vortex95

You are not paying attention. Do you not see the isolated supercell to your W on the HRRR fcst I sent? Very disappointing you are downplaying this event.

I don't know...

"If it's not in Weymouth, it doesn't matter!" What kind of attitude is that?

04z HRRR suggesting "cells ahead of line." So 1-2 punch for some areas. Also, nice sliver of high 0-3km CAPE SNE and a second area for NNE.

Hmmm, CoastalWx and WxWiz be like "???????." 5% tor risk? How can this be? No EML!

Ineedsnow will be "X-static" if we get 5% tor, but WxWiz will not b/c he can't chase!

We have a lot of climo for TC tracks globally, so AI has a lot to draw from, so it should do well here. Intensity, not so much. There are still so many things we can't model well for the stronger TCs. The inner cores of the intense ones are basically mesoscale features. RI occurs often in such a short time frame and it is hard for physics-based models to handle. For decades, we have seen a steady improvement in TC tracks, but very little for intensity until the 2010s, when significant improvement occurred. So again, AI should do well for TC tracks. And TC tracks often are latitude/longitude dependent, so this variable is a solid base indicator for AI to use.

This happens *every* time new technology becomes available. It is sold as the best thing since sliced bread w/ lofty claims. I recall in the 80s it was said "in 10-15 year advancement in computer modelling will make human forecasting obsolete!' Right, how did that work out? In fairness, computer modelling advancement has removed a lot of manual work a human used to do, but this is a *good* thing actually. Why?, b/c the shear volume of wx data now is enormous and keeps increasing. We need faster computers and AI to help sort and manage it all (do the grunt heavy lifting) so the human forecast can provide the insight/wisdom/understanding -- which AI does not have -- to give concise and clear messages to all. That's one reason why IDSS is so big in the NWS. Forecasts have become very good, but the real challenge is *communicating* it all effectively in this information overload and social media world. So I would not worry about AI "taking over," so to speak, anytime soon.

You are missing the point. Teleconnections deal w/ broad scale patterns. That only works/is useful to a point. When it comes to actual sensible wx in a specific area, that's an entirely different ballgame. How many times do we have a -NAO, and no big East Coast snowstorm occurs, as one example? I pointed out the mesoscale details being everything for all to read here, and learn something from it. You look on social media, and far too often you see people hyping certain patterns, and stating this or that will happen in X or Y region w/o stating any caveats or how it's not so simple/linear as a 1-1 correlation. People on this forum, some are still learning and other are eager to learn more, and furthering the discussion is of value net-net. And in New England, many just think about the outbreak sequence June 8-9, 1953. That is a high-end, outlier exception, and that can skew the perception of things from a local bias POV.

Charleston IL spinner PT.

Spinner paradise!

So here are the last 3 runs of the HRRR 12/18/00z all valid 6pm Thu. Which one will be right? It's impossible to tell. Small details on the mesoscale can make a big difference as to storm mode. Either way, the timing for this is ideal. 0-3 km CAPE from central NH to CT is high as 150 at 6pm. And that is big for spinner potential and 150 is more than enough. Perhaps this is what we will get Thu. Li'l spin-ups w/ nice crisp titled CB vistas! Punxsutawney PA spinner 6/14/26. https://www.facebook.com/reel/1355336163205600

Working overnights so hence my delayed response. I say to CoastalWx and WxWiz, SPC added a 2% tor risk in New England in the DY2 1730z update!!!! (adding the exclamation points to be like Indeedsnow - LOL). "As I had been saying for tor risk..." 18z HRRR VT 22z Thu. 3 discrete supercells. PWM, PSM, and one just N of Woburn MA (where I am from). Scott needs a plan to chase!

Yes, but when it comes to sig tor events no matter where you are, the very small details are everything. One parameter off, or not sufficient, can make a big difference. For instance, dew points are few degrees lower in in BL than fcst. That raises the LFC, decreases 0-3 km CAPE, and makes supercells more outflow dominant, which curtails the high-end tornado potential or any tornadoes at all. There is very little difference between a sig tor event and a weak tor event if you look at the large picture. It can look perfect synoptically, but it comes down to the mesoscale in the end.