When I use a word,’ Humpty Dumpty said in rather a scornful tone, ‘it means just what I choose it to mean — neither more nor less.
— Lewis Carroll: Through the Looking Glass
From the video transcript emphasis, mine:
Um you know we we toss around these uh terms like thousand year event um and it makes it sound like well okay it was just really bad luck that’s not what it means.
When we say this was a Thousand-Year event um means we shouldn’t have witnessed it if we lived for a thousand years.
Um Methuselah of biblical fame should not have witnessed an event of this sort um the only reason we’re witnessing itis because it’s no longer a thousand years uh event it’s maybe a five or ten year event…
https://youtu.be/ISwpB9SEg9c
Stalled thunderstorms aren’t rare. How long have we had the technology to measure a pocket of localized rainfall such as this, perhaps 200 years? How long have we had the granular coverage, 100 years? Radar, 50 years?
I’ll leave it to the readers to further discuss the idiocy of the Mann’s quote above.
Incidentally I live in Fort Lauderdale. It rained a lot.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Anyone reading this article might come away with the impression that Mann has just invented the ‘1000-year event’ concept off the top of his head. He hasn’t.
The term has been in widespread use for decades, including by the US Geological Survey; as has the method used for determining the probabilities.
This NOAA explainer uses rainfall, but the method can equally be applied to floods, storms or temperatures, etc, so long as sufficient observational data exist (and no, it doesn’t have to be 1,000-years of data).
Leaving aside the dodgy distribution fit on the NOAA web page, the article was about Professor Mann’s extremely poor explanation of 1,000-year events.
100-year and 1,000-year floods have been discussed here every time Australian floods are brought up.
Then what are we to make of Charles’s comment?
Where do the data points come from to estimate the probability distribution?
Why is it phrased as a question, though? The first commenter on this thread clearly thinks 1,000-year event probabilities require 1,000-years of data. Understandable, because that is the implication of Charles’s remark.
I read it as a comment regarding the amount of data available. Others seem to have interpreted it differently.
Perhaps Charles can answer your question during the day in the US.
And I answered that first question on the thread. There are historical plus geological records of droughts, floods, and earthquakes going back thousands, even tens of thousands of years.
There are few places in the world where there has been a rain gauge capable of measuring a one day event such as this for over 200 years, less than 100 years of good coverage, and less than 50, maybe 70 of radar for confirmation.
It can’t be extrapolated statistically if it was never observed in the first place.
The data are not being extrapolated Charles; the probability of extreme occurences is. Probability distribution statistics like this are widely applied in many scientific fields.
Likely from weather models.
You can’t perform a distribution analysis on an unknown distribution.
This isn’t how often this rain occurs anywhere. This is how often it occurs in this specific geographic region with these specific meteorological characteristics. It is a black swan event. Short term weather models are pretty good. I could see that working. I’m just guessing that’s how it is done.
“Short term weather models are pretty good.”
No, they’re not.
Extreme short rain events and only extreme short term rain events.
Without 1,000 years of data to validate the premise, predicting/labeling anything as 1,000 year events are simply false prophecies.
i.e., Prophecies with the same level of value as all of the rest of climate science climastrological mutterings.
Especially, since mankind keeps building large water obstructions and new channels over the selected locations.
With land receiving the rain in constant flux, there is no merit in such claims.
Most of the Florida or even Gulf coast locations identify their events as so many inches of rain per hour and how many hours rain falls at that rate. Which is what drives their real estate zoning maps.
Any location that receives very high inches per hour rainfall are liable for any weather event to rain at that rate for much longer periods of time.
The USA and Nevada built a highway through the local desert.
Henderson is a higher elevation that Las Vegas.
Yet, every few years people are surprised that the highway seriously floods as thunderstorm runoff happily runs down the highway in volume, flooding everything near the road.
I have no idea how anyone reading the article honestly could come up with that conclusion.
“Thank you Katie, always a pleasure”….pass the sick bucket.
I particularly liked his line “there us a certain amount of resilience in Earth’s climate” He goes on the say, ‘if we act now to reduce carbon emissions’ ,,,,,,,etc.. Makes you wonder what efforts were employed 1000 years back to reduce carbon emissions or maybe 12,000 years back when the resilience of Earth’s climate really came under threat. Maybe that is why there are no SUVs to be found from 12,000 years ago? The pre civilisation man/woman realised they wasn’t civilised and got rid of such things?….
He hasn’t got a new book coming out by any chance? Oh! yes he has, I remember Katie mentioning that, only a couple of times though along with a nice close up of the cover. She is so appreciative to ‘have him’ her words.
NB It is only a one in 1000 year event Katie, but beware, Mann could be back in ten years or if she is really lucky (sic) five.
Since nature appears to be entirely random then there was a time when it was exceedingly touch and go as to whether it would be Michael E Mann or someone entirely different born in 1965. Unfortunately for us he won. That is a once in a very very very big number event which we are paying a very very very heavy price for now.
I sincerely hope there are many parallel universes out there where nobody has ever heard of this pathetic publicly exposed liar and trickster.
Mike Mann saying something scientifically correct would be a real thousand year event, ie. we won’t see it in his lifetime or even 10 lifetimes of him and his snake oil salesmen confederates.
from BBC this morning:“Rishi Sunak sets up review to tackle ‘anti-maths mindset’
Oooooh, that’s a hard one, who might the first student be…..
The real deep cynics amongst us might suggest that if Rishi Sunak and the entire UK Government got out of its ‘anti-everything’ mindset, The People might follow the example and some actual good might come.
But UK Gov has been in that mindset for the last 20+ years thanks to Tony & Cherie Blair = why we’re here now.
But no, they pass the buck – exactly what Sunak is doing there
Similar and related to why UK Gov wants all the ‘early retirees’ (age 50 > 65) in this country to go back to work.
Why should they, they’ll be taxed, stressed, ripped-off and regulated so hard as to wreck their physical & mental healths and they’ll be completely no better off financially.
Potential math students are saying the same: Stuff you Sunak
You gotta love the guy for his eternal search for stupid.
If there are a thousand locations where we can see thousand year events, we’d expect to see one a year somewhere. There are more locations than that. There are more types of events too.
One thing about Mann is that his ability with statistics is poor. His ability with advocacy is pretty good though.
I always assumed that a PhD in any science would be required to master several courses in statistics. It’s a subject I don’t understand but I can appreciate its tremendous power.
I studied electrical engineering, and a single course in statistics was required for a bachelor level degree.
I believe mechanical and chemical engineers had to take a bit more.
Mann is a charlatan performing in a farce.
You know what is a once in a thousand year event? Mickey writing a paper that makes sense.
More like one in a million, and then, only is Mike the Mannsplainer interprets it – something that will occur only once in a million years, aka never (unless he is a vampire).
Michael Mann has no degree in meteorology but his comments on weather are taken as the gospel by those that worship at the idol of CO2. The same cretins will be found responding to Joe Bastardi’s comments on climate by saying he only has a degree in meteorology there by implying he knows nothing about climate.
One common characteristic of all leftist activists in any field is a total lack of self-awareness!
Al Gore and Bill McKibben both have a BA in English from Hah-vid- which makes them climate authorities. /sarc
Someone stated above that Mann started out in solid-state physics; if this is correct, at some point he would have been likely required to take Statistical Mechanics (Stat-Mech). I know of at least one school where it is used to weed out masters candidates who don’t belong there.
Bah, Mannhole has explosive verbal diarrhea again! I live in S Florida and these kinds of events are at least annual or biannual in some small zone. Two phenomenon produce these vents, either a stalled thunderstorm, which is constantly dissipating and forming; or a training line of small storms following the exact same geographic pattern coming in off the ocean.
I have personally witnessed this kind of 18+ inch rainfall in less than 12 hours at least a dozen times in the past 20 years, and there have been more, but not in my personal observational sphere.
Sorry Mikey, but no.
A “1,000 Year” weather event has a 1 in 1000 chance of happening EVERY YEAR, not something that you only see once every 1,000 years.
As with his statistical malfeasance, one can draw two conclusions: (1) He is an idiot who doesn’t know what he’s talking about; or (2) He is deliberately trying to deceive and spread alarmist nonsense.
In either case, this is not a man anyone should be listening to.
“Thanks for that question Katie” (I primed you just nicely).
And a chance to call carbon, pollution.
And he calls himself Professor!
Nope. Sorry, but can’t bring myself to press play on the video. Just the still image is enough to make me want to slap that smug expression off his face. Has he paid Tim Ball’s court costs to his family yet?
As many others have pointed out a 1 in 1,000 event does not mean we have to wait 1,000 years or more for it to happen. Assuming adjacent years are independent of each other, there is an approximately 63% chance of a 1 in 1,000 year event occurring in any given 1,000 year period.
If we take the very narrow view that the 1 in 1,000 event is for a single place (define as you see fit, but something smallish like a county or city) then it’s a 0.1% chance of happening in any given year. But as we are running this for many years we need to consider the cumulative probability over successive years (or season, or multi-day event, whatever statistic the claim is based on).
For these kinds of math problems it’s simpler to calculate the probability of it not happening and then subtract from 1. So if we make a quick excel sheet that multiplies 0.999 by 0.999, and then the product of that by 0.999 for each successive row for 1,000 rows, we get the running probability of it not happening for each year, which we can reverse to the probability of it happening.
After doing so we find that for this very narrow view of a 1 in 1,000 event in a single relatively small pre-specified place we find the number of years to get a 10%, 25%, or 50% chance of it happening is 105, 288, and 693 years respectively.
But we know the climate alarmist view would not take such a narrow view, they would claim any 1 in 1,000 year event as evidence. As this story is for South Florida we’ll limit our considerations to anywhere in the US. As South Florida is roughly half a state, lets for the sake of argument say there are 100 possible places in the continental US for a 1 in 1,000 year event of similar geographic scope to occur.
Now if we are being honest (which presumably is our goal given we are challenging Mann’s statements), there is also likely a spatial auto-correlation (as a few other commenters have pointed out) where a single weather pattern is likely to increase the chance of a 1 in 1,000 year event in more than 1 half-state at a time, so we would need to temper our calculation accordingly. That is no simple task to do well given regional patterns, influence of various cycles (ENSO, PDO…), seasonal differences, etc. and well beyond the scope of this comment.
But as an example, and if my dusty/rudimentary/3-decades since learning it probability math is correct, the chance of a 1 in 1,000 year event occurring after only 10 years, if we are watching just 20 independent places, rises to a little over 18%. If we expand to the full 100 possible half-state sized places (i.e. ignoring spatial effects) there is an 86% chance of a 1 in 1,000 year event occurring within only 20-years.
Now, if like number of possible places to find a 1 in 1,000 event, we do the same adjustment for the many possible weather variables tracked (single event precip, monthly precip, daily maximum temp, seasonal average temp, number of days hotter than X, length of “fire season”, etc.) we almost guarantee a 1 in 1,000 year event happening somewhere in the continental US for some weather statistic every few years.
The simplistic view shown here by Mann is the reason that Vegas exists – probability is not intuitive, cumulative probability is “hard”, and conditional probability is almost always ignored (gamblers fallacy, etc.).
I worked for a company that made gambling equipment for a few years. I helped to develop new slot machines. The company had on staff a gentleman who held a doctorate in mathematics, specializing in statistics.
It was his full time job to create payout charts for the various games that we made so that the casinos could calculate in advance the average earnings they would get from each game.
The machines could be adjusted to affect the odds of any given event coming up.
The state of Nevada had regulations stating how the odds of winning were to be displayed.
There was a manual for us developers on how to use various ‘C’ functions in order to get as close to true random as possible.
Speaking of “true” randomness. Most electronic games can only get pseudo random events. They have algorithms that create a list of seemingly random numbers. The big problem with this, is that each time you run the algorithm, you get the same list. So they use a “seed” value, that can be used to pick where in the string you start drawing numbers. Now you have to try and get a random seed, for each game played. One common trick is to take the number of micro-seconds since the machine was last re-booted, and then take the 16 or 32 least significant digits of that number as your seed.
Even that method has holes, if you are not careful.
There’s a story going around, about a programmer who worked on an electronic poker game. He programmed a laptop with the same random number generator that was used in the game.
He then went to the casino and played that game. He entered in the cards that he drew, and then had the laptop search the “random” number sequence for the pattern of cards in his hand. From that he was able to make a reasonable guess as to what the next card to be drawn, would be.
He was eventually busted, there’s a clause in your contract that you aren’t allowed to play the games that you worked on.
The solution that the programmers came up with once they learned of this weakness, was to take the remaining cards and re-randomize them after each card was drawn.
Great assessment!
https://simplemaps.com/data/us-cities
With over 108,000 cities in the US, when you multiply that huge number by the minuscule 1 in 1,000 year event probability at any 1 location…….you have the LIKELIHOOD of it happening at numerous locations each year.
If we took a more conservative number like 10,600 observing stations:
https://www.weather.gov/iwx/coop_station
We would still expect several 1 in 1,000 year events in most years for INDIVIDUAL locations.
When you expand the geographical region of the event, you greatly increase the number of events to compare with and it takes an even more extreme event to be a 1 in 1,000 event.
However, the numbers above based on numerous stations make for a very flawed, region expanding comparison when they expand to areas with a completely different climate.
We can’t expand the area to the entire US to compare with because 95% of the US would never have the same extremely favorable tropical/oceanic air mass in place that caused this.
An event like this in Bismark, ND for instance would be more like a 1 in a million event.
Throwing 95% of the country in that has almost no chance of this happening will make it look more unusual than it really was.
What a f’ing moron. It is hard to believe people actually believe his bs. Maybe it is an indication that there are more morons out there than I thought possible. Of course I have no data to support that.
It really boils down to this. Cliff Mass can be counted on as an elite source for using objective, authentic science.
https://cliffmass.blogspot.com/2016/03/the-golden-rule-of-climate-extremes.html
The Golden Rule
Considering the substantial confusion in the media about this critical issue, let me provide the GOLDEN RULE OF CLIMATE EXTREMES. Here it is:
The more extreme a climate or weather record is, the greater the contribution of natural variability.
Or to put it a different way, the larger or more unusual an extreme, the higher proportion of the extreme is due to natural variability.
As a youth growing up in south-central Kansas in the 1960’s we spent many evenings from April till June outside with walkie talkies listening to the “tornado watchers” on the edge of town. Tornados were constant, emminating from the Colorado “dry line”. Now, living in Wisconsin it appears weather stations are trying to jack the number and frequency of tornados and other “extreme weather” like small creek floods to a high level such that they can announce unprecedented frequency of extreme weather events. What tornados they get (or report) here in Wisconsin would never ever have been considered tornados in Kansas. High winds and some rotation. Maybe EF0’s or less. A couple weeks ago they reports dozens of these on a single day, with almost zero damage done.
Also, I lived through the “New Ice Age” scare of the ’70’s. It’s clear to me – and probably millions more like me – that there’s a shortage of sever weather event to attribute to “climate crisis”, and the weather statisticians are trying to make up for that.
He Reminds of those people that sit in casinos observing which slot machine has nt paid out for the longest time. Then play that machine. Oh and parts failure analysis is not analogous to to climatological analysis. We have access and observational and design data on manufactured parts; no such access to same degree of precision with climate.
Mr. Layman here.
I haven’t read all the comments so maybe it’s been covered.
I think it was in the context of flood plains where someone explained that a “1000 year” or a “(fill in the blank) year” flood doesn’t mean it only happens once in that number of years.
It’s a term used by those in the field that understand what it means but is easily misunderstood by the layman.
Sort of like “horsepower” in cars. If a 200 horsepower car can go 0-60mph in 5 seconds, that doesn’t mean that if it was pulled 200 horses it could go 0-60 in 5 seconds or even reach 60mph at all. But, with horsepower, the layman understands. With “(fill in the blank) year” stuff the layman doesn’t.
Seems that Mickey is also a “Mr. Layman here”.
“When we say this was a Thousand-Year event um means we shouldn’t have witnessed it if we lived for a thousand years.”
OMG! No wonder his “novel” hockeystick statistics was so awful and easy to debunk. He does not understand the nature of probability. There is a probability (albeit small), that its possible to have two or three or more 1000yr events within a decade or two, or to not have another 1000yr event for several thousand years. That there would be one exactly every 1000yrs would have the lowest probability of all!
I get a kick out of news articles that claim that some meterological record was broken, and that this is evidence of climate wierding.
In the U.S. alone, there are tens of thousands of meterological records that could potentially be broken. This means that we EXPECT many records to be broken in any one year. If theree aren’t, it would indeed be wierd.
“When we say this was a Thousand-Year event um means we shouldn’t have witnessed it if we lived for a thousand years.”
Sorry, a 1000-year event is defined as occurring 2–3 times a 1000 years. Mann is an idiot. S 10-year flood occurs 2–3 times in 10 years. Duh. Simple that he clearly does not know.
Did anyone ask Mann to produce the thousand years of daily rainfall data for Ft. Lauderdale that supports his claim about thousand-year-events? Shame.
And all of the king’s horses and men wouldn’t put Humpty back together again.
They might try if there were Subsidies to be mined in the process. They would just never be able to finish…just keep paying those subsidies