weathafella

EPS essentially says we get a lot more favorable after next week.

It’s AI version is much better

Skynet following uncle—Tuesday colder with even Brett happy. Also much less lasting and amplitude of warmth thereafter.

Uncle could be drunk but it has a solid and more widespread system for Tuesday.

Very windy here but mostly in gusts-not really sustained.

The problem is that the next 1-2 weeks feature opportunities only when northern stream systems pass south of us bringing enough of the Canadian icebox to snow on SNE. Then said systems blow up in the Atlantic but don’t curl up closing the door on the cold and allowing warmth to return in relatively short order. I did see improvements in the 11-15 on last night’s eps so January should offer hope of it holds.

Hard to have any idea which side of the boundary we’ll be on. I don’t ever recall such volatility in the guidance.

Also, some of the torchy guidance has a Canadian high in a nice position. The warmth seems at least partially suspect. We’ll see.

AI looks much less torchy for gfs. Skynet has a nice borderline warning event on the 28th but you’re too far south verbatim.

Exactly! As events occur, deep learning should ultimately improve the results.

The thing about 2014-15 is that 100+ happened in a 6 week period. Absolutely unprecedented for coastal sne.

Winter finally ends April 17th…

Can you provide data to prove me wrong? Give me low first halves and final totals vs climo. Also, to clarify what you quoted-I want to be sure you understand that I’m saying a low snow first half correlates with a low snow winter.

Let’s hope uncle is drunk today….

AI GFS is better. However the more suppressed look of the op run is probably a good sign.

I don’t think the statistics on Nina Decembers mean anything especially with a totally unscientific sample size. Also, let’s say everyone is below normal through December and a whopper dumps 1-2 feet starting New Year’s Day afternoon-but the arbitrary cutoff says ratter. Silly. What’s perhaps more relevant about the statistic is that low snow in the first half of winter is often a good harbinger regardless of the enso state. Of course the notable exception is 2014-15 when BOS had single digits for the season in mid January and finished with their all timer at 110.6”

So odds are highest in general for January being above average given it is the coldest month of the year climo wise.

Anthony isn’t posting. Nuff said.

Lol…their lifespan was 30 something.

Ok op huggers-today’s euro is snowy for many of us.

Paul, I think your approach narrows the possibilities. I mean isn’t AI inherently “better computing” considering the concept is deep learning? I think AI can be tweaked when we see where it needs to be quickly. As a crude example, my phone quickly learns how I use “muthufukka”. Phones five years ago took much longer and my next phone should even learn it faster.

Why do you have disdain for AI weather models?

I think there’s opportunity every 3/4 days. Obviously this first one is warm but based on guidance I’m seeing we’ll have lots of chances. We should be a lot more active and the hope is the boundary is favorable for us. It will be sometimes but we obviously are playing with fire a lot.

Inverted flake light snow now.

The ensembles in the 8-15 range are much more favorable vs the ops. It seems silly get tall wrapped up in op forecasts in a difficult pattern on clown range.