Slashdot Mirror
An Early Warning System For Earthquakes
Iphtashu Fitz writes "Would 15 seconds be enough warning time to prepare for an earthquake? It certainly wouldn't be long enough to evacuate from where you live, but it may be just long enough to get out of a building or brace yourself in a doorframe or under a solid desk. Italian scientists may have discovered a way to measure the initial shockwave of an earthquake two seconds after it starts, and from it predict the extent of the destructive secondary wave that will follow. It typically takes twenty seconds for the secondary wave to spread 40 miles, so sensors that can transmit warnings at the speed of light may provide just enough warning before a major quake for people to brace themselves. Even more importantly, such a warning could allow for utilities like gas companies to close safety valves, preventing potential fires or explosions in the aftermath of the quake."
(I'm not an expert on earthquakes, but 40 miles seems like a long way for the earthquake to travel.)
Shaking things up in the news today?
There would still be gas in the main lines, how would shutting a safety valve keep a broken pipe from leaking gas already in it?
Quiz: True or False -- On a scale of 1 to 10, what is your middle name?
My geology professor has gone over it again and again, Radon could potentially be an early warning system in some cases. Naturally the stuff leaks out of the ground, before an earthquake more of it is supposed to come out due to the shifting of the ground up to the earthquake. This could be months, days, minutes that this is detectable. Someone with a better understanding please correct anything and add to this.
What do earthquakes need to be warned early about?
"... brace yourself in a doorframe ..."
this is a myth. The only thing this acomplishes is broken fingers.
It stems from an observation from a red cross worker after a earthquake in mexico.(I think 1950ish.)
That archtecture of the entrance way was an adobe arch. Arches are very strong, as opposed wooden square door frames.
The Kruger Dunning explains most post on
Would 15 seconds be enough warning time to prepare for an earthquake?
Nope. But a few hours to a few days would be lots better.
is barely enough time to evacuate your bowels - let alone prepare for a large quake.
I thought the high speed trains in Japan would stop in the event of an earthquake (before the earthquake actually hit them), http://en.wikipedia.org/wiki/Shinkansen
"In the event of an earthquake, an earthquake detection system can bring the train to a stop very quickly"
Anyway, the idea of a broadcast system to warn of an earthquake is pretty obvious, the engineering task of doing it right without false positives is pretty difficult I bet.
ElarmS.
There are no hard-fast rules here. In many simple-wood frame houses here in the USA doorframes are usually a couple of 2x4's nailed together. However that is not to say every doorframe is that way. A bunker doorframe would do nicely, however not everyone has such a thing
..of crap. 15 seconds my ass.
:)
When you've lived in San Francisco and/or Tokyo as I have, you move without thinking.
And if that means climbing over you to get to the exit, then buddy you better duck, 'cause I'm coming thru
just huge strobes-- don't get on the overpass-- and the underpass...
*** HALT MFKR ***
or enter chasm
15 seconds enough to keep you from becoming a autobutter sandwich or a car contained base jumper?
every day http://en.wikipedia.org/wiki/Special:Random
While the article is quite light on details, much work has been done in this area, by groups such as ElarmS in California, if your interested in the methodology take a look at Allen's paper "Rapid magnitude determination for earthquake early warning (a 7 pg. PDF) which is reasonable understandable by lay persons if you skip through the math, yet still informative for people in the field.
Apocalypse Cancelled, Sorry, No Ticket Refunds
I kept a set of decorative cutlery on my headboard. Never once was injured during an earthquake, just suffered some facial lacerations and lost the sight in my left eye.
"Sometimes 15 seconds is all it takes to significantly increase your chance of survival."
You are right but only in some instances. I was at work in a factory in Emeryville, CA in 1989 during the Loma Prieta Earthquake. I was a mile or so from the Cypress Street Viaduct when it collapsed. 15 seconds in that case meant nothing at all. All we could do was hold on and hope the building didn't fall in on us. When you are really terrified it's nearly impossible to do much in 15 seconds.
Here in Greece we constantly suffer from nasty earthquakes. A company led by a well-known seismologist here sells products like the one mentioned. It also sells metallic boxes where you can hide inside during a quake. Too bad I haven't got any of these products, they could save my life one day. However, I believe there is no better protection against quakes than living in a flexible wooden house that 'moves' together with the seismic waves as they pass, instead of these stupid concrete boxes that break apart because they tend to resist against the seismic waves (and we all know nothing artificial can resist natural forces for too long, except for those things that are inspired by nature itself, like wooden houses). Also note that Ancient Greek temples never had any problem during 3000 years of earthquakes. Today building companies seek to maximise profits by keeping costs down, without researching how simple solutions like a stone over another stone sticked to it with some earth can drastically improve the behaviour of our houses to earthquakes.
"... brace yourself in a doorframe ..."
this is a myth. The only thing this acomplishes is broken fingers.
A very dangerous myth, too. Most of the deaths in Loma Prieta may have resulted from this myth.
There were 57 deaths attributed directly to the earthquake, and 42 of them were in the Cypress Street Viaduct collapse.
At the start of the earthquake, the drivers stopped. Because of the myth, most of them tried to stop under the arches. When the strucutre collapsed, the arches came all the way down to the pavement, pancaking the cars beneath them, while the regions between the arches had enough space that it was possible, in many cases, to survive the collapse itself.
- - - -
Of course a lot of the deaths there are attributable, not just to the quake, but also to governmental interference with volunteer rescue attempts.
Most of those who survived the initial collapse were still trapped in their cars or the structure itself. When the quake hit virtually all of the the nearby citizens dropped what they were doing (along with any inter-group animosity) and immediately began rescue efforts. (A notable part of this was workers at a nearby warehouse improvising an elevator using a dumpster and a forklift.) The pulled quite a few out of the collapsed structure's "sandwitch" in the first half-hour or so (at considerable risk to themselves, especially given the risk of further collapse or rolling debris due to aftershocks). Then the authorities arrived.
The police kicked them out and cordoned off the area to await the official first responders. They eventually arrived - around sundown. Then they had insufficient light (given the power failure) and mainly waited around further for portable lighting to arrive. It was several hours before rescue attempts, with a smaller force of official rescuers, resumed. (Of course by then the "golden hour" had long since expired and those who had been in shock were now dead or beyond hope.)
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Importantly, while 15 seconds is probably not enough time for something like a high speed train to stop, I'd much rather be travelling at 50mph rather than 200mph when the earthquake hits ;-)
I thought I read about this somewhere, or maybe it just came up in conversation, but I can't find the reference...
Anyway, the idea is this: If you have a laptop with a jolt/bump sensor (I have an IBM at work that does this, I'm sure others do to), you voluntarily run some software that knows where your laptop is by its IP address and when a shock hits it, it sends that info to a central server. Normally, it will just be noise... random shocks and drops coming from all over... but when all of a sudden the server receives reports from hundreds/thousands of laptops in one area and that area grows larger quickly, that data could be used to detect an earthquake.
Is anyone doing this? Could it be a possible early warning system? That would seem to be a fairly trivial piece of software to write, at least on the client end... the shock sensor goes off, ping the home base. Making sense of the data at the other end might be harder, but seems possible.
-S
--- What parts of "shall make no law", "shall not be infringed", and "shall not be violated" don't you understand?
This whole idea sounds pretty shaky.
Have you read my journal today?
This is somewhat like a project I did in my undergrad software engineering course. We spent the whole semester in a mock scenario role playing the software development cycle. We did everything up to implementation. Held requirements meetings with the Government of 'Claremount' - a fictional country prone to earthquakes and tsunamis (and this was before the big one that got all the attention), analysis of their needs, design of a detection grid and alert systems. Created UML diagrams for a java implementation, just didn't get to into the details. Thought about major classes, methods, attributes etc. Anyhow, it was an interesting project - and even more interesting to see how they will/are do(ing) it in real life.
"The computer revolution hasn't started yet. Don't be misled by the enormous flow of money into bad defacto standards fo
If things start shaking, I follow the advice on my safety pamphlet: go straight to the kitchen (room in the house with the least things to fall from above in my circumstance), get under the table, and stay there until the shaking stops plus a few minutes to ride out the aftershock and any settling of objects in the cabinets. Unfortunately the earthquake always seems to happen when I've in the bath or otherwise naked. I remember being scared out of my wits during my first earthquake (grew up in Illinois) and wondering "Now how will this look if the police come and find a stark-naked dripping white corpse under my kitchen table. The neighbors will be talking for months."
Help poke pirates in the eyepatch, arr.
...or you could and up being a crippled stud.
reminds me of the tornado warnings in middle america/central plains.
by the time the Hams radioed that one was on the ground, it was confirmed, and a signal was sent to ativate the system, the tornados were often over.
as my dad used to say, survivors will be notified.
we just listened to the hams our selves.
will there be a consumer version of this?
So you do drills. You practice moving without panic when an alarm goes off. Maybe you were terrified when your building started shaking, but when was the last time that ringing bells paralysed you with fear?
Solid engineering seemed to save this one. From University of Illinois professor Nathan Newmark, nonetheless.
http://bssa.geoscienceworld.org/cgi/content/abstra ct/95/2/708
I actually had the honour of proof-reading this paper before it was published. One of the authors is my wife's uncle.
This system is already in place and working today. It is based around a network of buried sensors that allow the accurate location of the epicenter within just a few seconds. The system is used to shut down high-speed trains, etc, before the damage-causing vibrations arrive.
Ringing bells alone won't paralyze you with fear (at least not normal people...), but when they mean that there's an earthquake and that the building you're in might collapse on you then they sure could.
For example, if you know somethings coming, even if it is only with a few seconds notice it would help alleviate the fear of the unknown that is so prevalent in humans. Plus, 20 seconds notice would be time enough for bridge operators to turn on warning lights and stop the traffic over bridges and under overpasses. That alone could save many lives in the event of a particularly nasty earthquake.
It's only paranoia if your wrong...
From TFA:
Primary waves travel around six kilometers [four miles] per second, covering around 60 kilometers [40 miles] in 10 seconds. Secondary, or S, waves, which are usually more destructive, travel more slowly, around 3.5 kilometers [2.2 miles] per second, covering only around 17 kilometers [11 miles] in 10 seconds. Therefore, a city located around 60 kilometers [40 miles] from an epicenter would have around 15 seconds of lead time to prepare for an earthquake's impact, the time difference between the arrival of the first P wave at a recording station near the epicenter and the arrival of the S wave at the city itself
So even if the sensor gave its warning the moment the fracture occurred, and it took zero time to send it, it would only give 6.19% more warning of the S wave than the arrival of the P wave itself. Add transit time from the depth of the epicenter and the distance of the nearest sensor from it, plus the two second delay while it computes the need to sound the alarm, plus the speed-of-sound delay from the alarm to your ear, plus the time it takes to recognize that the alarm is an earthquake warning, and you'd have to be pretty far away for the alarm to be more useful than just taking cover when the P wave hits.
Seems to me that earthquakes already have a faster warning system built into them - at least for warning humans - than any system that could be built on this discovery.
Now for warning our automation (such as the applications suggested in the story), which has inadquate "senses" for earthquakes but speed-of-light communication, electronic reaction times and controls mechanical processes for which a few seconds of warning might mean the difference between safe shutdown and major calamity, this could be great.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
The "We have 15 seconds before the shockwave travels 40 miles." Well, that's great. Except that it's only relevant IF the sensor is right at the epicenter.
In the real world, we don't know in advance where the epicenter will be. How many of these will we need to set out to get a reasonable expectation that we will catch the next earthquake? You'll need a fairly extensive net to be able to pinpoint the center. How much will this cost to set up and operate?
Over 150 workers, many near the center of a 300,000 square foot manufacturing plant full of very loud machinery, a hundred feet from the nearest exit with indirect narrow isles leading to the door, lighting provided only by emergency floods. Warning lights and bells for 15 seconds may seem like a lot to you, but in that situation it would have made absolutely no difference to those of us in that building. It would have been just enough time for a few to make it to the door and out. All the practice and drills will not prepare you for the real thing when it happens. You sometimes just have to take your chances where you stand. Sometimes the danger comes too fast for any reaction at all.
"when was the last time that ringing bells paralysed you with fear?"
When was the last time you rode out an earthquake in a building with no place to hide? I'm sure you are a very brave person but it is easy to just talk about fear.
If you really wanted to do that, you'd do it at the substation level. But I doubt you would. Substations already have circuit interrupting switchgear, houses have fuses and breakers, outlets in particularly hazardous locations have GFIs. Electricity won't leak out and start a roaring inferno like your gas service could.
I am not a crackpot.
The USGS has an extensive monitoring network setup near Parkfield, CA and a couple years ago a nearby 6+ quake had no warning signs. A few false alerts by Scientists, as they have had in Parkfield, CA, and the warning system will be scrapped or ignored.
Your workplace sounds like a really well-designed death trap. The warning is a good idea regardless of whether or not your particular place of work is designed to kill as many people as possible in case of earthquake-induced collapse.
Sir Arthur: Explain again how sheeps' bladders may be employed to prevent earthquakes.
Sir Bedemir: Oh, certainly, sir.
"If you mess with us, we're going to take you on, even to our utter destruction, whatever occurs." - Ralph Yarro (SCO)
The IEEE article is about what amounts to a strain gauge, which (if confirmed) tells you that something is about to crack but doesn't tell you how far and how widely the fault is going to unload itself. There's some reason to suspect that when an earthquake starts it doesn't "know" how big it's going to be.
Once the p-wave hits, though, you know what kind of ground acceleration to expect.
>a flexible wooden house that 'moves' together with the seismic waves
Unless it slides off the foundation. You need some rigidity.
Wood can be good, steel can be good, even reinforced concrete can be good. The thing you need that no building provides enough of is "damping", energy dissipation, what a shock absorber does. Imagine a car with no springs, then imagine a car with only springs, and you've got two lousy rides. The architectural equivalent of a shock absorber is a material which drags its feet when you deform it. Steel is wonderful for this: an earthquake can lose a lot of energy deforming a beam and then straightening it out again. Concrete will absorb energy by developing small cracks: the problem is that after enough shaking the small cracks join and get large and you have a structure of rebar holding gravel. The rebar can bow outward like a Chinese lantern and the floor collapses, unless the rebar was installed in a helical pattern in which case it may hold the gravel in place long enough to evacuate the building. Look closely at freeway construction in earthquake zones, and you'll see dense rebar that winds around the center of a column.
Wooden houses are great because an earthquake can wear itself out scraping plywood sheathing against studs.
You may be able to leave a two floor residence in 15 seconds - if you are prepared to start running every time of day or night. As it is, you probably wouldn't even realize what is happening before the wave hits. In highrise office buildings, there is no chance you can get out of elevator in time.
You work at a desk right? Some of us lesser folks work in factories. Factories full of machines aren't the tidy little places we would all like them to be.
You win this discussion. I'll just lean back in my chair and wait for the New Madrid fault to break south of where I live now. I hope I hear those warning bells first.
Um ... shouldn't WE be the ones being warned ?!?
Why don't the Italians just ring up the Japanese and ask them? They seem to be 4 months ahead. http://www.digi-help.com/disaster/japan-quake-warn ing-system.asp
Since March 30 of this year, Japan's Meteorological Agency has been operating a nationwide system [Japanese] to measure P-waves and estimate the earthquake's strength before the S-waves hit. While they say it's still experimental, it's been brought up in the news several times, and has in fact predicted [Japanese--partial list only] several significant earthquakes successfully, though it's put out a few false alarms as well. (One false alarm is listed as having been caused by a lightning strike, and they wrote that they deliberately accept such false alarms to maximize the pre-earthquake warning time for real earthquakes, rather than wait for additional data to come in that would delay the warning.)
With respect to the Shinkansen, I'm pretty sure they take advantage of this system, as do at least some other railways in the Tokyo area (I don't recall which). The data is also supposedly sent around to places like city halls, schools, etc.
The big problem with systems like these is that you can't just attach them to loudspeakers and whatnot, because whether it's not a false alarm or not, such broadcasts would easily lead to panic and stampedes that could cause more injuries and even deaths than the earthquake itself. So don't go looking for big "Earthquake Warning" boards next time you're in Japan, because you won't find them--the agency is being very careful with who they give the data to, at least for now.
It really doesn't matter if the building code specifies that all buildings are to be machined from 1-piece cast aluminum blocks.
Pay the inspector, and you can live in a house of bricks held together with dried cow dung.
That will add stress to the track.
Say what you will, this site is interesting:
http://quake.exit.com/
Commodore64_love: I don't comprehend people who're so frightened of death that they'll bankrupt themselves to stay alive
heh, us desk people work in carefully planned deathtraps as well. At least a factory has a potential exits (emergency exit, normal exit, window) and office on the 12th floor, however, has an emergency escape (good luck running down 11 stairs in 15 secs) and elevators (DON'T!) we had a fire drill where the first X people outside were paid a certain ammount of money. Normally we have a 75 second complete evac time, for a 6 floor building. With people actaully pushing and shoving to get down, it got up to a whooping 180 seconds. And that's not counting the time for a full sweep of every floor, or with actaul smoke.
I can still remember the Great Dudley Earthquake of 2002, I don't think there would be any point in installing such a monitoring system in the UK but I did briefly consider moving the bookshelves and couple of hundred hard back books from above my head.
15 seconds is obviously maximum amount of warning this system can provide so in practice even with the system most people would get much less than this. Although in theory people can take some appropriate action in 15 seconds to mitigate the earthquakes effect on them in practice the most likely result would simply be panic. After all you may have 15seconds or you may have 3seconds to act so if you're in a building with one doorway along with dozens of other people I doubt there is going to be much in the way of orderly queuing going on.
On the other hand automated systems which can react within milliseconds will benefit enormously from this, gas can be shut off, mobile phone networks can switch to emergency modes, factories, freight terminals, power stations and other large automated systems can can switch over to safe modes. Even in the case of false alarms it will be obvious withing 30seconds anyway so normal work can be immediately resumed.
...seem to be shifting the grounds for discussion.
Why not just fit all cities with huge airbags on every street corner, and between skyscrapers?
When an earthquake is detected, BAM! All the airbags inflate, securing the city and protecting the citizens instantly.
Cress, cress, lovely lovely cress
Just enough time to kiss your ass goodbye. :)
Serving your airship needs since 1995.
I worked in a scary manufacturing/research environment for a while: we had *lots* of big fluorine lasers. There were two alarms: fire, and fluorine. When the fire alarm went off, people sat around for a moment, looked at each other, waggled eyebrows, looked around some more, then moseyed over to one of the windows on the manufacturing floor to see if there were visible flames, and then slowly, reluctantly, walked over to the door and went outside. When the fluorine alarm went off, people dropped everything and ran as fast as they could. We were required to have our desks or workspaces "within the distance of one breath" of the nearest door, and that was a lot less than fifteen seconds. Sure, you *can* hold your breath for that long if you have warning, but if you just exhaled and the alarm went off... not so great.
(Fluorine, by the way, smells a little like elmer's glue, in my opinion.)
Anyway, I know this isn't practical for enormous skyscrapers, but if you design a single-story building and lay out the production/workspace areas with a little care, I think it's perfectly possible to have the building cleared in fifteen seconds.
Nostalgia's not what it used to be.
The USGS started Radon measurements right after the Chinese and Russians claimed some encouraging results in the early 1970s. The results have been inconclusive. There have been a dozen sizeable quakes in the San Andreas area where the USGS has its sensors- all inconclusive.
Besides radon is a slow signal. In the few promising results, the radon starts increasing weeks before an event, but with no clear signal pointing to the day or hour.
One serious problem with tsunami-size quakes, M7.5 and larger, is to accurately measure the magnitude of the large quake. Regular seismographs saturate at high magnitudes- that means that an 8 looks just like an 8.5 looks just like a 9 for conventional single-station magnitude calculations. There is a more accurate calculation called moment-magnitude, but that requires collecting data from at least a hemisphere of stations, preferrably a full global. That requires an hour, half an hour for the wave to to reach enough stations, and recording it for several tens of minutes to obtain essential low-frequency information. This assumes instantaneous telemetry to a central location and a good automatic computing program. An offshore quake can send water waves to the nearest shore in a quarter of that time. The 2004 Sumatra Boxer Day tsunami was a slow motion horror film. The first magnitude reports were around 8, but kept on growing for all morning as more precise computations were performed. It didnt help that computer programs were finicky at the time requiring human data editing and the humans in North America were sleeping off their Christmas parties away from the office. (The pacific tsunami center has a seismologist on 15 minutes paging duty, but didnt have the best computer programs at the time.) So there is a fair amount of research on quick-large-magnitude determination. Seismologist are hoping for a characteristic signature in the first 30 seconds of a seismogram. A hopeful method compares a new large quake to a previous large quake from the same area of the world. Plus you want seismographs at minutes of wave propagation distance from every offshore fault capable of a great quake. Such stations were supposedly in place for a 2006 Java great quake, but didnt work properly. The governement notification system didnt work at all. Fortunately there wasnt a tsunami.
Tsunami notification doesn't rely solely on seismic data. Special buoys measure the wave as it passes by. But they buoys are too far away for the nearest shorelines, so a quick seismic method is still desired.
The peak energy of quakes decreases in frequency (dispersion) as it goes off into the distance. Buildings and mud layers have resonant frequencies which can effective triple the force of the wave as it passes by. The Loma Prieta "sweet spot" was not close-in in Silicon Valley mudflats, but about twice that distance in Oakland and San Francisco Marina.
You can shutdown the subways, shut off refinery valves, park computer disk arms, and walk to a safer location.
So the rescue works can go in the right direction. Cant depend on media to all be working then.
The southern california early alert system (for organizations only) tries to compute and notify location ASAP.
How would you pass the information across the mass public in 20 seconds???
Just using a simple radio transmitters connected to pinball machine tilt switches, and place them in a grid one every 10 miles or so, you would be able to get a very advanced warning.
You'd want to agrigate the warning system into one or more centralized alert systems so that they can wait for multiple even triggers before sounding an alert.
Just as the 15 second warning system describe in the parent article here, the further from the epicenter the more warning you have but with this your advanced warning comes at 186,000 Mph instead of speed difference of the two wave type which are both travaling around 700Mph.
I am always doing that which I can not do, in order that I may learn how to do it. - Pablo Picasso