Remember this from 23 May 2025?
NOAA’s outlook for the 2025 Atlantic hurricane season predicts a 30% chance of a near-normal season, a 60% chance of an above-normal season, and a 10% chance of a below-normal season.
The agency is forecasting a range of 13 to 19 total named storms . Of those, 6-10 are forecast to become hurricanes, including 3-5 major hurricanes (category 3, 4 or 5; with winds of 178 kmh/111 mph or higher).
NOAA has a 70% confidence in these ranges.
As of today September 27th, 2025 zero hurricanes have made landfall in the United states during the 2025 Atlantic hurricane season. The season runs from June 1st to November 30th. So, we have just 2 months to go and we have already passed the peak climatology for hurricanes in the Atlantic.
The peak of the Atlantic hurricane season is climatologically around September 10th, a date when ocean and atmospheric conditions are typically most conducive to storm development and intensification due to warm sea surface temperatures and low wind shear. While storm activity can begin before and after this date, with the overall season running from June 1st to November 30th, the period between mid-August and mid-October generally experiences the most activity.
While the season has produced five named tropical storms and three hurricanes so far with the names Andrea, Barry, Chantal, Dexter, Erin, and Humberto ….none classified as hurricanes, with Category 1 or higher (with sustained winds of 74 mph or more) have struck US soil.
And, it doesn’t look like Humberto will make US landfall either.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Take note that NOAA makes mid-season “adjustments” to their formation and landfall predictions. For reliability only the first release counts. The rest are bets after the first die has fallen.
Is there any tracking of how they’re doing historically? Like when they say 70% chance of above normal, how often are they actually getting it right?
Feature, not bug.
I have only followed their predictions informally, not doing any real analysis. But for years it has been a standing joke with me at least, at how the initial predictions are waaay off always on the high side, and then the mid-season “adjustments” tend to incorporate reality instead of fantasy…with absolutely no self-awareness of how far wrong they were, at least not publicly.
I think they believe that if they forecast fewer storms and hurricanes, they’ll get less money, and that more fear-mongering means more money. I’m afraid they may be right…
I’ve been tracking Michael Mann who fancies himself as tropical storm 🌀 predictor. His record is not very good. He shouldn’t predict to a tenth of a storm, that is for sure…
My crystalball predictons are 3 times more accurate than Mann’s.
And I can only afford a plastic crystalball.
I’ll make a prediction for 2026: 20, +/- 15
2023 got the high side correct, almost.
2014 got the low side correct, almost.
That aside, I have in on my bucket list to view a 0.3 hurricane. 🙂
Here is a table of CSU, NOAA, and TSR seasonal forecasts.
https://www.aoml.noaa.gov/hrd/hurdat/SeasonalVerification.html
They should probably remember that Yogi Berra quote “It’s tough to make predictions especially about the future”
It’s too tough for them to predict the past.
None of their models can do it retroactively.
Even keeping data from the seems to be to much for them.
They only(pretend to ) have arctic see ice data until 1979 (while they have data above the ozone holes above the artic until 1969 🙂 )
They only had burned acreage data until 1981 during the Blinken presidency.
And the Scientific American told us that the ice age scare is based on a single Newsweek article.
If you don’t and won’t know about the past you can not predict the future.
Well, if that’s really what Scientific American published, then they had it wrong. It is a known fact that Leonard Nimoy (of “Mr. Spok” Star Trek fame) narrated a 1978 TV show titled “In Search Of The Coming Ice Age”.
The Web never forgets . . . here is a link to a YouTube version of this video, and in the info block section underneath that one can find another link to a transcript of the show:
Perhaps in recognition of Yogi’s statement, NOAA sought out help with making their seasonal hurricane forecasts starting in June 2022 with implementation of their Weather and Climate Operational Supercomputing System (WCOSS). WCOSS is two supercomputers, nicknamed “Dogwood” and “Cactus”, each operating at an incredible 14.5 petaflops speed . . . that’s 10^15 floating point operations per seconds each! (source: https://www.noaa.gov/news-release/noaa-completes-upgrade-to-weather-and-climate-supercomputer-system ).
The current government contract award to General Dynamics Information Technology for the WCOSS supercomputers and their designs, deployments, and management is stated to be $505 million over a potential 10-year period. (ref: https://www.gdit.com/about-gdit/press-releases/noaa-awards-general-dynamics-high-performance-computing-contract ).
Of course, the quality of the output from supercomputers modeling weather or “climate”, particularly tropical storms, is only as good as the “science” that goes into programming them and the “data” they are fed. In this case, it appears the NOAA supercomputer outputs are pretty much equivalent in uncertainty to just predicting the expected number of storms using an Excel program on a $500 laptop computer using simple statistics (mean and 2-sigma values from the last 4 hurricane seasons).
“The agency is forecasting a range of 13 to 19 total named storms . Of those, 6-10 are forecast to become hurricanes”
Well, Humberto is storm number 8, so that looks on track. With 3 hurricanes so far, they might have overprerdicted there.
Do you think they just take the last 20 years and take the average?
Tropical Depression NINE, just north of Cuba on Saturday, looks to be a rain-maker along the coast and inland to Charlotte. Looking at the Atlantic back toward Africa – nothing. With a storm starting there, it takes a week for arrival off the US coast. Closer starts are possible.
On the opposite coast, rain is about to shut down the 2025 fire season in Washington State. Cliff Mass has a post.
TD#9 does stand a chance of landfall in the US, at least according to my sources – but more likely to be a TS, not even a CAT1. What does seem likely is a load of rain along the Carolinas and big surf on Cape Hatteras.
I haven’t recorded it, but I think NOAA made the shift to “higher than normal” in 2006, after the US got pasted in 2005 (Katrina, Rita, etc). If they’ve predicted a “lower than normal” season since then, I must have missed it.
I am reminded of the time when we were living in Phoenix, and there was a storm coming over from CA – local forecasters, er, “weather readers” were forecasting “chance of showers” – but the weather in the mountains of CA was record breaking snowfall. Needless to say, the storm was a big ‘un, with tons of snow in the mountains and torrential rains in the desert. After that, every cloud seen coming from the West was “the storm of the century” for the rest of the season.
As of today, almost all forecasts for #9 have it turning out to the Atlantic.
A 60% probability of being above normal with a 70% confidence level is about as meaningful as saying it will be dry tomorrow if it doesn’t rain. If you can’t make a prediction with a properly determined confidence level of at least 90% just admit you have no idea what will happen.
There have been people out there who claimed to be clairvoyant,psychic etc
who performed above 60 and 70% – and yet it was considered a total failure,
just as a lucky streak for gamblers.
But in terms of weather and climate it is called science.
My guess is that the decline in psychic people can be explained with the increase in climate experts.
These people chose the saver profession.
Above and beyond the stated 60% probability an above-normal season and the stated 70% confidence in that statement, look at the stated uncertainties about their mean predictions:
— total named storms (16 +/- 19%),
— hurricanes (8 +/- 25%),
— major hurricanes (4 +/- 25%).
Just wondering how these all mash up to yield a reliability of prediction? /not-sarc
Leading with “zero hurricanes have made landfall” is bit disingenuous when the forecast being criticized states specifically “NOAA’s outlook is for overall seasonal activity and is not a landfall forecast.”
In defense of the above article, most inhabitants of North and Central America are concerned about hurricanes that make landfall . . . much less so about those that don’t.
Currently, one is still able to see the humor in (criticize?) NOAA’s shortfall in their “seasonal activity” predictions with only about 4 weeks to go before the 2025 Atlantic hurricane season ends.
Shouldn’t that be six named tropical storms?
There are three kinds of people in the world: those who can’t count, and those who can…
you forgot to name the fourth kind
I’ve told you a million times, stop exaggerating!
It’s good to be in the minority.
My interpretation of the quoted statement:
. . . five named tropical storms plus three that progressed to become hurricanes (equals 8 total, accounting for names with first letters A-H).
But yes, named tropical storms Fernand and named hurricane Gabrielle (ref: https://en.wikipedia.org/wiki/2025_Atlantic_hurricane_season ) were not listed in the above article . . . probably just a “typo”.
And for clarification, the hurricanes so far have been Erin (Cat 5), Gabrielle (Cat 3) and Humberto (Cat 5).
At one time, they probably did forecasting, now it’s just guessing.
They seem to predict “above normal” pretty much every year.
Eventually they’ll be right, so they can at least take heart.
Is the paucity of hurricanes this year caused by “climate change?”
. . . undoubtedly, since climate change is really just nature doing “her” thing with normal +/- variations as “she” allows.
As to the relative “paucity” of tropical storms and hurricanes, that is just the failed result of humans trying to model the somewhat-stochastic process we call nature.
fer sure! the climate in Washington DC seems to affect everything
The great thing about overproduction of climate change scare content is that it is now self identifying jobs and functions to be replaced by AI and stable quantum computers. That includes all the meaningless functions too.
Why do these people still have jobs? They’ve predicted the same numbers for the last 5 years and have been wrong.
One time in grad school on a winter day I asked a colleague how cold he thought it would get that night. He said 30° and I kinda chuckled because it was currently 32° with an 12° dew point, clear skies, calm wind and the sun had just set. Moreover MOS was going 22°. Well wouldn’t you know it some clouds rolled in earlier than expected and the low was only 29°. The next morning he shows up and and says to me “who’s the better forecaster”. The point of this story is don’t criticize a forecast till the verification period is over cause it can bite you in the a$$.
Anthony keeps harping on US landfalls, but neither the NOAA or CSU seasonal forecasts make predictions about landfalls. You can have a very active Atlantic hurricane season and zero CONUS landfalls. This year we’ve had a persistent trough over the eastern US and the subtropical high has shifted eastward resulting in the bulk of Atlantic tropical activity recurving east of the Eastern seaboard. So that means fewer landfalls despite how much activity there is.
Landfalls is the only meaningful number when comparing to past records. Pretty sure they didn’t have satellite monitoring over 100 years ago.