Scarlet Pimpernel

Haha! Yeah, kind of reminiscent of the "Groundhog Day" Feb. 2 snow we got right before the two big HECSs shortly after! I don't know how many people were really caring much about that at the time, even though we got ~5" out of it (and it was very pretty, similar to today's snow)! Not saying I think next weekend and the following week is a "guarantee" by any means, but man, that signal for *something* big is there. If it weren't for the fact that my power would probably go out in this situation, I'd think (if it can't be all snow) it would be cool to get heavy snow followed by glacier/ice followed by wrap-around snow, then very cold! That would be damn near epic. Hell with it, I've got plenty of blankets and will have a fully charged phone, screw it if the power were to go!

A few photos (one from last night, two from today)...

Nice photo! I think that buck is saying, "Dafuq you looking at??"

Right around 7.0" as of 11:00AM, Chevy Chase/Bethesda. Some light snow falling.

Every day is a good day for bacon!

Bit late reading through the most recent stuff here, but yeah I think that's snow followed by a boat load (iceberg load?) of sleet and ice then some snow at the end and very cold. Would be quite spectacular if very destructive. There have been some crazy solutions appearing the past couple of days!

Hell I still haven't totally recovered from the Indians losing in extra innings of game 7 vs. the Cubs two seasons ago...or for that matter also losing game 7 in extras vs. the Marlins in 1997!

True story: had a dream last night where I woke up, it was sunny, and I was walking around outside with almost all the snow melted except a few patches. It's almost as if snjokoma infiltrated my dreams LOL!

6.8" measured around 8:30am here in Chevy Chase/Bethesda. Light snow falling...beautiful out! Got several photos.

Right around ~2.5" Chevy Chase/Bethesda area. Nice sized flakes ripping for awhile, has let up a bit now. Side roads are long gone, even some major roads (E-W Hwy, Conn Ave) are not exactly easy traveling.

Haven't actually measured anything yet, but here in Chevy Chase/Bethesda area, ground is mostly coated, including pavement. Snowing lightly and has been for a little while now.

True...the heartbreak events tend to stick with you, I agree. But man, these reverse busts are awesome when they do happen! Others have said this I think already, but this event kind of makes me think of Jan. 30, 2010.

Have fun with it, Supernovasky! I think we've all earned it, indeed! And if things are how they appear now, this won't be the last warning-level potential you experience this year!!

One caveat...if you look *very* closely, you'll notice the one pixel right immediately over DCA has only 17.8".

LOL! Agreed, I wouldn't buy the extreme amounts the NAM is spitting out (unless every other model does same thing today). But I do like the overall increases in QPF and better looks that guidance is showing. I think LWX right now is OK with low end warning extending into immediate metro DC area and increased advisory amounts north of there...seems reasonably conservative and realistic for right now pending later guidance.

He's thinking, "meh...bring on the MLK storm next weekend!"

I see that, and my apologies, you in fact didn't specifically say the GFS for that case (2016 blizzard). It just happens to be one of the models I saved output for (being an historical event, I had to keep various model output!) so I showed it. But even still, the models that were pumping out high QPF were correct overall in the end I recall. Which was nearly all of them by that point! Maybe we didn't get the ultra-"extreme" values that some meso models were pumping out, but the very high totals were quite good. Now, that said, I'm not in any way saying that the higher QPF models for this weekend's event will be correct (but I sure hope so!). I was only pointing out for that one event since you had mentioned it, too.

Yeah, kind of noticed that about the FV-3. Whether that's anecdotal for this season or only looking at specific events, not sure. I do seem to recall for the December storm, it held on to the idea that the DC/Balt metro areas would be in for moderate to more significant snows for quite some time. Even after nearly all other guidance was clearly going with a more suppressed solution (including, I think, the currently-operational GFS?). Then the FV-3 jumped on the suppressed idea finally. For awhile, it was like a nerve-wracking test of that model for that system...if it continued to hang on to the higher QPF/snow totals here, but we got the suppressed solution in reality, oh boy, that would not have looked good. Fortunately, we didn't have to worry about in the end...that time!

I don't think the GFS for the Jan. 2016 blizzard was exactly "the least" as I recall. In fact, I thought it was the Euro that had a run (or 2?) right on the eve of the storm that kind of burped and cut back precip totals, which turned out not to be right. That DCA recorded "only" 17.8" snow doesn't matter (they always seem to be less at that airport), the entire region got 20-30" (including in the District itself, not DCA), with some places pushing 3 feet or more. Here's an image of the 48-h total QPF from the GFS, from the 12Z run the morning right before the storm began. It has ~2.50"+ liquid equivalent over the area, and was fairly consistent for some time before this cycle too:

Yup...looks like this is mostly a late afternoon through tomorrow night deal, not so much during the day.

We can also, depending on the antecedent cold, end up with quite a good thump of front-end snow, then perhaps ice/sleet, then a dry slot or drizzle or light rain that doesn't totally destroy all the snow that had fallen. And then, as you say, it gets cold afterward typically. A good example of solid front-end followed by drizzle was Feb. 2014...many areas got a foot, even more, overnight. It then warmed into the mid-30s and we had light drizzle most of the remainder of that day, and we actually got some CCB wrap-around snow at the end late day and evening.

I think this was sort of already answered...sorry, just catching up in this thread...but I'll add a couple of things if I may. Basically, ensembles are perturbed members of the base model, and run typically at lower resolution from the operational deterministic version. How the perturbations are done varies from one model ensemble to the next (e.g., GFS, ECMWF, CMCE, etc.), but they all try to do the same thing...that is, show a range of possible solutions by accounting for various errors (the perturbations) that could be in the regular deterministic run. In *theory*, you would like to get all possible ranges of solutions, but of course in *practice* that's not possible (would be essentially infinite solutions!). So various techniques have been developed by different centers...the temperatures may be perturbed, or the pressure/height field, or the physics of the model itself. I can't recall how each model center does theirs offhand. If you have more ensemble members in the suite, that should give you a better idea of the spread around the mean. But just given the fact that you can only have a finite number of ensemble members, and the fact that they're all based on the "same model", you do tend to get what Bob and others call "hive mentality"...i.e., there are many times the ensembles tend to "look a lot like" their parent deterministic model. However, in general, the ensemble mean is a far better estimate for medium to longer range (I think most consider them useful up to 72 hours out?) than a single deterministic model and a single ensemble member. To also partly answer another of your questions, I don't believe there's any tendency for one ensemble member to be "better" than the others over time (though of course, a single one can end up being correct in the end). The ensemble spread is a good thing to look at too, not just the mean. If the spread is very tight around the mean, that lends more confidence than if they're all over the place. Likewise, it can give you a clue as to whether the parent deterministic model is an outlier, such as if it's on the extreme edge or outside the envelope of the ensemble spread. Oh, and when people talk about the "control" run, that's basically a low-resolution unperturbed version of the deterministic (so it's essentially the same as the ops, but lower res). Sorry for the long-winded explanation, but hope it helps! And hope I'm reasonably accurate here, others probably know far more on the nuances of the ensembles.

Now I've got that scene from "Casablanca" going through my head, the German vs. French national anthem part in Rick's Cafe. Anyhow on topic...I have no idea how good the ICON is, but at least it's showing a similar trend to what we saw in this afternoon's model cycle runs.

You sure you didn't just pull a fake punt (a'la FSU vs. Auburn in the 2014 BCS game...awesome play and turning point that led FSU to win that game!):

Fuzzy math...fake arithmetic!