An array of underwater sensors off Vancouver Island's west coast measured the progress of a tsunami approaching British Columbia's seaside communities of Tofino and Ucluelet early Tuesday morning.
It turned out the tsunami amounted to a barely visible ripple, just a few centimetres in height.
Based on the data from the sensors, instantly transmitted to the Pacific Tsunami Warning Centre, scientists quickly knew the earthquake off Alaska's coast, 10 kilometres under the sea, was a big one. But the estimation of how big the tsunami would be was far less precise.
The tsunami warning was cancelled when it became clear that there was no threat to coastal communities. But the warning was a reminder of the risks and vulnerabilities of a region that has long been warned to brace for a major earthquake and its aftermath, including potentially devastating tsunamis.
For the thousands of British Columbians, roused from their beds on Tuesday by sirens, text alerts or neighbours pounding on their doors and urging them to rush to mustering stations on higher ground, the news two hours later that the whole thing was called off was abrupt relief.
Scientists say the difference between what was feared and what actually transpired was a result of the limits of the highly specialized equipment, precise to the point of being able to detect waves far away from a coastline, but unable to quickly and exactly predict how those waves might build.
"With tsunamis, we are able to predict when they will arrive very, very well. We can do that down to the 10-minute interval. We know it takes pretty much three hours from the earthquake in Alaska to start hitting Tofino and Victoria in B.C.," geophysicist Mika McKinnon says.
"However, it is very, very difficult to tell how big a tsunami is going to be. And the reason that it is difficult to tell, is because we don't know how much water was displaced. We can look at an earthquake and say, 'That's a really big earthquake, that's underwater, in an area that usually has vertical movement that would move the sea floor up and down – it probably created a big tsunami.'"
The quake with a magnitude of 7.9 struck at about 1:30 a.m. Pacific Time. It was centred 278 kilometres southeast of Kodiak, Alaska.
Earth sciences Professor Brent Ward of Simon Fraser University said the quake was a strike-slip variety, where the plates slip sideways past each other.
"To get a tsunami, you have to have vertical movement of the sea floor and that more often occurs in what we call a thrust fault … where one of the plates is moving over top of the other." Prof. Ward said when a plate moves up very quickly, that displaces the water above it, setting off a tsunami.
"It looks as though this earthquake was triggered on what we would call a transcurrent fault in the oceanic plate that is going underneath North America," he said.
After those early seismographic measurements, deep-ocean buoys called dart buoys measure pressure of the tsunami going by, Ms. McKinnon explained. But out in the open ocean, a tsunami is only a couple of centimetres high. It is not until the tsunami gets closer to shore, and hits tidal buoys and tide gauges, that more accurate estimates can be made.
"Until a tsunami starts coming on shore, we don't know how big it will be," Ms. McKinnon said.
Sensitive pressure detectors, connected to a network of underwater research observatories, showed that the wave heading toward the shore was a barely visible ripple compared w the winter storm waves that closed beaches in the region last week.
Kate Moran, president and CEO of Ocean Networks Canada, a University of Victoria agency that has multiple ocean observatories, said the system of sensitive pressure detectors, connected to a network of underwater research observatories, provided a clear picture of what was coming an hour before the tsunami arrived. The recorders, installed in 2009, first detected the powerful earthquake at 1:35 a.m., and the first tsunami was picked up by the first sensor at 3:55 a.m., with an hour of travel time before it would arrive on the beaches of Vancouver Island.
"That information allowed us to give an estimated time of arrival of coast, data automatically sent to the coastal tsunami warning," Dr. Moran said.
Seismic information gets fed into tsunami-modelling tools that use information about past earthquakes and tsunamis, coastlines and other factors to predict when a tsunami might reach land, Ms. McKinnon explained.
"But it is very difficult to tell how big a tsunami might be at that stage, because we are doing it based entirely on a model and an earthquake."
When a tsunami gets closer to shore, it can do something called "shoaling" – bunching up or shortening, so that a wave that was long and low, and moving at the speed of a jetliner, becomes shorter, higher and potentially more devastating.
"It happens near the shoreline, it only happens near the shoreline, and we don't know how big it's going to get until that starts happening," she said.
There aren't that many tidal buoys along the northern coast of British Columbia and southern coast of Alaska, in part because the conditions are so rugged, she said.
So emergency management authorities have a difficult choice.
"You have to make the decision: Do you risk having people lose trust and lose faith in your ability to predict the science? Or do you risk people dying, when you could save them? Here in B.C., the choice is, we are going to try to save people," Ms. McKinnon said.
With a report from The Canadian Press