By Rose Eveleth
Posting in Design
Complicated computer models can tell us where wildfires will spread, when two planets might collide and how high our oceans might rise due to melting ice caps. But we still can't predict earthquakes.
At least 50 people died in a 5.7 magnitude earthquake that hit China this morning, and another 20,000 homes were damaged. The quake hit Yunnan, and the nearby Guizhou provinces, home to some of China's poorest people. Of course, they didn't see it coming.
From who will win the most Olympic medals, to Nobel Prizes, humans - and the machines they've designed - have gotten pretty good at predicting things. Complicated computer models can tell us where wildfires will spread, when two planets might collide and how high our oceans might rise due to melting ice caps. But we still can't predict earthquakes. What gives? Shouldn't we be able to figure out where the plates will give and when?
Well, you're not the only one asking. Science wants the answer to that question too. To find out, researchers recently looked at the three deadliest earthquakes in recent history: Wenchuan, China in 2008, Leogane, Haiti in 2010 and Tohoku, Japan in 2012. They found that all three occurred in places where even the hazard maps - out best guesses for earthquake prediction - didn't expect them to be.
“We’re playing a complicated game against nature,” the study’s first author, Seth Stein of Northwestern University, said in the press release. “It’s a very high stakes game. We don’t really understand all the rules very well."
This is a big deal not just to save lives, but to save money too. Countries each spend a certain amount of money to prevent damage from natural disasters. Knowing where to spend that money could save both lives and the economy.
The inability to predict earthquakes often confuses us non-scientists too. “The public is accustomed to the uncertainties of weather forecasting, but foreseeing where and when earthquakes may strike is far more difficult. Too much reliance on earthquake hazard maps can have serious consequences," Mian Liu, another researcher on the study, staid in the press release.
In Italy, six seismologists were even tried for manslaughter in court for failing to predict an earthquake. The New York Times explains why that trial has outraged scientists all over the world:
Earthquakes differ from other types of natural disasters. Meteorologists can track a hurricane with precision, but seismologists cannot predict exactly when and where an earthquake will occur. Scientists have condemned the Italian prosecution for this reason, saying the defendants are on trial for failing to do something that is impossible.
For the researchers, there were a few standout reasons that hazard maps failed. First, earthquakes are complicated. Second, historical records in most places don't go back long enough to get a good sense for patterns - since earthquake trends tend to emerge only on long time scales. The combination of bad and incomplete data, with a complicated system, means that hazard maps are guesses at best. The study writes:
As a result, key aspects of hazard maps often depend on poorly constrained parameters, whose values are chosen based on the mapmakers' preconceptions. When these are incorrect, maps do poorly. This situation will improve at best slowly, owing to our limited understanding of earthquake processes.
What can we do about it? Well, we can wait to gather more data. Which means waiting for more earthquakes to happen. We can also test the maps, the authors say. Which, of course, isn't so easy either. Even they admit that, "There are likely, however, limits on how well hazard maps can ever be made because of the intrinsic variability of earthquake processes."
Image: San Fransisco after the 1906 earthquake, Wikimedia
Sep 6, 2012
http://www.nikeshoxjoggesko.com/nike-h?yh?lte-sko-c-395.html Chaussures Shox homme, Je parle quelque chose.Il ya une norme quantit de composants d'ikke indstille la contrainte de Penser Chaussures un produit jetable, y COMPRIS:
http://www.nikeairmaxsko.com/nike-free-sko-c-251.html Ce qu'un livrer fornemmelse confortable pour les femmes de faire certains qu'ils sont en mesure d'obtenir la opfyldelse de la manire dont ils font l'Essai. da utilisant dans le, men de maintenir stabil, Nike Shox femme peut-Tre la vrede Toute nouvelle aussi semblent selv transformer da plus l'attention dans la JournE. Le dveloppement de Nike Shox briser la mdecine conventionnelle de la parole, pas ternel mirakel du prjudice et peut ventuellement venir Quant la poursuite des Recherches soir par Jour.
http://www.nikeshoxjoggesko.com/nike-shox-sko-c-283.html La Chaussures nike shox r4 en prouv pour obtenir probablement la ud efficace qu'un seul au sein de votre marts. Shox R4 consacrer le meilleur plateau distribue kr
http://www.nikeshoxjoggesko.com/nike-air-free-c-352.html intervaller for dine f
The author leads with mentioning the three most recent damaging quakes: in China, Haiti and Tohoku, Japan. The date for the Tohoku quake is given as 2012 but it was actually 2011, March 11th. I'll never forget because I was in it. As for predicting, all of the posts are right on, in that in the current state, there is just not enough data and too many variables. As advanced as the computing and data crunching technology is today, the science is still by orders of magnitude deficient in ability to gather and analyze enough data in real time to predicting earthquakes with useful accuracy. There's a long way to go before that becomes a reality.
As Dr_Zinj says, heterogeneity of the earths crust is one difficulty, as are the data problems Dr. Stein raises, but the more fundamental problem IMHO, is that large earthquakes almost always begin at 10-15 km depth, and we make measurements at the surface. Thus between our instruments and the point where the earthquake will begin, are 10-15 km of wet, heterogenious rock. And because drilling is prohibitively expensive, we know little to nothing about the physical state of the rock below just a km or two. So the problem is very different than most of the other physical problems that are cited as examples of successful prediction. We don't know exactly what the material is, we don't know its physical state, we can't get any instrument closer than 10-15 km, and between our instrument and the action is hundreds of cubic km of wet heterogenious rock that damps out, or completely absorbs, the information needed about what is happening at depth. Very hard problem -- will probably take a conceptual breakthrough -- a completely different way of looking at the data that can be obtained.
Earthquakes originate when rocks far underground under stress suddenly break, and release that stress by moving in a quick jerk. No place underground in the world is the same as any other place. Which means that every peice of ground has a different breaking point. Since it is impossible to know the composition of all the Earth's crust from the surface to 50 miles down, nor exactly how much stress any one peice is under, it is impossible to know at what point that particular peice of ground will fracture.
A lot of earthquakes are hundreds of miles deep. Even if you know where and when an earthquake can strike, there are too many variables to determine how much the ground will shake.