2024 Atlantic Hurricane Season

[Seminole]

Some interesting snippets regarding the 2022 Tongo volcano and how it may be related to what is going on with the ITCZ and Hadley Cell. Solar Activity also cited as a possible culprit. 

https://www.zerohedge.com/weather/atlantic-tropics-are-completely-broken  

[GaWx]

 

[GaWx]

COLORADO STATE UNIVERSITY FORECAST OF ATLANTIC HURRICANE ACTIVITY FROM SEPTEMBER 3–16, 2024
We believe that the most likely category for Atlantic hurricane activity in the next two weeks is below-normal (60%), with near-normal (30%) and above-normal (10%) being less likely.
(as of 3 September 2024)
 

https://tropical.colostate.edu/Forecast/2024-0903.pdf

[Retrobuc]

I'm a lurker, but I do financial modeling/forecasting for a living.  Since every model called for a hyperactive season, how much do you feel groupthink plays into this?  Meterology has become a business and it may be safer to stay with the pack than it is to venture out on your own.  Its easier to be wrong with the group than to be wrong as an outlier.  Interested to know what people here think.

For the record, I said this season would be 13/8/2 in the contest.  I feel pretty good about those numbers.

[Windspeed]

I'm a lurker, but I do financial modeling/forecasting for a living.  Since every model called for a hyperactive season, how much do you feel groupthink plays into this?  Meterology has become a business and it may be safer to stay with the pack than it is to venture out on your own.  Its easier to be wrong with the group than to be wrong as an outlier.  Interested to know what people here think.
For the record, I said this season would be 13/8/2 in the contest.  I feel pretty good about those numbers.

Better yet, what was your reasoning in only going with 13 named storms in the face of overwhelming model and climate parameter support for a hyperactive season? I mean, you're sharing your numbers while knowing full well the majority of the best TC climatologists on the planet are scrambling for answers. 13 named storms is barely average.

[Seminole]

18 minutes ago, Retrobuc said:

I'm a lurker, but I do financial modeling/forecasting for a living.  Since every model called for a hyperactive season, how much do you feel groupthink plays into this?  Meterology has become a business and it may be safer to stay with the pack than it is to venture out on your own.  Its easier to be wrong with the group than to be wrong as an outlier.  Interested to know what people here think.

For the record, I said this season would be 13/8/2 in the contest.  I feel pretty good about those numbers.

Factors like solar activity and volcanic activity are not data points used in the models. Both impact weather. 

[Retrobuc]

25 minutes ago, Windspeed said:

46 minutes ago, Retrobuc said:

I'm a lurker, but I do financial modeling/forecasting for a living.  Since every model called for a hyperactive season, how much do you feel groupthink plays into this?  Meterology has become a business and it may be safer to stay with the pack than it is to venture out on your own.  Its easier to be wrong with the group than to be wrong as an outlier.  Interested to know what people here think.
For the record, I said this season would be 13/8/2 in the contest.  I feel pretty good about those numbers.

Better yet, what was your reasoning in only going with 13 named storms in the face of overwhelming model and climate parameter support for a hyperactive season? I mean, you're sharing your numbers while knowing full well the majority of the best TC climatologists on the planet are scrambling for answers. 13 named storms is barely average.

I went with those numbers based on another post that showed deviation from the previous high forecast calls.  I took that as a baseline and then discounted from there simply using a calculation, not one bit of empirical weather data. 

[WEATHER53]

51 minutes ago, Retrobuc said:

I'm a lurker, but I do financial modeling/forecasting for a living.  Since every model called for a hyperactive season, how much do you feel groupthink plays into this?  Meterology has become a business and it may be safer to stay with the pack than it is to venture out on your own.  Its easier to be wrong with the group than to be wrong as an outlier.  Interested to know what people here think.

For the record, I said this season would be 13/8/2 in the contest.  I feel pretty good about those numbers.

Models are worse than 20  years ago in the winter in mid Atlantic largely due to attempting to be too precise and offering up every possible outcome.

Many of the indexes don’t work anymore in the changing environment and many newish  indexes are still erratic and unproven 

[Windspeed]

I went with those numbers based on another post that showed deviation from the previous high forecast calls.  I took that as a baseline and then discounted from there simply using a calculation, not one bit of empirical weather data. 

If I put you on the spot, well, you just made a post inquiring pros and cons about stepping away from the pack. So without empirical datum, you just played devils advocate and went with average numbers knowing if a 20-25% failure rate of climate precursors resolves, you are sailing to victory and looking quite brilliant and bold. lol... I appreciate the honesty versus jumping on the WAM strength/OS Index/Tonga (*barf*)/+NAO etc., etc... things we are trying to resolve in situ.

I should also point out that you can still have a hyperactive season and only have 13 named storms based on metrics being ACE. Granted, that means the insanity of nearly all TCs being long-lived hurricanes or powerful majors. At any rate, I digress. I am pretty sure we know what you meant by hyperactive.

[Retrobuc]

1 minute ago, Windspeed said:

If I put you on the spot, well, you just made a post inquiring pros and cons about stepping away from the pack. So without empirical datum, you just played devils advocate and went with average numbers knowing a 20-25% failure of climate precursors, and you are sailing to victory and looking quite brilliant and bold. lol... I appreciate the honesty versus jumping on the WAM strength/OS Index/Tonga/+NAO etc., etc... things we are trying to resolve in situ.

I should also point out that you can still have a hyperactive season and only have 13 named storms based on metrics being ACE. Granted, that means the insanity of nearly all TCs being long-lived hurricanes or powerful majors. At any rate, I digress. I am pretty sure we know what you meant by hyperactive.

I'm not a weather forecaster and dont pretend to be.  I follow this page because it provides interesting information that you cant get elsewhere.  So, my method (naïve forecast)  may not be something supported by Colorado State or an official agency, but it shows that trends and mathematical data can be just as useful as maps and analogs.  I did not play devils advocate, I felt that everyone being 50% higher than the highest ever peak was far too bullish.  Looking back at history, previous bullish calls missed by 25-30%.  Therefore, I went with the lower baseline and discounted slightly more.  Dont get me wrong, I could care less whether I am right or wrong.  I am more interested in methodology and how pros use this to correct their modeling.

[Windspeed]

I'm not a weather forecaster and dont pretend to be.  I follow this page because it provides interesting information that you cant get elsewhere.  So, my method (naïve forecast)  may not be something supported by Colorado State or an official agency, but it shows that trends and mathematical data can be just as useful as maps and analogs.  I did not play devils advocate, I felt that everyone being 50% higher than the highest ever peak was far too bullish.  Looking back at history, previous bullish calls missed by 25-30%.  Therefore, I went with the lower baseline and discounted slightly more.  Dont get me wrong, I could care less whether I am right or wrong.  I am more interested in methodology and how pros use this to correct their modeling.

Fair enough. I think we're all just a bit stressed trying to resolve what is going on versus the 20-25% failure mode. What factors are essentially countering other overwhelming favorable data points.

Additionally, all hell could break loose in two weeks. We get complacency. Peak merely got delayed until mid-to-late September this year when all these countering factors subside.

[GaWx]

20 minutes ago, Retrobuc said:

I'm not a weather forecaster and dont pretend to be.  I follow this page because it provides interesting information that you cant get elsewhere.  So, my method (naïve forecast)  may not be something supported by Colorado State or an official agency, but it shows that trends and mathematical data can be just as useful as maps and analogs.  I did not play devils advocate, I felt that everyone being 50% higher than the highest ever peak was far too bullish.  Looking back at history, previous bullish calls missed by 25-30%.  Therefore, I went with the lower baseline and discounted slightly more.  Dont get me wrong, I could care less whether I am right or wrong.  I am more interested in methodology and how pros use this to correct their modeling.

In April CSU predicted ACE of 210. Their prior highest April progs were 160-183 (five years). Of these five, all progs ended up too high with even the closest being 34 too high! (That day I predicted 176 ACE based on it being 34 under 210.) The five averaged a whopping 85 too high for ACE! So, if their April of 2024 forecasted ACE ends up 85 too high, it will end up at 125.

[Normandy]

The all hell breaking loose will come.  But numbers into the 20s is gone.  The season can’t produce enough quantity.  However I’m very nervous as there still is a lot of energy I’m the tropics that have yet to be evacuated, and eventually the storms will have to form to take care of that.  Once they form they could be absolutely explosive.  Got to wait and see but this is hella interesting 

[GaWx]

1 minute ago, Normandy said:

The all hell breaking loose will come.  But numbers into the 20s is gone.  The season can’t produce enough quantity.  However I’m very nervous as there still is a lot of energy I’m the tropics that have yet to be evacuated, and eventually the storms will have to form to take care of that.  Once they form they could be absolutely explosive.  Got to wait and see but this is hella interesting 

The Euro Weeklies sort of agree. More later.

[WxWatcher007]

I’m not sure if it’s just a delayed signal that’ll get washed out on the EPS, but 12z was quite robust in the Gulf with the current Caribbean disturbance. Seems to pull it north or NE eventually too with the eastern U.S. trough next week. 

[CurlyHeadBarrett]

59 minutes ago, Windspeed said:

If I put you on the spot, well, you just made a post inquiring pros and cons about stepping away from the pack. So without empirical datum, you just played devils advocate and went with average numbers knowing if a 20-25% failure rate of climate precursors resolves, you are sailing to victory and looking quite brilliant and bold. lol... I appreciate the honesty versus jumping on the WAM strength/OS Index/Tonga (*barf*)/+NAO etc., etc... things we are trying to resolve in situ.

I should also point out that you can still have a hyperactive season and only have 13 named storms based on metrics being ACE. Granted, that means the insanity of nearly all TCs being long-lived hurricanes or powerful majors. At any rate, I digress. I am pretty sure we know what you meant by hyperactive.

Like 1996

[Kevin Reilly]

1 hour ago, Normandy said:

The all hell breaking loose will come.  But numbers into the 20s is gone.  The season can’t produce enough quantity.  However I’m very nervous as there still is a lot of energy I’m the tropics that have yet to be evacuated, and eventually the storms will have to form to take care of that.  Once they form they could be absolutely explosive.  Got to wait and see but this is hella interesting 

Oh it will December to March Mid Atlantic and Northeast winter season is going to be historic!!

[salbers]

FYI - both Spire's Hi-res model and the HRRR do some spinup of the Gulf low during the next 48 hours, keeping it more to the north. ICON brings it more to the south like yesterday's run.

[cptcatz]

1 hour ago, GaWx said:

The Euro Weeklies sort of agree. More later.

I'm not sure I trust anything anymore. Here's a post from 2.5 weeks ago calling for a good chance of an MDR storm just two weeks down the line (lining up to the past few days).

That being said, the wave in the caribbean is looking quite juicy right now so we'll see what happens. 

[GaWx]

 Today’s Euro Weeklies added a new week, Sep 30-Oct 6, and that week is quite active vs climo. In addition the week before is also progged to be active as has been the case on many runs. 

 So this run suggests that activity will pick up markedly next week though would still be below avg. After a similar subsequent week, activity is suggested to pick up substantially in late Sep and continue into early Oct vs climo:

9/3/24 Euro Weeklies (EW) mean ACE projections as % of climo:
9/2-8: 20% per yest.’s run (meaning ACE of 3) 
9/9-15: 60% (meaning ACE of 10)(climo peak week)
9/16-22: 70% (meaning ACE of 11)
9/23-29: 130% (meaning ACE of 17)(vs peak wk climo of 16)
9/30-10/6: 140% (new week)(meaning ACE of 13)

 So, today’s EW progs by week are 3-10-11-17-13 meaning two most active weeks are at the end despite climo having dropped off substantially from peak by then. That would get 2024 to ~110 ACE as of Oct 6.

[CurlyHeadBarrett]

11 minutes ago, GaWx said:

 Today’s Euro Weeklies added a new week, Sep 30-Oct 6, and that week is quite active vs climo. In addition the week before is also progged to be active as has been the case on many runs. 

 So this run suggests that activity will pick up markedly next week though would still be below avg. After a similar subsequent week, activity is suggested to pick up substantially in late Sep and continue into early Oct vs climo:

9/3/24 Euro Weeklies (EW) mean ACE projections as % of climo:
9/2-8: 20% per yest.’s run (meaning ACE of 3) 
9/9-15: 60% (meaning ACE of 10)(climo peak week)
9/16-22: 70% (meaning ACE of 11)
9/23-29: 130% (meaning ACE of 17)(vs peak wk climo of 16)
9/30-10/6: 140% (new week)(meaning ACE of 13)

 So, today’s EW progs by week are 3-10-11-17-13 meaning two most active weeks are at the end despite climo having dropped off substantially from peak by then. That would get 2024 to ~110 ACE as of Oct 6.

I want to believe it, but my head is saying no.

[GaWx]

Just now, CurlyHeadBarrett said:

I want to believe it, but my head is saying no.

 If it doesn’t get more active that would be fine with me. And of course it may not. But the Euro weeklies have done well so far this season. They had very high activity in late June and early July. Also, they did very well with the recent quiet period. They’ve been calling for a substantial increase in late Sep for a full week. Of course even if it gets more active they could still mostly avoid land.

[CurlyHeadBarrett]

9 minutes ago, GaWx said:

 If it doesn’t get more active that would be fine with me. And of course it may not. But the Euro weeklies have done well so far this season. They had very high activity in late June and early July. Also, they did very well with the recent quiet period. They’ve been calling for a substantial increase in late Sep for a full week. Of course even if it gets more active they could still mostly avoid land.

 

So in other words, a 2022/2016 redux

 

[GaWx]

39 minutes ago, CurlyHeadBarrett said:

So in other words, a 2022/2016 redux

 

The Euro Weeklies’ ACE climo base is 2004-2023 (running 20 year). What I discovered is interesting for 2004-23 vs 1991-2020 climo for ACE:

- 8/26-9/1: decreased ~1/8

- 9/2-9/8: little change (tiny increase)

- 9/9-15: little change/still peak week

- 9/16-9/22: increased ~1/7; now as active as 9/2-8; old climo was 1/10 less active than 9/2-8

- 9/23-29: increased 28%; slightly more active than 8/26-9/1; used to be just over 1/4 less active than 8/26-9/1

 - 9/30-10/6: increased 1/4

 So, peak day for ACE appears to be ~2 days later. With a drop in late Aug and a rise last 1/2 of Sep through early Oct, the season is significantly more backloaded ACEwise 2004-23 vs 1991-2020. Total seasonal ACE is 131 vs 122.

[Normandy]

Alright let’s bRIng in some positive energy and focus on the now.  Our wave in the Caribbean is at least firing convection and showing signs of life.  If we get a good d max, maybe we can finally get something to pop.  It’s got a small window before the Yucatán but let’s see if it can take advantage

[TheDreamTraveler]

Definitely one of the more interesting seasons we've had in a long time. A record breaking category 5 at the beginning of July that smashed basically every single record known to us and then...nothing. Even into the peak of the season where we were forecast to have a very hyperactive season. I'd love to know all the ins and outs in depth of why this is happening. I've read some thoughts on it but I'm sure there's more things. I swear I see people say "oh the waters are warm its gonna be active!!" even though you can have a dead season with record warm waters if other factors don't line up. The atlantic hurricane mechanisms are very deep and varied.

[CheeselandSkies]

[WxWatcher007]

This is a worthwhile read. That massive area of subsidence over the Gulf and western Atlantic may also be why the western half of the basin has been bereft of activity despite several opportunities—as we’re even seeing now with the Gulf disturbance we’ve been casually watching.

 

[Windspeed]

Klotzbach weighs in again.

[Seminole]

15 hours ago, GaWx said:

In April CSU predicted ACE of 210. Their prior highest April progs were 160-183 (five years). Of these five, all progs ended up too high with even the closest being 34 too high! (That day I predicted 176 ACE based on it being 34 under 210.) The five averaged a whopping 85 too high for ACE! So, if their April of 2024 forecasted ACE ends up 85 too high, it will end up at 125.

How in the world do CSU have any credibility at all with that poor a track record? They consistently over estimate and the media hypes their over estimate every year.