Inkwell: Authors and Artists

    
Inkwell usually hosts discussions about books, but this conversation
is a little different. The recent Great East Japan Earthquake and
tsunami in Japan (http://en.wikipedia.org/wiki/Earthquake_Japan_2011)
was, according to Wikipedia, "one of the five most powerful earthquakes
in the world overall since modern record-keeping began in 1900,"  The
9.0 magnitude quake actually shifted the earth on its axis and moved
Honshu 7.9 feet east, and triggered a major tsunami. This massive quake
triggered much interest in, and discussion about, earthquakes in
general. Here on the WELL, we have a forum dedicated to discussion of
earthquakes, and it's been busy.

We've invited the forum host, geologist Andrew Alden, to spend the
next couple of weeks in Inkwell talking about this recent quake, and
earthquakes in general. According to Andrew, quakes are "natural, more
widespread than people think, challenge society because of their long
cycles exceeding human lifetimes, are survivable and interesting, can
topple the mightiest nation if not handled well, and are not well
reported by American media." What can we learn from Japan's example?
From Sumatra's example?

Andrew Alden has been writing about geology for About.com since 1997,
but in fact his involvement with seismology and the public began on the
Well five years earlier as host of the Earthquakes conference, whre
his baptism by fire was the Northridge earthquake in 1994. He has a
B.A. in geology and worked for the U.S. Geological Survey for a span of
eight years that included the Coalinga earthquake and the eruption of
Mount St. Helens. He experienced the 1957 Daly City quake, the 1989
Loma Prieta quake, and countless smaller jigglers in the San Francisco
Bay area. He lives in Oakland, not 3 kilometers from the Hayward fault.

Mark McDonough will lead the conversation with Andrew.  Mark has been
a geology enthusiast since childhood, when he actually enjoyed being
dragged to quarries and fossil beds in upstate New York by his
"rockhound" father.  After completing about 2/3rd of a geology major,
he realized that his skills in allied sciences like math
and physics fell a wee bit short of those required by a professional
scientist, but he's kept in touch with the field ever since.  He has an
interest in the history of mining and has managed to visit many
Western mining sites without once falling into a shaft.  He is also
interested in catastrophes of all kinds - earthquakes, asteroid
strikes, tsunamis, you name it.   He maintains a somewhat rarely
updated web page on the subject at
http://www.well.com/~mcdee/Catastrophe.html

Other relevant links: 
Earthquakes at About.com:
http://geology.about.com/od/earthquakes/Earthquakes.htm 
World quake map: http://earthquake.usgs.gov/earthquakes/recenteqsww/

Welcome, Andrew and Mark!
  

    
Howdy!

Andrew, to start things off here, just how big was the recent quake in
Japan compared to past earthquakes in the US (for example, the San
Francisco quake of 1906)?  

Are there any places in the US subject to quakes that large?
  

    
Sizing earthquakes isn't an exact science, even though we are always given
precise-looking numbers for them. The most basic scientific measurement is
the magnitude, which measures purely the energy involved. That's like
ranking basketball players by their height alone. We'll get into that more
over the next two weeks, I'm sure, but for now I'll stick to magnitude.

The Japan quake is being called a magnitude 9.0 event at the moment. Things
change as the seismologists get more complete records from around the world;
for instance the Sumatra quake of 26 December 2004 was upgraded a couple
tenths of a point. But I think 9.0 will stand. The U.S. has had a quake
larger than that -- the 1964 Alaska (Good Friday) event, which was a 9.2.
The 1906 San Francisco quake doesn't even make the list of the biggest
quakes, being something like 7.8. But I would argue that it was bigger in
the human sense.

We also had a comparable earthquake in 1700, in the Pacific Northwest. The
tectonics of the Pacific Northwest is the same as Japan, and the same as
Sumatra. So what we saw "over there" is very relevant to the United States.
The 1700 quake is estimated to have been a magnitude-9 event, and what
happened before will surely happen again.
  

    
Yes, there was a certain amount of criticism of the Japanese for not
realizing that Fukushima was subject to enormous earthquakes and
tsunamis.  If I recall correctly, the last major earthquake of that
size in the area was almost 1,000 years ago.  Granted, we have the
excuse that only Native Americans were living in the Pacific Northwest
in 1700, but I've never talked to anyone from that part of the country
who had any idea that it was subject to major earthquake hazards.

Are there any other earthquake hazard areas in the US that might be a
surprise to people?  I felt a couple (very small) when I lived in
Providence, Rhode Island.
  

    
I may live in an especially well informed cocoon, because I know that almost
no place in the US is NOT affected by earthquakes. But I do think
nonetheless that awareness is growing. It ought to be part of every state
history course in Massachusetts, for instance, that Boston was widely
damaged in 1755. Virginia had a damaging magnitude-5 event in 1897.
Charleston, South Carolina, had a major event in 1886, estimated at M 7.3.
The Missouri-Arkansas-Tennessee-Kentucky region was devastated by a series
of magnitude-7 and 8 events 200 years ago; in fact, this coming Thursday
there will be a huge earthquake drill involving those states and seven more,
stretching from Oklahoma to South Carolina as part of a commemoration: see
<http://www.shakeout.org/centralus/> for that.
  

    
Yes, I think that geology buffs do live in a well-informed cocoon on
these sorts of issues.  I lived in Boston for a number of years and
never ran into anyone who had any idea there could be a serious
earthquake there.

I think more people are aware of the New Madrid earthquakes of 1811-12
- those Missouri-Arkansas-Tennessee-Kentucky quakes.  Ironically, some
of that awareness comes from a bizarre episode about 20 years ago when
someone widely convinced people that another huge quake was going to
happen any day.

Which raises the whole issue of earthquake prediction.  Is it still
pretty much useless?  

It seems I only run into two sorts of predictions - very specific ones
by cranks that turn out to be wrong, and very general ones from
scientists which are presumably more accurate, but almost equally
useless ("there is a 50% chance of a magnitude 7 or greater earthquake
in [wherever] over the next 40 years").

Any signs of progress on that front?
  

    
Seismology is still a young science, and a difficult one especially if you
want to help people. Many seismologists point to that motive, and I think
even the most mathematically oriented theorists acquire it once they've been
through the process of turning earthquakes into data. So let's grant that
every seismologist really wants to predict earthquakes, not just model the
deep Earth and map the interior and so on. But there are only whispers of
progress in truly predicting earthquakes. The search party has checked out
the obvious places, so to speak, and has since spread out into the
underbrush looking for clues.

Where seismology has made a big difference is in mapping active faults and
measuring plate motions all around the world. That's good, basic data. They
claim that earthquake forecasts, rather than predictions, are now useful --
as you say, X percent chance of a magnitude-Y event during the next Z years
in area W. But even those are primitive science, no offense meant to the
seismologists because it's the best they can do. And we know that the
forecasts fail, in that most of the notable quakes have never been
forecasted. That is a discouraging way to conduct science, and most
practitioners are resigned to incremental advances.

That attitude, by the way, accounts for some of the vehemence with which
seismologists respond to would-be predictors. Seismologists have had their
hopes cruelly dashed for a hundred years. I won't speak for them, but I hate
to see lay people clutch at rumors and espouse the most obvious flim-flam
when it comes to quake prediction. I hate it because I sympathize.
Earthquakes are nightmarish things, coming with no warning and disrupting
the foundation of the solid ground itself! Earthquakes are like the very
threat of death; they're practically designed to create anxiety. For a
highly abstract science, seismology is fraught with strong emotions.
  

    
Yes, and anyone who's experienced an earthquake can understand why. 
I've been lucky - I think my biggest was a 3.5, but even that was a
little unsettling.

Speaking of magnitude, the recent Japanese quake was a 9.0.  Is that
about as big as they get?
  

    
They get a lot bigger, but it doesn't look like much on the magnitude scale.
The largest quake on record was 9.5, the 22 May 1960 Chile earthquake. Oh,
just half a unit greater -- but the magnitude scale is logarithmic. That
number represents an exponent. So the latest Japanese event, at 9.0, is a
small fraction of that in terms of energy, about one-fourteenth the size.
The 1964 Alaska and 2004 Sumatra quakes had five times the energy of the
March 2011 event.

Whatever. They're all horrible.

The highest realistic magnitude possible on Earth? Well, let's look at
Chile, the biggest. Its rupture zone was estimated to have been 1000
kilometers long, and there's about 1500 km of possible rupture available on
that edge of the plate boundary. If you crack all of that at once, you've
got one and a half times the energy, or another 0.2 magnitude units, for a
theoretical maximum of 9.7. That's close enough for me. You simply can't
line up enough lithospheric rock and stress it to rupture to get a larger
event on this planet.

But frankly, if you're near the thing when it lets loose, a magnitude 7
event is just as bad as a larger one. The experience is that the bigger
events last longer, so the rubble bounces more.

You also have to recall that the March quake was complicated, as it was in
Indonesia, by a monstrous tsunami. It will take some concerted research to
separate their effects. We will find, as we did in 1906, that buildings
could survive shaking but not the aftermath -- fire in San Francisco's case,
flood in Japan's.
  

    
It seems like we've been getting an unusual number of big quakes
recently.  There was Chile in 1960 and Alaska in 1964, and then nothing
of that magnitude for decades.  Suddenly we've had two major quakes
fairly close together.

Is it possible that really major earthquakes do come in groups, or is
it more likely to be just the random bunching of fairly rare events?
  

    
It's even worse than that: we had no magnitude-9 events at all before 1952
(there was a 9.0 in Kamchatka that year). But yes, truly random events are
not evenly spaced -- if they were, that would be a pattern. We have no way
of saying anything about it given the shortness of the record, which is
barely over 100 years.

What I said earlier about the imprecision of earthquake measurements applies
doubly to events before the advent of seismometers. We look at those
earthquakes through the evidence of written accounts and that's pretty much
it. Descriptions like that can tell us about the intensity of an earthquake
-- how *bad* it is at location X rather than how *big* it is in an absolute
sense -- and we can only make estimates, with a large amount of uncertainty,
about its magnitude.

Intensity, Burton Richter once said, is like how strong a signal you get
from a radio station, whereas magnitude is the power the station broadcasts
at. You can see how fuzzy the numbers might be when you try to estmate
magnitude from written accounts of intensity.

Seismologists are scientists, and they are deeply familiar with the concept
of uncertainty. We can say exactly how fuzzy the numbers are, and we must be
constrained by that fuzziness. It's like the same problem in astronomy, when
a telescope gives us a fuzzy image. The scientist knows that trying to
sharpen that picture yields untrustworthy artifacts. All of this goes
profoundly against human nature; we're wired to see patterns in uncertain
conditions. So in sum, lay people see patterns in randomness that feel
valid, while seismologists look like fuddy-duddies for holding back and not
seeing the face on Mars, as it were. If we had descended from cats instead
of monkeys, we might be better scientists. But I suppose we'd be prone to
different illusions. Anyway, we have no good evidence to say anything about
long-term global trends in earthquakes.
  

    
If you're reading this discussion but you're not a member of the WELL,
you can still ask a question or add a comment by sending via email to
inkwell at well.com.
  

    
When you say no 9.0 events before 1952, you mean since reasonably
accurate and modern records were kept, right?  How far back can we go
in terms of estimating the magnitude of historical earthquakes?
  

    
> If we had descended from cats instead
of monkeys, we might be better scientists.

That gets my vote for quote of the week.
  

    

Welcome, Andrew and Mark!  Thank you for this topic.

> If we had descended from cats instead of monkeys, we might be better scien-
> tists.

I just love that.

I am a California coastal resident for all of my 57 years, and I have lived
through a couple of major quakes, literally hundreds of smaller but
perceptible quakes and thousands we never even noticed.  I suppose we're all
pretty philosophical about it, but I prefer this to the certainty of
crippling weather every year in other regions, etc.

I want to put in a good word for Andrew as a geology guide.  He took a
carload of WELL neighbors on a tour of the Bay Area a while back, showing us
some of the oldest exposed rocks in this amazingly complex region.  He's very
good at explaining this stuff!

Can you tell us a bit about why the Pacific Rim is referred to as the "ring
of fire" and why there are more quakes and vocanoes here than around the
Atlantic?

(P.S.  John McPhee's "Assembling California" concludes with a lengthy account
of the 1989 Loma Prieta quake and how it went down in various corners of the
affected area.)
  

    
Like-minded slippage!
  

    

Like many others, I'm always interested in reading about "what are the
odds it's going to happen to me". But I'm also aware that we've had quakes
that were on an unknown fault lines and I've even read in the media (I
know, I know) that quakes can occur not on a fault line presumably due to
other geological forces.

Is the later true or just media mis-reporting?

Another question that I tried to research and failed was to try to see how
close my house is to a fault line. I could not find detailed enough maps
to indicate that. Are such maps available online?
  

    
I am going to spam the world with that cats/monkeys quote, with
attribution of course.
  

    
Mark: Seismographs were first devised in the late 1800s, and only after 1900
was the most rudimentary worldwide network possible. Ever since then, we've
caught all the great (M 8 size) events. Anything older you might call
prehistoric. But our estimates of earlier events get better as we learn
about the typical behavior of different fault zones and early building
types. As ancient records are translated and analyzed, we're getting a
better idea of longer-term seismicity. It's not a dead end for research,
especially for records like the Chinese ones.

Still older earthquakes are studied with methods similar to those that
archeologists use: digging trenches across active faults and dating the
traces of old quakes. Paleoseismology is a field with great potential, and
California does a lot of it. We can go back as far as carbon-14 and trench
safety allow, potentially tens of thousands of years. That ought to be
enough to capture the typical earthquake cycle even in the quietest active
regions--assuming our working theory is valid.

David: To visualize the "ring of fire," start with the picture of a big pot
of soup simmering on a stove, with the water stirring by convection and
islands of scum floating on top. The stove burner is old and funky, so one
side gets a little more heat than the rest. Most of the time the soup
surface is clean on that side, but every now and then the scum breaks up and
wanders over it. Very simple picture, in fact too simple.

Take that pot and turn it into a sphere instead, covered with soup. Turn the
heat way down and blow on the soup instead. The tectonics of the real Earth
is governed by cooling of the top more than heating from below, but you'll
still get clumpiness, with the scum (continents) drifting around and joining
and splitting while the patches of clean surface (oceanic crust) grow and
shrink. The Pacific, today, is a very large patch of oceanic crust that
rises into existence at a long seafloor spreading zone near South America,
spreads apart in both directions, and plunges back down at the edges. The
plunging part is where long volcano chains typically form. So (going
clockwise) you've got volcanoes in New Zealand, eastern Indonesia, the
Philippines, Japan, Kamchatka, the Aleutian Island arc, the US Pacific
Northwest, and Central and South America. That's the ring of fire.

The Atlantic is a much simpler ocean, one that's opening up passively as the
continents on either side pull apart. There's none of the subduction that
feeds volcano chains anywhere in the Atlantic, except in the eastern
Caribbean. It gets earthquakes and also tsunamis, some of them quite large,
but they aren't like those you get around the Pacific.

I cover the basics of plate tectonics on About.com at
<http://geology.about.com/od/platetectonics/a/About-Plate-Tectonics.htm>
  

    
Is the structure of Iceland different than in the Caribbean?
  

    
Some slippage going on there while I was composing the last remark...

jmcarlin: Earthquakes can happen without the ground breaking; in fact most
of them don't break the ground. So in that sense you can have quakes without
"faults." And here's the thing: mapping the Earth's rocks and faults and so
on is damnably hard. Everything is covered with soil and landslides and
cities and water. The rocks themselves have to be studied under the
microscope and the newer generation of electronics. Making a geologic map is
a tremendous effort, and each generation in a field area upgrades and
reinterprets the map, and even so the result is a blend of data and
carefully directed imagination. So it's no surprise that faults can elude
detection until they rip apart in an earthquake.

But in pure, simple physics you can't release energy within a solid body
except along a planar surface of some kind. Unless you have an explosion, of
course, but they have a very distinct seismic signature.

About maps of faults, only California (that I know of) requires active
faults to be mapped, which it's been doing since the 1970s under the
Alquist-Priolo law. Those are online at
<http://www.quake.ca.gov/gmaps/ap/ap_maps.htm>

I'd love to know if other states do anything similar.

Peter: Iceland is very different; it sits right on the spreading line in the
center of the Atlantic. It's the opposite of a subduction zone. There are
two theories on why it's so volcanically prolific. One is that there's some
sort of really deep source of fresh magma that comes up in a "plume"; that's
the hotspot theory. The other is that a large piece of unusually "fertile"
(that is, lava-yielding) rock is sitting there where subduction left it long
ago. Hotspots are the textbook consensus, but I subscribe to the
alternative, as longtime About.com readers may know. I treated the subject
most recently in a book review:
<http://geology.about.com/od/bookreviews/fr/Plates-Vs-Plumes-By-Gillian-
Foulger.htm>
  

    
>with the scum (continents) drifting around and joining
and splitting 

You calling my continent scum?

But seriously, that is one wicked lucid explanation!
  

    
Thanks, Gary.

I guess the upshot of what I've been saying is that there are all kinds of
earthquakes. They all result from some sort of stress being relieved. More
precisely, they are strain being relieved. If you take a block of rubber and
deform it, whether by stretching or squeezing or twisting it, that
deformation is strain. The strain can be relieved elastically (if you let it
spring back) or plastically (if it somehow settles into that deformed shape
permanently) or brittlely (if it ruptures). Rock is the same way; everything
is at least a little bit elastic. We heard after the Japan earthquake that
the island of Honshu moved several meters east. What really happened was
that Honshu had built up strain, just like the rubber block, as the Pacific
plate moved against it over the centuries. The quake was the result of a
rupture between the Pacific plate, which is trying to move west as it's
being pulled down by subduction, and Honshu. That was brittle failure. Parts
of that fault between the two plates surely  had plastic deformation too;
that's what we see in faults that have been exposed on the surface. And
usually there's some elastic strain left over afterwards. Anyway, Honshu
didn't move several meters east so much as it rebounded several meters east.
  

    
OK, I have a bit of a trivia question for you - except the effects on
history were large.

The city of Lisbon, Portugal was virtually destroyed by earthquake,
fire, and tsunami in 1755.  This basically ended Portugal's reign as a
great power and some say set off a wave of skepticism about the concept
of a just and loving God, accelerating the Enlightenment.

Be that as it may, why Lisbon?  I just don't think of the Iberian
peninsula as earthquake country.
  

    
As an addendum to that, aren't the Azores are known for undersea
slides?
  

    

> in pure, simple physics you can't release energy within a solid body except
> along a planar surface of some kind.

Curius about this, if it isn't too much of a digression.