shake shake shake

September 30, 2008 8:31 AM

"The Quake-Catching Network is a collaborative initiative for developing the world's largest, low-cost strong-motion seismic network by utilizing sensors in and attached to internet-connected computers." The Economist's writeup notes that, since network communications are (sometimes) faster than the speed of sound in the earth's crust, a distributed network's observations of a temblor might reach a warning network before the quake itself reaches a traditional seismometer.

This appears to be the first use of Berkeley's Open Infrastructure for Network Computing as a sensor network, rather than for number-crunching.

posted by fantabulous timewaster (8 comments total)

...network communications are (sometimes) faster than the speed of sound in the earth's crust...

Sometimes?! Whoa. (Although what's a typical network speed in km/s?)
posted by DU at 8:43 AM on September 30, 2008

I'll bet there will be a lot of false positives from Italy, what with all that hot-blooded smashing of plates and all.
posted by StickyCarpet at 8:46 AM on September 30, 2008

Although what's a typical network speed in km/s?

Based on my ping times from St. Louis to servers in Sydney, Australia (~250ms round trip covering 14,577 km), I'd say about 116,000 km/s. The speed of light (or dang near) is pretty fast.

The (sometimes) caveat is probably because of things like sensor-equipped laptops using cellular network connections.
posted by jedicus at 9:08 AM on September 30, 2008

How about we all pick one time to vigorously shake our computers and give those Berkeley folks a taste of what they got themselves into when they recruited the average internet user for scientific purposes!

but seriously, it's cool - I like it. Neat idea.
posted by Salvor Hardin at 9:23 AM on September 30, 2008

DU, I get 100 millisecond round-trip ping times to the west coast, 2500 miles away.

The edge of a voltage change moves down a conducting cable at about half the speed of light (a "1 nanosecond delay cable" is six inches long). Optical fiber is a little faster, depending on the index of refraction of the glass. I'm 2500 miles = 13 light-milliseconds from Stanford, and I have a round-trip ping time of 100 milliseconds to qnc.stanford.edu and some other machines in California. So about half the transit time is signal propagation, and the other half is probably routing and processing delays.

jedicus correctly diagnosed the logic behind my (sometimes).
posted by fantabulous timewaster at 9:25 AM on September 30, 2008 [1 favorite has favorites]

I like the Quake Catcher Network so much I designed a logo for them for free. But I have a feeling they didn't like it, or maybe they aren't allowed to use it because of rules about accepting gifts or something.

Also, I can't be the only nerdly kid in earthquake country that has had this idea. I remember thinking that seismically aware automatic natural gas cutoffs would be a good idea. If I recall correctly I think at one point that major utilities were looking at developing an early warning network like this that would have used the electrical infrastructure or radio for signalling. You only need a few seconds worth of warning to get somewhere safe in a quake.
posted by loquacious at 10:01 AM on September 30, 2008

How come journalists are so eager to write "temblor" at every opportunity? It's not even a realistic word... you never roll into work and hear Larry saying "Man, Rick, did you feel the temblor last night?"
posted by crapmatic at 12:31 PM on September 30, 2008

The USGS set up an aftershock alarm for rescuers working in the collapsed Cypress Freeway after the 1989 Loma Prieta earthquake. The alarm gave rescuers 10 - 20 seconds warning.
posted by ryanrs at 6:51 PM on September 30, 2008