OceanStWx

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.

It takes some stones to forecast a significant period of 3:1 or worse ratio.

Even Kuchera isn't going to save you in an event like this. 33F snow is still going to give you a 7:1 ratio. We were about half of that at the office from 1-7PM when it was nearly all snow/frozen.

Tip is right, NW winds downslope for just about everyone across the region. That does a number on precip.

Today is fun, because now I can play Monday morning QB on myself trying to diagnose WTF went wrong.

Well I think a problem is that clown maps (especially these realistic looking ones) are a relatively new feature. Because of how they are generated, you rarely end up seeing more snow than they spit out.

We were trying to tease out elevation from people when they called, but there aren't many people like Gene who know their elevation and all their neighbors elevations too. But I think it's clear that played a big role when you look at our office total vs up on Lava Rock.

2.5” at home. 2.76” liquid in the gauge. Probably a half inch rain before snow started mixing in. 1:1 ain’t going to cut it.