the strong -NAO/-AO combo. This new +PNA was also showing up starting around the same time on yesterday's 12Z EPS (December 12th or so) as hinted at in your post yesterday. The following 0Z went away from it, but this run is the strongest yet. This is a pretty strong signal for 13-14 days out. Although Dec prior to this time looks to average out warmer than normal, this is the kind of setup that would have the opportunity to result in a cold SE December overall, just as was the case in the 3rd year cold ENSO analogs of 1910, 1917, and 2000. Let's hope future ensemble runs continue to show a new +PNA for around 12/12.

Followup with more details for NC major snow based on my own research: When I look at very heavy RDU winter storms (I chose the 21 6”+ for minimum needed, which means mainly or all snow in just about all cases) since 1950, a +PNA was favored a pretty decent amount: +PNA (+0.25+): 11 storms (52%) Neutral PNA (-0.25 to +0.25): 7 storms (33%) -PNA (-0.25-): 3 storms (14%) So, for RDU for heavy snowstorms, there’s a partial correlation with a +PNA. Note that this isn’t for mainly ZR/IP, which keep the total under 6”. These probably prefer a -PNA to neutral PNA. I don’t know for sure though. I think what it boils down to is that classic major SE Miller A snowstorms prefer neutral to +PNA as opposed to CAD based Miller B icestorms, which probably prefer neutral to -PNA. For the +PNA big snows, the mean trough normally needs to be fairly close to the Mississippi River so that it doesn’t suppress the storm track too far south. Here's the PNA and the NAO for the 21 Raleigh 6"+ SN/IP storms since 1950. (I'm calling neutral PNA/NAO to be from +0.25 to -0.25): -1/19/1955: +0.1 neutral PNA; -0.8 -NAO - 12/11/1958: +0.7 +PNA; +0.0 neutral NAO - 3/2-3/1960: -0.8 -PNA; +0.1 neutral NAO - 3/9/1960: -0.5 -PNA; -0.2 neutral NAO - 2/26/1963: +1.3 +PNA; +0.2 neutral NAO - 1/25-7/1966: +0.2 neutral PNA; -1.2 -NAO - 2/9/1967: +0.5 +PNA; +0.5 +NAO - 3/1/1969: +1.3 +PNA; -0.6 -NAO - 1/7-8/1973: +0.3 +PNA; -1.1 -NAO - 2/18-9/1979: -0.0 neutral PNA; -0.1 neutral NAO - 3/1-2/1980: +0.2 neutral PNA; +0.4 +NAO - 3/24/1983: +0.7 +PNA; +0.1 neutral NAO - 2/6/1984: +0.7 +PNA; +1.1 +NAO - 1/7-8/1988: +0.7 +PNA; +0.5 +NAO - 2/17-8/1989: +0.0 neutral PNA; +1.6 +NAO - 1/24-5/2000 (“Crusher”): +0.8 +PNA; -0.5 -NAO - 1/2-3/2002: +1.2 +PNA; -0.5 -NAO - 2/26-7/2004: -0.1 neutral PNA; -0.5 -NAO - 12/25-6/2010: -0.3 -PNA; -0.9 -NAO - 1/17-8/2018: -0.1 neutral PNA; +1.2 +NAO - 12/9-10/2018: +0.9 +PNA; +0.9 +NAO Tallies: 1) PNA: +PNA: 11 (highest +1.3) Neutral PNA: 7 -PNA: 3 (lowest -0.8) **Average PNA for the 21 big RDU snowstorms: +0.4 2) NAO: -NAO: 8 (lowest -1.2) Neutral NAO: 6 +NAO: 7 (highest +1.6) **Average NAO for the 21 big RDU snowstorms: +0.0 Correlation of +PNA and surface temperatures DJF: https://psl.noaa.gov/cgi-bin/data/correlation/corr.test1.pl?iregr=1&var=Air+Temperature&level=Surface&mon1=12&mon2=2&iy[1]=&iy[2]=&ilead=0&ilag=0&type=2&timefile=&customtitle=&labelc=Color&labels=Shaded&cint=&lowr=&highr=&scale=&switch=0&proj=USA&xlat1=&xlat2=&xlon1=&xlon2=&custproj=Cylindrical+Equidistant&level1=1000mb&level2=10mb&Submit=Create+Plot ------------------------------- Data sources: 1) RDU 6"+ snowstorms: https://www.weather.gov/wrh/Climate?wfo=rah 2) Daily PNA: https://ftp.cpc.ncep.noaa.gov/cwlinks/norm.daily.pna.cdas.z500.19500101_current.csv 3) Daily NAO: https://downloads.psl.noaa.gov/Public/map/teleconnections/nao.reanalysis.t10trunc.1948-present.txt

The +PNA has the highest correlation to SE US snow overall for any non-seasonal binary index. That's especially the case the deeper in the SE one goes. This includes major NC snowstorms, which I researched earlier. This idea excludes ZR. Based on my own past research, this strong correlation also applies to cold, again especially deeper into the SE. Regarding seasonal, El Niño has the highest correlation for both in the SE. It is no coincidence that a more frequent +PNA is favored by El Niño over other ENSO.

1. Speaking of "natgas", which is tagged in both tweets, it is currently actually down 3% vs the Friday close. So, from that market's perspective at least, the intensity of projected December cold for the E half of the US overall vs how it looked Friday at 12Z isn't exciting fwiw. If it were, that market would almost definitely be up now. 2. I'm wary of those tweets that talk about cold in the E US and also tag "natgas". When they do that, they're often trying to get natgas prices to spike up meaning there may be an ulterior motive to such tweets. 3. That all being said, three of the eight third year cold ENSO analogs had quite a cold December in the SE US: 1910, 1917, and 2000. So, that tells me that a cold December in the SE is a decent possibility. Whether or not it is December, I think there's a good chance that one of the three winter months will be solidly cold. Edit: Natural gas, which was down 3% earlier as mentioned, reversed to go higher at 8AM as ensembles turned colder late in the runs at 0Z/6Z. Edit #2: As of 2:20 PM, it is up a whopping 8% (session high) on continued support for a cold December from the 12Z runs! That's quite impressive.

This afternoon with a highest only near 50 was cold here even compared to midwinter afternoon averages that are well up into the 50s. Another nice taste of winter.

Climo says extremely unlikely for a TCG that late in the season, but fwiw I count a rather notable 6 of the 52 (12%) 12Z EPS members with a TS in the W Car between 11/27 and 12/3. That compares to only 1, 3, and 0 on the prior three long range runs. That is not exactly what one would expect to see that late in the season on the EPS. On the GEFS or GEPS, 12% wouldn't be notable. Til 12Z today, this was purely a GFS out in fantasy land storm. But the 12Z CMC had something and now the EPS is possibly doing a little sniffing fwiw.

First freeze at KSAV (32) and KSVN (31) although I don't think I was quite that cold. This is more typical of midwinter though first freezes are more often than not in November in this area even during the warmer era since 2000 (14 of the last 23 Novembers at KSAV). To show how misleading is KATL, they just had their first freeze way up there, too!