The earth truly moved for Marlborough in Sunday's severe quake, but scientists concede the fault responsible might never be identified.
GNS Science revealed tonight that even though most residents would not have noticed, Sunday's 6.5 magnitude quake moved parts of Marlborough 5 centimetres to the east.
But most people would have been too distracted by the shaking of the earthquake to notice, they said.
The National Institute of Water and Atmospheric Research ship Tangaroa spent this morning surveying Cook Strait's sea-floor around the area where the big quake and numerous aftershocks have been centred.
The quakes appear to be coming from an area between the Vernon and Needles faults where little mapping has been done, and it was hoped today's work would identify sea floor disturbance that would help show where the fault line is.
Marine geologist Scott Nodder said the results needed to be analysed, but preliminary indications were that there had been no change to the sea floor. That meant it was likely that the fault was quite deep, as their measurements looked at the top 50 metres of sediment.
Niwa would now work with GNS to get more exact data about the location of the quakes and aftershocks before considering whether to undertake further mapping with more specialist instruments.
Analysis by GNS among the "grand central station" of faults on the Cook Strait seabed has narrowed the location of the fault rupture to a 19-kilometre northeast-southwest slash in the earth's crust, with the southernmost point coming close to the Clifford Bay coast.
It lay between the London Hill and Hog Swamp faults - both obscure on-land faults not known to extend offshore.
"This new data is not a clean fit with either fault, so it raises the possibility that it could be a previously unknown fault," said seismologist Stephen Bannister.
The London Hill fault was thought to rupture once every 3500 to 5000 years, he said.
GNS seismologist John Ristau said GNS has now deployed temporary sensors to more locations to help pinpoint the locations of the quakes more precisely. That could help rule out either Hog Swamp or London Hill as the culprit.
However, he conceded that they may never know exactly which fault caused the earthquakes.
"Unfortunately that's the way it works out sometimes, but right now we're still hopeful."
Work to identify the source of the quake was important as it could help scientists determine the level of stress caused on other nearby faults capable of creating much larger events.
It would take a couple of weeks to get data back from the new sensors, Ristau said.
Nodder said the Tangaroa also ran lines through the Cook Strait Canyon to see if the quakes had caused any landslides.
"To see if any of the sea floor had been mobilised by the shaking and if anything had slumped off."
That would help determine levels of tsunami risk from quakes in the area, he said.
WHAT ARE THE ODDS?
The probability of a magnitude 6 or greater quake striking in the Cook Strait continues to decrease as more time passes since Sunday's quake.
The latest probabilities released yesterday are:
■ Likelihood of a magnitude 6 of higher quake hitting within the next 24 hours: 3 per cent (4 per cent on Tuesday)
■ The next seven days: 10 per cent (13 per cent on Tuesday)
■ The next year: 30 per cent (33 per cent on Tuesday)