18 comments

[ 5.2 ms ] story [ 65.0 ms ] thread
Seattle is another area I would worry about as far as preparedness for a Major earthquake. Most of the new skyrises are probably safe, but the Alaskan Way Viaduct would most certainly collapse (it was nearly destroyed due to a moderate eathquake 10 years ago, but due to local politics no suitable replacement/repair has been made), as would most of the waterfront, whose foundations have been eaten away by something I've heard called 'gribbles.' Additionally I would worry about many of the much older buildings downtown, which have also been damaged before by moderate earthquakes.
Sell short catastrophe reinsurance bonds?
Tectonic plates can remain irrational longer than you can stay solvent.

Or something like that.

(comment deleted)
Hello from the East Bay, home of the Hayward Fault Line:

The last five major events were in 1315, 1470, 1630, 1725, and 1868, which have intervals of about 140 years (note that 2008 is 140 years from the major 1868 event).

If a major earthquake were to occur on the fault, damage would be catastrophic. More than 1.5 trillion U.S. dollars in property exists in the affected area, and more than 165 billion US dollars in damage would likely result if the 1868 quake were to recur.

The estimated probability of a major earthquake on the Hayward within the next thirty years was estimated at nearly 30 percent, compared to about 20 percent for the San Andreas Fault, which can have larger earthquakes but farther away from a significant portion of the urbanized parts of the Bay Area.

(various rips from Wikipedia)

Yup, and one of the major concerns in the East Bay are so called "soft story" buildings: homes and apartment buildings with ground floors largely open to create garage and retail space and therefore lacking adequate shear structure to resist earthquake forces.

It's a well documented problem, with huge loss of life and housing stock predicted in the event of a large earthquake on the Hayward fault. Not sure exactly what measures other cities are taking to improve the situation, but you can learn more on the City of Berkeley site:

http://www.ci.berkeley.ca.us/contentdisplay.aspx?id=622

There's a list where you can look up whether your building is structurally vulnerable or has been retrofit.

(comment deleted)
"The average time between magnitude 8 and larger Cascadia earthquakes is about 240 years."

The outliers on that graph make me extremely skeptical of an average[1]. Eyeballing it[2], the most common delta seems to be close to 300 years. Doesn't using the mode make the most sense here, from the perspective of highest-probability outcome?

[1] Should we just outlaw the average? Perhaps a real statistician can advise.

[2] Earthquakes:

  -7800 -7200
  -7100 -7000
  -6750 -6500
  -6200 -5800
  -5500
  -5150 -4800
  -4450
  -3900 -3800 -3750 -3500 -3300 -3100
  -2800 -2500 -2400
  -2200 -1900 -1600
  -1500 -1200
  -1100 -900 -800
  -500 -300
  -100 200
  500 600
  700 900
  1200 1400
  1500
  1700
Deltas:

  600 100 100 250 250 300 400 300 350 350 350 550 100 50 250 200 200 300 300
  100 200 300 300 100 300 100 200 100 300 200 200 300 300 100 100 200 300 200
  300 200
Frequency distribution of deltas:

  50 1
  100 9
  200 9
  250 3
  300 12
  350 3
  400 1
  550 1
  600 1
I actually think the average makes sense here. The mode will give you '300', but you have 18 that have occurred at 100/200.
I see, makes sense.

There's outliers, but not that many and not that far off.

Using the 1.5 * IQR (Interquartile Range) rule of thumb, there don't appear to be any significant outliers.

    Q3 = 400
    Q1 = 200
    IQR = 400-200 = 200
    1.5 * IQR = 300
    300 + 400 = 700
    200 - 300 = 0 (can't have negative years)
All of the data points fall within that 0-700 year range.
What really matters is the conditional probability of earthquake given the time since the last earthquake. Based on the numbers that you just gave, about half the time when you make it to 300 years, the next one goes within 50 more years.

But the current best guess, as documented in http://www.nature.com/news/2010/100531/full/news.2010.270.ht..., suggests that the actual risk in the next 50 years is about 37%.

I've known about this one for years, and it is one of the reasons I don't want to move back to the Pacific Northwest (I grew up in Victoria).

It was only in the last few decades that geologists realized the size of the earthquakes that the area experiences. Building codes have not caught up. And a lot of the population centers in the area, including Seattle, Vancouver and Victoria, have important construction predating any useful building codes.

A few areas have notable risks.

The scariest is Victoria. It is on Vancouver Island, which typically drops 10 feet during a Cascadia earthquake. Most of the city is under 20 feet above sea level. Add a tsunami, and a lot of that could be covered. (I've read the projections for a tsunami from the Cascadia quake are about 30 meters. But I have no idea how big the tsunami would be within the Straits of Juan de Fuca, which is what Victoria would be exposed to.) Normally one would worry about the fact that densely populated areas (like James Bay) are largely built on landfill, which behaves poorly in earthquake. But that's incidental compared to being unexpectedly under the new sea level. I don't know exactly how many people would be under the new sea level, but I think it is about 100,000. (You can see the distribution of people in http://www.victoria.ca/cityhall/pdfs/sustainability-census-p.... All of Oak Bay, most of Victoria, a significant portion of Esquimalt, all of View Royal, some of Saanich, and a small part of Langford would be under water.) Given the unavailability of higher ground and low water temperature, I don't expect a lot of survivors.

Next look at the delta area of Vancouver. It is called the delta because it is the river delta from the Fraser River. Which means dirt with a high water level. Take dirt and water, and shake. What do you get? Mud! The delta area of Vancouver is perfect for http://en.wikipedia.org/wiki/Earthquake_liquefaction. There are about 100,000 people living there.

Now consider Mt Rainier. http://www.geotimes.org/apr04/feature_MountRainier.html documents some of the lahar risks that it faces. Odds are that a Cascadia earthquake wouldn't trigger one of those lahars with 150,000 people in its path. That will happen some day regardless. But if it did, and you had a lahar on top of a major earthquake, well, nothing good would happen.

Japan lost perhaps 10,000 people in this earthquake and tsunami. I firmly believe that a Cascadia quake would kill at least an order of magnitude more people in Victoria alone. And, unlike Japan, there is no real disaster preparedness ethic to mitigate the follow-up to the disaster.

Sorry if this is a stupid question, but would an 8-9 Cascadia quake have implications for the Yellowstone supervolcano?
That's something I sometimes (perhaps stupidly) worry about too. Some earthquakes go off over in Asia and set off earthquakes in California and Pacific Northwest and set off earthquakes in Utah and then the supervolcano in Yellowstone goes off...
In case it's not obvious, when people talk about faults being "N months pregnant" it doesn't mean it's extremely likely there will be a large earthquake in the next week/month/year, but rather on the order of magnitude of 50 years. The San Andreas in California has been considered overdue for a major earthquake for decades as well.