Slashdot Mirror
Scientists Predict Earthquake's Location and Strength
A new study has been published in Nature Geoscience (abstract) detailing how scientists correctly anticipated the location and strength of an earthquake earlier this year. On September 5th, a 7.6 earthquake rocked Costa Rica's Nicoya Peninsula. That region had seen earthquakes of (roughly) magnitude 7 in 1853, 1900, and 1950, so "geoscientists had forecast that a magnitude 7.7 to 7.8 quake should occur around the year 2000, plus or minus 20 years."
"The Nicoya Peninsula is prone to earthquakes because it's an area of subduction, where the Cocos Plate is pushing underneath the Caribbean Plate, moving at a rate of about 8.5 centimeters per year. When regions such as this suddenly slip, they produce a megathrust earthquake. Most of the world's largest earthquakes — including the magnitude 9.0 Tohoku-Oki quake in Japan in 2011 and the magnitude 9.15 Sumatra-Andaman earthquake in 2004, both of which produced devastating tsunamis — fall into this category. .. The close study of this region allowed scientists to calculate how much strain was building in the fault and in May 2012 they published a study in which they identified two locked spots capable of producing an earthquake similar to the one in 1950. In September of that year, the landward patch ruptured and produced the earthquake. The offshore one is still locked and capable of producing a substantial but smaller earthquake, an aftershock with a magnitude as high as 6.9, the researchers say."
"When asked how they discovered this information, scientists said 'we just asked yo mamma where she'd be on the fifth, ohhhhhhhhhhhh!' Upon hearing this, passers by offered this reporter gifts of ice for that 'sick burn.'"
Great, so live in a tent in your back yard for your entire adult life then?
With a window like that, any fool can "correctly predict" just about anything.
Beware the stock market crash of 2034, plus or minus 20 years. Don't say you haven't been warned.
I predict hoots of derision heaped upon this story.
Sig Battery depleted. Reverting to safe mode.
Seriously, this sounds like a math problem for a middle school math class:
Fred found out that an earthquake happened in 1850, 1900, and 1950. When should Fred expect the next earthquake to happen?
A. 2000
B. 1951
C. 2013
D. This is a bogus question because it makes unfounded assumptions as to how earthquakes work without any true understanding as to the underlying mechanics
Over 20 years ago I predicted that in late December, 2013 a group of scientists would announce that they had successfully predicted an earthquake. Unfortunately due to my data from back then having disappeared, you'll just have to take my word for it.
I deny that I have not avoided attaining the opposite of that which I do not want.
just don't try this in Italy.
Okay, in fairness, I went back and read TFS, and it looks like the real problem here is Soulskill's poor writing skills. A transitional sentence was needed to indicate that the "+/- 20 years" prediction was an old prediction, and that scientists did a new analysis in 2012 that suggested that a large earthquake was imminent (and which happened not long thereafter).
Thanks for this. Anybody know if the paper talked about Type I and Type II errors with their method?
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
No. Two different studies. The old one gave a 40 year window. The one done in May of 2012 predicted an imminent earthquake due to two potential lock spots, the earthquake then happened in September of 2012.
Forecasts are based on trend.
That is all.
The Kruger Dunning explains most post on
From this article:
The researchers published a study a few months before the earthquake, describing the particular locked patch with the clearest potential for the next large earthquake in the region. The team projected the total amount of energy that could have developed across that region and forecasted that if the locking remained similar since the last major earthquake in 1950, then there is presently enough energy for an earthquake on the order of magnitude 7.8 there.
Because of limits in technology and scientific understanding about processes controlling fault locking and release, scientists cannot say much about precisely where or when earthquakes will occur. However, earthquakes in Nicoya have occurred about every 50 years, so seismologists had been anticipating another one around 2000, give or take 20 years, Newman said. The earthquake occurred in September of 2012 as a magnitude 7.6 quake.
I don't see any reference to "imminent".
This is key. To look at the previous pattern of earthquakes and forecast a quake, is not prediction. It is guessing.
To predict, you have to show a relational mechanism that is associated with a coming earthquake. Unfortunately in both weather and quakes, the governmental "science" agencies have pursued a statistical model, which would lead to the conclusion that earthquakes cannot be predicted. Then, as it appears happened with Jack Coles, they equate prediction to crying fire in a crowded theatre, and put those who attempt science in prison.
Thus the science is prevented from advancing.
To predict a quake, you might drill in strain sensors, or study slip, or look at 1-.01 hz radio, or look at ants. All of which is early science at its best, and hogwash at its worst... but when the government prosecutes science, then you limp along as well as you can.
Correct Horse Battery Staple: 72 bits of entropy. Enter "Correct H" into google. When it generates the phrase, that's
" I went back and read TFS,"
hay, no fair! I wrote a scathing post comparing your post to monkey poo only to notice at the last minute you went and read the article then posted something intelligible!
Jerk~
The Kruger Dunning explains most post on
Some people will still not evacuate.
Happiness in intelligent people is the rarest thing I know.
Ernest Hemingway
This is beyond useless.
I live in a country with significant seismic activity.
We know for hundreds of years that every 30 to 50 years a big (7.6 - 7.8) earthquake will happen. We even know where the epicenter will be. We know how it will propagate.
It's been 36 years since the last one so many people who live in problem areas are beginning to get nervous. But we don't know when it will happen. Could be tomorrow or ten years from now.
The study in the article comes from a government-funded agency which operates within a public university, which, in Costa Rica, effectively makes the researchers public employees.
I think what this post is really about is how rudimentary geoscience is compared to similar fields like oceanography, astronomy and epidemiology. If it's a major story to see that there's a 50 year pattern on earthquakes in a particular area, that's not so great.
In August 2001, a FEMA training session discussed the three most likely major natural disasters they would have to deal with in the next decade. They were a terrorist attack in New York, a hurricane hitting New Orleans, and an earthquake on the San Andreas line.
Two of those happened. But nobody is really surprised they were 66% accurate, because with timespans like this, 66% accuracy is trivial. A damn comedy website managed to predict Fukushima to within 200km (they predicted a TEPCO-run reactor on the opposite side of Japan would be struck by an earthquake and tsunami).
Why? Because it turns out that when you look at the situation, the history and the statistics, it's pretty easy to predict the future as long as you don't give a strict time. Just look at science fiction, or even general fiction, to see how people who did their research were able to come up with scenarios plausible enough to actually happen.
Now, the difference here is supposedly that instead of looking at the history and statistics, they were looking at geological data. If they can improve on this technique to get even just a one-year timespan for an earthquake, that would be groundbreaking (pun totally intended). Or if they can use it to predict an earthquake where there isn't a history of earthquakes. As it is, it's useless, but hopefully they can improve it.
Yes, and government funded science also makes advances in weather forecasting and tornado analysis, despite the official positions of NOAA leadership as well. As I said, the science limps forward as best it can.
Correct Horse Battery Staple: 72 bits of entropy. Enter "Correct H" into google. When it generates the phrase, that's