eduggs

We're speaking past each other and probably not reading full posts. Problem is simple. The seasonable modeling bias across the full spectrum of guidance beyond 7 days is too cold in the east and too warm in the west. People aren't making the adjustment. So again, the problem is the correlation between a modeled "good look" and future outcome, superimposed on top of a questionable correlation between a loosely defined "look" and local non-mountain snow. The solution is to 1) focus closer in and 2) pay more attention to specific ensemble modeled features as opposed to mathematically averaged parameters.

Call it modest or moderate then - whatever you like; semantics. And lets define it as the correlation between certain geophysical parameters and something like temperature or east coast cyclogenesis. Because snow is truly a local realization. The other problem is temporal in nature: the correlation between the "look" of the future "pattern" today and the eventual weather outcome. These things are not so well correlated out beyond 10 days because modeling has minimal skill at this range. So really what you're talking about is the correlation between a characterization of the "pattern" today and the weather today... which is admittedly a somewhat stronger correlation. Though still not great with today's "pattern" vs. weather outcome a good example. However, that's still a very different thing than the future correlations that everybody refers to when they talk about a "good look."

I just think "patterns" and "good looks" are vastly oversimplified. They are phrases coined from Mt. Stupid for those familiar with Dunning-Kruger. Everybody loves to think these concepts are strongly correlated with local snow. I concede there is a correlation, I just think it's relatively weak. I believe snow is a local phenomenon that is often given or denied by nuance in the evolution of features. Even if "patterns" are definable to a degree that allows meaningful local correlations with weather outcomes, we cannot see these patterns coming any better than we can see, for example, specific longitudinal jet structures and height fields. So my feeling is that most are sniffing around in the wrong places. I would think experiences like this winter would cause some to see the light... but many stubbornly persist.

For the love of god please retire this phrase. Times mentioned this winter: 8274 Times the look materialized: 0.5

We're celebrating snow flurries on an above freezing, mid-January day I sure wish I had flurries here in NJ

So far this winter feels relatively active for such a snowless winter. In my memory a lot of our dud winters have long dry periods. The reason being that typically in Jan we can sneak an inch or two on the front end or back end.

CMC offers some hope for the end of week event, but I think it's going to be really hard to keep that one south of us. The ULL and primary are headed to the Lakes come hell or high water. Starting to lose hope that we get lucky for that one.

I suppose it's possible I'm in favor of anything that brings us snow. Raindrops and snowflakes need nuclei - usually dust particles - to crystallize around and transfer vapor into the solid phase. I believe salt particularly near coastal regions, e.g., LI, coastal SNE, are very efficient at facilitating crystallization. My guess is this would promote flurries vs. drizzle but it's possible it could allow precipitation to form below the normal crystallization zone.

I didn't see any models with measurable today - could have missed it though. Although I did notice the HRRR indicating flurries tonight. Glancing at soundings, the snow looks to be forming at a low level. Maybe models were assuming no crystallization at those temperatures. It could be the salt nuclei facilitating snow growth and crystallization. Hopefully flurries/snow showers continue all day into tomorrow.

I haven't noticed it moving closer on guidance over the past day. The NAM had 1 run that brought a some snow to most of LI, but since then everything focused on EMA. This maybe could have been a coastal grazer for us with a different upper level evolution. That said, the HRRR suggests that maybe flurries and snow showers are possible for eastern areas for the next day or so.

As soon as people stop obsessing about MJO, PNA, ENSO, EPO, WAR etc, they will become better forecasters. We simplify weather prediction into these loosely defined entities. And then we talk as if they are real things that have a causal impact on outcomes. It's almost like we anthropomorphize them. The problem remains that models are no better at predicting future states of these indices than they are predicting 500mb height fields. So even if MJO, PNA, NAO are loosely correlated with regional weather outcomes, we cannot accurately predict these beyond 7-10 days. And these indices have no causal power. They don't force anything to happen. They are simply averaged numerical approximations of geophysical parameters that are all connected and interwoven into global ocean and atmospheric circulation patterns. They reflect the current state of the global ocean-atmosphere couplet as opposed to predicting its future state.

Even the CMC and GFS runs that were bringing snow to NYC N&W had a primary SLP into the Lakes. That's not a "pattern" change. I'm rooting for snow as much as anyone, and it could happen next week. But people claiming to see positive changes coming are seeing a mirage. They simple don't learn from their experiences.

If this keeps being said, eventually it's bound to be true. But right now it's just wishing and hoping because nobody has a crystal ball. Also, right now we have a high amplitude, east coast trof with significant 850mb and 500mb temperature and height anomalies moving through the southeast and mid-Atlantic. How does that fit with our current "pattern?"

If you loop the 500mb heights on the any of the mid-range models, places like Atlanta are constantly at 570dm or above. Typically for us to get a good wintry event, regions in the southeast like Atlanta, Knoxville, and Charlotte need to get down into the 540-546 range. Otherwise the primary storm track will be north or west of us. The only time that region is modeled to get into that 500mb height range over the next 7 days is actually right now, which is why this weekend was the closest we will likely get to a snowstorm in a while, and why parts of SEMA and coastal ME might get some snow this weekend.

For the late week event next week, the EC is doing what we've come to expect with a SLP into the Lakes and 570dm heights surging towards us. But as far as I can tell, it is the only hold out amongst the midrange models in terms of showing some wintry precipitation nearby.

I'd take those OP GFS and CMC runs through the end of the mid-range. Couple of wintry threats there. Ensembles say these likely end up weak, dampened waves or cutters. Let's see if the EC shows any movement in the same direction.

We could have gotten a coastal snowstorm this weekend if the multiple shortwaves interacted just a little differently. The trof axis has good amplitude and orientation and there is sufficient cold air available. Snowstorms are local phenomena that depend on nuances in features and their evolution. Yes the background airmass and longwave jet structure is important, but it's usually the minor details that deliver snow.

I think you attribute too much causal certainty to ENSO, MJO, and other climate indices. Most of the correlations are weak, and the indices themselves, broadly defined. There is a lot of randomness in weather outcomes.

It amazes me that people keep looking at ensemble charts out beyond 14 days. They have effectively no skill. We have seen it over and over, yet people don't learn the lesson. If there's light way out there at the end of tunnel, you won't see it coming until it's much closer. Often it will just surprise you in the mid-range.

I don't remember any winters like this. Back then there were low snow years and warm periods, but the temperature still went below freezing regularly. This winter is not comparable in that regard.

I'm feeling uneasy about my predictions. We'll probably need a big event to get close.

You can squirm and twist all you want. But facts are facts.

Not correct that there's no cold available. <10C 850mb temps to the Carolinas and sub 530dm heights at 500mb throughout the mid-Atl on Saturday. We even have a very high amplitude longwave trof with a nearly perfect alignment and orientation. Our problem is in the nuances of the height structure... the evolution of the shortwave interactions. Snow vs. no-snow outside the mountains is almost always due to the nuances of mid- and upper-level features.

If you loop like 168hr - 216hr on the 12z GFS North Atlantic view it looks like two hurricanes moving westward in September. I don't remember seeing such well defined mid-level lows in that region during the winter, but I don't usually look for it either.

Not trending colder per say. Clearly the initial phase and SLP track into Ontario has trended sharper with a quicker scouring of antecedent cold. But one thing that's changed from 3+ days ago is that there's a fresh cold air source to the west that isn't being immediately pushed out by the next wave. Days ago a modeled negatively tilted ULL was pinching off and being separated from the longwave flow, resulting in a complete emptying of subfreezing surface temps east of the Mississippi. With CAA behind the "front," at least there's a longshot chance of developing a new SLP on what's left of the baroclinic zone that could potentially produce snow somewhere along the east coast.