Slashdot Mirror
Can Earthquakes Be Predicted Algorithmically?
An anonymous reader with this story about a practical application of big data analysis as applied to the trove of sensor readings taken by satellites and by ground-based senosrs. A company called Terra Seismic says that earthquakes can be predicted 20-30 days before they occur, by sifting data for thermal, ionic, and other abnormalities in areas where quakes are considered likely. Says the linked article: "The company claims to have successfully predicted a number of earthquakes. For example, on 5th of April 2013, the firm issued a forecast for Japan. On 12th April 2013, an earthquake hit the identified area and 33 people were injured. On 4th June 2013, the firm again made a prediction for an earthquake in North Italy. On 21st June, an earthquake hit the identified area. On 3rd March 2013, the firm issued a forecast for an earthquake in Iran. Again, after 35 days, an earthquake hit the identified area."
how about a list of times they predicted an earthquake and nothing happened.
without context its meaningless
Even if it were used, people would likely begin to ignore it.
Silence is a state of mime.
It would be like predicting election results. Am I right?
Yes i am interested in what they successfully predicted.
I am also interested in how much they predicted and also crucially their unsuccessful predictions.
I'm glad someone's finally taking a crack at it.
At least the theory seems well grounded.
Sensitivity and specificity or it didn't happen.
Once you have enough of the right type of information, anything can be predicted. Earthquakes follow the laws of physics, so if you gather the data you need you can predict them. Much easier said than done, but still possible.
Betteridge's law.
Sure, I can say that I made predictions about earthquakes two years ago. I'd be impressed if they said today that in 10 days that there would be a magnitude 7.5 in a specific area and be right.
Charged and convicted scientists because they didn't predict a deadly earthquake. Just saying, don't sell it to them or they might bite you in the arse.
Fracking = earthquake. It's a rather simple algorithm.
At the Kardashian faultline within the next 5 years. The constant over-fracking of the posterior end of the fault is creating an enormous crack running vertically to a depth of 8cm.
Would love to know the data set used and accuracy of the predictions.
I can easily, in a few minutes, write a linear estimator that shows perfect correlation with anything you'd like. However, the problem is that same linear estimator is incredibly divergent (a.k.a unreliable) beyond the boundaries of which it shows correlations. You want me to believe you've got a big data formula that will predicit earthquakes? Make a few predictions and be proven right with large magnitude earthquakes.
The previously published algorithm expressed as pseudo-code is as follows:
if(fracking) then earthquake_probability = 90%;
TL;DR No.
Slashdot, fix the reply notifications... You won't get away with it...
FTA:
With the recent Nepal earthquake claiming more than 6,000 lives, many of us have often wondered why earthquakes cannot be predicted the same way as Tsunamis or cyclones are predicted?
This already tells a lot on how much the authors of the article know about forecasting vs predicting - this opening line is wrong in so many ways. Tropical cyclones (which grow into typhoons aka hurricanes), tsunamis, tornadoes and other such natural events can not be predicted any more accurate than earthquakes.
Tropical cyclones can be predicted with a similar accuracy as earthquakes: this are the key areas, and they happen with that frequency. That's how much you can predict a cyclone to happen: Hong Kong is affected by about eight tropical cyclones per year, and about two a year will give rise to a T8 or higher signal. That's predicting: we've had years with five such typhoons hitting, and years without any hitting the city. When a cyclone forms (which is never predicted, only observed as it happens - like an earthquake is observed as it happens), meteorologists indeed are able to forecast with reasonable accuracy where it will head, and what strength it takes. This usually leaves a few days for people to react.
Tsunamis can be predicted with even less accuracy: when an earthquake or similar event has happened the presence of a tsunami can be measured, and a quick forecast can be made of when and where it will hit shorelines, and an alert may be issued. This leaves usually a few hours to half a day for people to react.
Tornadoes form without much warning, leaving often mere minutes for people to get out of the way and into shelters - if the alarms sound at all. They, too can not be predicted.
Earthquakes happen so fast, and end so fast, that there is nothing to forecast, no alarm to sound when it happens. By the time an alert is out, the quake is pretty much over.
And there we have the difference between prediction and forecasting. Forecasting is a lot more accurate by nature, as it is reacting to what is already happening, and works quite well for following slow processes such as the formation of a tropical cyclone. I'm used to know about an incoming typhoon a few days ahead, so plenty of time to prepare. Forecasting earthquakes, well, that doesn't work like that.
The company claims that earthquakes do not "happen fast", but that things start to happen a good while (20-30) days before the actual quake hits. These telltale signs can be picked up and (apparently) with the proper analysis be used to predict the quake. Or forecast, if you're picky. The point is that they claim that there is time to issue a useful warning.
If construction was anything like programming, an incorrectly fitted lock would bring down the entire building...
I own a camping trailer with a stove, fridge, heater, etc. Have a generator to power my home and enough frozen and canned foods to last for weeks. If or when there's a big quake on SoCal I'd be OK.
Why does /. even post this crap? There's nothing in the article that shows any evidence that this is real.
Just another day in Paradise
If they could predict the close of the SP500 tomorrow, they would be millionaires in short order. That will never happen of course since financial systems are much more complex than geologic systems.
Do you think perhaps Siesmologists have not done this already, with better worldwide data, and better resources, and found that it does not work ?
Saying there will be an earthquake in a band of 30 days and getting it right 1/3 times is useless, if the mass evacuations cost more in time, money and lives than the earthquake itself ...
And eventually people are going to accuse them of crying wolf and ignore them ...
Puteulanus fenestra mortis
This already tells a lot on how much the authors of the article know about forecasting vs predicting - this opening line is wrong in so many ways. Tropical cyclones (which grow into typhoons aka hurricanes), tsunamis, tornadoes and other such natural events can not be predicted any more accurate than earthquakes.
That's not true.
Tropical cyclones can be predicted with a similar accuracy as earthquakes: this are the key areas, and they happen with that frequency. That's how much you can predict a cyclone to happen: Hong Kong is affected by about eight tropical cyclones per year, and about two a year will give rise to a T8 or higher signal. That's predicting: we've had years with five such typhoons hitting, and years without any hitting the city.
There's a difference between predicting, forecasting, and wild ass guessing. Earthquake prediction is wild ass guessing, as we neither have the technology nor data to form an accurate model. Long range tropical predictions (like those cited a the beginning of the tropical storm season) are wild assed guesses based on projected long range weather patterns and statistics; there's no physical model involved.
Short term tropical cyclogenesis predicitons on the other hand are pretty good these days. Ana was predicted, for example, several days before the storm actually formed. Once a storm has formed, again, predicting how much it strengthens and where it's going to go also as a high degree of skill. Of course, models aren't perfect, but trying to equate tropical storm prediction and earthquake prediction is comparing apples to oranges.
When a cyclone forms (which is never predicted, only observed as it happens - like an earthquake is observed as it happens),meteorologists indeed are able to forecast with reasonable accuracy where it will head, and what strength it takes.
This is incorrect. Ensemble model runs usually predict formation several days in advance.
Tsunamis can be predicted with even less accuracy: when an earthquake or similar event has happened the presence of a tsunami can be measured, and a quick forecast can be made of when and where it will hit shorelines, and an alert may be issued. This leaves usually a few hours to half a day for people to react.
Tsunami predictions aren't that great because we have a severe lack of information that would be critical in determining when an where a tsunami will form/strike. It wasn't even until relatively recently that we had decent resolution sea floor maps. We don't have real-time monitoring of the ocean floor, or even the ocean surface. Without that information, it basically becomes a wild ass guess.
Again, this is comparing apples and oranges. For TCs, we have lots of data and real-time monitoring feeding well established phsyical models. For tsunamis, we have poor modeling and almost no real-time monitoring. For earthquakes, we have no model and no real-time monitoring (seismographs don't tell you jack about the state of the crust).
Tornadoes form without much warning, leaving often mere minutes for people to get out of the way and into shelters - if the alarms sound at all. They, too can not be predicted.
If you're talking about predicting exactly where they're going to form, then you're somewhat correct. However, we know at least a day in advance of regions where tornadic conditions are highly favorable. Storms that are likely to drop tornados are monitored and their path is predicted, with warnings going up well in advance. Again, it's not perfect but it is in no way comparable to earthquakes where we have no idea what's going.
Earthquakes happen so fast, and end so fast, that there is nothing to forecast, no alarm to sound when it happens. By the time an alert is out, the quake is pretty much over.
And there we have the difference between prediction and forecasting.
No, th
True positives, true negatives, false-positives and false-negatives all contribute to the significance calculation. The problem is that damaging earthquakes are rather rare and it doesnt take much in the way of a missed prediction to clobber the statistics.
I remember the Chinese claims of preduiction from folk observations and Russian claims from seismic velocity speed changes from stress microcracks in the 1970s. Although they may have "predicted" one or two qquakes, they did not work very well when intensely studied.
The most likely methods seem to be based on previous seismicitty- future quakes will occur where previous quakes have occurred, e.g. tectonic bondaries. There are tighter algorithms such as fore-shock and after-shock statistics: a larger quake will occure the week following another quake 5% of the time.
A Russian group had another interesting seismicity method. They drew cirlces around exisiting seismic activity of radius corresponding to quake size. These regions had increased seismic risk. this method sort of predicted the 1989 Santa Cruz quake from Tahoe area activity. The USGS studied this algorithm, but it hasnt been too fruitful.
Seismologists have crushed rocks in labs and heavily instrumented likely fault areas. These have found some physical purcursors of quakes, but not reliable forecasts yet. The most famous and costly experiemnt was a segment of San Andreas near Garfield that broke four times reviously in apparent twenty year cycles. It was 13 years late the time it was instrumented in the 1980s and 1990s. The USGS was heavily criticized on devoting so much money to a single experiment.
Pattern recognis more pragmatic. It doesnt have to show a physical mechanism (though may point to such). No pattern recognition algorithm has worked in the past. But new ones developed ofr data mining show have their try.
Yes - and.... These seismic activities take place - which they measure. Are they not Forecasting which one will become an Earthquake? A tropical storm begins to form (Prediction by your definition) - and then teams begin to Forecast its path, how strong, and other damages (storm surge etc).
Does a "tropical storm" form beneath the earth - these gases, heat, micro-quakes. And from that information Forecast "this will be an earthquake, it will take place [here], and damages are likely to be [y]" ?
Scientists have been predicting a big quake in California for years. It will be an interesting future. "Make sure you have your hard hat and survival gear today, the quake outlook appears to be 70% and could reach magnitude 4.4 for those in the valley"
I think it would be really neat if one day, there are systems and machines that can gather enough data about what's happening beneath the earths' crust, around the tectonic plates (especially high risk areas) that scientists can create models to forecast possible earthquakes and magnitudes for a given area, using probabilities just like in weather forecasting.
It makes sense that earthquakes don't JUST happen out of nowhere; perhaps there are a lot of realistic signs going on that science can use to see the likelihood of a nasty quake over the next n days. It might depend on strain or pressure, movement, how much slipping is happening, etc. If there is eventually thorough and accurate enough measurements available, perhaps then there will be some way of knowing the chances of an earthquake before one might happen. It probably wouldn't be a simple yes or no, but a weather forecasting approach could work. This company may not have enough data or they aren't using it correctly, so we might have to wait a while to iron out wrinkles.
Al Gore is predicting climate AND seismic events now?