weatherwiz

Kevin gets dry slotted though most of QPF probably done

Can't wait to see how amped the NAM turns out to be based on the look so far.

Isn't Google DeepMind the one which has had extremely encouraging results, however, also described as "the model is able to produce great results, however, the model has no clue how it derived the results". There was some article or paper on this a year or two ago.

doing both

ughh something wonky must be going on with bufkit profiles from PSU. keep forgetting to make a backup bufget list from iowa state.

my thinking too...hope that is overdone. doesn't look like a ton of convection associated with this across the SE

yup...and too much data created to "save people time" while degrading forecasting skill. Hey, why do I have to spend an hour or two assessing all the critical factors which influence snowfall totals when I can just pull up a 10:1 map, call it a forecast, then blame the model for being wrong when it doesn't pan out. Or significant tornado parameter map is showing widespread values of 4-5...let's go with tornado outbreak! its gross.

Forecast models haven't gotten worse, what has gotten worse over the last decade is forecasting skill and a large part of that is due to all of these ridiculous products which have supplemented the actual forecast process. So many convective events in the Plains hyped up because of supercell composite and significant tornado crap, snow maps, etc. It's all run to snow maps, QPF output, etc. which more often than not will not produce an accurate assessment and then when they don't verify, its cry and blame models rather than taking personal responsibility.

I wouldn't be surprised to see the euro come back the other way a bit tomorrow but something like this run may not be too far off from what occurs

the 12z GFS bufkit for ORH is odd. I don't really put too much stock into the precip type but it has this look as heavy ZR. It's more than likely a lift thing but it's still funny to see. but then has this as heavy snow

The storm but I also think I could get into that range. I think a lot would have to go right to really get a widespread area of 6-7-8-9” or so. But depending on how this evolves there could room for those higher totals a bit farther north into parts of NH

I’m going to say 4-6” is probably ceiling with this. Maybe some 7-8” reports if things really fall right and depending on the angle of the ruler into the snow. But I will wear a Santa hat and do cartwheels if I can get 4-5” Dec 2

Heavier rates along the Mass Pike might be enough to pump enough moisture higher up into the DGZ and get better growth going

yeah 700mb is quite warm. snow growth would be pathetic except perhaps towards SNH or so and just north

Already huge differences by hr 60-66 at H5 compared to the 12z run yesterday. Not much difference versus 0z or 6z but its crazy to see such drastic differences in 24 hour

My guess is this ends up being something flatter which would benefit most in the Ptype department but I don’t think you’d be looking at any beefy totals. Probably borderline warning criteria for some towns. I mean which is still great for early Dec…but you’ll see a lot of disappointed people in the totals because they didn’t live up to clown map expectations

I doubt its coming in this amped. Not even going to have any serious thoughts on this until Sunday morning or early afternoon.

Looks like 18z gfs going to be quite amped...shocker.

What differences between the GFS/Euro for next weekend GFS is like a carbon copy of the forecast for Tuesday, meanwhile the Euro looks more like a re-developing clipper with not much srn stream going on. But the two models aren't even on the same planet with how 500 looks across the country

getting snow flurries here!

Yup...more than likely the future...and then society is doomed once this stuff crashes because nobody is ever going to be taught basic understandings, principles, and concepts anymore (talking beyond weather here) and taught "how to build a model or algorithm" and things crash...nobody is going to know what to do because they don't have a computer telling them how.

There are definitely alot of factors. But if you're still producing scenarios where solutions are inconsistent on a run-to-run basis and you have all sorts of jumping around, then what is the value in there? Unless like Ray said, "its used as another tool". But then all its doing is more than likely increasing uncertainty, not decreasing it.

That AI model stuff is such a waste of resources. But some storm will happen that an AI run "nailed at D10" (even though it jumped around a million times inbetween) and parades are being conducted and research papers flying saying "AI is the next thing".

I would wager we see the GFS tone down a bit as we get closer. I also wouldn't be surprised if we saw some other guidance even trend towards the GFS, only for guidance to back down Sunday evening.

Been weird though because the GFS has underdone some significant differences with H5 evolution. It's also been trending stronger with that trailing shortwave which maybe is acting to amp it up a bit? We're also kind of entering the time period where the GFS seems to have a bias towards this regard. Wouldn't be surprised to see it backed off some Saturday or even Sunday.