All Activity

Might be a little too new to do that just yet. The documentation for using their API's needs to get added into the training so they can understand when queried. A fully trained programmer familiar with Google Cloud could probably figure it out. That is not I.

Raines to Maines

What’s the KFS AI have?

Have AI to write the code for you. All you need is the logic.

18z euro was decent for Friday - looks like a 1-3”/2-4” type thing.

Would love to figure out how to run the latest Google weather models but above my level in programming, and there's costs involved. These are supposedly the most advanced AI based weather models available.

The ultimate inhibitor is going to be initialization. That’s always the main issue. That’s something that even AI can’t solve easily.

It's going to take a significant resolution upgrade in order for the AI models to make the jump. Outputs look smoothed, especially at long range

Lmao.

It’s only got a couple of years. just for fun, read the AI 2027 report Wont matter eventually.

I hope you're right. If you draw a line from Worcester to Methuen, it passes straight through Lowell. The GFS sounding at 00 has the temp at 32F around 950.

TRI will finish the month of November ~roughly +0.9F above normal. It was a wild month with wild rides. The highest high temp for the month was 71F. Four days later the high was 38F! The month featured a few days where wind chills were in the teens during the day. I have one tree which normally keeps its leaves until late December. It's leaves are all gone now - thankfully! The cold snaps and wind did the trick. We did manage about 0.5" of snow mid-month. I suspect our winter will mirror this pattern - extremes!

That’s like 4 hours later than euro had it last night

This would have been a great time for a Charles Town trip. Win big and score big!

I’m assuming AI will eventually learn from its errors. But it’s prob gonna be a bumpy road.

The frontal passage is spiking the temperatures by 10+ degrees within 15 min. Mixing out the inversion + downsloping.

I'll never forget when you went to Donners Pass and experienced the Reality of eleven feet of fresh snow! You got to watch the entire event, too!

looks pretty meh for the first two weeks

Check around the fins of the head towards the coil/starter are for a nest, had one nibble a small wire years ago, regarding moth balls had no luck, in irrigation boxes, they were cuddled up on top of the moth balls! BACK TO WEATHER......

Pretty bad medium range fumble by the Euro-AI. A bit more concerning was just how long and arduous the cave back towards the globals was. Once outputs began to trend, corrections were steady and incremental but far too slow for me to take runs seriously for several days. Model outputs aren't very useful if it's obvious they're playing catch up. The AI Models as they currently stand are a great new tool, but I don't think they're ready to compete with the globals... yet. I guess I currently treat them akin to the Canadian, maybe a tick below. However, I think that in a couple years we may have significant disruption of the current model heirarchy.

I will take either background state as long as it is weak - Nino or Nina. It seems like the last few winters have had a dominant ENSO state, but also strong elements of the opposite background state. I think that is probably due there not being a strong difference between the colder Nina water and the warmer water around it -> What is the word for that? In other words, the lack of sharp differences creates kind of a muted ENSO effect. Is the IO still slated to be quiet?

One word when storing equipment . Moth balls. Well actually two words...

Yeah that's my worry. I have two snow blowers for two different locations. Of course where I expect the most snow that one's not running.