eduggs

The 18z ICON was also very cold and south. The surface freezing line roughly bisected the City throughout most of the event and the 850mb freezing line was just to the north. Dutchess county north is probably mostly snow on that run. Waiting for the 0z...

Indices don't modulate weather. Weather does what it does. Indices are a statistical way of observing and measuring it. The idea that climate indicates influence weather is a common causal fallacy.

Indices don't cause weather or modeling changes. Forecasted indices themselves are models. We might be seeing a trend. But then synoptic guidance will shift in tandem with climate indices.

NAM still rides up the Apps with the SLP but it's a lot slower. A couple of surges of precip. out ahead of the system are frozen from the MA up through our area. But the 500mb depiction is still unfavorable for snow. Ice maybe.

He is one of the best posters for his enthusiasm and interest in the hobby. I commend him for that. But the counterpoint to his poorly supported ideas was lacking. My criticisms are more for the others than for him. Many of us are "trained", red tag or not. And I'm sure he can take the criticism. Forecasting zilch when everything else is 2-4" for days is ballsy. When you stick your neck out that's what should happen. It teaches you to respect the models.

I agree with almost all this post. And it was a little harsh, I guess to make up for the gentler critiques that have been lacking. The models saw the dry air too. They just model its effects a lot better than we do. I can remember just as many virga events from the 1990s at Tip can. But that doesn't mean I'm going to ignore 4 days of consistent modeling.

I can't speak for your specific location, but surface temps. were definitely an issue (as expected) for many locations in the area based on official and unofficial obs. Lots of 0.2 - 0.3" QPF totals with mostly sub 32F at 925mb.

I completely agree. And obviously they never get it completely correct. Atmospheric physics and its cascading causal effects are incredibly complicated. Unfortunately humans are not very good at diagnosing when and how models make errors (at least in real time). It used to be easier when models had consistent biases and poorer skill. But it gets harder and harder as models improve. And worse, there is a lot of confirmation bias and problems associated with small sample sizes. If we think we identify a model error once and then the model busts as we expected... we often incorrectly conclude that we correctly identified the error and can apply it to future scenarios. But more times than not it was just random chance and coincidence. The models are extremely complicated. In my experience, the best contributions that human forecasters make over weather models happens at the very local level... sub-regional or mesoscale. That's where consistent biases associated with local temperature or topographical effects can be identified. At the synoptic scale, I really don't believe humans can outforecast models in most cases.

Seriously though. This board is the last bastion of knowledgable amateur meteorologists. We rely on you guys to call out the hype and the gut-instinct and the non-scientific forecasts. Twitter is riddled with it. Everybody thinks their an expert. People get really confused.

Actually the precise opposite is true here. A model blend worked out really nicely I think. Trying to out-think the models burned a few people. And it was relatively consistent for days. As a result, there were some really good amateur forecasts on this forum.

No. I'm not saying that everyone agreed with a forecast of C-1". I mean that very few people expressed the view that that was a non-scientific and unreliable forecast... basically a guess. Kind of like a forecast of persistence or default.

Models get it "wrong" almost always because of a combination of the relative coarseness of data ingestion and parametrization. But model physics is incredibly sophisticated these days. We can create a very accurate representation of geophysical parameter evolution. Forecast errors are almost always attributable to input errors (outside of the meso/micro scale).

There's too much gambling and gut "calls" is "meteorology." It does a disservice to professional forecasters who have a lot riding on their decisions. Stupid "calls" should be called out for what they are. And as far as I can tell (been following this thread for days), you guys really dropped the ball in this regard. My call was model consensus. It worked out very well this time and almost always beats human forecasters.

I just hope he learns from it. The model algorithms that estimate QPF, although the least reliable parameter, still "understand" and incorporate physics vastly more accurately than amateur or even professional meteorologists.

Was that C - 1" forecast I saw for all of CT ever updated? That rivaled the worst forecasts I've ever seen. Ballsy speculation but stupid forecast. Thank god that guy doesn't forecast for anything that matters. Classic case of trying to beat science/math with emotion. Guidance showed several tenths frozen behind the front for days. Overall this was pretty well modeled from the mid-range. Congrats to those who scored a bit this morning. Here in SENY it was generally 1 - 5" depending on elevation. A nice little event in any month.

Overall this event was pretty well modeled even from several days out. It was highly trackable and ended up like a standard, prototypical snow event for this area. I'm glad a few spots hit 4", which had been suggested by the guidance as a possibility since last weekend. It looks like it was a general c - 2 inches for the coast and river valleys with 2-5" above 400 feet or so.

This event was elevationally dependent as expected because of the slope of the thermal front. As is typical, surface cooling lagged mid-level cooling substantially. So elevated areas changed over and began accumulating much sooner than lower areas. And lower/coastal areas had only a brief period of accumulating snow when temps. were marginal. The problem with the snow maps isn't an issue of "sticking" with "warm ground." The issue is that the algorithms tally up all modeled "snow QPF" and display that total. Snow accumulation is different than total modeled precipitation in frozen form. Moderate snow sticks at 33 or 34F with dense overcast. Truly heavy snow sticks at almost any realistic temperature. But even steady light snow typically requires <33F to accomplish more than coating the cars. And the vast majority of snow obs. are light in any given winter.

There's a cold pocket in the mid-levels and a very steep lapse rate. Elevated areas are getting going much earlier than low elevations, as advertised.

Elevated areas of the Hudson Highlands appear to be snowing. Temps between 35-40F depending on elevation.

The radar looks good down in the mid South but a lot of that is virga down in AR, TN, and KY.

NAM is definitely a shift. It is wetter and focuses more QPF with the initial fronto push NW of the cities. Eventually that all shifts east but in sporadic bands. It is also several hours faster with the initial surge. Changes in QPF distribution and timing this run suggest the event is still in flux, as usual. At first glance this looks more favorable for inland locations from EPA - NNJ - SENY and maybe slightly less favorable for coastal NJ and LI. But I still think there is room for a scenario that gives a couple of inches to almost the entire region. The thermal boundary is in a relatively favorable position. The more precip. on the cold side the better.

All guidance has been shifting towards the ICON for 2 days now. It has consistently shown limited post-frontal precip. west of I-95 from MD to EMA. Underrated model. We'll see how it shakes out. There could still be surprises.

It looks like the GFS basically held course. Roughly 0.2 - 0.4 frozen.

The 0z NAM looks pretty unimpressive. Dry and SE. The ribbon of moderate post-frontal precip. from MD through EPA, SENY, and SNE consistently shown for the past few days is not very impressive as depicted. This might be good for the immediate coastal areas and Long Island, but there's not much NW of there. I hope it comes back some tomorrow.

The 06z and 12z had widespread 10+ mm snow QPF with pockets of ~18mm for a long swath from MD up through New England. The 18z is widespread 6+ mm with an area of 10mm mostly in NE. It's not a huge change but a step down from being on the wetter side of the envelope. It has also shifted SE with the axis since yesterday. The heaviest band is now close to the I-95 corridor and close to consensus.