Because we don’t know enough about plate tectonics. “We basically have a 50-year-old science, compared with most other sciences, which are hundreds of years old,” says Chris Goldfinger, director of the Active Tectonics and Seafloor Mapping Laboratory at Oregon State University. “We’re pretty low along the learning curve, and the most elusive thing of all is prediction.”
A few methods are in the works, such as using historical records to identify quake timing patterns. Alas, the records are too recent. Some earthquake cycles span thousands of years; in most places, geologists have only 100 years of hard data.
Another option is to detect a precursor to an earthquake, like a change in groundwater levels, electromagnetic energy, or wildlife behavior. “We keep catching glimpses of things after the fact and going, ‘Hey! Do that again somewhere,’ ” Goldfinger says. So far, nothing has emerged as a verified quake predictor because no early-warning signs have been repeated.
Option three is to study how stress is transferred between earthquakes, since it’s known that when a quake occurs, destructive energy is transferred to nearby faults. Geologists still can’t say if a fault will go off tomorrow, but they can say that its chances of failure have increased.
Since we aren’t there yet—and may never be—science’s best advice remains: Be ready. “Even though paleoseismology hasn’t yet answered the $64,000 question, it does help people do the much more practical thing, which is to prepare for a quake when it does happen,” Goldfinger says. “Have it be a bad day instead of a catastrophe.”