Scientists find four faults caused shake
Four and possibly more faults are now believed responsible for the September 4 earthquake.
The closest to Christchurch is thought to be between West Melton and Burnham.
The trigger for the magnitude 7.1 quake has been revealed as a hidden, "blind thrust", fault lying at depth in a broad northeast to southwest zone centred on the focus of the shake close to Charing Cross.
When that fault ruptured it quickly activated the sleeping Greendale Fault slightly further south, ripping open a nearly 30km-long gash across the Canterbury Plains.
Detailed investigations by GNS Science using satellite radar-imaging, GPS [global positioning system] calculations and computer modelling now show the likely presence of two other blind thrusts angled across the Greendale Fault.
The fault to the southwest of Hororata is thought to have ruptured underground third in the sequence, followed quickly by the fourth, around the West Melton-Sandy Knolls-Burnham area.
The scientists say this latter fault helps explain the intense shallow aftershock activity near Rolleston and across to Lincoln.
At this week's GeoNZ conference in Auckland, GNS seismologist Dr Caroline Holden and colleague, geophysicist Dr John Beavan, outlined their findings.
Holden said the 7.1 magnitude was accounted for by an about 6.3 magnitude quake lasting two to four seconds on the Charing Cross fault, followed by a 6.9 shake lasting seven to 18 seconds on the Greendale Fault and a close to 6.5 jolt of 15 to 18 seconds on the feature near Hororota.
Beavan's calculations were slightly different, with the initial quake of magnitude 6.5, followed by shakes of 7.0 and 6.2. He believed the last rupture, on the fourth fault, generated a quake of about 6.5.
"We have got more work to do towards the eastern end of the Greendale Fault. Something else is going on there.
"It was a very complex earthquake. Why was it so complex? Are most earthquakes like this ... or is there something special about the tectonics in the Canterbury region?"
The fault near Burnham had not shown up in satellite imagery because it was a subtle feature, Beavan said.
"There are probably several bits of fault, little segments that break, but I suspect they are not as important as the other faults. There may be other ones too."
GNS seismologist Dr Bill Fry said the hard volcanic rocks of Banks Peninsula could have affected the way the earthquake was felt in Christchurch.
His analysis showed that from about 16 seconds into the quake, energy was radiating outwards from the peninsula.
"It looks like a lot of energy from the original two ruptures was going into Banks Peninsula and reflected back up into Christchurch."
Another GNS seismologist, Dr Martin Reyners, said a zone of shallow aftershock activity around Rolleston and Lincoln appeared to stop at the edge of the peninsula.
"As you approach Christchurch the aftershocks get shallower and shallower.
"That shallowing most likely reflects a change in the crustal structure there," he said.