Author
|
Topic: Supervolcanoes and the gambler's fallacy (Read 4726 times) |
|
David Lucifer
Archon
Posts: 2642
Reputation: 8.94
Rate David Lucifer
Enlighten me.
  
|
 |
Supervolcanoes and the gambler's fallacy
« on: 2005-12-11 12:28:58 » |
 |
I was reading an old Zentastic blog entry today (and regretting I didn't introduce myself to Shannon Larrat when I saw him at TV04 in Toronto). Here's what he had to say about supervolcanoes.
Quote:
Anyway, other than that, while processing images this morning I watched Supervolcano which asks the question of what would happen if Yellowstone erupted (as it does about every 600,000 years, with the last eruption being 650,000 years ago)... It's quite possible for there to be an eruption large enough to kill a billion people and [whatever number comes after 'trillions'] of animals. It's happened before and statistically it must happen again.
I was thinking about it, and I kind of hope this happens. Now, before you tell me how terrible that is, let me tell you a fact: you are going to die soon, and so am I, and there's nothing that can be done to stop it — at best it can be delayed. So really, for that brief period that I'm alive, I'd like it to be one of the most interesting periods in human history, if not in geological history as well... So far odds are looking good that I will get my wish.
|
Leaving aside whether it would be cool to see a billion people die violently in our lifetime is he right about the facts? The eruption is 50,000 years overdue so we're bound to see it pretty soon, right? Well, no. I think he commits something like the gambler's fallacy here. If it was a periodic event like a comet he may have a point, but in that case it wouldn't be overdue. It is probably more like the old faithful geyser that erupts every hour or so. If so then it could be modeled as a poisson process which would imply an exponential distribution for events (also used for mean time between failure of electric components like light bulbs).
If my calculations are correct, a 600,000 year mean time between super eruptions means that the annual probabilty is 0.000001155. (Correct me if I'm wrong) If Shannon lives another hundred years he only has 1 chance in 8656 of seeing Yellowstone blow up. Not great odds by any definition (for better or worse).
|
|
|
|
|
David Lucifer
Archon
Posts: 2642
Reputation: 8.94
Rate David Lucifer
Enlighten me.
|
|
Re:Supervolcanoes and the gambler's fallacy
« Reply #1 on: 2005-12-11 14:07:20 » |
|
[Shannon replies...]
To put the supervolcano into gambling terms, it's more like finding the
ace of spades in a deck of cards. If each turn you get to pull a card
and then discard it if it's not the ace of spades, then each turn that
goes by, the higher your odds get of finally pulling that ace.
While it is true that ON AVERAGE every year has a 0.000001155 chance
(I'm assuming your math is right), that's just the average. As I understand
the geology, the odds of an eruption start low and go up slightly every
year, which is what makes them go off "like clockwork" rather than with
a more random distribution. Thus I am far more likely to see a supervolcano
here than I would if I'd been born 500,000 years ago.
Or to give you another example, imagine a balloon. Blow it up to capacity
and then deflate it. Then repeat. Then repeat. The first few times the odds
of the balloon popping are slim. However, the more time that goes by, the
greater the odds are of the balloon popping on that inflation cycle.
Shannon
---
Shannon Larratt - http://www.zentastic.com/
"Democracy is two wolves and a lamb voting on what to have for lunch.
Liberty is a well-armed lamb contesting the vote." --Benjamin Franklin
|
|
|
|
|
David Lucifer
Archon
Posts: 2642
Reputation: 8.94
Rate David Lucifer
Enlighten me.
|
|
Re:Supervolcanoes and the gambler's fallacy
« Reply #2 on: 2005-12-11 14:24:47 » |
|
http://www.livescience.com/forcesofnature/ap_050408_supervolcano.html
Quote:| Hank Heasler, park geologist and the park's coordinating scientist for the volcano observatory, said any claim that Yellowstone is overdue for an eruption is based on simplistic reasoning, drawn only from when the three previous known eruptions occurred. |
Apparently there have only been 3 known eruptions, occuring 2.1, 1.3 and 0.64 MYA so the two known intervals between eruptions are 800K and 660K years. Not enough data to predict the next one with any confidence.
Scientists seem to agree with my assessment that "the odds of another catastrophic volcanic eruption in Yellowstone within anyone's lifetime are extraordinarily remote"
|
|
|
|
|
David Lucifer
Archon
Posts: 2642
Reputation: 8.94
Rate David Lucifer
Enlighten me.
|
|
Re:Supervolcanoes and the gambler's fallacy
« Reply #3 on: 2005-12-11 15:11:32 » |
|
[Shannon again...]
Oh, I agree with all those statements below. All I'm saying is the odds are
higher now than in the past -- still FAR from a sure bet though!!!!
|
|
|
|
|
Sat
Acolyte
Gender: 
Posts: 50
Reputation: 4.60
Rate Sat
Freelance Pedestrian
|
|
Re:Supervolcanoes and the gambler's fallacy
« Reply #4 on: 2005-12-11 16:58:10 » |
|
You guys got me interested in this subject today so I did some searching and found a few things:
According to the Christer at this page: http://www.revlu.com/ystone.html , Yellowstone's gonna blow as part of God's great Amegeddonomatic Master Plan... maybe with some help from terrorists... He also notes that, "Geysers have a sulphur smell." Evidently this means that they're Satanic or something... wow I sometimes emit sulfer smells too.
I found this image of hot springs in YellowStone:
There's obviously something wrong in this picture. In my neck of the woods hot springs are usually full of naked people tripping on mushrooms.
According to the dude over at http://www.bowjamesbow.net , "SuperVolcano has a strong Canadian connection. Michael Riley, born in London, Ontario, stars in the CBC series This is Wonderland. A number of other Canadian actors star, and a lot of the movie was filmed in Vancouver." - http://www.bowjamesbow.net/2005/04/18/when_vesuv.shtml
Note that both Lucifer and Shannon are Canadian. There's obviously something unsavory going on here.
Yellowstone is cool, but Evidently there are Supervolcanos are all over the place! Steve Quayle says, "super volcanos are found in Italy, South America, the United States and New Zealand...". There's one at Mount Taupo where there's supposed to be a big eruption every 2000 or so years. http://www.stevequayle.com/News.alert/05_Yellowstone/050404.Merapi.html
Last time I checked the World didn't end 2000 years ago, but I could be wrong. Add a few more supervolcanoes into the equation and we'll up the chances of overall doom ,death, destruction and... some other word that starts with a 'd'.
The clear headed folks over at http://armageddononline.tripod.com say, "...we are overdue for annihilation." According to them, "Scientist have discovered that the ground in Yellowstone if 74cm higher than in was in 1923 - indicating a massive swelling underneath the park. The reservoir is filling with magma at an alarming rate. The volcano erupts with a near-clockwork cycle of every 600,000 years. The last eruption was more than 640,000 years ago..."
The guy at armageddon online also has this to say, "I've opened the Armageddon Store selling great books, videos and DVDs related to the end of the world." - http://armageddononline.tripod.com/volcano.htm
In any case it makes for fun scary stuff. If people gets a little to used to being scared by terrorists and black people and asteriods and such, the media can glom onto this stuff.
|
Ascend Through Modification
Church of Virus Web Hosting Donations
|
|
|
Blunderov
Archon
Gender:
Posts: 3160
Reputation: 8.91
Rate Blunderov
"We think in generalities, we live in details"
|
|
Re:Supervolcanoes and the gambler's fallacy
« Reply #5 on: 2005-12-11 17:23:34 » |
|
Quote from: David Lucifer on 2005-12-11 15:11:32
[Shannon again...]
Oh, I agree with all those statements below. All I'm saying is the odds are
higher now than in the past -- still FAR from a sure bet though!!!!
|
[Blunderov] There is another very significant factor in play though and I think it swings the odds in a way that is not good:
http://www.nasca.org.uk/Strange_relics_/reversal/reversal_latest/reversal_latest.html
Polar reversal could be imminent...
"... Seismic and volcanic activity would also increase quite dramatically"
[Bl.] Meanwhile global warming is taking place simultaneously..
http://www.nasca.org.uk/Strange_relics_/reversal/magma_tides/magma_tides.html
"The Magma Tide connection.
Magma Enigma.
Below the surface of the Earth flow vast magma tides. Some believe it is these that are capable of toppling the world on its axis.
These tides could even be influenced by Global Warning".
[Bl.] Just the other day there was a largish earthquake in Kenya, not something that happens every day. Earthquake activity has been increasing steadily throughout the world for some decades now. So has volcanic activity.
"I see the bad moon arising.
I see trouble on the way.
I see earthquakes and lightnin’.
I see bad times today.
Chorus:
Don’t go around tonight,
Well, it’s bound to take your life,
There’s a bad moon on the rise.
I hear hurricanes ablowing.
I know the end is coming soon.
I fear rivers over flowing.
I hear the voice of rage and ruin.
Chorus
All right!
Hope you got your things together.
Hope you are quite prepared to die.
Looks like we’re in for nasty weather..."
On that cheery note, bye for now.
|
|
|
|
|
David Lucifer
Archon
Posts: 2642
Reputation: 8.94
Rate David Lucifer
Enlighten me.
|
|
Re:Supervolcanoes and the gambler's fallacy
« Reply #6 on: 2005-12-11 18:45:50 » |
|
http://www.ma.hw.ac.uk/icms/ems/volcano/keynote.html
Quote:
Some of the earliest work applying statistical methods in volcanology involved looking at the expected repose interval between volcanic eruptions. The repose interval is typically defined as the time elapsed between the start of successive volcanic eruptions. The start time and duration of some types of eruptions, such as Vulcanian explosions, are easily characterized from seismic records. In other cases, activity ramps up over time and the exact start time of the eruption can be ambiguous. Wickman [1966] looked at the repose intervals between volcanic eruptions for several volcanoes, plotting these as complementary cumulative distribution functions, to assess the likely duration of quiescence in volcanic systems. In some cases, such sequences of repose intervals are stochastic and have Poisson distributions; their survivor functions are exponential. This means that the system has no memory. The probability of a volcanic eruption in a given time interval is the same as in any other time interval, regardless of the time elapsed since the start of the last eruption. Dividing the probability density function for repose intervals by the complementary cumulative distribution function (the hazard function) for these series of repose interval data yields a flat curve.
|
This supports what I was saying. However it goes on to mention that the repose interval data of some volcanoes fit a log logistic distribution which could support Shannon's intuition, but depending on how the competing processes interact the probability of eruption can also decrease over time.
|
|
|
|
|
Bass
Magister
Posts: 196
Reputation: 6.08
Rate Bass
I'm a llama!
|
|
Re:Supervolcanoes and the gambler's fallacy
« Reply #7 on: 2006-11-29 20:40:42 » |
|
Shannon's really a barrel of laughs, isn't he? Can't argue with what he's saying though...
As for your calculations, Lucifer, I can offer nothing. So, i'll just offer some opinions.
I saw a programme about the Yellowstone supervolcano once. It was quite frightening; apparently the amount of SO2 released would bring about worldwide freezing, and so on. America would be the worst affected though; an eruption would mean the destruction of the majority of the USA's agricultural land.
But maybe an eruption could be just what humanity needs...something to make nukes and our million other pet horrors seem obselete, if only for a little while.
It wouldn't be like a disaster movie. Some despot would capitalise, I don't doubt.
However, back to the point. I don't think Shannon's that far off in his timescale. I remember something about the actual volcano moving East, because of the westward movement of the USA itself. They can trace the old craters back through heat imaging and so on, and there's always a distinct distance between previous eruption craters, if memory serves. That's how they came up with the 600, 000 year idea, I guess. It seems to be fairly systematic and predictable. The geologists in the programme were very excited; they seemed to be waiting for the eruption. It frightened me when I saw it; but then again, that was before I knew what Trident was.
|
|
|
|
|
David Lucifer
Archon
Posts: 2642
Reputation: 8.94
Rate David Lucifer
Enlighten me.
|
|
Re:Supervolcanoes and the gambler's fallacy
« Reply #8 on: 2006-12-03 20:00:23 » |
|
Quote from: Bass on 2006-11-29 20:40:42
Shannon's really a barrel of laughs, isn't he? Can't argue with what he's saying though...
|
Actually I can and I did.
Quote:
However, back to the point. I don't think Shannon's that far off in his timescale. I remember something about the actual volcano moving East, because of the westward movement of the USA itself. They can trace the old craters back through heat imaging and so on, and there's always a distinct distance between previous eruption craters, if memory serves. That's how they came up with the 600, 000 year idea, I guess. It seems to be fairly systematic and predictable. The geologists in the programme were very excited; they seemed to be waiting for the eruption. It frightened me when I saw it; but then again, that was before I knew what Trident was.
|
Like I said above "Apparently there have only been 3 known eruptions, occuring 2.1, 1.3 and 0.64 MYA so the two known intervals between eruptions are 800K and 660K years. Not enough data to predict the next one with any confidence." That is not systematic or predictable.
|
|
|
|
|
|
Logged
