By GeoLurking, January 8, 2014 (republished)
Curmudgeon “an ill-tempered person full of stubborn ideas or opinions”
Well, if the shoe fits, I guess I’ll wear it. But… I don’t come about it lightly. To me, stuff has to make sense. One thing I abhor is mindless ranting that is specifically intended to scare people. What is even worse is when sound science and research is twisted to fit this purpose. Usually you see this as “Argumentum ab auctoritate.” One of the more popular tactics is to roll out someone carrying credentials and point at them as justification for the fear mongering because something they have said supports your alarmist claim… even if it doesn’t. Recently commenter Robin Hull Pointed out a fresh new article by Independent.co.uk titled “Risk of supervolcano eruption big enough to ‘affect the world’ far greater than thought, say scientists.” Yeah, like we haven’t heard that before. (yawn). Before I continue, many thanks to Robin for bringing it up. At the time I was freaking out by the actions of my dogs and needed the distraction so that I could get my mind off of what it was they were alerting on. (I had already checked the yard). Reading the article, this is what I found. “The eruption of a “supervolcano” hundreds of times more powerful than conventional volcanoes” Really? Okay, lets go with that.
First of all, the term “supervolcano” is a media construct. Before they started flailing about, wildly chanting the term, it didn’t exist. It’s purpose, like all alarmism, was to get attention. Since the article is about the old standard “spook everyone” subject of Yellowstone, that means that it is a product of the Yellowstone hotspot… because that is what feeds the system. Just in case you are new to this, the entirety of the Snake River Plain was formed by periodic Large Caldera Events and Volcanic Fields that leveled any mountains that were in the way, leaving a relatively flat plain in it’s path. Since the heavy work was done, the Snake River had a pretty easy path to form in as a drainage path for the surrounding watershed. Now, referring back to the independent article. “…they have been responsible in the past for mass extinctions, long-term changes to the climate and shorter-term “volcanic winters” caused by volcanic ash cutting out the sunlight.” Um… the jury is still out on that. The BBC, who ran their own version of the supervolcano article, had this linked as a related article: “Toba super-volcano catastrophe idea ‘dismissed”. In that article, they look at research centered around Toba, a system that has at times been claimed to have nearly wiped out mankind about 75,000 years ago. From the BBC version, which is a bit less alarmist: “”We think Yellowstone currently has 10-30% partial melt, and for the overpressure to be high enough to erupt would take about 50%.” Now lets take a look at those numbers and see how they square with the Independent’s statement of “it would take at least a decade or so for the magma pressure within a caldera to build up to a point where an eruption is likely.”
The last eruption was 600,000 years ago. In that time, Yellowstone has accumulated “10 to 30% partial melt.” (ignoring the numerous caldera filling resurgent activity eruptions, so that adds a quite bit of uncertainty). 30% in 600kyr yields about 1% additional melt for every 20,000 years. In order to get to the 50% melt accumulation where the authors believe that the pressures would be high enough to be in danger of erupting, we need another 20%. At the 20kyr per 1% rate, that’s an additional 400,000 years. If you assume that the 10% current melt is correct, you get 60,000 years for a 1% increase. To get to 50%, you need another 2,400,000 years.
Hmm…. One of the more common statements is that Yellowstone is “overdue” since the spacing on the last three or so eruptions is about 600,000 years. One thing I have found is that volcanoes don’t play the stats game very well at all. If your prognostication uses only 3 events, your stats are worth crap. Taking the eruption dates from a larger history of the hotspot reveals that the average interval is actually about 500,000 years. Worse eh? Not so. That list includes many caldera filling eruptions. The only criteria is that it made a tuff deposit somewhere. Since it is composed of 31 data points it is a bit more robust. The one sigma (standard deviation) is 648,588 years. The 95% confidence interval is 271,684 to 728,316 years. If you want to be more realistic, and only use honkingly huge events, the average gets closer to 5.4 million years or so, but that is from looking at it a few months ago and is from memory.
Caveat: I am not a geologist, nor am I a statistician. My stats experience is from failure analysis and process control of electronic equipment. If you think you can do a better job, by all means, please do. Even if your numbers are different than mine, at least you are using your own mind and looking at the data for yourself. Fair warning, not only do volcanoes not care about the Gaussian distribution, they don’t care much for Poisson distributions either. Hekla and Katla showed me that.
In keeping with the apparent tradition of articles about the Yellowstone Supervolcano… the obligatory picture of Grand Prismatic Spring.
Image source: Wikimedia Commons
Pretty isn’t it? I think that’s the reason most of them use it. This is not the yellowstone caldera. This pool is about 26 meters by 96 meters. The actual Caldera of Yellowstone is several orders of magnitude larger. In my opinion, this spring/pool formed from a maar like explosion. It’s also what I expect to occur again at Yellowstone should activity actually begin creeping upwards. About 160,000 years ago, a maar like detonation formed West Thumb. Yet one more event in the many that filled in the ancient collapse caldera. Magma + Water below the supercritical pressure = boom.
The track of the Yellowstone hot spot: Volcanism, faulting, and uplift
Pierce and Morgan (1992) Geological Society of America, Memoir 179
Supervolcano eruption mystery solved James Morgan BBC News.
Risk of supervolcano eruption big enough to ‘affect the world’ far greater than thought, say scientists Steve Connor The Independent.
This is still as true as it was several years ago. The story that Yellowstone could build up to an eruption fast still runs today. And it is still used every year as the volcano that
sells papers threatens a super eruption. But it threatens nothing of the sort. In fact, the chances are that it will never have another major eruption. The hot spot moves on, and at some time will make an attempt further east. Or perhaps, as we don’t really understand where the hot spot came from, we don’t understand how it will end either. Perhaps it will just peter out. No, the real dangers are elsewhere.
We have had a VEI7 about once every 300 years. But they don’t happen to schedule, and the risk doesn’t depend on when the last one was. We can say that there is a 0.3% chance of a VEI7 each year. But even though we haven’t had one for 200 years, the chance remains 0.3% per year. That is high enough that risk analysis should take it into account, but without panic. VEI6’s are more common and there is a 1-2% chance of one during the next year. Bigger than an ordinary VEI7? A Toba-like eruption, perhaps once very 100,000 years – 0.001% per year. A flood basalt eruption? Once every 10 million years – 100 times less likely.
Now changes to the Earth may play a role: where major ice caps are melting, the chances of a large eruption increase. It is quite plausible that Vatnajokull will be the site of a large eruption within the next 200 years, once the ice has gone. But not certain, and Iceland does not do VEI7’s. Never has. Now Antarctica may be a different beast, once it’s ice has gone. But until this happens, Geolurking remains a more reliable source of information than the Express. Albert
And finally, a movie of all USGS recorded earthquakes since 2000. Spot the Bardarbunga eruption, when Iceland flickers.