Why is the 2024 Atlantic hurricane season underperforming expectations?

[MarioProtVI]

It remains to be seen how much 2024’s forecasts are off by, but one thing is for certain is we are likely not seeing any hyperactive forecasts (even in favorable years) for the foreseeable future based on possible losses of funding similar to 2013’s backlash. Perhaps for their sake we hope it’s not that bad of a bust, but we’re in the dark here at this point.

Honestly, if we're heading towards a future where every agency constantly predicts near-average or slightly above-average numbers -- either due to fear of losing funding or being called a "bust", or because seasons like 2022 and 2024 made them too scared about underestimating effects of indicators that we don't understand well, while overestimating effects of indicators that we do understand "well" (e.g. SSTAs, ENSO) -- that sounds not only dull, but more importantly, wildly unscientific and not beneficial for enhancing human knowledge of hurricane seasons in the long term.

When a surprise like 2024 happens, the most important task should be to study it, understand it, and apply it to the future so that we can predict more seasons (including hyperactive ones) with greater confidence. It shouldn't be to run away from it for fear or repercussions.

Edit to add: Think of it this way -- if we take it as granted that seasons with forecasted hyperactivity often end up busting, on the flip side, seasons that actually ended with top-tier ACE and/or impacts were rarely forecasted accurately in advance, either. Of the top 10 ACE seasons, 5 occurred in the era with professional forecasts:

• 1995: CSU August forecast 16/9/3, actual 19/11/5 227.1 ACE
• 1998: CSU August forecast 10/6/2, actual 14/10/3 181.8 ACE
• 2004: CSU August forecast 13/7/3, actual 15/9/6 226.9 ACE
• 2005: CSU August forecast 20/10/6, actual 28/15/7 245.3 ACE
• 2017: CSU August forecast 16/8/3 135 ACE, actual 17/10/6 224.9 ACE

Every single year outperformed CSU forecasts in all of NS, H and MH -- every single year, every single metric. For every single one of these seasons, CSU "missed" at least 2 NS, 2 H and 1 MH. The average misses were 3.6 NS, 3 H and 2 MH. And in the only such year after CSU started forecasting ACE, the ACE forecast ended too low by a whopping 90 (quite plausibly the same amount that 2024's ACE forecast would miss the mark by, but in the opposite direction).

It should be obvious that there's great value in being able to predict such 180+ ACE or 200+ ACE seasons in advance when they actually occur. Just because we haven't been successful at doing that (with false positives and false negatives in both directions) doesn't mean we should stop trying, or that seasonal forecasters should be discouraged from trying.

I should’ve added the other side of the coin there as well and you seemed to do that for me 180+ or 200+ ACE seasons have only seem to come when not really expected - kinda like how until Lee last year there was a pattern of C5 NHC forecasts not verifying because the storms peaked a bit lower as a high-end C4. I do hope that in the case we have a 200+ ACE season in the future that is forecasted to be that high, we have at least improved in detecting factors that while to the eye may look favorable but in reality actually unfavorable, as 2024 has demonstrated.

[Emmett_Brown]

2 important things from the CSU 2024 forecast seem to have played out;

1. Much of the Ace would occur in the western part of the basin
2. There would be an above normal chance for these storms to impact land areas

[tolakram]

How many times have I said in multiple threads that you can't predict an outlier? I think there is a bit of something involved when a significant number of agencies predict a record breaking season. If predictions were all AI, and actually intelligent, then the final prediction would be scored based on statistical likelihood of occurring, which I believe would be quite small. Outlier prediction is HARD. With a human involved I respectfully think that a forecast should have come with at least a disclaimer.

[CyclonicFury]

While there's still more than 2 months left in the season and I could see a significant storm or two develop (with the possibility of a CAG system coming up), I think it's clear by this point that this season can objectively classified as a "bust", at least in terms of raw numbers. 2024 was widely expected to be one of the most active seasons on record, and it's increasingly likely by the day to end up the least active since 2015. Since the bell ring day on August 20, the Atlantic has produced a measly 2 named storms, 1 hurricane and 6 ACE (11% of normal). Even 2013 had more activity during this time period. ACE production will likely remain very low for at least the next week or so, barring an unlikely Gordon comeback.

In some ways, this season's underperformance is more shocking than 2013, especially with the formation of the earliest Category 5 on record. At this point, I'll change the entire thread title.

[Sciencerocks]

 https://x.com/IPTCWCDirector/status/1835767165313773702

[Sciencerocks]

While there's still more than 2 months left in the season and I could see a significant storm or two develop (with the possibility of a CAG system coming up), I think it's clear by this point that this season can objectively classified as a "bust", at least in terms of raw numbers. 2024 was widely expected to be one of the most active seasons on record, and it's increasingly likely by the day to end up the least active since 2015. Since the bell ring day on August 20, the Atlantic has produced a measly 2 named storms, 1 hurricane and 6 ACE (11% of normal). Even 2013 had more activity during this time period. ACE production will likely remain very low for at least the next week or so, barring an unlikely Gordon comeback.

In some ways, this season's underperformance is more shocking than 2013, especially with the formation of the earliest Category 5 on record. At this point, I'll change the entire thread title.

It has produced quality and the gulf is in fact living up in many ways to the predicted favorability. aka 3 landfalling united states landfalls of hurricanes. Sure, a cat or 2 lower then expected in those hurricanes but the fact that they're waiting until the gulf is part of the reason behind that. We just aint developing long lasting systems like Ivan, Frances, Georges, Allen, Isabel, Ike, Gloria, 1938, etc... Because the MDR and caribbean after Beryl on the other hand is a complete and utter bust.

Reminds me of the difference between 2004 and 2005. 2005 had a poop mdr, Dennis and emily waited until they got into the caribbean to blow up, while the 2004 season was probably one of the top 3 cape verde season in the historic record for producing frances, ivan and charley. 2005 in my opinion had explosive development in the caribbean, gulf and western Atlantic above 20 north that made up for it, whereas this season is literally dead outside of some maybe slight above avg favorability in the gulf.

My opinion is everywhere outside of the gulf is literally capped in an inversion and can't take advantage of the record water temperatures. Just goes to show that sst's isn't everything and this season proves it.

[Category5Kaiju]

While there's still more than 2 months left in the season and I could see a significant storm or two develop (with the possibility of a CAG system coming up), I think it's clear by this point that this season can objectively classified as a "bust", at least in terms of raw numbers. 2024 was widely expected to be one of the most active seasons on record, and it's increasingly likely by the day to end up the least active since 2015. Since the bell ring day on August 20, the Atlantic has produced a measly 2 named storms, 1 hurricane and 6 ACE (11% of normal). Even 2013 had more activity during this time period. ACE production will likely remain very low for at least the next week or so, barring an unlikely Gordon comeback.

In some ways, this season's underperformance is more shocking than 2013, especially with the formation of the earliest Category 5 on record. At this point, I'll change the entire thread title.

I sincerely believe that Beryl and this season as a whole will be the topic of some pretty extensive studies. We learned from 2022. We learned from 2013. I guess 2024 will just be another unexpected bust year that we can learn from.

I think the concept that I’ve personally learned is that 2024 basically broke the idea that early action in the MDR, regardless of strength, is indicative of a fierce season later on. I sincerely believed that Beryl was a major harbinger of things to come. But clearly, something happened in the basin that completely smashed expectations. How the Atlantic was more favorable for activity in June/July than now is quite an intriguing thought

[WeatherBoy2000]

While there's still more than 2 months left in the season and I could see a significant storm or two develop (with the possibility of a CAG system coming up), I think it's clear by this point that this season can objectively classified as a "bust", at least in terms of raw numbers. 2024 was widely expected to be one of the most active seasons on record, and it's increasingly likely by the day to end up the least active since 2015. Since the bell ring day on August 20, the Atlantic has produced a measly 2 named storms, 1 hurricane and 6 ACE (11% of normal). Even 2013 had more activity during this time period. ACE production will likely remain very low for at least the next week or so, barring an unlikely Gordon comeback.

In some ways, this season's underperformance is more shocking than 2013, especially with the formation of the earliest Category 5 on record. At this point, I'll change the entire thread title.

I sincerely believe that Beryl and this season as a whole will be the topic of some pretty extensive studies. We learned from 2022. We learned from 2013. I guess 2024 will just be another unexpected bust year that we can learn from.

I think the concept that I’ve personally learned is that 2024 basically broke the idea that early action in the MDR, regardless of strength, is indicative of a fierce season later on. I sincerely believed that Beryl was a major harbinger of things to come. But clearly, something happened in the basin that completely smashed expectations. How the Atlantic was more favorable for activity in June/July than now is quite an intriguing thought

I disagree that Beryl completely destroyed the idea of early mdr activity as being a sign of an active season. The correlation between more active seasons and early mdr activity still exists, 2024/2013 are just notable outliers (Bonnie 2022 was not technically mdr storm despite the attempt). 2024 will likely go down as the one season where record warm ssts and -enso didn't automatically translate into at least an active season. Kinda like how last year was the one exception to the rule that strong el nino = inactive Atlantic season.

[Category5Kaiju]

While there's still more than 2 months left in the season and I could see a significant storm or two develop (with the possibility of a CAG system coming up), I think it's clear by this point that this season can objectively classified as a "bust", at least in terms of raw numbers. 2024 was widely expected to be one of the most active seasons on record, and it's increasingly likely by the day to end up the least active since 2015. Since the bell ring day on August 20, the Atlantic has produced a measly 2 named storms, 1 hurricane and 6 ACE (11% of normal). Even 2013 had more activity during this time period. ACE production will likely remain very low for at least the next week or so, barring an unlikely Gordon comeback.

In some ways, this season's underperformance is more shocking than 2013, especially with the formation of the earliest Category 5 on record. At this point, I'll change the entire thread title.

I sincerely believe that Beryl and this season as a whole will be the topic of some pretty extensive studies. We learned from 2022. We learned from 2013. I guess 2024 will just be another unexpected bust year that we can learn from.

I think the concept that I’ve personally learned is that 2024 basically broke the idea that early action in the MDR, regardless of strength, is indicative of a fierce season later on. I sincerely believed that Beryl was a major harbinger of things to come. But clearly, something happened in the basin that completely smashed expectations. How the Atlantic was more favorable for activity in June/July than now is quite an intriguing thought

I disagree that Beryl completely destroyed the idea of early mdr activity as being a sign of an active season. The correlation between more active seasons and early mdr activity still exists, 2024/2013 are just notable outliers (Bonnie 2022 was not technically mdr storm despite the attempt). 2024 will likely go down as the one season where record warm ssts and -enso didn't automatically translate into at least an active season. Kinda like how last year was the one exception to the rule that strong el nino = inactive Atlantic season.

Sorry, I should've phrased that better. What I meant to say was I, as a hurricane enthusiast, have been so used to the idea of early MDR action meaning very active season later on (especially with regards to Cape Verde hurricanes) that I basically took Beryl for granted as a lock for this season to be generating strong Cape Verde storm after strong Cape Verde storm later on. Clearly, what happened later this year with the dearth of such wasn't exactly something that my mind took as even remotely possible back in July.

[Teban54]

One thing for sure is that 2024 will add a whole lot of fuel to the fire during future "season cancel seasons".

In future years, regardless of what happened during earlier parts of the season, how favorable any indicators look like and what the experts forecasted, I'm sure people will be talking in mid-August about how the season would vastly underperform, especially if models also show nothing, despite the fact that such an outcome is statistically very unlikely. 2024 only strengthens this argument, even more so than 2013 (and perhaps 2022) did. Yes, such statements proved correct in 2024 (at least so far), but in the overwhelming majority of years they would eat crow.

[Teban54]

Reminds me of the difference between 2004 and 2005. 2005 had a poop mdr, Dennis and emily waited until they got into the caribbean to blow up, while the 2004 season was probably one of the top 3 cape verde season in the historic record for producing frances, ivan and charley. 2005 in my opinion had explosive development in the caribbean, gulf and western Atlantic above 20 north that made up for it, whereas this season is literally dead outside of some maybe slight above avg favorability in the gulf.

An interesting thing with the two bolded parts that you said. Despite a very active early Caribbean season in 2005 with Dennis and Emily, the Caribbean then went onto a long hiatus during peak season. No TCs existed in the Caribbean between July 18 and October 1, 2005; and this streak was only broken by Stan, which was a minimal TS for less than 6 hours just before its Yucatan landfall.

The next TC in the Caribbean formed on October 15, after almost three months of minimal or no activities in the Caribbean... And that was Wilma. Three more named storms formed in the Caribbean after that, including another major (Beta).

[Ptarmigan]

This season has almost an almost 1989 feel to it. Predictions for this year were for an extremely active season. However if you look at 1989, the bulk of the storms were mid-September on with Hurricane Hugo in mid to late September being quite the beast. Early fall fronts, followed by tranquil weather for a few weeks until activity fired back up. The MJO also kicks up mid-September which will positively influence convection the remainder of the Fall.

1989 was quite backloaded. Texas got hit by Hurricane Chantal and Jerry in 1989. Jerry is one of the latest hurricane to make landfall on Texas. A hurricane hit Galveston in November of 1839.

https://www.wpc.ncep.noaa.gov/research/txhur.pdf

[Ptarmigan]

Speaking of baffling, per a recent Phil Klotzbach tweet, in the last 123 years (since 1900) only eight Atlantic Basin seasons have seen 3 or more US landfalling hurricanes by the date of 9/11
1915, 1916, 1933, 1934, 1985, 2004, 2005, 2020 and now 2024 makes 9 seasons.

Check out the back to back years that show up not once, not twice, BUT THREE TIMES in this small number of instances... now THAT's baffling So this season may not reach pre-season forecasted storm count or ACE totals (we shall see) but the word BUST is getting a little smaller and a little less bold as we move on here ..

So far, the forecast 2024 is a bust. In terms of landfall, 2024 has been active in terms of landfall. That by itself is an amazing feat.

[aspen]

One thing for sure is that 2024 will add a whole lot of fuel to the fire during future "season cancel seasons".

In future years, regardless of what happened during earlier parts of the season, how favorable any indicators look like and what the experts forecasted, I'm sure people will be talking in mid-August about how the season would vastly underperform, especially if models also show nothing, despite the fact that such an outcome is statistically very unlikely. 2024 only strengthens this argument, even more so than 2013 (and perhaps 2022) did. Yes, such statements proved correct in 2024 (at least so far), but in the overwhelming majority of years they would eat crow.

Yeah after this year, people are going to be a lot more skeptical of seemingly favorable indicators and claims of an upcoming switch flip. 2024 more than any other year showed that you can have some unexpected suppressive factors come in at the last minute (the +NAO didn’t start developing until Beryl was dissipating IIRC) and shut down a season that may have very favorable “typical” indicators. Plus, it was the one year that the July-into-August lull we’ve been seeing for the last few seasons was actually a sign of a bust.

Also to add onto the MDR hurricane indicator discussion: in most instances, something like Beryl would be a very strong indicator of an active season. Even a weak TS in the MDR is enough. From 1980 to 2021, a total of 13 seasons (1989, 1990, 1995, 1996, 1998, 2003, 2005, 2008, 2013, 2017, 2018, 2020, 2021) saw at least one NS formation in the MDR or Caribbean south of 20N in July, which averaged out to 18/9/4 and 161 ACE. Changing to criteria to limit it to July systems formed from AEWs anywhere in the basin or AEW/MDR hurricanes still yields a mean of 17-18 NS and ~160 ACE. Meanwhile, seasons with July development from non-AEW sourced averaged out closer to the 1991-2020 climo mean (~13-14 NS and ~110-120 ACE). Therefore, getting a 35 ACE Cat 5 MDR/Caribbean hurricane in July would have been an indicator for an active season…if it wasn’t for the +NAO and hyperactive AEJ. So this all just makes Beryl’s existence even more baffling.

[ConvergenceZone]

I really think they NHC should issue a statement at the end of the season as to what went wrong with their crazy prediction? I know I've seen theories, but I always take them with a grain of salt unless they come directly from the NHC. I think they should explain why the season was a bust(number wise).

[Hammy]

I really think they NHC should issue a statement at the end of the season as to what went wrong with their crazy prediction? I know I've seen theories, but I always take them with a grain of salt unless they come directly from the NHC. I think they should explain why the season was a bust(number wise).

NHC does not predict seasonal activity, they do nothing beyond the 5-7 day period. But they do usually issue a season summary. Here's 2013's for instance https://www.nhc.noaa.gov/data/tcr/summary_atlc_2013.pdf

[al78]

Many seasonal forecasts are generated from statistical models which are trained on data and teleconnections going back many decades. The underlying assumption, aside from relationships being linear, is that the hurricane season will respond to large scale parameters as it has done so over the last 70 or so years. If a year like 2024 comes along which sticks two fingers up at the large scale teleconnections, the statistical forecasts will bust, but it is also assumed that such years are very rare (if they weren't then statistical relationships would either pick up years like this or the correlations would be much weaker than they are).

[ConvergenceZone]

I really think they NHC should issue a statement at the end of the season as to what went wrong with their crazy prediction? I know I've seen theories, but I always take them with a grain of salt unless they come directly from the NHC. I think they should explain why the season was a bust(number wise).

NHC does not predict seasonal activity, they do nothing beyond the 5-7 day period. But they do usually issue a season summary. Here's 2013's for instance https://www.nhc.noaa.gov/data/tcr/summary_atlc_2013.pdf

Hammy, you are right, I was confusing them with NOAA. Then NOAA should be the ones issuing the statement as to what went wrong and why the predictive models got that produced those insane numbers got it so wrong this year.

[Cyclenall]

After Beryl there were indications that July would be quiet for 3 weeks so right there was the first big strike, there were comparisons starting with 1933/2005 during Beryl but once mid-July was shown to not produce any sauce I knew it wasn't going to be in the historic tier (a Top 5 season out of contention). I still thought a 2004/2010/2017 type season was possible but then once Ernesto came along with its sprawl and sputtering nature I suspected a less impressive show. By early Sep its clear this will be a historic bust for number of named storms. I was hoping to just see a short lived TD in mid-Sep then have the basin just shut down for the giggles. Each TC starting with Debby was more underwhelming compared to expectations. I think this may be the most bizarre hurricane season on record.

[Hypercane_Kyle]

I remember after 2013 it was thought that such an underperforming season would be once-in-a-lifetime anomalies. Now we have a second, potentially even more historic underperformance (counting the chickens before they're hatched, I know) it's clear there's a mechanism that is impacting hurricane seasons we simply do not understand and models aren't able to capture correctly.