OceanStWx

Maybe a tickle slower, but very similar look to 00z going from neutral to negative tilt across Appalachia.

Last I saw was Feb 2021. I haven't seen any evaluation suggesting that testing is not going well and I don't foresee any slippage due to that (though red tape could always cause a delay). Evaluation also suggests that it performs as well if not better than the operational GFS for most parameters. The most consistent complaints are regarding TCs and low CAPE biases, so not really relevant to cool season cyclogenesis. I view it as the operational GFS is still a useful data source, but I maybe weight v16 a little more given it's had a pretty successful year+ of testing.

I'm actually very interested in the parallel GFS this year vs last year. This will actually include some physics updates vs just changing the computing core of the model. So v16 may bring actual changes to model accuracy. I'm also not going to evaluate one model vs another based on clown maps. They were all pretty poor in that regard for the 5th/6th. If I remember correctly the Euro was first to move towards a more amplified and juicy system then which ultimately was correct. Unfortunately the antecedent air mass was too crap to support widespread accumulating snow.

The deterministic Euro is essentially just another ensemble member. The ECMWF (entity not the model) has chosen to vary both initial conditions to be close to the best guess starting state of the atmosphere, and the actual model itself to be close to but not perfectly match the control run. Their goal is to capture both uncertainty in the initial state of the atmosphere but also uncertainty in the model.

Every ensemble system is different, but in general it is just tweaks to initial conditions. You're not going to find scenarios where member 36 is going to score a coup because it handles shortwaves running into strong highs better or something.

Whole shortwave is north of 12z, and a little more negatively tilted compared to 12z.

Nah, more like tomorrow evening.

And really from subtle changes 18z to 00z. If anything the GEFS mean confluence looks stronger on the 18z run, but the 00z run just has a slightly farther south shortwave. Similar strength run to run.

18z GFS was a northern outlier within its ensemble system, so a southward correction doesn't surprise any. And honestly for all we know right now this could be a southern outlier on the 00z GEFS.

Can I borrow this for my AFD tonight, full credit of course.

I'll be the turd in the punchbowl. Cold storms are where the Kuchera ratios fail most. It's designed for near freezing cases, to account for mixing and lower ratios, but there is actually no limit to how high the ratio can get as temps below 500 mb get colder. And we know that snow ratio doesn't improve linearly with cooling temps. Take the mid to high teens, let's say average of 18. Let's assume that's the max temp below 500 mb too. That means Kuchera is like 18:1. That's not a realistic ratio for the duration of an event.

Big spread on the north side of ensemble systems.

GEFS members are pretty solid for having some QPF even up this way. I feel like both ensemble system are saying similar things, despite the deterministic runs being quite different. The ensembles are probably a nice compromise between the progressive GFS and usually too amped Euro.

Yeah, that kind of ice would be too brief. You need those weenie waves to give you more subtle overruning over several days.

Bahahaha Get my standby generator installed 30 days later.

Scooter streak in the ski pants?

The only reports from the 1998 ice storm were west of ORH, Paxton north through Ashby.

There was definitely beam attenuation going on with BOX. Winds were starting to turn west of north (which is why the dome was attenuated in that direction), so the storm was going to begin to wind down. Like Will mentioned though LWM still had 0.34" between the hours of 5 and 7, so it's not like it was spitting out. It just was shitty snowflakes and a shitty environment to keep them frozen for very long.

Right. Kind of had a partial pupu platter and could never complete the order. We had rates, but not the dendrites or low level cold. A little low level cold (elevation) went a long way in this event to help save things. Valleys needed one or the other (dendrites) to come through and neither did.

Check out the GYX evolution. Whole sounding (including the DGZ) moves up and to the right, indicative of WAA (which you can see in one of the middle columns "inferred temp advection"). Right after the flip to snow you probably only had a 1-1.5km deep near freezing layer vs 3km later.

Most likely answer is that when the changeover occurred the thermodynamic profile didn't look like what we've described. The DGZ was likely lower in the atmosphere, and over time the WCB forced it to rise. Those mid level temps were way colder at 12z (despite surface temps being near 40) than they were at 00z (despite surface temps being near 33).

It's not worse, just different. 15 years ago when I was in school, models were all 80 km resolution. You would have to tease out banding potential and snowfall forecasts were more broad, with "locally higher amounts". Now models are finer resolution which is prone to more wobbling when they can resolve features down to 9 or 13km. Banding doesn't need to be teased out, but can be modeled, and "locally higher amounts" are explicitly forecast by guidance. So when those shift around it's far more noticeable.

I wasn't positive of that, but intuitively it makes sense. If you create a crystal at -5C it's going to take less heat to warm it up to 0C than if you create a crystal at -12C. Even if your crystal structure may be more delicate.

Really? I thought they were frcsting 5" or so And I ended up with 2.9" It was probably more, but I was fast asleep when the snow ended and the board was a wind blown/compacted 0.4" after my 2.5" midnight ob.

Trying to dive deep to figure out what the real failure point was here, because I don't really think it was forcing. The QPF was pretty good, right ballpark anyway, clearly there was strong f-gen in the right location relative to the storm. I keep coming back to our 00z sounding last night with DGZ up around 500 mb. That's not a great spot and indicative of a pretty torched air mass. Dropping less than dendrite ratios into a deep near freezing layer primed snow to melt at the surface. Even when rates were decent. There was a lot of 1/2 to 3/4SM snowfall that normally would accumulate if your flakes weren't already near slush by the time they reach the ground. Honestly the 04.12z NAM forecast soundings weren't bad in the vicinity of GYX, and it should've been a bigger red flag to me. We were running with ratios around 7:1, but they needed to be lower than that given the environment we were working with. Elevations did better because it shrunk the near freezing layer that flakes had to fall through and kept near surface temps slightly cooler.