How good are we at predicting earthquakes? Further, how much do we even understand about the processes that make the ground shift under us?
The record says it all as far as our ability to prognosticate goes. We didn't see the Indian Ocean earthquake of 2004 and its ensuing tsunami coming. We didn't predict the massive offshore Japanese earthquake of last year or the Chilean one the year before. Seismologists are terrible at predicting earthquakes. Leading earthquake researchers are not shy to admit that accurate forecasting is currently almost impossible.
We know where earthquakes are likely to strike, and we know how to measure them as they happen and study them later. Why then do we find it so difficult to predict these events?
Earthquake epicenters are usually 20 miles below the surface, severely limiting our ability to observe precursor conditions. Twenty miles of solid rock is further than we have ever drilled down, and it is opaque to most measurements. An analogy would be like trying to predict weather without being able to see satellite images of clouds. You would have to predict the weather solely based on people looking up at the bottoms of the clouds above.
Much like weather, the motion of the earth's interior is also far too complicated to ever fully predict well. Asking whether a quake is likely to strike on a specific day in six months is like asking whether it will be raining on a specific day in six months.
Even our basic understanding of the earth's crust and the rocky layers just beneath it (called the lithosphere) is very recent. Your parents (or grandparents for younger readers) likely weren't taught this theory (called "plate tectonics") when they were in school.
Until roughly 60 years ago, nearly everyone believed that the earth's surface was basically immobile. However, geological observations in the first half of the 20th century revealed that plates (pieces of crust -- i.e., enormous sheets of rock -- that span continents) slide around, on and under the earth's surface.
In the following years seafloor spreading was also found, giving evidence of a method for creation, or upwelling, of new surface. If surface is being created, but the size of the earth stays the same, it follows that surface must also be destroyed, or pushed back under. Around the same time as seafloor spreading began to be observed, subduction, or pushing of surface back down into the interior of the earth was also discovered.
Faults are the places where crust is forced back down into the earth. One type, called transform faults, are places where pieces of crust crash together and force each other up to form mountains, or simply grate together as they shift about to relieve pressure. Another type, called subduction zones, are where one plate slides underneath another. (These can cause enormous mega-quakes, and one is located off the coast of the Pacific Northwest.)
In short, nearly all earthquakes are caused by plates crashing together, scraping each other, being pushed skyward or pressed down into the earth's interior.
Thus, seismology is a relatively young field, hindered by the difficulty in performing direct observations and measurements. An incredibly complex, mostly hidden chain of events far underground obscures the world beneath. All of these factors cause seismologists to have a hard time understanding and predicting quakes.
But, there are larger issues at stake in the Italian trial.
More troubling than the misunderstanding of scientific capability demonstrated by the Italian verdict is the trial of science. Scientists can only offer their very best hypotheses, and for a field such as seismology, there are literally no guarantees. The idea of criminal and financial liabilty for best guesses is ludicrous.
What's next, putting the weatherman in prison for not predicting a tornado?