OceanStWx

I actually don't hate your location.

Even that GFS sounding for BOS is sneaky. Technically accumulates no snow because pytpe doesn't spit any out, but the boundary layer warmth is only about 1400 ft thick. That's pretty close to not being able to fully melt. And if you are snowing hard enough it gets even tougher to keep up with the melting.

It could definitely be a rate dependent pytpe for those on the edge.

Some of the NAM soundings are a little MAUL-y for a time overnight. Near and shortly after 06z could be interesting. DGZ is maybe a shade high in the column, but overall not a bad overlap with omega.

Eye popping, but the reality is that locally we need to look lower in the atmosphere to correlate LLJ and surface gusts. 925 mb is a better match. See below. I don't think this is an overreach for a high wind watch (AS MODELED).

I'm intrigued with 925 mb tickling 80 knots from PVD up through BOS. Even with some weak lapse rates and deep surface to 925 mb layers that should be good enough for 45+ knot gusts. SSW LLJs do tend to have more red flags than a SE or NE jet, but this is not a marginal wind max.

The good news is that temps warmed many places once the snow stopped, and quite a bit melted off the trees. Could've been a lot worse had we actually cold advected overnight.

I drove past a downed line in North Yarmouth that had started a fire in the shrubs along the road. Alternating blue and orange flames was a pretty cool sight at 11 pm.

Man that's a gorgeous band of snow moving north through CAR.

My PFM (point forecast matrix) had AUG going from 45 to 34, LEW from 46 to 34. Not bad. I had PWM go from 45 to 39 by 21z, but just didn't take it down far enough.

I was like Bruce Willis on the runway waving the caution flags furiously that models were driving temps down to 35 or 36 and reality was likely going to end up lower than that.

This new formatter is so literal. It just takes the highest and lowest amounts rather than trying to pick a "most likely" snowfall bin. I usually hand edit to something more reasonable, like someone in Androscoggin was more likely to 6-10" than 3-11".

Welp, another 3.7" today brings me to 62.2" for the season.

700'? You're effed. Seriously though, I went "pessimistic" and used Bufkit max temp in profile snow ratios (since this will be a relatively warm storm). With that and compaction it still spits out between 2.5 and 10" at GYX across the suite of models. The wild part is that most of the damage is done in 3 to 4 hours across the area.

I'm definitely taking them up, the question is how far. It's going to be a really close call for GYX.

So I can't find anything besides the RAP that gives me less than 8" in AUG. And that's because the RAP cuts off after 3 hours of snow before 00z.

Negative tilt, left exit, a little warm season juice from down south...

I kinda feel like we're going to slapped up here. Someone just north of this low is going flash over to paste, and it could be GYX/LEW.

Western Maine mountains should get crushed based on the low development and LLJ.

Or later this week.

http://hudson.dl.stevens-tech.edu/sfas/ It actually overpredicted for BOS and PWM which a model almost never does.

Do you remember any specifically? We have National Blend waves now (which does include some Euro). They verify pretty well so far. Surge we're fairly limited in GFE (ESTOFS or ETSS) but we can see others on the web and adjust forecasts accordingly. The Stevens Institute SNAP-EX model does pretty well.

I can (have someone else ) create an AWIPS smart tool using 925 wind, LR, and depth to spit out a wind gust forecast grid. It's a lot harder to calculate pressure gradient across an area and do the same thing. So that's what made picking the variables tough. But those three variables got me to about 80% of the variance in gusts.

I IDed events using 850 > 50 knots, and then used the 925 wind to do the multiple regression. Because honestly we're never getting the 850 wind to the surface. Oddly the 925 wind correlated better at PWM than the 500 mb did. I figured the closer to the ground you get the better, but not necessarily the case. I also really need a level that models produce data for, because otherwise the regress is no good. I need something I can use to produce a forecast.

I think I grabbed 73 events with a CHH 850 wind of 50 knots or greater. I looked at 500 m wind, 500 m lapse rates, and depth. The multiple regression equation eked out 0.577 correlation. So those variables barely explained half of the variance in observed wind gusts (+/- 2 hours from balloon launch) at CHH. So there's something else factoring in for Outer Cape wind gusts. I think my method works quite well inland though, especially for near coastal areas (gets a little squirrely in terrain when 925 depths get shallow).