Sunday, January 26, 2014

January 28-30 Potentially Historic Southeast Snowstorm Update

This is an update to the earlier post on the January 28-30 Potentially Historic Southeast Snowstorm event.

We'll begin with the NAM model, which has actually beefed up its snowstorm amounts since its previous run. The NAM model now brings snow amounts near 2 FEET in the hardest hit areas, namely eastern South Carolina into eastern North Carolina. We see widespread amounts in excess of 12 inches slathered across the Southeast from Pensacola, Florida to southern Georgia and into the Carolinas. Accumulating snow also hits far southern Mississippi and Alabama, even extending into Louisiana. The NAM indicates the maximum amount possible in its solution is 21.2 inches, which would most likely hit somewhere in the Carolinas. I maintain my position on the idea that the NAM is too high with snow amounts, which would leave the Southeast paralyzed for not days, but possibly over a week or two.

The 18z GFS model really beefed up its snowfall amounts, bringing amounts as high as 22 inches to eastern South Carolina into eastern North Carolina. The model wants to begin a swath of accumulating snow in eastern Texas, including the city of Houston, before blowing up totals to near 12" amounts in southern Mississippi and Alabama. From then, significant snowfall pummels the states of Georgia, the Carolinas and even Virginia. Once again, I'm a little skeptical of the high snowfall amounts. Since models have been having trouble handling the energy responsible for this snowstorm potential since the get-go, confidence in these extreme snowfall totals is not that high.

We now go to the Canadian GGEM model's projection, which is far less bullish than the NAM or GFS models. The GGEM begins with a bit of snow in southern Mississippi and southern Alabama, only beginning to bring up snow totals in far eastern South Carolina and really getting going in extreme eastern North Carolina. I don't think we'll see amounts this low- while I'm not buying the extreme snow totals forecasted by the already-discussed GFS + NAM, I don't think we'll see these low totals either.

Lastly, we take a look at the European ECMWF model. The ECMWF starts out with accumulating snow in eastern Texas before weakening, only dropping dustings of snow throughout southern parts of Louisiana, Mississippi, and Alabama. We then see 2-4" totals popping up in Georgia, before 6" amounts form in far eastern South Carolina. The bullseye here hits the far eastern areas of North Carolina, where the ECMWF lays down 13.3". I'd like to see this model beef up amounts across the board and draw the higher totals back to the northwest, like the GFS and NAM do. I expect we will see 6" totals in Georgia, with totals near a foot (possibly just over that benchmark) in the Carolinas.

To sum up, I expect we see a swath of 1-3" snows from eastern Texas to southern Louisiana, and into southern Mississippi. From there, accumulations of 4-8" look to pop up in southern Alabama and the panhandle of Florida, before accumulations really ramp up in Georgia, South Carolina and North Carolina, where I expect 8-12"+ to fall.



Eric (weather advance) said...

Nice post Andrew, but I do want to give my thinking on things, see what you think. FIrst of all, some of that precip in the GFS & NAM runs isn't all snow. You see, the model automatically assumes all of the wintry precipitation that falls, whether that be snow, sleet, &/or freezing rain is snow & there's also an assumption of 1:10 ratios, that are likely to variate substantially during the event from higher ratios on the northwest side of the low pressure system where the air is drier & higher to the south & east closer to the low pressure track where the snow is slightly melting as warm air is advected aloft. The big point I want to get across though is (although you can disagree w/ me here) that I am going to side much more with the GFS than the ECMWF , CMC, & even to a certain extent the NAM at the moment. The reason why has a lot to do with the handling of the subtropical jet energy which is currently nearing the Baja Peninsula. The CMC & ECMWF you should note take this energy & try to crash it southward into Mexico, but this solution seems like absolute garbage if you understand the pattern here @ hand. In case you haven't seen on twitter, I've posted the observed trends with these models handling the subtropical jet energy (which originated via a convectively coupled Kelvin Wave that got going near the Maritime Continent a few weeks ago) & how they have trended significantly to the northeast in just the past 2-3 days from at one pt keeping the low held way down in the eastern Pacific now have been forced to push the low towards northern Mexico.

Andrew said...

Nice disco, Eric: you certainly raise some valid points. There will likely be points of mixing throughout this storm which ought to cut down on those snow totals, as you alluded to- I would expect totals to be a bit lower overall if snow ends up falling more in that 8 or 9:1 ratio range in response to that mixing in of ZR and sleet.

Eric (weather advance) said...

You can thank these errors in the CMC & ECMWF to their slow bias in the SW US & I think this bias in the models is @ the root as to why they're no way near aggressive with snowfall in the southeastern US as the GFS. Even the NAM at one point started to get very aggressive on the 12Z run, but backed off at 18Z & that's simply due to the fact that its trying to bury the subtropical jet energy southward into Mexico. Now, what's really interesting is the GFS from saturday looks a lot like the new CMC & ECMWF solutions, which suggests that these models are lagging behind the GFS (for once.) The placement & speed of this subtropical branch energy is the key part to this whole set-up which would be all it takes to set-off a historic winter storm in the southeast US. However, because of the slow bias of the CMC & ECMWF, they simply hold the energy too long over the SW US & allow a new ridge to build over top the storm system in advance of the next oncoming trough that will establish a west coast trough which was well predicted in this timeframe using the Lezak-Recurring Cycle. The implication here of the ECMWF & CMC holding back the subtrop energy & trying to unrealistically crash it southward into Mexico is that it doesn't phase with the northern branch energy & Alberta Clipper. Thus we are left with a positively tilted trough axis & frontal boundary that can still produce significant some winter weather across the southern US. However, I think the really telling aspect that the CMC/ECMWF are wrong is that if you compare their 500mb of the operational to the ensemble, you should see how much further to the NE the model's own ensembles are with the subtropical jet energy, which suggests that the operationals are likely way too far southwest & out to lunch (esp. considering huge differences already exist @ just over 48hrs) & we are likely going to see these models become more aggressive with time & trend closer to the GFS. The most interesting aspects of this whole ordeal is some of best analogs like Feb 1973 & even to a smaller extent Jan 2000 are completely crazy set-ups for the SE US. We also share a commonality of not just this storm, but in setting the +TNH record last Dec, we beat out the 1972-73 winter. Also, out of all the entire model suite, the GFS is the one that agrees most with its ensembles while significant disagreement is evident in models like the CMC for ex, which just adds credibility to the seemingly "outlandish" GFS. In addition, you should note that the GFS is usually the model we scold for being too flat, progressive, & far east with east coast troughs, however, its currently the western outlier. Just some food for thought