2022 Global Model Runs Discussion (Out thru day 16)

[skyline385]

Happy hour GFS incoming

[boca]

I guess New Orleans thus run

[aspen]

Weakest GFS run for the 0/60 EPac AOI in a while. Central America is completely covered in a giant messy CAG that tries to produce something on the EPac side.

Also a weak system off the coast of the Carolinas pops up again

[DorkyMcDorkface]

Some activity on 12Z EPS

https://uploads.tapatalk-cdn.com/20220609/54a7da6c94c96e760b6c2b6bcf5c0b44.jpg

There is a weak disturbance that forms in the WCAR on the 12z operational Euro that swings up into the Yucatan. This is probably what those scattershot EPS members are hinting at.

[cycloneye]

Some peeps are making fun of GFS but look what happened last year.

 https://twitter.com/DrJeffMasters/status/1534918201410207744

[caneman]

Some peeps are making fun of GFS but look what happened last year.

https://twitter.com/DrJeffMasters/status/1534918201410207744

Agree, let's stop allowing the modeling wars. At least the GFS sniffs stuff out. As always, the best is the blend. If anyone states otherwise, they are willfully uneducated or somehow think it's a game. Those type posts are borish and annoying.

[skyline385]

Some peeps are making fun of GFS but look what happened last year.

https://twitter.com/DrJeffMasters/status/1534918201410207744

Agree, let's stop allowing the modeling wars. At least the GFS sniffs stuff out. As always, the best is the blend. If anyone states otherwise, they are willfully uneducated or somehow think it's a game. Those type posts are borish and annoying.

Been saying that for a while. With forecasting, an overtuned model is far better than one which is too late to the party

[NXStumpy_Robothing]

00z GFS continues to model development in the western Caribbean at around +180, but it's an outlier; only one other model from this morning's runs and the 12z runs yesterday show this solution. Other models have the precursor disturbance moving into the Pacific before developing.

I'm not sure why they run these models more than 10 days out, unless there is an African wave train with very stable upper air pattern locked in place they aren't worth much.

Late reply, but I just wanted to point out that model runs certainly have a usage beyond a week or so out; it's just not at the level that makes screenshots of fantasycanes plowing into Lake Charles 2 weeks out all that useful. At that point, so many different decisions (for lack of a better term) have already occurred in the atmosphere that pinpointing any given disturbance's intensity and track is an exercise in folly. All you can say for the previously-addressed scenario is that the GFS is forecasting a storm developing and then moving towards the Gulf Coast at some point past 200 hours or so. Hell, if you factor in the model biases' in and the level of error that even 5 days out gets you, it's entirely possible that the model is too far east and too willing to fall into a convective feedback loop, though the GFS is particularly good at sniffing out TCG events regardless. Besides, the last two years of the GFS's performance in hurricane season has been among the best when compared to other models (and even holding up fairly well against composite forecasts), so discarding it outright is silly.

Back to the point of this reply, though; ultimately, the value of these long-range runs is to explore the synoptic development of the atmosphere and for noticing long-term trends, as well as utilizing the expansive ensemble suite in order to note the conditions that give rise to the events that unfold on the deterministic run. It's important to have these longer runs in order to notice the shift in a model's depiction of the future in order to properly analyze the likelihood of a given scenario occurring near the time listed. Changes in ridge/trough flow or orientation, trade bursts, MJO progression, Kelvin wave forecasts, stratospheric/tropospheric polar vortex alterations, synoptic trends: this is where the long-range runs are their most useful and why they are important to produce. Just because they're not that useful for pinpointing the location or strength of a given synoptic/mesoscale feature does not mean that they are worthless; if they truly had little value, then the long-range portions of the models likely wouldn't be run in order to save computing power to focus on the shorter timeframe.

Anything involving the exact placement and category of a forecasted hurricane weeks out being highlighted is generally the result of people either posting the runs to laugh at the long-range, to drum up hype for a modelcane that probably won't come to fruition, or to panic over something that won't happen (for those less weather-savvy). While it is annoying to see these frames pop up on social media for the last two options (and in some cases, even the first one), it's just a symptom of focusing on the wrong part of the model at the wrong point in the run. Hopefully we can get weather enthusiasts on social media to lessen their posting of long-range model runs if they're not going to focus on where these runs thrive.

P.S. - I hope this didn't come off as demeaning or as derogatory towards you as I'm imagining; I just hate when people post long-range snapshots of various model runs just to dunk on the long-range in an emphasized area that doesn't correlate to what we actually should be looking at these models for. I'm also really appreciative we live in a time where we can see these longer runs multiple times a day across a variety of products, and so the constant dunking on models that go beyond 240 hours is a bit of a pet peeve of mine. Hope you're safe and unaffected this hurricane season!

[skyline385]

3 runs in a row with almost the same end solution which were all 260-270 hours out, is the GFS okay
Seriously though, there are so many variables at play this far out that I cant imagine how it got three very near solutions in a row

[Category5Kaiju]

3 runs in a row with almost the same end solution which were all 260-270 hours out, is the GFS okay
Seriously though, there are so many variables at play this far out that I cant imagine how it got three very near solutions in a row
https://imgur.com/OCooGwz

Something interesting I noticed too, but the GFS is ever so slightly less aggressive with the future EPAC system compared to earlier runs from several days past. It's also not really pushing back the timeframe of this Atlantic system; if anything, it seems to have gotten a tad earlier (June 17ish compared to June 22ndish beforehand). It sure loves the idea of spinning up a formidable hurricane in the Western Atlantic by the end of next week.

[aspen]

3 runs in a row with almost the same end solution which were all 260-270 hours out, is the GFS okay
Seriously though, there are so many variables at play this far out that I cant imagine how it got three very near solutions in a row
https://imgur.com/OCooGwz

Make that four runs in a row now.

This is also the strongest run for that Carolinas/SEUS tropical storm the GFS has been predicting for late next week.

[NDG]

I think we have a new crazy uncle and is American this time, the GFS.

[wxman57]

I think we have a new crazy uncle and is American this time, the GFS.

Note that it is 1 day later, and the GFS develops it 1 day later.

[MHC Tracking]

Very likely our GFS WCAR fantasycane won't develop in the long run. Genesis being steadily pushed back with every recent run. Only GEFS continues with widespread ensemble support. A more realistic TCG pathway would come from a MCS moving offshore of NC/SC as a result of the current ridging pattern, which favors southeastward moving MCS. GFS has depicted this in the past few runs.

[skyline385]

Weird GEFS run, just a few runs back the members were getting clustered towards Mexico from the death ridge but they are back out again

[Stormcenter]

Something is going to develop in the next 10 days it’s a matter of how strong will it be. IMO

[South Texas Storms]

Something is going to develop in the next 10 days it’s a matter of how strong will it be. IMO

Yep, on the EPAC side. Development looks unlikely in the Atlantic over the next 10 days.

[Hurricane Mike]

I'm watching with slightly, very slightly, more interest.

The GFS has shown this system for some 17 out of 20 runs. That's remarkably consistent, and as far as "fantasyland" goes, 200 hrs is a lot closer than 384 hrs.

Unlikely, but starting to arouse my interest a bit more.

[Category5Kaiju]

I'm watching with slightly, very slightly, more interest.

The GFS has shown this system for some 17 out of 20 runs. That's remarkably consistent, and as far as "fantasyland" goes, 200 hrs is a lot closer than 384 hrs.

Unlikely, but starting to arouse my interest a bit more.

Plus, it's not May. Seeing that is May definitely should raise doubts, but time in time again, Junes in the Atlantic have had their fair share of potent GoM systems. Alex 2010, Audrey, and Alma as major examples.

[wxman22]

The Icon now shows the tropical storm the GFS has been developing off the SE coast.Has the caribbean system also as a disturbance.