What's All This Quakey Shakey Stuff, Anyhow?

Oct. 24, 1991

You only have to read the newspaper occasionally-any newspaper-to be aware of the earthquakes that occasionally rock and rattle various areas of California.  People who live in the hills above Santa Cruz are VERY aware.  For several years, I've proposed that we Californians need a good earthquake detector.  Then suddenly I came up with 3 good ideas-three ideas with different levels of sophistication and usefulness.  When a major scientific panel came up with an idea quite similar to my best idea, I asked the editors at Electronic Design to expedite this column.

Originally, a friend of mine, Carl Nelson, proposed to make a quake detector/recorder out of 10 cans of beans.  You put one right near the edge of the pantry shelf, one further back, etc., and the last one you GLUE down to the back of the shelf.  If the front one rolls off, you had a little quake; if the last one is smashed by falling beams, you know that was The Big One-and everything in between.  But, you need to do this in at least 2 dimensions, in 4 directions, and most people would find this bulky, expensive, and dangerous.

So my first invention, costing as little as $1, is to set up a stack of 30 pennies, a stack of 25 pennies, and stacks of 20, 10, and 5 pennies, on a standard sheet of paper.  After a quake, you can see which of the piles fell, and in which direction.  I set up one of these at home and one at work.  During the big 7.1 quake, at home, the stack of 30 pennies fell, and the 20, but not the 25.  At work, everything fell except the 5.  So, for a small investment, you can compare notes with your friends.  It's not absolutely calibrated, but you can get a ballpark indication of the amount of shock in your area.

I was still not completely satisfied, so I designed another detector-recorder.  It's not exactly as technically correct or elegant as a seismograph built with levers and 10,000 turns of wire, as shown in The Scientific American.  But you can use the little circuit shown below, to be triggered by a "swaying pendulum," and make a stretched-out pulse of perhaps 60 or 120 seconds.

This pulse can turn on your tape recorder or your camcorder for a minute or two to detect the creaking or rattling of things in your house.  If you hear sirens, then the neighborhood may be on fire.  If it also turns on a little radio, you can hear the radio announcer say, "Oh, my Gosh, that's a big one." If all the books fall out of your book-case, that will all be recorded, too.  I just got it built up this weekend.  I haven't yet seen how touchy it is about false alarms, but I can live with that....

On a more serious level, I realized that every time I climb under my car to work on something, or jack up the car even to change a tire, or climb up a ladder, I wish I could get an early warning of a far-off quake.  When the workers at the Cypress Structure in Oakland were using their heavy equipment to demolish the ruined concrete, they had the benefit of a little radio transmitter that would detect any quake down by Loma Prieta, the epicenter.  A little radio receiver at the work site would give them 10 or 20 seconds of warning to jump off their crane, to get away from the dangerous piles of rubble, and increase their safety considerably.  I don't know who engineered this, but as far as it went, it was a darned good idea.  It must have made those brave workers feel a little less nervous about a quake sneaking up on them.

Well, I want to take that idea and expand it.  I want to add in a bunch of features and make it available to everybody:

  •  I want to have a whole bunch of quake detectors, scattered all around the San Francisco Bay Area, located at interesting sites near and along each major fault.  (Los Angeles can have their own network, and so can Tokyo).
  • I want all of the sensors to transmit the warning of any significant quakes to a suitable central station, which can process the information and send it out immediately on two or more radio stations or TV channels.

The first radio station can monitor ALL quakes, large, medium, or small, and broadcast the information, and you can tune it in if you're going up on a ladder or under your car, or if you are just interested or curious.  This will be a clear channel that will broadcast nothing but earthquake information and other related emergency info (tsunami reports, etc.), with an occasional tone to let you know you really are tuned in.

The second radio station would monitor all BIG quakes.  Then you might just leave this station on all the time, even when you're asleep, so if it wakes you up with a computerized voice saying "Big one, Loma Prieta, Big one, BIG ONE...." you would have some warning to get under a table or a doorway, to grab a flashlight or a video camera, or to head for your kids' bedroom, or whatever you have decided in advance to do.

This second channel would be available to any other radio station, for a small charge, so it could break in and give you advance warning, even if you're listening to a ball game, or the opera, or whatever.  It could add onto a TV station, and break into the regular program material.  The number of useful, life-saving possibilities is large but finite.  You can use your own imagination.

Let's say you're listening and you hear about a quake starting in Sonoma.  If you live in San Francisco, you know that's a few dozen miles away, so you have a number of seconds (at the rate of about 1.8 second per mile) to get in a safe mode.  If you live in adjacent Napa county, you might have only a few seconds.  So, the broadcast would have to tell you the location of the strongest shock, from a standard list of places, and you would have to be prepared to act suitably and instantly, depending on where YOU are located.  You have to plan in advance, and you have to recognize all of the places on the standard list.

The sensors have to give you a very quick indication of what is the significance of the magnitude of the quake.  I read in one popular scientific tabloid that some scientists were planning to set up a network of sensors and (digital) computers to analyze the data, and then give a warning in just "15 or 20 seconds." Don't look now, guys, but most of the people closest to the quake will be hit with the shock in only 5 or 10 seconds.

A perfectly computed analysis in 15 seconds would be no help at all.  The key to the sensor is to have some wide-range logarithmic amplifiers and sensors that can put out appropriate signals for every size of tremor.  Back in October of 1989, a friend of mine was standing in his garage, about 3 miles from the epicenter.  He said, "In the first second, it knocked me on my butt".  If a quake of that magnitude comes along, you can have pretty simple sensors, linear and logarithmic amplifiers, and discriminators or comparators that will get out the message "in the first second," which is the right kind of warning to save lives.

What if the quake's center is so deep underground (as it was in the Loma Prieta quake of 1989) that the shock waves hit areas several miles away from ground-zero, at just about the same time as it arrives at the sensors at ground-zero?  Easy-drill a deep hole and put a few sensors down deep, to give advanced warning.  Every mile of depth can help another few vital seconds.

Obviously, there are lots of practical and legal considerations.  What if some driver gets the message and jams on his brakes, causing a pile-up worse than the quake damage?  What if people get nervous and panicky and the first thing they do is call their lawyers?  Obviously, the practical considerations aren't trivial.  Still, if we plan a little, this system can be much better than no warning at all.

When I had this radio-warning idea back in December of 1989, I took it immediately to our Patent Committee at National.  They considered the scheme and decided it wasn't related to any of National's business, so they told me I could do anything I wanted to with it.  I sent a technical note to a couple of the major radio and TV stations and newspapers in the Bay Area.  The silence was almost overwhelming.  But in late August of 1991, a National Research Panel of experts from the National Academy of Sciences proposed a similar plan to assemble a system of seismometers, computers, and radios to give people an early warning*.  I immediately wrote off to the chairman, Mr. I. Selwyn Sacks of the Carnegie Institution in Washington.  Soon I hope to hear more about their proposals.  But you have already heard about my plans and proposals.

So, if this system gets going and it saves your life someday, you can buy me a beer.  Fully-paid licenses, of course, are available at a very reasonable fee.

You may not have many earthquakes in your area, but you have to agree, it's certainly a fascinating and challenging set of problems to think about.  Maybe you have ideas that are better than mine, or that would improve mine when added to it.  You don't have to sit on top of a big fault to have good ideas.

All for now./Comments invited!/RAP/Robert A. Pease/Engineer

Address:  Mail Stop C2500A National Semiconductor P.O. Box 58090 Santa Clara, CA 95052-8090

*"Experts Push High-Tech Quake Detection System," Warning of even a few seconds could save lives; San Francisco Chronicle, August 28, 1991, page 1.


To join the conversation, and become an exclusive member of Electronic Design, create an account today!