2020 Global Model Runs Discussion (Out thru day 16)

[TheStormExpert]

The chances this is just another in a long line of GFS W Caribbean phantoms have increased markedly during the last 24 hours. I’d say it is now at 2:1 in favor of it being a W Caribbean phantom vs it being at 2:1 in favor of it being for real just 3 days ago.

I'm not so sure. GEFS now shows a signal starting in 168 hours, and the CMC and Icon are showing signs of latching on too. It would seem the GFS only wants to show the system when it is not unreasonably pumping out epac storms. Would be interested to hear your reasoning though.

1. GFS runs had been showing 50%+ with a H for several days. Now none have done so recently.

2. GFS has had a history of fake Oct W Caribbean storms for at least several years.

3. The GEFS appears to be too genesis happy in relation to the operational.

4. The Euro suite continues to have very little. The 0Z EPS has backed off that slight uptick the 12Z EPS has and is now back to hardly anything.

5. The ICON and CMC have a history of being genesis happy. For example, the ICON had a fake 94L storm hitting NC on many runs whereas no other model had this.

6. I’m not so sure either. I’m not saying zero chance, especially in deference to the climo of the ENSO based-analogs. But the trend of the GFS toward no development in the W Caribbean is obvious.

7. The bias on weather forums is toward more development chances than reality because most members want development because no development is boring vs development being interesting. This is notwithstanding exceptions like 57 and Frank2, who tend to downplay most of the time.

We have a full blown La Niña and off the charts TCHP going for us this year. It would be off if something doesn’t eventually come out of the Western Caribbean. That region has been basically untouched territory in terms of seeing a monster hurricane in October for YEARS now.

[DestinHurricane]

1. GFS runs had been showing 50%+ with a H for several days. Now none have done so recently.

2. GFS has had a history of fake Oct W Caribbean storms for at least several years.

3. The GEFS appears to be too genesis happy in relation to the operational.

4. The Euro suite continues to have very little. The 0Z EPS has backed off that slight uptick the 12Z EPS has and is now back to hardly anything.

5. The ICON and CMC have a history of being genesis happy. For example, the ICON had a fake 94L storm hitting NC on many runs whereas no other model had this.

6. I’m not so sure either. I’m not saying zero chance, especially in deference to the climo of the ENSO based-analogs. But the trend of the GFS toward no development in the W Caribbean is obvious.

7. The bias on weather forums is toward more development chances than reality because most members want development because no development is boring vs development being interesting. This is notwithstanding exceptions like 57 and Frank2, who tend to downplay most of the time.

All fair points, especially 7 haha. Agreed the gefs May be genesis happy, especially since the upgrade. To counter though, the timeframe has moved up somewhat consistently. The gfs also has given into a clear bias this season to over produce in the epac, which, if those storms did verify, would directly suppress development in the Caribbean. However if those currently depicted are phantoms, then it would be falsely be underdoing the Caribbean development.

Also, how do we define phantom? Is a phantom any storm the gfs depicts at a given time that never pans out, or is it a system the gfs depicts in a pattern that never pans out? If the gfs models a pattern that has, say, a 30% chance of producing a system, and ends up correct but a storm never develops, does that count as a phantom?

All in all, I feel like my chances are higher than yours that this isn’t a phantom (by the latter definition above), and maybe a little higher than yours in development at all, but I haven’t left my comfortable spot on the fence yet.

IF there ends up being no named storm by 10/5 in or near the W Caribbean/Gulf in the Atlantic basin, I’m going to retrospectively call it a “GFS phantom”, even if one forms later, which I’m still betting on based on ENSO based analogs (current weak to moderate La Niña following a weak to moderate El Niño the prior fall/winter.)

Keep in mind that one week ago today that I first brought up these particular analogs implying a significantly higher than normal chance for a CONUS H hit from the Caribbean once we get to early to mid October and I’m still not backing down from that idea. So, I’m not downplaying just for the heck of it like I think a few like to do. But I’m also not a cheerleader who likes to “upplay” (is that a word lol?) like many members like to do on any weather board.

Why GFS phantom and not EPS or CMC phantom? They all show it too.

[LarryWx]

All fair points, especially 7 haha. Agreed the gefs May be genesis happy, especially since the upgrade. To counter though, the timeframe has moved up somewhat consistently. The gfs also has given into a clear bias this season to over produce in the epac, which, if those storms did verify, would directly suppress development in the Caribbean. However if those currently depicted are phantoms, then it would be falsely be underdoing the Caribbean development.

Also, how do we define phantom? Is a phantom any storm the gfs depicts at a given time that never pans out, or is it a system the gfs depicts in a pattern that never pans out? If the gfs models a pattern that has, say, a 30% chance of producing a system, and ends up correct but a storm never develops, does that count as a phantom?

All in all, I feel like my chances are higher than yours that this isn’t a phantom (by the latter definition above), and maybe a little higher than yours in development at all, but I haven’t left my comfortable spot on the fence yet.

IF there ends up being no named storm by 10/5 in or near the W Caribbean/Gulf in the Atlantic basin, I’m going to retrospectively call it a “GFS phantom”, even if one forms later, which I’m still betting on based on ENSO based analogs (current weak to moderate La Niña following a weak to moderate El Niño the prior fall/winter.)

Keep in mind that one week ago today that I first brought up these particular analogs implying a significantly higher than normal chance for a CONUS H hit from the Caribbean once we get to early to mid October and I’m still not backing down from that idea. So, I’m not downplaying just for the heck of it like I think a few like to do. But I’m also not a cheerleader who likes to “upplay” (is that a word lol?) like many members like to do on any weather board.

Why GFS phantom and not EPS or CMC phantom? They all show it too.

The GFS had been quite bullish with many runs having a full blown hurricane on many runs while the EPS had little.

[TheStormExpert]

I'm not so sure. GEFS now shows a signal starting in 168 hours, and the CMC and Icon are showing signs of latching on too. It would seem the GFS only wants to show the system when it is not unreasonably pumping out epac storms. Would be interested to hear your reasoning though.

1. GFS runs had been showing 50%+ with a H for several days. Now none have done so recently.

2. GFS has had a history of fake Oct W Caribbean storms for at least several years.

3. The GEFS appears to be too genesis happy in relation to the operational.

4. The Euro suite continues to have very little. The 0Z EPS has backed off that slight uptick the 12Z EPS has and is now back to hardly anything.

5. The ICON and CMC have a history of being genesis happy. For example, the ICON had a fake 94L storm hitting NC on many runs whereas no other model had this.

6. I’m not so sure either. I’m not saying zero chance, especially in deference to the climo of the ENSO based-analogs. But the trend of the GFS toward no development in the W Caribbean is obvious.

7. The bias on weather forums is toward more development chances than reality because most members want development because no development is boring vs development being interesting. This is notwithstanding exceptions like 57 and Frank2, who tend to downplay most of the time.

EPS probability for a TD is almost near 60%. It has been trending up and I am curious to see if it continues. I do agree I would like to see more deterministic runs pick it up. We shall see

If ensembles are bullish then it’s best to follow them even if the deterministic run isn’t, right?

[Spacecoast]

1. GFS runs had been showing 50%+ with a H for several days. Now none have done so recently.

2. GFS has had a history of fake Oct W Caribbean storms for at least several years.

3. The GEFS appears to be too genesis happy in relation to the operational.

4. The Euro suite continues to have very little. The 0Z EPS has backed off that slight uptick the 12Z EPS has and is now back to hardly anything.

5. The ICON and CMC have a history of being genesis happy. For example, the ICON had a fake 94L storm hitting NC on many runs whereas no other model had this.

6. I’m not so sure either. I’m not saying zero chance, especially in deference to the climo of the ENSO based-analogs. But the trend of the GFS toward no development in the W Caribbean is obvious.

7. The bias on weather forums is toward more development chances than reality because most members want development because no development is boring vs development being interesting. This is notwithstanding exceptions like 57 and Frank2, who tend to downplay most of the time.

EPS probability for a TD is almost near 60%. It has been trending up and I am curious to see if it continues. I do agree I would like to see more deterministic runs pick it up. We shall see

If ensembles are bullish then it’s best to follow them even if the deterministic run isn’t, right?

Is 60% probability for a TD really bullish, considering we have a full blown La Niña and off the charts TCHP?

Why only 60% for a TD, and not 90% for TS?
What do the models see that seem to inhibit development?

[boca]

Aren’t we in the suppression phase of the mjo right now?

[SFLcane]

Aren’t we in the suppression phase of the mjo right now?

Ensembles are not showing development right now more towards 7-8 days when the Atlantic will be close to moving into a more favorable phase.

[otowntiger]

I'm not so sure. GEFS now shows a signal starting in 168 hours, and the CMC and Icon are showing signs of latching on too. It would seem the GFS only wants to show the system when it is not unreasonably pumping out epac storms. Would be interested to hear your reasoning though.

1. GFS runs had been showing 50%+ with a H for several days. Now none have done so recently.

2. GFS has had a history of fake Oct W Caribbean storms for at least several years.

3. The GEFS appears to be too genesis happy in relation to the operational.

4. The Euro suite continues to have very little. The 0Z EPS has backed off that slight uptick the 12Z EPS has and is now back to hardly anything.

5. The ICON and CMC have a history of being genesis happy. For example, the ICON had a fake 94L storm hitting NC on many runs whereas no other model had this.

6. I’m not so sure either. I’m not saying zero chance, especially in deference to the climo of the ENSO based-analogs. But the trend of the GFS toward no development in the W Caribbean is obvious.

7. The bias on weather forums is toward more development chances than reality because most members want development because no development is boring vs development being interesting. This is notwithstanding exceptions like 57 and Frank2, who tend to downplay most of the time.

#7 for SURE

otowinger is a big bear too

Thanks! That’s good company! I agree with Larrywx! Lots of hype from the gfs only to have it die down to just about nothing. But I’m not ruling out more tropical development- almost assuredly there will be more before the season is up- just not sure what, how much or where. And as you might expect I’m not buying the hype .

[SFLcane]

[SFLcane]

No hype.. This is the model discussion thread that’s what we do discuss and post models.

[Shell Mound]

1. GFS runs had been showing 50%+ with a H for several days. Now none have done so recently.

2. GFS has had a history of fake Oct W Caribbean storms for at least several years.

3. The GEFS appears to be too genesis happy in relation to the operational.

4. The Euro suite continues to have very little. The 0Z EPS has backed off that slight uptick the 12Z EPS has and is now back to hardly anything.

5. The ICON and CMC have a history of being genesis happy. For example, the ICON had a fake 94L storm hitting NC on many runs whereas no other model had this.

6. I’m not so sure either. I’m not saying zero chance, especially in deference to the climo of the ENSO based-analogs. But the trend of the GFS toward no development in the W Caribbean is obvious.

7. The bias on weather forums is toward more development chances than reality because most members want development because no development is boring vs development being interesting. This is notwithstanding exceptions like 57 and Frank2, who tend to downplay most of the time.

All fair points, especially 7 haha. Agreed the gefs May be genesis happy, especially since the upgrade. To counter though, the timeframe has moved up somewhat consistently. The gfs also has given into a clear bias this season to over produce in the epac, which, if those storms did verify, would directly suppress development in the Caribbean. However if those currently depicted are phantoms, then it would be falsely be underdoing the Caribbean development.

Also, how do we define phantom? Is a phantom any storm the gfs depicts at a given time that never pans out, or is it a system the gfs depicts in a pattern that never pans out? If the gfs models a pattern that has, say, a 30% chance of producing a system, and ends up correct but a storm never develops, does that count as a phantom?

All in all, I feel like my chances are higher than yours that this isn’t a phantom (by the latter definition above), and maybe a little higher than yours in development at all, but I haven’t left my comfortable spot on the fence yet.

IF there ends up being no named storm by 10/5 in or near the W Caribbean/Gulf in the Atlantic basin, I’m going to retrospectively call it a “GFS phantom”, even if one forms later, which I’m still betting on based on ENSO based analogs (current weak to moderate La Niña following a weak to moderate El Niño the prior fall/winter.)

Note that analogs, even for “hyperactive” seasons, have not applied particularly well to 2020, given that we have ended up with far less ACE than ENSO and overall NS would have implied. Compared to climatology, this season has seen much less ACE than other years that also featured moderate Niña conditions and above-average NS. In this case, ‘20 has certainly been very anomalous, but in a bearish rather than bullish manner. We have seemingly seen more shear from TUTTs than overall conditions would have suggested, perhaps due to the influence of the very warm subtropics, which has resulted in sinking air over the MDR/Caribbean. At this rate, I think that ‘20 may well follow its bearish trend and end up with no additional MH activity, contrary to climatology for high-activity seasons coincident with Niña conditions.

[otowntiger]

No hype.. This is the model discussion thread that’s what we do discuss and post models.

When I said hype- it was more directed towards model hype, if that makes sense. But I do enjoy watching models as they try to sort things out which of course why I am always hanging out here.

[cheezyWXguy]

All fair points, especially 7 haha. Agreed the gefs May be genesis happy, especially since the upgrade. To counter though, the timeframe has moved up somewhat consistently. The gfs also has given into a clear bias this season to over produce in the epac, which, if those storms did verify, would directly suppress development in the Caribbean. However if those currently depicted are phantoms, then it would be falsely be underdoing the Caribbean development.

Also, how do we define phantom? Is a phantom any storm the gfs depicts at a given time that never pans out, or is it a system the gfs depicts in a pattern that never pans out? If the gfs models a pattern that has, say, a 30% chance of producing a system, and ends up correct but a storm never develops, does that count as a phantom?

All in all, I feel like my chances are higher than yours that this isn’t a phantom (by the latter definition above), and maybe a little higher than yours in development at all, but I haven’t left my comfortable spot on the fence yet.

IF there ends up being no named storm by 10/5 in or near the W Caribbean/Gulf in the Atlantic basin, I’m going to retrospectively call it a “GFS phantom”, even if one forms later, which I’m still betting on based on ENSO based analogs (current weak to moderate La Niña following a weak to moderate El Niño the prior fall/winter.)

Note that analogs, even for “hyperactive” seasons, have not applied particularly well to 2020, given that we have ended up with far less ACE than ENSO and overall NS would have implied. Compared to climatology, this season has seen much less ACE than other years that also featured moderate Niña conditions and above-average NS. In this case, ‘20 has certainly been very anomalous, but in a bearish rather than bullish manner. We have seemingly seen more shear from TUTTs than overall conditions would have suggested, perhaps due to the influence of the very warm subtropics, which has resulted in sinking air over the MDR/Caribbean. At this rate, I think that ‘20 may well follow its bearish trend and end up with no additional MH activity, contrary to climatology for high-activity seasons coincident with Niña conditions.

If you compare ace by date instead of named storm count, it is currently above average. Using named storm count alone would imply we are very near the end of our season since the names we recently passed don’t usually occur until November, if ever. However, we are still only about 60% through the season when activity is taken into account

[cheezyWXguy]

12z gfs shows the system starting to take shape at hour 216, but keeps it very weak due to consistent land interaction

[Spacecoast]

12z gfs shows the system starting to take shape at hour 216, but keeps it very weak due to consistent land interaction

Looking a little like Opal after hour 216.
Quite a different setup than the 6z run....

[SFLcane]

12z CMC a bit stronger..

[SFLcane]

GFS waking up?

[boca]

Even the models have a protective storm deflector from the Florida peninsula

[gatorcane]

The GEFS starting at 180 hours, not far off from the CMC, just more east.

[SFLcane]

12z GEFS is quite active.