GaWx

From KCHS first issued this afternoon: ATTENTION THEN TURNS TO THE VERY END OF THE FORECAST PERIOD WHERE THERE IS INCREASING CONFIDENCE IN A WIDESPREAD PRECIPITATION EVENT IN ASSOCIATION WITH A COASTAL LOW TRACK, POTENTIALLY BEGINNING AS EARLY AS TUESDAY. THIS EVENT WOULD BE UNFOLDING IN THE PRESENCE OF THE ARCTIC AIRMASS NOTED ABOVE, YIELDING THE POTENTIAL FOR WINTER WEATHER. MODEL GUIDANCE, BOTH OF THE DETERMINISTIC AND ENSEMBLE VARIETY, SUGGESTS A THREAT OF WINTER WEATHER FOR THE FORECAST AREA AND THE REGION AS A WHOLE. IT MUST BE STRESSED THAT CONSIDERABLE UNCERTAINTY REMAINS, ESPECIALLY REGARDING SUCH IMPORTANT ITEMS AS POSSIBLE PRECIPITATION TYPES, POSSIBLE PRECIPITATION AMOUNTS, AND SPECIFIC IMPACTED AREAS. Our local TV forecasters are wisely not yet mentioning this possibility due to too much uncertainty, especially with wintry precip being rare here (it’s been 7 years without even a trace).

Based on my analysis today of individual members’ qpf, 2m temps, and snowfall, I now know for sure that the WB snowfall on the EPS, like on the GEFS, excludes IP and ZR. My biggest concern for here is for significant ZR potential. The 12Z EPS and 18Z GEFS means have ~0.14”/0.10” of qpf in the form of 10:1 calculated snow, which is itself very high for an ens mean for snow a week out for this area. But what these aren’t showing is that the mean qpf in the form of IP/ZR on the 12Z EPS is an additional whopping ~0.4” and on the 18Z GEFS is ~0.3”. This means that mean IP/ZR producing qpf is ~3 times that for snow on these two runs! So, just looking at snow means for especially from E NC through far N FL, where the IP/ZR is a bit more prevalent on the 18Z GEFS, is deceptive for this threat. I very roughly estimate that the IP/ZR qpf breakdown in the deep SE on the 18Z GEFS is 75% ZR. That implies that the mean ZR qpf could very well be 0.25-0.3”, which is quite high for an ensemble mean a week out. Hopefully there wont be a bad icestorm! But unfortunately for the deep SE, that is a significantly higher probability than a big snow based on these latest two ens means. Hopefully that will change.

18Z GEFS mean snow:

3 of the 50 EPS members show the heaviest snowfall in my area in ~185 years after I decreased two of them from the 10:1 calculation, which would be too high in those two cases.

12Z 1/3/24 EPS 7 days before 1/10 storm (ignore VA/KY and further north snow as that was from an earlier storm): 12Z 1/14/24 EPS 7 days before projected 1/21 storm:

I’ve saved a lot of EPS mean snow maps with lots of snow and this one easily takes the cake for the heaviest overall from E half of NC to the Gulf coast, especially this far out! This is massive for being 7 days out, especially in those areas. Some of the members start as snow as early as late afternoon on Mon (1/20).

1/7/1988. 4.2” officially of mainly sleet at KATL and maybe a little less in Dunwoody. Stayed on the ground for a long time! Opposite (sort of) occurred 2/1979. Was expected to be mainly snow and ended up almost all sleet (4.2” also at KATL!!) after no more than first hour or so of snow.

Thanks. Here’s my personal experience. I was in Dunwoody for the 2/2014 storm. It started as rain that quickly changed to ZR. Fortunately, the ZR lasted only a few hours and was fairly light. It changed to a long period of sleet (~1”, my 3rd highest there) followed by snow to end it. But the southside (airport to Tony) eastward to Augusta wasn’t so lucky as they had majority ZR and major problems. The Eisenhower Tree at Augusta National fell because of it. We’re you then in ATL area?

Thanks. Regardless, the Euro was too cold further out or 7-10 days before last week’s storm. Remember those runs with single digits at ATL, for example? Also, storm tracks for that storm out 7+ days were much further SE than what verified. But we’re now getting right to day 7/crucial forecast time. It did much better for last week’s storm at 96-140, consistent with what you said. But we’re not there yet. By late tomorrow we will be getting to 140. This as you’ve implied is a much different setup with a much colder Siberian high that is progged to be 20 mb stronger to our north/NW vs the weak and weakening high for last week’s storm.

No. It and the models as a whole are cold biased, which is why the storms often track NW and temps are often not as cold as one gets closer. But there are always exceptions. Nobody knows if this will be an exception, but it is a much different setup vs last week’s storm.

12Z Euro as you said now has a winter storm back (1st since 12Z run on 1/12): Snow: Sleet: ZR:

GFS runs with significant SE coastal wintry precip within 1/21-2: Jan 11 6Z and 12Z Jan 12 6Z and 18Z Jan 13 0Z, 6Z, 12Z, 18Z Jan 14 0Z, 6Z, 12Z So, 11 of last 14 and 8 in a row! The 12Z CMC has ZR later, on 1/23-4, with it still falling some areas especially just inland at the end (fortunately only light along coast SAV northeastward):

Here’s 12Z GFS sleet for all to look at and especially for @dsaur

See above as I added it centered on AL/GA/SC.

12Z GFS fwiw (taking with a huge grain) as a result of 24 hours straight of ZR has close to the worst icestorm on record in the SAV area with 1.45” (Hunter) and 1.63” (airport further inland) at reporting stations and a general 1.25-1.65” over much of the area away from the islands. The 6Z GFS also had a lot of ZR with 1.30” at airport and 0.66” at Hunter. Statesboro had a whopping 1.89”! Needless to say, I might be taking a 2nd Jan trip to snowy Atlanta (in this case for two reasons) if something like this were to somehow materialize:

12Z Icon is much drier than most other models for 1/21 (similar to 0Z run) but it is likely delayed but not denied as the run stops too early to see. It has light snow moving ENE over extreme S LA from a W GOM development while the SE is very cold due to a 1040 mb Siberian high right there: Note the split flow with the very cold WNW flow into the Midwest while at the same time there’s moistening WSW H5 flow over the SE going over the lower level very cold dome:

1) EPS 0Z 1/3/24 EPS (7 days before 1/10 storm): amounts were dropping then: Compare that to 0Z 1/14/24 EPS (7 days before potential 1/21/24 threat): much higher most of GA, SC, NC especially E NC to FL panhandle and with substantial increase since 12Z: 2) GEFS 0Z 1/3/24 GEFS (7 days before 1/10 storm): amounts were dropping then: Compare that to 0Z 1/14/24 GEFS (7 days before potential 1/21/24 threat): much higher ALL of GA, SC, NC, and FL panhandle and with substantial increase since 12Z:

Big increases on 0Z ensemble snow means throughout SE on GEFS, EPS, and GEPS for 1/21-2 vs 12Z: EPS: 12Z 0Z CMC ens: 12Z 0Z

Tony, It’s because those half county areas are on the border between one category and the adjacent one. They’re barely in the higher category. For example, that one county N of Columbus with the higher category is likely only barely above 1” vs their neighbors barely under 1”.

GEFS: 0Z vs 12Z 12Z 0Z big increase in S VA, NC, SC, GA, and Gulf coast!

GFS runs with significant SE coastal wintry precip within 1/21-2: Jan 11 6Z and 12Z Jan 12 6Z and 18Z Jan 13 0Z, 6Z, 12Z, 18Z Jan 14 0Z So, 9 of last 12 and 6 in a row! Normally when the models tease this area the NW trend takes the wintry precip away as it gets closer. That has yet to happen at all in this case. The coastal wintry GFS runs for 1/21-2 started at just over 10 days out. We’re now down to 7.5 days out and there’s no NW trend yet. The prior threat trended NW enough to end the coastal threat by 6 days out. If this doesn’t trend NW by during the Wednesday runs, this would be looking like a totally different situation. It already looks suspiciously different and I wonder if it is due to the Siberian high/extreme cold making this a different situation. 12/1989 and 1/2018 also had extreme cold to the north feeding those coastal storms. Edit: 2 of the last 6 Euro 0Z/12Z runs have had it then, both on 1/12: 0Z and the insane 12Z. The two on 1/13 were dry. The 0Z 1/14, just out, is also dry. No CMC runs through 0Z 1/14 have had it yet. So, since 1/13, GFS has been on an island. @Awesomesauce81 @gtg947h

I’m not really familiar with WxModels to be able to compare them.

0Z GFS has 12/1989 or 1/2018: chances of this actually occurring with these amounts on or near the coast…history says small; but note that these types of runs haven’t yet gone away

0Z GFS

0Z GFS