OceanStWx

Very clear from looking at fuzzy clustering that the EPS members overwhelmingly the ones supporting a low tucked closer to the coast.

So I was catching up with office verification stats yesterday for this winter ( @tamarack and @dryslot may be interested) and found some pretty interesting results for GYX. While our probability of detecting warning events is good (88%, goal of 90%), our false alarm for warning events (events that are warned but don't happen) is 39% compared to a goal of 20%. That's bad. Really bad actually. In addition, the "lead time" for those false alarms (time from warning issuance to the rough expected time of warning snowfall) was over 27 hours! Meaning we're issuing warnings about a full shift too soon, and suffering the consequences because of it. We are missing key hi-res guidance opportunities in the 12 to 24 hour period that could help our forecasts of QPF and ptype. Conversely, our numbers for a watch are great, our false alarm rate is actually lower (35%)! The numbers really tell me that models are great at sniffing out events at long ranges, but not great at sniffing out details of those events at those same ranges.

It's either that or somebody was taking pot shots at the building.

Nothing extreme here, but the window to the office shattered because I think the combination of metal frame contracting and flexing in the wind. Security has been compromised.

I mean we really haven't had a closed mid level low pass SE of us this season, so it's hard to not call it nickel and dime. Not quite the same story for the mountains and CAR. The closest we came was 1/20, but that was a terrible forecast regionally and ended up well below forecast amounts.

I've always termed snowfall in my AFDs as light (1-3), moderate (3-6), and heavy/significant (6+). I agree once we hit 12+ I start pulling out words like major.

I haven't used that at all. My raob program is not loving the format it gives you though, so I would have to dig into it a bit more.

I bet they had some good downslope winds on the backside of the Whites.

That's is as ridiculous as their cold and snowy Northeast forecasts, but an exact number of tornadoes? Come on.

W.T.F? https://www.accuweather.com/en/weather-news/accuweathers-us-tornado-forecast-shows-4-states-face-the-highest-risk-this-year/70007449

I grabbed the old April 1934 surface map to reverse engineer the mid level jet responsible for the 231 mph gust on MWN. The pressure gradient was 7 mb from PWM to Northfield, VT. The max gust was 103 m/s, so the "slower" wind speed would mean a 135 knot jet around the inversion. That's a pretty stout LLJ, but not impossible given the pressure gradient in place. Unfortunately we don't know what the vertical profile looked like that day.

I mean the more I think about it the more I'm blown away that clown maps essentially ended the FV3 implementation. That probably goes to show you how many HIGH-END users look at the snow maps.

Now the units are going to be a little unfamiliar, but they work out. I used the pressure differences between HIE and IZG, and NAMnest wind speeds (65 knots) to estimate how winds would accelerate from the NW side to the SE side of MWN. 100720 Pa - 100480 Pa = 1.225/2(33.44^2 - V^2) Solving for V = 53 m/s = 103 knots So you can see how pretty typically mid level winds can translate into some much higher gusts. Considering that local effect were likely more extreme than HIE-IZG and 148 knot gust is not surprising.

I did a little nerd work on the MWN gust yesterday:

It may be good news that it appears as we near the cool season, because that's when they typically turn some of those parameterizations on. That could mean an easy fix.

They will not. The GFS and GEFS are frozen on development. If something were to actually break down in either they would fix it to keep operational, but no improvements are being done. It's a bit ironic that a problem with the clown maps is one of the reasons why it is being delayed. But after that and the low level cold bias are fixed, it will take another 30 day stability test until it's operational. So we're basically April at the earliest.

So that FV3? Um... https://www.weather.gov/media/notification/scn19-12gfsv15.pdf

We were just saying it was almost like the low level stability finally mixed out late in the afternoon and allowed the gusts to really power to the surface.

There's a difference between 45 mph on NW flow and 60 mph on NW flow. It may not seem like much, but you can see a different level of damage as winds ramp up those 15 mph.

I’m sure they were accumulating rime, it’s a lighter ice than clear ice, but still adding weight. Even if you take the lower bound of the EF scale that’s 115 mph. And given Mansfield topped 130, I’m betting Sugarloaf was in that range too.

I feel pretty comfortable with an estimate around 135 mph.

I’m just as fascinated when this stuff happens too.

You can see the inversion starts just above summit level, 750ish, and that’s what squeezes the flow.

Inversion.

Most likely. We don’t have hard copies of all their records, but it’s highest since 2005. And highest in February. So decades is likely a safe call.