Seismologists have never had a better understanding of earthquakes. But tragedy after tragedy shows that quakes still surprise and choke people with their mercurial behavior. Precise predictions of when and where earthquakes will occur and how mortal they may be, are not yet possible. However, if scientists could chronicle how quakes grow, they may better predict how powerful they are.
The most powerful quakes are far from immediate. They can last for minutes, making them less like a single underground blast and more like a series of outward-moving explosions. A new study published Wednesday in Science Advances explains that the outward path of the explosions varies depending on the force of the earthquake.
[ Like the Science Times page on Facebook. | Sign up for the newsletter Science Times. ]
A one-digit jump in earthquake size means that 32 times more energy is released. Many factors determine the risk of an earthquake, but small increases in size can make the difference between just harmful and catastrophic. If finally earthquake sizes were found early, it would give quick and more accurate warnings to peoples still to shake.
Diego Melgar, a professor of seismology at the University of Oregon, explained that this connection was not what he and the other co-author of the paper, Gavin P. Hayes of the United States Geological Survey, originally sought. Instead, they had collected data from earthquake databases to get the most accurate simulations of the most powerful quakes.
"And along the way we just stood over something interesting," said Dr. Melgar: a key moment in time that frames the future of an earthquake.
The team took a close look at 3,000 earthquakes recorded by the agency's seismometers. The data captured by these sensors can show the energy saving of an earthquake over time far from the source. The researchers also dug through 30 earthquake values of GPS station data, where an antenna bolted to the ground follows the development of the fracture close to the earthquake.
Based on previous work, the team described how large earthquakes develop. Immediately after they begin, they grow chaotic for a few seconds, a pandemonium that lasts longer for prolonged quakes. The rupture organizes itself – for reasons that are currently unclear – into something resembling a pulse, an annular area that moves outward from the source of the earth over time.
This pulse ring indicates where the rock breaks or slides. A thinner pulse is less likely to continue to grow into a larger event, while a thicker pulse is more likely to do so. The team claims that because of these differences, the dimensions can be used to determine the final size of the earthquake mid-earthquake.
Since the 1980s, seismologists have discussed if such an option is possible. Some said the final sizes could be calculated right at the birth of the earthquake, while others suspected that seismologists should wait for the break to cease. Others like Dr. Melgar and Dr. Hayes, falls somewhere in between.
Stephen Hicks, a seismologist at Imperial College London who was not involved in the research, said the data suggested that the correlation between fracture evolution and final size was not a case. The way in which large quakes accelerate can be a recurring feature among the overall chaos.
Men-Andrin Meier, a seismologist at the California Institute of Technology, said his own research also showed that it was possible to determine final sizes of the mid-earthquake. But he differs from Dr. Melgar and Dr. Hayes about how fast into the break the final magnitude can be calculated. Their new paper puts it in about 10 seconds, but Dr. Meier says this depends on size and can vary wildly.
A limitation of the new model is that it assumes an average earthquake behavior. In fact, every earthquake still has a personality, says Dr. Melgar. The development of certain earthquakes may not suit expected patterns, making difficulty calculations of final sizes more difficult or, in some cases, easier.
Dr. Melgar also recognizes that powerful quakes, especially those over the size of 8.5, are less frequent than their weaker counterparts. More data on large temblors, from simulations or real events, is needed to shore up this model.
"It's a good speculative idea, we just have to fill it in before we can have great confidence in it," said John Vidale, professor of seismology at the University of Southern California.