eduggs

NYC: 7" EWR: 13" MMU: 23" ISP: 29" SWF: 32"

We'll see how the next heavier batch of precipitation plays out. But so far the Canadian seems to have verified better than the NAM and GFS with forecasted snowfall distribution. In fairness, the 3rd party algorithms that calculate snow based on the raw output may be to blame more than the model itself.

In some aspects, yes. But in elevation-dependent spring snowstorms, the spots over 1000ft in Putnam and Eastern Dutchess counties often cash in. The CMC and EC model depictions leading into this event didn't show that characteristic pattern. Instead they showed an elevation effect superimposed on a low level easterly flow shadowing effect.

As long as mid-levels don't torch, snow vs rain in NNJ, SENY, and NW CT will likely come down to localized mesoscale effects. Precipitation intensity will play a role. I'd also be interested to know how different the topographical maps input into the models are for the GFS, NAM, and CMC. I wonder if small differences in the discretized topography could be affecting the slight downsloping and upsloping leading to the critical 1 or 2 degree differences. This is a very marginal snow situation so there could be significant differences with elevation and location.

Don't try to catch a falling knife. Still lots of run to run changes and this is 7+ days out. This look for the late week storm next week is MUCH better than the mega cutter from a few days ago, but we should wait to see how it shakes out over the next few days before mentally investing too heavily. We have 2 trackable winter events over the next 10 days! Not bad after a long warm season.

That was a helpful nudge south, colder, and a hair stronger on the 0z GFS. The boundary layer temps are still on the higher end of guidance, but it gets snow to the hills a little bit closer to the City.

The (18z) GFS soundings are torched, NAM is cold AF, and the Canadian is in between. One scenario is a non-event outside the elevated interior, and the other is an advisory snowfall including slush covered roads to the City's near suburbs. The surface low position initially near Lake Erie historically suggests a warm outcome. And synoptically I keep expecting the meat of this event to shift into NY and western NE. But there is a pretty potent pocket of vorticity with steep lapse rates, an unusual angle of approach, and relatively cold surface air is nearby. There's is also a mesoscale banded feature at the tail end that may or may not mean anything. This is an interesting event to track It would be great for this to move in early and provide some rare weekend, daytime snowflakes.

All mid-range models are showing snow now for the N&W suburbs, and probably at least a little into the City. Boundary layer temps look warm, and you probably have to hedge warm when there's a surface low near the Lakes. So this favors elevated areas, but it could also favor areas where banding sets up. There's a mesoscale inverted trof type featured hinted at the tail end. This event has an unusual upper level progression - really kind of like a redeveloping clipper low.

Dozens of national meteorological organizations around the world as well as a few collaborations (e.g., ECMWF) have created weather forecast models. Germany's ICON is probably the 4th or 5th best mid-range model in the world. Hardly garbage. It's sister model, a hi-res version of ICON is the best short range model across much of Europe, particularly the Alps. It is a rapidly improving model suite.

The CMC looks impressive at 84hrs. H5 looking more potent than other models and previous runs.

The 0z ICON looked pretty good for snow on Sun-Mon. Ended up roughly similar to the 18z GFS, but with slightly delayed timing. Say what you want about the ICON, if nothing else it's useful for confirming the other models and identifying trends.

What it shows is an overreliance on indices and the danger of obsessing over "patterns" when local weather is determined by a specific and unique combination of features at any given moment in time.

This next 10 day timeframe feels to me like a period when the higher elevations from CPA through the Poconos, Catskills, and Taconics/Berkshires of NY/MA could get unexpectedly smoked. Like a late-evolving, high-elevation warning event. There are lots of little waves in the flow, marginal cold air, and a tendency towards a lack of progressiveness. What will likely skunk the coastal plain is a complete lack of antecedent boundary layer cold air. There will likely be a northeasterly cold air drain at times, but early indications are that the lower 1000ft or so will remain torched. I expect lots of model variability and maybe a few surprises.

I can't remember such a convoluted height field in decades of model watching. Probably there have been other such periods, but none that are memorable to me. On the NA chart, there are too many shortwaves and pockets of vorticity to count, haphazardly evolving in sometimes opposing directions. I can't imagine the mid-range models will demonstrate high forecast skill during this period.

I think this is a great point. It also makes tracking this week more fun as long as we don't look completely out-of-the-game heading into Tue. or Wed.

If we pretend this run is real we have 3-4 wintry "threats" to track in the next 10 days. If the GFS is real we have a week long deluge. The real reality will likely incorporate elements of both plus some unexpected changes.

The block is blocking on the EC. We have cold/coldish air!

Regardless of the surface outcome, I like this 12z EC run so far. It's the first OP run in a while that I've been excited about. Maybe it's the Irish Coffee I just drank putting me in an optimistic mood... not really sure.

12z EC shifted a little south with the weak upper level wave sliding east at the end of the week, reversing a northward trend of the past few cycles. This will likely leave precip. just south of us, but I prefer this scenario to the GFS washout. It's also a colder solution, which IMO increases the changes of wintry precip. on Fri or Friday night. Right now I'd prefer to get the cold first and worry about the precipitation later.

Should we expect otherwise? It would be a fluky coincidence to have two relatively unchanged runs at the surface 7+ days out. Small changes aloft can lead to large changes locally at the surface.

1 week of near-continuous precipitation (mostly rain) on the GFS is I think I would be mildly excited about snow potential if I lived in the higher elevations of northern or western New England. But if I kind of squint at the charts I can even imagine a long-duration fantasy snow scenario for the Cities. To get any local snow out of the medium range I think we would need the cold air source to push a little further south than currently modeled.

I still think this is true - but it would be late week not mid-week. The CMC, ICON, and EC bring frozen precip. near or to the area. It's exciting to get a wintry threat as close as 5 days out! That said, the GFS remains much warmer and wetter and the other models have been shifting northward for a few cycles. My gut says this ends up rain with a few high elevation cat paws or sleet pellets. I expect high inter-cycle and inter-model variability in this regime.

Are we in a stable "pattern" now that we expect to cleanly change out of? It looks to me more like continuously evolving climatic and meteorological characteristics that change on a daily and even intra-daily basis. Despite what some people have convinced themselves, there is no guarantee of a new or favorable "pattern." Snow chances will likely increase as we move deeper into winter. But the atmosphere doesn't know what NAO, PNA, AO, and La Nina mean. These are not tangible things that can be switched on and off. There is no discrete era of snowfall that can be promised or delayed. And to the degree that some meteorological characteristics are loosely correlated with weather outcomes, our ability to forecast them beyond 10 days is extremely limited, to the point of frequently misleading.

I don't care how good the long range ensembles look. Multiple rain events in December with snow threats pushed out beyond 10-day fantasyland is never a good thing for snow enthusiasts. When the goods are relegated to the distance extended timeframes, there's a reasonable chance they will never come. Eventually it will snow. But right now I see as many signs to be concerned as to be excited. Hopefully that changes for the better tomorrow.

Humans love to categorize everything. We are compelled to separate things into boxes to understand a complicated world. But in truth, every 500mb height field is unique. There are no true patterns - only a collection of general characteristics that we lump together. I believe we do a disservice to ourselves by obsessing over these "patterns." And if long range 500mb anomaly charts ceased to exist, we would be smarter, happier people