It occurred in the area where the Sichuan basin and the Longmen Mountains meet. This is on the eastern edge of the Tibetan Plateau in an area deformed by the collision of the Indian and Asian tectonic plates.
The area is crisscrossed with fault lines.
Around the world, earthquakes like the one in China are associated with triggered aftershocks that are very large.
For the May 12 earthquake, the researchers performed analysis of co-seismic stress transfer onto Sichuan basin faults using broad ranges because at this time, exact values for all the various factors are unknown.
According to the researchers, this approach enables rapid mapping of faults with heightened rupture likelihood.
"We knew that the fault was there and we knew it was active," said Eric Kirby, associate professor of geosciences at Penn State. "I had done some previous work in the area, but I do not think anyone would have anticipated the size of this earthquake," he added.
The researchers, who include Tom Parsons, research geophysicist, US Geological Survey, Chen Ji, assistant professor of Earth sciences, University of California-Santa Barbara, and Kirby, used a model to see how the Sichuan earthquake, which took place on the Beichuan fault, affected other portions of that fault and others in the area.
They looked at physical characteristics of the faults including the directions and amounts of movement of the faults - whether and how much they moved up and down and side to side, and the estimates of the frictional resistance to motion along the fault.
"The Sichuan earthquake seemed to rupture on the northern portion of the Beichuan fault," said Kirby. "It does not seem to have involved the southwestern branch," he added.
According to the model, after the May 12 earthquake, stress increased on faults running parallel to the Wenchuan-Maowen fault and the two major faults that are perpendicular and to the north of the fault.
Some smaller faults south of the earthquake zone show a decrease in stress. However, according to the model, the majority of the faults in the area are still stressed.
"The model takes what we think we know about the faults in the region and asks what was the change to stress associated with the earthquake," said Kirby. "The model shows where an increase in the potential for failure may occur," he added.