psuhoffman

Wow kinda crazy how someone that just happened to be at WIntergreen just happened to join right now and their first post is about this. Crazy huh

thanks, and what would you suggest?

I meant the GFS. You said it’s not…ugh never mind

Yes I discovered it’s getting warmer. Right after Al Gore invented the internet.

It was south

No wait a minute… if he wants to give me credit for discovering it’s getting warmer…

I know it’s facetious but that 2014 storm it was in the mid 20s and it’s only warmed like 1/2 a degree since then. And the 2018 storm was even more recent and not really that borderline when the snow was falling.

Thanks Nostradamus

I wish that’s all that was wrong with him.

Right but the point is it was LESS WRONG then all the other models day 5-8 and so on a chart of verification scores it would look good.

But this point is actually a great example of what I meant by "when all we care about is how much snow falls in our yard it might not align with verification scores". Yes the AIFS had multiple runs in the day 5-8 range with HECS snowfall results for our area. But, those results were real, just displaced about 150 miles to the northeast. And no other model, at those ranges, were even close...the GFS didn't start showing those crazy snow totals until like day 4 out. So when compared to all the other models, which didn't have an HECS anywhere at all...the AIFS which had it but displaced a small amount too far southwest, the AIFS was by far the closest to the truth (the less wrong) model in the day 5-8 period. We look at them all wrong, in that we expect them to be exactly right at a range that there is almost no chance they will be. The AIFS showing a HECS somewhere in the northeast at all day 5-8 was a win for it...but we think it was wrong because the big snow ended up not over us.

Verification scores don't lie. However, when all we care about is how much snow ends up on our lawn, that doesn't always necessarily correlate to some hemispheric h5 or MSLP verification score! Also, sometimes we see a model leading the scores chart and think "that means it's right" when it really means it's slightly less wrong than the others.

This definitely has merit, thinking back in the late 90's and early 2000's stuff at day 3-5 was treated like we look at stuff day 7-10 now. We didn't even try to look at a specific storm threat past day 5, most of the models didn't even run past 144 hours and it was a complete waste of time. Usually by 72 hours we have pretty good idea what the major features will be, but now we also expect meso scale things to be right and that was never a thing in the past. And in marginal setups where a 1-2 degree difference is huge, expecting models to nail that is crazy. But some people do now. So maybe it's also a case of expectations increasing faster than the actual improvements which gives the perception things are worse. I do think there is some truth the the decreased consistency of bias errors but it's likely not as bad as I am perceiving it.

@Terpeast @WxUSAF Was the Palm Sunday blizzard of 1942 one of the HECS storms that would likely not have happened today? I know when this was a topic and that regression study was done Feb 1987 was a lost one and one other was mentioned, was it 1942? Baltimore got 22" but the temperature never got below 33 degrees the whole storm. Seems unlikely that would have worked out today with the roughly 3.5F increase in temps since then. That's kinda depressing...one of Baltimores biggest snowstorms ever would probably just have been a dismal rainy spring day if that same exact thing happened again today.

It really dampens the wave around 90-96 hours. It doesn't eject enough energy, most of it hangs back, the wave dies with no mid and upper level support at all as it slides east.