2024 Global Model Runs Discussion (Out thru day 16)

[mcheer23]

Is that including the west coast of Florida? Seems like with the current pattern we are safe.

The Florida peninsula should be safe by Thanksgiving.

Agreed.

[cajungal]

After Ida, I definitely don’t want a storm. But still shocked that the models are this dead through beginning of September. Odd that almost every forecast was showing an insane amount of storms and now that likely won’t verify. Would be a blessing if it continues.

[DorkyMcDorkface]

Looking for an ensemble member somewhere to indicate development? Ha!

Sorry but I don't recall anyone in this thread doing that? The maps are being posted because there are genuine signals containing numerous ensemble members showing storms in contrast to the operational guidance. No one is selectively cherry picking stray members and claiming there will be a storm based off that alone.

[Category5Kaiju]

Looking for an ensemble member somewhere to indicate development? Ha!

Sorry but I don't recall anyone in this thread doing that? The maps are being posted because there are genuine signals containing numerous ensemble members showing storms in contrast to the operational guidance. No one is selectively cherry picking stray members and claiming there will be a storm based off that alone.

Adding to your point, the ensembles in 2022 were very quiet at mid to late August. And look what happened.

If anything, the fact that ensembles actually show activity coming up really says something, even if operational models don’t at the moment.

[skyline385]

Regardless of what happens and we all know activity will resume in early September if not end August, it is certainly extremely unusual to see deterministic models this quiet post Aug. 20 in a record warm +AMO during a transitioning Nino year and if anyone is claiming that it was expected, that argument is certainly not in good faith because anyone who has been tracking the NATL for few years would know, the switch flip always happens around this time in NATL climo.

I still fully expect activity to resume in a couple of weeks and the hyperactivity threshold is still extremely easy to beat (we only need ~105ACE more, the average activity post Aug 20 in 1991-2020 climo is ~110 ACE) but this little lull is going to make it difficult for those near record breaking 220+ ACE numbers. We could still potentially get there if September peaks as it does and we get an extended CV season in October but its going to be difficult imo.

[Spacecoast]

Latest TC anomaly forecast derived from Euro EPS (consisting of 100 ensemble members plus a control member). Probability anomalies within a 20-year model climatology within the same 7-day time window:

Aug 26th through Sept 2nd:


Sept 2nd through Sept 9th: Depicting anomalously higher activity off northern African coast, and anomalously lower activity off southern coast.


Sept 9th through Sept 16th (peak):


Should I be skeptical of ensemble models because they aren't showing the 'switch'? We should be entering a phase where I am skeptical of models spinning up random majors bouncing around the Gulf.

One thing I am sure of....
Once this quiet period is over, the model runs will be... well...less quiet, perhaps significantly less quiet.
When that will happen is still TBD.

[Teban54]

Regardless of what happens and we all know activity will resume in early September if not end August, it is certainly extremely unusual to see deterministic models this quiet post Aug. 20 in a record warm +AMO during a transitioning Nino year and if anyone is claiming that it was expected, that argument is certainly not in good faith because anyone who has been tracking the NATL for few years would know, the switch flip always happens around this time in NATL climo.

I still fully expect activity to resume in a couple of weeks and the hyperactivity threshold is still extremely easy to beat (we only need ~105ACE more, the average activity post Aug 20 in 1991-2020 climo is ~110 ACE) but this little lull is going to make it difficult for those near record breaking 220+ ACE numbers. We could still potentially get there if September peaks as it does and we get an extended CV season in October but its going to be difficult imo.

While I don't think anyone was expecting what we're seeing... Here's some food for thought.

In earlier years, global models (especially GFS and, for a while, CMC) were notorious for spinning up phantom storms that didn't materialize or vastly underperformed in reality. It was not uncommon during late August: it happened in 2017 when models made Harvey and PTC 10 MDR majors, it happened in 2022 when they showed numerous MDR storms before Danielle and Earl that poofed one by one, and I'm sure there were other years.

(In fact, phantoms can still happen even later, well into peak season. 2021 had one such episode that resulted in people season cancelling while Larry was still a Cat 3.)

My point is: Suppose we take it as a given that no other storms form in August (which is not a guarantee). Would it fundamentally change things if operational global models were showing storms that end up being phantoms, compared to the status quo of not showing anything at all?

Personally, I feel the answer should theoretically be no -- what models are showing 10 days in advance has no effect on the conditions that actually materialize 10 days later. Yet, I can't help but feel that the discussion here would be very different if the phantom storms appeared, even if we end up being fooled.

[skyline385]

Regardless of what happens and we all know activity will resume in early September if not end August, it is certainly extremely unusual to see deterministic models this quiet post Aug. 20 in a record warm +AMO during a transitioning Nino year and if anyone is claiming that it was expected, that argument is certainly not in good faith because anyone who has been tracking the NATL for few years would know, the switch flip always happens around this time in NATL climo.

I still fully expect activity to resume in a couple of weeks and the hyperactivity threshold is still extremely easy to beat (we only need ~105ACE more, the average activity post Aug 20 in 1991-2020 climo is ~110 ACE) but this little lull is going to make it difficult for those near record breaking 220+ ACE numbers. We could still potentially get there if September peaks as it does and we get an extended CV season in October but its going to be difficult imo.

While I don't think anyone was expecting what we're seeing... Here's some food for thought.

In earlier years, global models (especially GFS and, for a while, CMC) were notorious for spinning up phantom storms that didn't materialize or vastly underperformed in reality. It was not uncommon during late August: it happened in 2017 when models made Harvey and PTC 10 MDR majors, it happened in 2022 when they showed numerous MDR storms before Danielle and Earl that poofed one by one, and I'm sure there were other years.

(In fact, phantoms can still happen even later, well into peak season. 2021 had one such episode that resulted in people season cancelling while Larry was still a Cat 3.)

My point is: Suppose we take it as a given that no other storms form in August (which is not a guarantee). Would it fundamentally change things if operational global models were showing storms that end up being phantoms, compared to the status quo of not showing anything at all?

Personally, I feel the answer should theoretically be no -- what models are showing 10 days in advance has no effect on the conditions that actually materialize 10 days later. Yet, I can't help but feel that the discussion here would be very different if the phantom storms appeared, even if we end up being fooled.

I agree with the sentiment in your post but here's the thing about it. Models would show phantom majors around Aug 20th because the conditions they saw were very favorable in the 10 days ahead as soon as the switch flipped. They are in the end numerical models without a mind of their own or any kind of reference to past seasons, they run off what they see every single run so the question then is, why aren't they seeing the favorable conditions yet this year and what is causing the delay? The current theory seems to be the high latitude waves combined with a heavy SAL season and a MJO stuck in phases 2/3 producing heavy easterly shear but all of this should be starting to decrease soon. The delay however will affect the activity of the season though as the NATL is using up peak season time. Like I said earlier, hyperactive season is still easy to beat but the more peak season time the NATL wastes, the less likely it is for those 220+ ACE numbers to verify. This of course does not say anything about the impact the season will have which has already been pretty significant.

[Teban54]

Regardless of what happens and we all know activity will resume in early September if not end August, it is certainly extremely unusual to see deterministic models this quiet post Aug. 20 in a record warm +AMO during a transitioning Nino year and if anyone is claiming that it was expected, that argument is certainly not in good faith because anyone who has been tracking the NATL for few years would know, the switch flip always happens around this time in NATL climo.

I still fully expect activity to resume in a couple of weeks and the hyperactivity threshold is still extremely easy to beat (we only need ~105ACE more, the average activity post Aug 20 in 1991-2020 climo is ~110 ACE) but this little lull is going to make it difficult for those near record breaking 220+ ACE numbers. We could still potentially get there if September peaks as it does and we get an extended CV season in October but its going to be difficult imo.

While I don't think anyone was expecting what we're seeing... Here's some food for thought.

In earlier years, global models (especially GFS and, for a while, CMC) were notorious for spinning up phantom storms that didn't materialize or vastly underperformed in reality. It was not uncommon during late August: it happened in 2017 when models made Harvey and PTC 10 MDR majors, it happened in 2022 when they showed numerous MDR storms before Danielle and Earl that poofed one by one, and I'm sure there were other years.

(In fact, phantoms can still happen even later, well into peak season. 2021 had one such episode that resulted in people season cancelling while Larry was still a Cat 3.)

My point is: Suppose we take it as a given that no other storms form in August (which is not a guarantee). Would it fundamentally change things if operational global models were showing storms that end up being phantoms, compared to the status quo of not showing anything at all?

Personally, I feel the answer should theoretically be no -- what models are showing 10 days in advance has no effect on the conditions that actually materialize 10 days later. Yet, I can't help but feel that the discussion here would be very different if the phantom storms appeared, even if we end up being fooled.

I agree with the sentiment in your post but here's the thing about it. Models would show phantom majors around Aug 20th because the conditions they saw were very favorable in the 10 days ahead as soon as the switch flipped. They are in the end numerical models without a mind of their own or any kind of reference to past seasons, they run off what they see every single run so the question then is, why aren't they seeing the favorable conditions yet this year and what is causing the delay? The current theory seems to be the high latitude waves combined with a heavy SAL season and a MJO stuck in phases 2/3 producing heavy easterly shear but all of this should be starting to decrease soon. The delay however will affect the activity of the season though as the NATL is using up peak season time. Like I said earlier, hyperactive season is still easy to beat but the more peak season time the NATL wastes, the less likely it is for those 220+ ACE numbers to verify. This of course does not say anything about the impact the season will have which has already been pretty significant.

Keep in mind that August 20 is likely a statistical average and not a precise cutoff date. If we assume that models can sneak out a change towards favorable conditions very quickly (which I still somewhat doubt), whether such an identification happens on August 17 or August 20 shouldn't matter -- so whether it happens on August 20 or August 23 shouldn't, either.

As others mentioned recently, EPS and GEFS both show the Eastern Atlantic shear drop off significantly around September 2, which also coincides with MJO forecasts. That's Day 12, outside the range of operational Euro (or at least the free version), and in the unreliable range for GFS. That may just be ensemble variance washing out the mean, but I think it's possible that the operational models simply aren't far out enough just yet.

[Stratton23]

That gulf disturbance may need to be watched a little more closely eary next week, the 12z CMC now closes off a weak low, prior runs were an open wave, very interesting

[ConvergenceZone]

I found a comparison between this time in 2023 vs. now using the NHC 7-day outlook graphic so I made my own. Quite a difference. After August 19 in 2023, there were 3 storms that formed in late August, 6 in September and 2 in October. That's 11 named storms after August 19th. This year, nothing. Unlikely we'll see any additional storms in August and early September is looking dead, too. We could reach the peak of the season (Sept 10) with only 5-6 named storms. I love it! Days off this time of year are a luxury.

I read Klotzbach's season update but I think something is being missed. Warm water - check. Low wind shear - maybe. As for a favorable environment - nope. Caribbean and Tropical Atlantic are dominated by dry, sinking air. It doesn't matter how warm the water is. If the atmosphere isn't favorable then hurricanes struggle to develop. I'm thinking my revised numbers of 18/8/4 may be too high. I'm just hoping for the current pattern to continue and any additional storms to stay out to sea. Perhaps those global models indicating no development are correct. Looking for an ensemble member somewhere to indicate development? Ha!

https://wxman57.com/images/NHC7Day.JPG

https://wxman57.com/images/NHC7Day.JPG

Tropical Atlantic Instability is way below normal:
https://www.ssd.noaa.gov/PS/TROP/TCFP/data/current/ts_al_tat_THDV.gif

Agreed, biggest forecasted Bust incoming.

[Teban54]

I found a comparison between this time in 2023 vs. now using the NHC 7-day outlook graphic so I made my own. Quite a difference. After August 19 in 2023, there were 3 storms that formed in late August, 6 in September and 2 in October. That's 11 named storms after August 19th. This year, nothing. Unlikely we'll see any additional storms in August and early September is looking dead, too. We could reach the peak of the season (Sept 10) with only 5-6 named storms. I love it! Days off this time of year are a luxury.

I read Klotzbach's season update but I think something is being missed. Warm water - check. Low wind shear - maybe. As for a favorable environment - nope. Caribbean and Tropical Atlantic are dominated by dry, sinking air. It doesn't matter how warm the water is. If the atmosphere isn't favorable then hurricanes struggle to develop. I'm thinking my revised numbers of 18/8/4 may be too high. I'm just hoping for the current pattern to continue and any additional storms to stay out to sea. Perhaps those global models indicating no development are correct. Looking for an ensemble member somewhere to indicate development? Ha!

https://wxman57.com/images/NHC7Day.JPG

https://wxman57.com/images/NHC7Day.JPG

Tropical Atlantic Instability is way below normal:
https://www.ssd.noaa.gov/PS/TROP/TCFP/data/current/ts_al_tat_THDV.gif

Agreed, biggest forecasted Bust incoming.

Saying a season that already had 5/3/1 and 55 ACE will be a bigger bust than 2013 is just absurd, IMHO.

[aspen]

I found a comparison between this time in 2023 vs. now using the NHC 7-day outlook graphic so I made my own. Quite a difference. After August 19 in 2023, there were 3 storms that formed in late August, 6 in September and 2 in October. That's 11 named storms after August 19th. This year, nothing. Unlikely we'll see any additional storms in August and early September is looking dead, too. We could reach the peak of the season (Sept 10) with only 5-6 named storms. I love it! Days off this time of year are a luxury.

I read Klotzbach's season update but I think something is being missed. Warm water - check. Low wind shear - maybe. As for a favorable environment - nope. Caribbean and Tropical Atlantic are dominated by dry, sinking air. It doesn't matter how warm the water is. If the atmosphere isn't favorable then hurricanes struggle to develop. I'm thinking my revised numbers of 18/8/4 may be too high. I'm just hoping for the current pattern to continue and any additional storms to stay out to sea. Perhaps those global models indicating no development are correct. Looking for an ensemble member somewhere to indicate development? Ha!

https://wxman57.com/images/NHC7Day.JPG

https://wxman57.com/images/NHC7Day.JPG

Tropical Atlantic Instability is way below normal:
https://www.ssd.noaa.gov/PS/TROP/TCFP/data/current/ts_al_tat_THDV.gif

Agreed, biggest forecasted Bust incoming.

Saying a season that already had 5/3/1 and 55 ACE will be a bigger bust than 2013 is just absurd, IMHO.

We’ve already surpassed 2013’s hurricane, MH, and ACE totals lol.

But I do think we will see NS totals busting. We’d need a huge burst in TCG to get to 20 NS, and I don’t see that happening. Not just because of the current AEJ/monsoon troubles, but because we haven’t been getting any shorties this year that would boost the NS count like in 2020/21/23.

[MarioProtVI]

I found a comparison between this time in 2023 vs. now using the NHC 7-day outlook graphic so I made my own. Quite a difference. After August 19 in 2023, there were 3 storms that formed in late August, 6 in September and 2 in October. That's 11 named storms after August 19th. This year, nothing. Unlikely we'll see any additional storms in August and early September is looking dead, too. We could reach the peak of the season (Sept 10) with only 5-6 named storms. I love it! Days off this time of year are a luxury.

I read Klotzbach's season update but I think something is being missed. Warm water - check. Low wind shear - maybe. As for a favorable environment - nope. Caribbean and Tropical Atlantic are dominated by dry, sinking air. It doesn't matter how warm the water is. If the atmosphere isn't favorable then hurricanes struggle to develop. I'm thinking my revised numbers of 18/8/4 may be too high. I'm just hoping for the current pattern to continue and any additional storms to stay out to sea. Perhaps those global models indicating no development are correct. Looking for an ensemble member somewhere to indicate development? Ha!

https://wxman57.com/images/NHC7Day.JPG

https://wxman57.com/images/NHC7Day.JPG

Tropical Atlantic Instability is way below normal:
https://www.ssd.noaa.gov/PS/TROP/TCFP/data/current/ts_al_tat_THDV.gif

This just seems like bait at this point. Each season is different and the switch flip comes around different dates, and your comment about ensembles seems a bit asinine considering the ensembles seem to be doing better at sniffing out genesis then the operational models which have explicitly shown to be having a poor performance with genesis (Beryl, Debby, Ernesto). Secondly, that THDV has been repeatedly proven since 2017 to be factually outdated and incorrect and a mere bait point to claim an inactive season or that forecasts will bust.

We have gone through this every single year where people start claiming bust around August 20 because models are dead until the first week of September, that something is wrong and/or is being missed and insist that for days on end. Then, right around September 1, activity explodes and models suddenly start latching onto the favorable conditions and most of the downcasters end up with egg on their face. I’ve never seen the downcasting be this high and part of that stems from the earlier season forecasts and people expecting either an August 2004 repeat or an Allen/Dean-type CV storm every August and if that doesn’t happen = bust or something is wrong.

Were the initial numbers a bit high? Certainly, but we have seen activity switch up fast and I see no reason that the Atlantic will only produce one meager storm between now and September 10 as you are claiming, especially with the base state in the Atlantic is in right now. That would be a 2022-level bust which is almost certainly not happening. The Atlantic is actually more moist then average and even moreso then August 2022 ever was, and the only reasons the Atlantic is quiet right now is due to the anomalous 200mb winds causing a very strong AEJ and easterly shear near Africa - which is expected to calm down per climatological reasons within the next week or so, and a CCKW currently in the EPac is expected to reach the Atlantic by that time too, also around the same time the currently-displaced-northwards monsoon trough begins to sag back down south.

I think the ~22-25 named storms is likely out of reach right now, but a more reasonable total of 18-20 / 10-11 / 4-6 and around 190-200 ACE seems far more likely, when accounting for how much we have left - we are already at 5/3/1 with 55 ACE, would need at least 14/7/4 and 155 ACE from here to November 30, which is definitely doable (for example if September is 7 (2022 for example managed 6 from Danielle to Ian so it’s achievable), October 4 (2019 managed 5 from Melissa to Rebekah and 2020 was a league of its own - achievable but maybe a bit more variable and depends on conditions) and November only 2-3 (also achievable). Additionally, 2017 finished 17/10/6 but just with a slight tweaking and slightly more favorable conditions towards and maybe a bit better handling by NHC, could’ve easily been 19/10/6 or even 20/10/6 (accounting for if POTC 10, the system near Florida in early October that looked like a TC formed and if Lee II was considered seperate from Lee I which it probably was but that’s besides the point) - and that season didn’t really take off until Irma on August 31 which models had struggled for a bit to lock onto in the days leading up to its formation. So, it’s achievable, but man we should be patient. I’d only consider something is seriously wrong if by like September 5 we’ve had only one more storm and it’s weak. But I just think the models are having a very hard time with the current conditions and perhaps extrapolate them going forward, and some model biases as well (GFS with its string out bias for example).

[IcyTundra]

I found a comparison between this time in 2023 vs. now using the NHC 7-day outlook graphic so I made my own. Quite a difference. After August 19 in 2023, there were 3 storms that formed in late August, 6 in September and 2 in October. That's 11 named storms after August 19th. This year, nothing. Unlikely we'll see any additional storms in August and early September is looking dead, too. We could reach the peak of the season (Sept 10) with only 5-6 named storms. I love it! Days off this time of year are a luxury.

I read Klotzbach's season update but I think something is being missed. Warm water - check. Low wind shear - maybe. As for a favorable environment - nope. Caribbean and Tropical Atlantic are dominated by dry, sinking air. It doesn't matter how warm the water is. If the atmosphere isn't favorable then hurricanes struggle to develop. I'm thinking my revised numbers of 18/8/4 may be too high. I'm just hoping for the current pattern to continue and any additional storms to stay out to sea. Perhaps those global models indicating no development are correct. Looking for an ensemble member somewhere to indicate development? Ha!

https://wxman57.com/images/NHC7Day.JPG

https://wxman57.com/images/NHC7Day.JPG

Tropical Atlantic Instability is way below normal:
https://www.ssd.noaa.gov/PS/TROP/TCFP/data/current/ts_al_tat_THDV.gif

Agreed, biggest forecasted Bust incoming.

Let’s wait until the season is over to declare the forecasts a bust or not.

[Category5Kaiju]

I found a comparison between this time in 2023 vs. now using the NHC 7-day outlook graphic so I made my own. Quite a difference. After August 19 in 2023, there were 3 storms that formed in late August, 6 in September and 2 in October. That's 11 named storms after August 19th. This year, nothing. Unlikely we'll see any additional storms in August and early September is looking dead, too. We could reach the peak of the season (Sept 10) with only 5-6 named storms. I love it! Days off this time of year are a luxury.

I read Klotzbach's season update but I think something is being missed. Warm water - check. Low wind shear - maybe. As for a favorable environment - nope. Caribbean and Tropical Atlantic are dominated by dry, sinking air. It doesn't matter how warm the water is. If the atmosphere isn't favorable then hurricanes struggle to develop. I'm thinking my revised numbers of 18/8/4 may be too high. I'm just hoping for the current pattern to continue and any additional storms to stay out to sea. Perhaps those global models indicating no development are correct. Looking for an ensemble member somewhere to indicate development? Ha!

https://wxman57.com/images/NHC7Day.JPG

https://wxman57.com/images/NHC7Day.JPG

Tropical Atlantic Instability is way below normal:
https://www.ssd.noaa.gov/PS/TROP/TCFP/data/current/ts_al_tat_THDV.gif

Agreed, biggest forecasted Bust incoming.

Marking this for November to see how well it ages.

[wxman57]

To be clear, I don't care about ACE, I'm looking at the number of storms. To reach 25 named storms we'd need to have 20 develop in September and October. That's highly unlikely. We are on track for 18 or fewer named storms, assuming we do reach a very active period the last three weeks of September. To have no disturbances and no likely development over the coming week this late into August is very rare. The environment across the tropics is clearly not very favorable for development. I suspect that will change by the second week of September as we move into the last half of the season (typically, numbers-wise). For now, I'm enjoying the quiet time and planning on taking another comp day this Friday and next Friday going into the Labor Day weekend. I think I have had only a single Labor Day weekend off since I started in 1980. That would be quite a treat.

[otowntiger]

To be clear, I don't care about ACE, I'm looking at the number of storms. To reach 25 named storms we'd need to have 20 develop in September and October. That's highly unlikely. We are on track for 18 or fewer named storms, assuming we do reach a very active period the last three weeks of September. To have no disturbances and no likely development over the coming week this late into August is very rare. The environment across the tropics is clearly not very favorable for development. I suspect that will change by the second week of September as we move into the last half of the season (typically, numbers-wise). For now, I'm enjoying the quiet time and planning on taking another comp day this Friday and next Friday going into the Labor Day weekend. I think I have had only a single Labor Day weekend off since I started in 1980. That would be quite a treat.

agreed and well said!

[CyclonicFury]

Something else that's important to consider: we've seen a lot of prominent Atlantic hurricanes form on short notice. Dorian and Fiona are two of the biggest examples. Dorian had practically no model support before it developed, but went on to become one of the strongest hurricanes in the basin on record by wind speed. Fiona also had very little model support in 2022, and there was speculation the Atlantic would have one of the least active seasons on record before it formed. A lot can change in this basin very quickly. It's easy to look at the operational models and assume the Atlantic will stay dead for the next two weeks, but I really don't believe that is a likely outcome. Once the AEJ weakens just a little, I think the MDR will come to life with significant activity. I don't think it will take until the middle of September to get another storm, in early September at the latest I think we'll see a substantial uptick.

[otowntiger]

To be clear, I don't care about ACE, I'm looking at the number of storms. To reach 25 named storms we'd need to have 20 develop in September and October. That's highly unlikely. We are on track for 18 or fewer named storms, assuming we do reach a very active period the last three weeks of September. To have no disturbances and no likely development over the coming week this late into August is very rare. The environment across the tropics is clearly not very favorable for development. I suspect that will change by the second week of September as we move into the last half of the season (typically, numbers-wise). For now, I'm enjoying the quiet time and planning on taking another comp day this Friday and next Friday going into the Labor Day weekend. I think I have had only a single Labor Day weekend off since I started in 1980. That would be quite a treat.

agreed and well said!

Especially the part about nothing happening this time of year being rare. I’d say that it is exceptionally rare to be occurring during what has been unanimously predicted to be a very very busy season.