First Legit Storm Potential of the Season Upon Us

[mahk_webstah]

32 minutes ago, ORH_wxman said:

Not very good in short term. It def kicked the operational’s butts though in that D3-5 period. 

Maybe that’s the take away. Good with the pattern in the medium range but not great with the details in the short range. If that’s the case, that would be a good thing to know going forward.

[dryslot]

You want it to get it right in the shorter terms though.

[dryslot]

Nobody is shoveling day 3-5 snow.

[Baroclinic Zone]

4 minutes ago, dryslot said:

Nobody is shoveling day 3-5 snow.

Emotional shoveling 

[mahk_webstah]

6 minutes ago, dryslot said:

You want it to get it right in the shorter terms though.

Absolutely but for meteorologists, and for very skilled and knowledgeable hobbies like yourself, you want to know which models to use at what point and in what context. So if that’s where the AI models are now then that’s where they are.

[dryslot]

Just now, Baroclinic Zone said:

Emotional shoveling 

Seen plenty of tears from a bunch on here that you would probably need a pump instead of a shovel, Maybe now this will subside some.

[dryslot]

6 minutes ago, mahk_webstah said:

Absolutely but for meteorologists, and for very skilled and knowledgeable hobbies like yourself, you want to know which models to use at what point and in what context. So if that’s where the AI models are now then that’s where they are.

TY, This is true, But we have many that are already decent in the mid term but very few getting it right in the short term.

[dendrite]

45 minutes ago, mahk_webstah said:

Maybe that’s the take away. Good with the pattern in the medium range but not great with the details in the short range. If that’s the case, that would be a good thing to know going forward.

That’s why I like them more for the extended and mid range. Granted that was an AI ens to begin with, but even with the “op” AI I’m still hesitant to use use them close in except for trends. Yeah, the coarseness tightens up as we get closer, but it hasn’t been to regular op gradients. But it’s a new tool and we’re all figuring out how to use it. 

[ORH_wxman]

49 minutes ago, dryslot said:

You want it to get it right in the shorter terms though.

Yes but we just don’t use it in the shorter term since we have great short term models. If we can rely on AI as a solid tool for medium range, then that has a lot of value. 

[ORH_wxman]

Fwiw, I think it’s good to talk about it a lot because the more we all recognize and point out where the AIs were successful, then we can start to pick up patterns….maybe they are good in some types of setups and crappier in others relative to the OP guidance.