BioEd Online

When the tsunami hit California's Crescent City, harbour-master Richard Young was sitting in his office, gazing out the window at the 200-boat harbour. He wasn't surprised when the first surge arrived at around 11:30 PST on Wednesday morning he'd been warned that it would come at about 11:38.

But when the peak surge hit 3 hours later and sent a river of water flowing at tens of kilometres per hour into the harbour, Young was caught off guard. "It looked like the tide coming in really fast," says Young. "The water would go from high tide to low tide in the span of maybe 10 minutes." The nearly 2-metre-high waves were bigger than he'd been advised, and he hadn't been aware that the surges would last for hours.

The West Coast and Alaska Tsunami Warning Center had issued a tsunami advisory for the coast of California at 3:31 PST, just 17 minutes after a magnitude-8.1 earthquake struck the Kuril Islands near Japan. By 6:41 in the morning, they called the advisory off almost 5 hours before the surge reached Crescent City.

Warning signs

All this occurred despite the fact that modelling work done the year before had shown that Crescent City was particularly susceptible to large waves from a Japanese earthquake, that the peak wave would arrive much later than the first, and that the area would be prone to ongoing tide fluctuations for several hours. Crescent City has long been a known hotspot of tsunami activity. In 1964, a wave there killed 11 people - the only example of tsunami fatalities in the continental United States.

Real-time models of this week's tsunami were also better than ever thanks to prompt utilization of data from 14 Deep-ocean Assessment and Reporting of Tsunamis (DART) buoys in the Pacific, says Vasily Titov, chief scientist at the US National Oceanic and Atmospheric Administration's (NOAA's) Center for Tsunami Research in Seattle, Washington. Titov says this tsunami provided scientists with their first opportunity to create real-time models that were updated and corrected as new data arrived.

So why was the warning called off? The decision was based on the small predicted size of the surge, says Paul Whitmore of the West Coast and Alaska Tsunami Warning Center. It was a classic example of officials struggling to balance the need for warning with the dangers of a false alarm, he adds. Titov and other experts say they agree with the tough decision, but the event has left some calling for changes in the US tsunami alert system.

Personal phone call

With the exception of Crescent City, predictions of tsunami wave heights along the California coast were well below 1 metre says Dave Reynolds, a meteorologist at NOAA's Weather Forecast Office in Monterey, California. "We consciously decided that it wasn't appropriate to warn the entire state of California for this event," says Troy Nicolini, a meteorologist at NOAA's National Weather Service office in Eureka, California. "So we did a very targeted, one-on-one personal notification."

Young says he got a call from Nicolini Wednesday morning, warning him that a 1-metre, 'non-damaging' tidal surge was on its way. "I guess there was some interesting miscommunication there," says Young, who estimates that Crescent City harbour sustained anywhere from $500,000 to $1 million in damages. If it had been high tide, the damage would have been even worse.

Whitmore admits that the Warning Center had underestimated the size of the waves that hit Crescent City by about two-thirds.

Costas Synolakis, director of the Tsunami Research Center at the University of Southern California in Los Angeles, modelled what would happen with an 8.8-magnitude quake in a similar location last year. "The models had predicted the oscillations, including max height," he says. But, he adds, there were a few surprises even after the model had been refined with up-to-date quake information Wednesday morning. "We didn't expect it to last for seven hours and we didn't expect the seventh wave to be the largest."

Swimming hazard

Reynolds appreciates the need to avoid creating a panic. He remembers tsunami warnings issued for Northern California in June 2005 following an earthquake off the coast near the town of Eureka. "People living up by Coit Tower in San Francisco were running for higher ground," says Reynolds. Coit Tower is located on top of a hill 83 metres above sea level. "We could see that a lot of people really didn't know what to do in this situation."

But in retrospect, Reynolds believes a mild warning should have been communicated to California in this instance. "I think they would have put out special warnings much earlier, given that there might be a hazard to swimmers and small boats in the water," says Reynolds. "That's what I would have liked to have done."

Whitmore says the problem is that they don't have an intermediate level of alert. "When we call a warning it's either all on or all off," he says. "There needs to be a level of warning that doesn't trigger that massive response an advisory that just tells people to stay off the beach and out of the water."

Synolakis says what happened in Crescent City also shows that scientists need to pay more attention to local topographies and how they might create wave resonances.

"Up to now, we've basically been paying attention to exposed coastlines," says Synolakis. "We always felt that inside harbours and bays were secondary. But this puts it into an entirely new perspective."

Visit our newsblog to read and post comments about this story.