GaWx

*Reposted to correct thread* The ACE calculations are in for the period 9/30-10/13 based on best track data: -2024 had a whopping 63.6, 3rd highest since 1851! This compares to 1991-2020 avg of only 12.3! Since 1851: -1893 had ~67 -2016 had ~65.5 -2024 had 63.6 -1878 was just below at ~59-60 Breaking it down into weeks back to 1951: 9/30-10/6: 1991-2020 avg 7.7 ~47.5 in 2016 32 in 2024 ~30 in 2015 So, 2024 had 2nd highest 9/30-10/6 ACE back to 1951 only behind 2016 and just ahead of 2015. 10/7-13: 1991-2020 avg 4.6 31.6 in 2024 24 in 2018 ~18 in 2016 ~11.5 in 2017 11 in 1995 10 in 2010, 2001 8 in 2020, 2014 So, 2024 by a large margin had the highest 10/7-13 ACE since 1851! Kudos to Euro Weeklies predicting weeks in advance significantly greater than avg ACE 9/30-10/13 (delay of peak by 4 weeks)!

*Reposted to correct thread:* The busiest week on today’s Euro Weeklies for ACE is unusually enough for this late, week 4: Nov 4-10. It projects as per the image below 2.7 times the 2004-23 mean ACE, which is 2. So, the prog is for 5-6 ACE for then. The prog for each of the two weeks prior is for ~5 each. So, 15-16 is progged for the 3 week period that’s often pretty quiet, 10/21-11/10. But with Nina and very warm Caribbean/Gulf, it would make sense. The most active areas per the run for 10/21-11/10 go from SW Gulf through W Car through far SE FL, and Bahamas/Bermuda. 11/4-10: 2.7 x mean 2004-23 ACE 11/4-10: most active areas are shaded

*Fwiw* JB keeps hinting at a cold Nov/Thanksgiving/Dec for the E US. I’m not putting much weight on that for obvious reasons. Also, on Oct 8th he ditched his +2 for Megapolis for Nov-Mar for a NN (+1). And that’s obviously subject to further cooling in later updates. Keep this in mind: Since at the very least 2014-5 (because they’re all circulating on the internet back to then), JB has yet to have a mild FINAL Megapolis winter forecast. With the number of mild winters (5 of the 10) there have been, that’s not a good record. The ones 3+ AN at NYC since 2014-5 for those who don’t know were 2015-6, 2016-7, 2019-20, 2022-3, and 2023-4. A. The 5 mild NYC winters (vs respective 30 yr mean): -JB had -1 for Nov-Mar 2023-4 vs actual of +4 (DJF N/A). So, he missed by -5 -He had +0.5 for DJF 2022-3 vs actual of +5. So, he missed by -4.5 -He had -2 for his final DJF 2019-20 (vs earlier fcast of +0.5) vs actual of +3.5. So, his final missed by -5.5. So, his earlier fcast was not as bad (miss by -3). -He had -2 for his final DJF 2016-7 (vs earlier fcast of 0) vs actual of +4. So, his final missed by -6. So, his earlier fcast wasn’t as bad (miss by -4). -He had -2 for both his initial and final fcasts for DJF 2015-6 vs actual of +5.5. So, he missed by -7.5. So, for the 5 mild NYC winters: -The avg of the absolute value of his final fcast misses was 5.5 and his avg bias was -5.5. The range was -4.5 to -7.5. -He revised 2 of the 5 significantly colder, which resulted in larger - misses. ————————— B. The 5 non-mild NYC winters: -2021-2 had -0.5 for final Nov-Mar (vs 0 for initial). The actual was +1 (DJF N/A). So, his final missed by -1.5 and his earlier fcast was slightly better (missed by -1.0). -2020-1 had +0.5 for his final DJF (vs +3 for his earlier). The actual was +0.5. So, his final was perfect! Also, his earlier fcast wasn’t so good (missed by +2.5) meaning his change was great! -2018-9 had -2 for his final/initial DJF vs actual of +0.5. So, he missed by -2.5. -2017-8 had 0 for his final DJF (vs +0.5 for initial). The actual was +0.5. So, his final missed by only -0.5. His earlier fcast was perfect. -2014-5 had -2.5 for his final DJF (vs -3 for initial). The actual was -4. So, his final missed by +1.5. This is the only one of the 10 that was actually cold, for which he missed too warm, and for which he revised warmer. This also was his coldest fcast. His earlier fcast was slightly better (miss by +1). So, for the 5 non-mild NYC winters: -The avg of the absolute value of his final fcast misses was 1. The avg bias was -0.5. The range: -2.5 to +1.5. -He revised 4 of the 5 (3 cooler and 1 warmer). ——————————— C. Summary of all of last 10 winters at NYC: -He did well when winter wasn’t mild as his bias for those was only -0.5 and avg absolute value of misses was only 1. His coldest fcast was for the coldest winter. -He did poorly when winter was mild with a bias/avg miss of -5.5 (avg absolute value of misses was 5.5). He didn’t predict a single mild winter. He was able to predict well the one cold winter. He’d be a much better forecaster if he’d improve on forecasting mild winters but that’s hard to do when he doesn’t forecast them. That’s why I route for him to predict mild and stick with it. -None of his 10 final forecasts were mild (warmest +0.5/NN) (5 were NN and 5 were BN) as they were from warmest to coldest: +0.5, +0.5, 0, -0.5, -1, -2, -2, -2, -2, -2.5 That means that IF his final fcast of +1 for 24-5 doesn’t change, it would still be the warmest since at least 2014-5 despite the cooler revision from +2 to +1. -6 of the 10 winters were +1 or warmer. The avg of the winters was +2 vs his avg of final fcasts of -1. So, he was too cold by 3 on avg for the 10 winters! So, a bias correction of +3 would probably be appropriate. -He predicted on avg -1.3 for the 5 warmest winters and -0.9 for the 5 coldest winters -He had revisions for 6 of the 10. Of these 6, 5 resulted in a worse forecast and only one a better forecast. Of these 6, 5 were revised cooler just like the one he just did for 2024-5. Only 2014-5 was revised slightly warmer. @bluewave @donsutherland1

Like clockwork and even earlier than I expected, predictable JB has just cooled his Nov-Mar forecast. Warning: this may just be a baby step on the way to more cooling…we’ll see. In 2020-21, he also had an initially mild E US that he drastically cooled in November. He has since at least 2014-5 not had a mild FINAL cold season forecast as 20-21’s was changed to NN (which in all fairness verified pretty well). Old one (8/22/24): New one (10/8/24): ~1F cooler from +2 to +1 in NE Megapolis, which isn’t trivial as +1 is NN to me meaning he’s ditched his somewhat mild NE fcast: @donsutherland1

1. Not officially because that (on ONI basis) is done retrospectively since several months of SST anomalies have to be -0.5C or colder in Nino 3.4 and it has just gotten there in recent weeks. 2. I actually prefer to look at it on a Relative ONI basis (RONI) because that much better takes into account recent warming of global tropical waters. Based on that, it entered La Niña territory nearly 3 months ago.

94L on 12Z UKMET: keeps it near steady state through entire run, which ends with it moving slowly (<10 mph) WNW well N of the Greater Antilles (I’d have to see H5 maps to determine if that would likely recurve safely from the US):MET OFFICE TROPICAL CYCLONE GUIDANCE FOR NORTH-EAST PACIFIC AND ATLANTICGLOBAL MODEL DATA TIME 1200UTC 13.10.2024TROPICAL DEPRESSION 94L ANALYSED POSITION : 17.9N 33.8WATCF IDENTIFIER : AL942024LEAD CENTRAL MAXIMUM WINDVERIFYING TIME TIME POSITION PRESSURE (MB) SPEED (KNOTS)-------------- ---- -------- ------------- -------------1200UTC 13.10.2024 0 17.9N 33.8W 1013 250000UTC 14.10.2024 12 17.6N 36.6W 1013 241200UTC 14.10.2024 24 17.2N 38.6W 1013 230000UTC 15.10.2024 36 17.0N 40.3W 1012 231200UTC 15.10.2024 48 17.0N 42.4W 1011 250000UTC 16.10.2024 60 17.0N 45.2W 1011 261200UTC 16.10.2024 72 17.5N 48.3W 1011 290000UTC 17.10.2024 84 18.5N 51.8W 1011 321200UTC 17.10.2024 96 20.6N 55.5W 1011 330000UTC 18.10.2024 108 22.0N 58.1W 1011 301200UTC 18.10.2024 120 23.2N 60.3W 1013 290000UTC 19.10.2024 132 23.9N 62.3W 1013 251200UTC 19.10.2024 144 24.0N 64.5W 1012 300000UTC 20.10.2024 156 24.2N 66.0W 1011 311200UTC 20.10.2024 168 24.7N 67.5W 1009 29

Note that the Euro doesn’t even have the cold Canada like some others. Actually, the E half is downright mild (for Canada)!

Except that normals have warmed so much that modestly BN isn’t all that cold anymore. And then wait til we get to 2001-30 normals!

Only 6 of the last 45 winters averaged -NAO (sub -0.25): 1984-5, 1986-7, 1995-6, 2009-10, 2010-1, and 2020-1. All 6 had averaged sunspots under 35 and were within about 2 years of solar cycle minimums. Coincidence? Some feel there is a partial negative correlation of winter -NAO and winter sunspots. I’m leaning that way.

Despite things tending to calm down starting in late Oct, I fully expect more storms due to La Niña and the very warm Atlantic. I just hope that they’re not nearly as impactful especially in the CONUS as the westerlies come further south per climo. We’ll see. Impacts on the US generally go down late in Oct despite exceptions like Wilma and Sandy. GFS does tend to have a lot of phantoms in the long range in the W Caribbean/Gulf, regardless. So, we’ll have to see about those as I do expect more activity there.

Hopefully no more hits this season! Whereas nobody knows what will happen, climo is our ally now as usually things calm down overall a lot past mid Oct. So, fingers crossed, especially since average ACE has increased for late season.

This is well known. The first one that was seeded was one in 1947 that was moving NE OTS, meaning a seemingly safe one to seed. However, shortly afterward, it turned sharply (135 degrees to a westerly heading) and subsequently hit Savannah. There’s the possibility that the seeding weakened it enough to lower the mean steering level, which conceivably could have helped lead to more influence from a high to the north. I believe that that sharp turn and subsequent landfall lead to a stoppage of that program. But I feel that bringing up these past experiments now is a red herring because they’re not occurring now. Marjorie T. Greene lied about this for pure political purposes, of course. She decided to just make up that “they” (meaning Democrats) were controlling Helene. The best thing would be for the media to ignore her.

Maybe the model average but definitely not all of the major globals had a too far left bias with these three. This was especially the case with UKMET. UKMET was: -best with Ian with it near actual landfall point well to the SE several days in advance. I documented this in detail in the Ian thread for future reference. The GFS and CMC were horrible with them way too far left, including some Panhandle landfalls just a couple of days out! Euro was also too far left but not by nearly as much. Icon was 2nd best to UKMET with a smaller too far left miss overall. - UKMET a good bit too far right for Milton (along with the CMC) a few days in advance. GFS overall missed too far left about the most of the globals. I need to go back and analyze more closely but my gut feel is that Euro may have done best overall with perhaps Icon in 2nd - UKMET and Euro did well overall with Irma in advance, better than GFS, as I clearly recall

I follow you as I was thinking about that myself, but this was all I had available (and also is an amazing record, regardless). The total for 10/9-10 at St. Petersburg was 18.85”. So, I also dug further into the area’s monthly records for the heaviest events listed in that other link’s heaviest daily records from past years back to 1916 to try to get a feel for whether or not adjacent days would have added enough for a full event’s total to reach Milton’s 18.85”. Based on that extra digging I concluded there’s a very high likelihood that no full event rainfall reached that. The closest event actually looks like storm #1 of 1915, which gave them 15.45” on Aug 2nd. But for even that one there are no indications that the event total exceeded 18.85”. Thus it is my belief that Milton not only produced their heaviest daily rainfall on record but also their wettest full event rainfall in recorded history.

To add to tornadoes, winds, and storm surge, the rainfall in central FL was highly anomalous considering the storm wasn’t large or slow moving. The rainfall rates were extreme with St Pete getting 5” in one hour! This lead to a record for any single day (records back to 1914)! The flooding in central FL was awful, including at Daytona, and still is ongoing.

For the last ~30 winters: Dec a good barometer for rest of season at NYC during La Nina - Dec 3” or under (6 seasons), rest of season averaged 10.6” (range 2.3”-17.7”). Note, however, that 16-17 was only barely above 3” (at 3.2”) and the rest of the season had 27”. - For >3” in Dec (8 seasons), rest of season averaged almost 3 times as much, 30.1” (range 21.6”-64.1”)! - So, the high end of the Dec 3” or less rest of season range (17.7”) is actually lower than the low end of the rest of season range (21.6”) for Dec greater than 3” seasons during La Niña!

Thanks, Don. So, the Oct fcast for DJF is clearly the warmest of these going back to 2021-2. The +1 to +2 in the current one covers nearly all of the US, which is larger coverage than all of these. Note also that none had anything >2. I wonder if that is due to a cold bias. Note also that E Canada is warmer than 2021-2 and 2022-3 although it is less warm than 2023-4. So for the E US: -2021-2 ended up too warm, especially Midwest/Lakes -2022-3 ended up not warm enough especially NE -2023-4 ended up not warm enough Midwest/NE

I know it’s not apples to apples since these are probabilities vs anomalies, but am I correct to say that the comparison implies that 2024-5 is probably the warmest Oct (Sep) forecast by a good margin back at least to 2021-2 for the US as a whole?

I posted earlier about the record high rainfall at St. Petersburg for any day in history back to 1916. Here’s the official NWS release about the 18.54” which has history slightly further back (to 1914). This is interesting because they noted the previous record from 8/2/1915, 15.45”. That extreme an amount seemed to be very localized based on a much lower amount at Tampa and other cities. It was a result of storm #1 that had a landfall at Titusville with it recurving well E of St. Pete: https://en.m.wikipedia.org/wiki/1915_Atlantic_hurricane_season#/media/File%3A1915_Atlantic_hurricane_1_track.png RECORD EVENT REPORT NATIONAL WEATHER SERVICE TAMPA BAY FL 400 AM EDT THU OCT 10 2024 ..WETTEST DAY ON RECORD AT SAINT PETERSBURG ..RECORD DAILY MAXIMUM RAINFALL SET AT SAINT PETERSBURG A RECORD RAINFALL OF 18.54 INCHES WAS SET AT SAINT PETERSBURG YESTERDAY. THIS BROKE THE OLD DAILY RECORD FOR OCTOBER 9TH OF 1.56 INCHES SET IN 1953. THIS ALSO SET A NEW DAILY RECORD FOR ANY DAY OF THE YEAR BREAKING THE PREVIOUS RECORD OF 15.45 INCHES SET ON AUGUST 2, 1915. RECORDS IN SAINT PETERSBURG BEGAN ON AUGUST 1, 1914.

Don, how does the US map compare to the Oct 1 C3S forecasts for DJF in recent years? Do you have access to those maps to do a comparison?

This tornado outbreak was noted as having the most warnings for FL of any day in their recorded history. It was also noted as having the 2nd largest number of warnings for any day for any state, second only to AL in the great 2011 outbreak: https://en.m.wikipedia.org/wiki/2011_Super_Outbreak

I agree that NHC did a pretty bad job with the projected track of Helene through GA (disappointing because they usually do much better) but I’m saying that because the global model guidance was overall actually not bad. In contrast several globals did a great job in consistently calling for the center to track well E of the NHC’s W GA track in going NNE into NE GA, which is what actually occurred. In contrast I thought the NHC was excellent with Milton, which is typically how they do.

Whereas the Euro preseason forecasts for ACE missed way too high, the shorter term Euro Weeklies did very well. They forecasted the very active early season, the inactive peak season, and the very active last couple of weeks.

Check this amazing stat: St Petersburg got over 18” of rain yesterday!! How big is that? Since records started in 1916, the prior record high for any one day was 9.61” on 9/5/1933 (from storm #11 that hit SE FL as a MH and then moved just NW of Tampa). They got 7.76” on 6/23/1945 from a H that had a landfall 100 miles to the N. They received 7.42” on 9/4/1935 from the Labor Day Hurricane moving N just offshore. None of these days are even close to what Milton gave them! And steady rain is still falling there! https://www.currentresults.com/Yearly-Weather/USA/FL/St-Petersburg/extreme-annual-st-petersburg-precipitation.php

“126 tornado warnings were issued today by NWS offices in Florida This is a record for the State of Florida It's also the 2nd highest number of tornado warnings issued in the U.S. for a single calendar day - trailing only the 2011 Super Outbreak (Alabama)” From hylandwx’s Twitter (Pat Hyland)