Geologist hat on.
Feb. 26th, 2011 02:46 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
kaberettDisclaimer: this isn't actually my field; I'm specialising in - essentially - how water (and carbon dioxide and other volatile components) affect melting processes, and magma & eruptive processes. Having said which, I shall continue.
There is no good reason for people to die in earthquakes.
Earthquake prediction is impossible and unnecessary. The way to save lives is to apply rigorous building standards. There are hospitals in California that don't even notice magnitude 7 quakes: not a bottle falls over, they are that well isolated. The same size quake in Iran will kill hundreds of thousands of people. This is not inevitable and is not necessary; it all comes down to the ability to afford - and afford to enforce - regulations for quake-proof buildings.
There are good reasons for people to settle above faults: fault planes are composed in part of a zone of very fine-grained ground-up mess, where rock's been pulverised in earthquakes past. Any pore space in this material is rapidly closed up. So when water comes into contact with a fault plane it will frequently rise and appear at or close to the surface, at or close to the surface expression of the fault. So people come and live near it. And then the fault goes. And the earthquake kills people.
I say "earthquake prediction is impossible". This isn't quite true: there's a number of tools we can use to work out roughly how frequently major earthquakes will occur on a particular major fault - but this is with uncertainties of the order of years, which while good in geological terms is no use whatsoever if you want to evacuate people.
Fundamentally, we simply do not understand how stress is stored, or even whether an earthquake represents the release of all the stored stress or only some of it (this is "strong" versus "weak" faults, and a lot's been written on the subject, particularly regarding the San Andreas Fault). Our predictions of how frequently a fault will go are best in places like Greece, where there are frequent magnitude 6 faults and a wide variety of methods (largely to do with sea level) that mean it's possible to make reasonable estimates of the long-term slip on the faults averaged with time. For the fault zones with larger, more irregular earthquakes, we can't do that. And people are realising that the big-and-showy events aren't the only way that stress can be released: now we're looking for it, we increasingly see evidence of "slow" or "silent" earthquakes, where instead of releasing the accumulated stress in fractions of a second, it takes hours or days.
But all of this is a distraction from the fundamental point that people do not need to die in earthquakes. And nations with the money to do something about it damn well ought to be.
There is no good reason for people to die in earthquakes.
Earthquake prediction is impossible and unnecessary. The way to save lives is to apply rigorous building standards. There are hospitals in California that don't even notice magnitude 7 quakes: not a bottle falls over, they are that well isolated. The same size quake in Iran will kill hundreds of thousands of people. This is not inevitable and is not necessary; it all comes down to the ability to afford - and afford to enforce - regulations for quake-proof buildings.
There are good reasons for people to settle above faults: fault planes are composed in part of a zone of very fine-grained ground-up mess, where rock's been pulverised in earthquakes past. Any pore space in this material is rapidly closed up. So when water comes into contact with a fault plane it will frequently rise and appear at or close to the surface, at or close to the surface expression of the fault. So people come and live near it. And then the fault goes. And the earthquake kills people.
I say "earthquake prediction is impossible". This isn't quite true: there's a number of tools we can use to work out roughly how frequently major earthquakes will occur on a particular major fault - but this is with uncertainties of the order of years, which while good in geological terms is no use whatsoever if you want to evacuate people.
Fundamentally, we simply do not understand how stress is stored, or even whether an earthquake represents the release of all the stored stress or only some of it (this is "strong" versus "weak" faults, and a lot's been written on the subject, particularly regarding the San Andreas Fault). Our predictions of how frequently a fault will go are best in places like Greece, where there are frequent magnitude 6 faults and a wide variety of methods (largely to do with sea level) that mean it's possible to make reasonable estimates of the long-term slip on the faults averaged with time. For the fault zones with larger, more irregular earthquakes, we can't do that. And people are realising that the big-and-showy events aren't the only way that stress can be released: now we're looking for it, we increasingly see evidence of "slow" or "silent" earthquakes, where instead of releasing the accumulated stress in fractions of a second, it takes hours or days.
But all of this is a distraction from the fundamental point that people do not need to die in earthquakes. And nations with the money to do something about it damn well ought to be.
(no subject)
Date: 2011-02-26 09:08 am (UTC)