OceanStWx

The ground is warm? We’re actually making some real frost depth up here right now with the lack of snow cover and cold temps. Ground temps aren’t going to limit anything.

I just bake them into the forecast. It is pretty routine to see models overamplify in the short term (days 2 or 3) and then in the last 24 hours just slide a little or flatten out. So then I am left sitting at the WFO with the snow totals...

Scooter might let his kids out of the basement, it's starting to converge on a nice forcing signal around his backyard.

In theory you don't have to have a meteorology degree, but there is a list of required coursework that more or less means you would have received a BS in meteorology or atmospheric science.

Speaking not of ALY, but the NWS more broadly, I think there is a tendency to get sucked in by the QPF. The pressures of short staffing combined with added duties beyond the forecast are leading to more and more step by step building of the forecast. It's great for duplication and amending the forecast, but it can lead to less critical thinking at times. For instance, does the QPF max location make physical sense? Is the storm moving slow enough for widespread double digit snowfall? etc. That's why I like coming here and talking about the weather. It forces me to think about why I'm doing what I'm doing in the grids. And when in doubt I can always start high and adjust higher based off the KFS.

I already have the hair.

If you take a look at the latest NAM/GFS in Bufkit, even out Scooter's way it's like an hour of maybe catpaws and then it starts ripping.

That's more or less where I drew my max amounts too.

I capped the max around 15:1. One thing I definitely am not comfortable with is having a third of an inch of QPF and using 20:1 ratios to get me a warning. That's John McClane on the runway stuff right there.

I am a little worried I'm maybe too heavy near the coast, and that the best banding ends up inland a bit. I'm also not quite as bullish as my neighbors on snow ratios (closer to BOX than anyone else).

Okay, I updated the snowfall. Probably going to piss off @wxeyeNH and make @dryslot happy.

Poor guy just wanted to make a better snow ratio. I actually wouldn't hate the equation if it just capped ratios on the cold side. You could have a scenario where the lower end of the DGZ (-12C) is the warmest temp in the column, and the forecast ratio would be like 22:1. We know those single digit snowfalls can often be Arctic sand.

For reference, all offices produce these maps the same way. WPC has a super ensemble and they ship us all the members snowfall distributed into bins. You end up with some version of a curve, and we picked the 90/10th percentiles off that curve essentially. So this is probably a result of some wild Kuchera members.

You have to look at where those 1" QPF maxes are located too. Do they make physical sense. The Reggie has them on the edge and equatorward of the DGZ saturation. The highest QPF is going to be underneath where there is saturation, not adjacent to it. It's why the highest snow totals are typically west of the QPF max as modeled.

Just wait until I get a hold of them in the next hour. Those ranges are based off of the WPC super ensemble after all, but there are definitely issues with that. If Kuchera is part of that distribution it is going to shift the 90% to a pretty high number.

I'm having a hard time finding a Cobb total above 10 inches. The closest really is the GFS around BED right at 10 inches.

I love the GFS ripping 4 inches in one hour for PYM as the band rips through.

Running it back to 00z last night, the Euro definitely initialized too high with heights in the shortwave diving through the Rockies.

Ripping after 5 or so. 00z looks like the sweet spot for PVD.

If you are being charitable about what accumulation is, the GFS is 9 hours at ORH. 90% of it is 5-6 hours.

We have to keep in mind the regime we're in right now too. The Pacific jet is in retraction right now, which is more prone to unpredictability in the models.

It’s a different physics package so I wouldn’t expect one to predict the other.

Like actually QPF or just a low location that you like? Because the 12z run was pretty paltry.

We know that the ensembles in general are all underdispersive (the real outcome sometimes isn’t within the ensemble envelope). But it is quite something to see no EPS members with a significant QPF right now. Normally you would think a few juicy members would be in the 50.