Next step: plot locations of 1906, 1960 and 2010 earthquakes on Google Earth and see if the 1906 quake's epicentrum is positioned a little south of the one of 1960...
The source of data for this plot is the wikipedia list of earthquakes mentioned in other comments.
A few observations: the latitude v. longitude plot is just a map of Chile. It is distorted since the longitude range is just 5 degrees and the latitude range is 35 degrees, and it is plotted on a square. But you get the picture. The outlier is on Tierra de Fuego.
The Year v. Magnitude plot is interesting. First, historical earthquakes (before 1900) are all Magnitude 8.5 or stronger. Also, 5 of them earthquakes are exactly 8.5 magnitude, 4 of them before 1900. This is because historical magnitudes are estimated from accounts of the damage and not from seismograph measurements. Also, earthquakes weaker than 8.5 "didn't make history". So there is a selection bias in the old data, similar to the ones one find in astrophysics where the farther out one looks the bighter an objects needs to be to be seen.
But now look at the Year v. Latitude/Longitude plots. The 5 oldest earthquakes for which we have location information occur within one degree of longitude and 5 degrees of latitude of each other. In fact 4 of them occur at about the same latitude. They are all at least 8.5 in magnitude. 3 of them happened in Concepción, one in Santiago and one in Valdivia.
Looking at the Magnitude v. Latitude/Longitude plot, the very strongest earthquakes (magnitude greater than 8.5) are not exactly close to each other, relatively speaking. The 1960 earthquake is not in a cluster at all. En un viejo país ineficiente, algo así como España entre dos guerras civiles, poseer una casa y poca hacienda y memoria ninguna. -- Gil de Biedma
I read (on paper) this morning that the past earthquake was expected because the 1922 earthquake (8.5) at Vallenar and the 1960 earthquake (9.5) at Valdivia have been correlated to the same thrust fault. The middle part hadn´t undergone any measurable displacement by these historic two quakes - that area was lodged.
On such an area stress builds up the highest. And the recent earthquake is situated smack in the middle between these previous two - so in that respect, this earthquake was expected...
So just for fun: next step is putting in the location of the major thrust faults - but that will have to go 3D, because depth becomes a factor.
Next step: outlining the areas of displacement per quake at major faults, and combine again with time intervals. Then we can sensibly start correlating earthquake frequencies and start extrapolating which areas are at increased hazard.
Next step: start a geohazard company and sell the information. :)
But hindsight is 20/20... En un viejo país ineficiente, algo así como España entre dos guerras civiles, poseer una casa y poca hacienda y memoria ninguna. -- Gil de Biedma
