---

Indian Ocean tsunami
offers valuable lessons

The Pacific Tsunami Warning Center has no responsibility for the Indian Ocean and has no instruments there, but the center has been criticized for its performance during the southern Asian tsunami. The feeling almost everywhere is, "They knew. They should have warned those people."

Unfortunately, they didn't know. The earthquake was so large that answers to such fundamental questions as "How big?" and "Where?" were misleading. Without tide gauge information, the earthquake appeared to be unthreatening everywhere except Sumatra.

It takes an hour to compute earthquake magnitude because the seismic waves used take that long to reach enough seismometers. But useful warnings depend on speed, so PTWC first makes a quick-and-dirty estimate, called Mwp, from the first two minutes of motion on a seismogram. Their first bulletin, only 15 minutes after the earthquake, reported a magnitude of 8.0. It was woefully too small.

PTWC's epicenter put the earthquake off the northwestern coast of Sumatra. A local tsunami might have been generated, but it would have hit already. Since Sumatra stood between the epicenter and the Malay Peninsula, there would be no tsunami on the Asian mainland. Waves might have been projected across the Indian Ocean to the southwest, but they were unlikely to be large, and they would take seven hours to reach Madagascar. It seemed there was time to follow events.

But this was a remarkable earthquake. The epicenter, the point of initial rupture, was at the southern extremity of the break. Rupture had continued for six minutes, extending northward for 750 miles. Since Mwp used only two minutes of information, it ignored most of the earthquake and was far too small. Worse, the earthquake was not centered on its epicenter off Sumatra, but 300 miles around the curve of the Nicobar Islands to the north. Instead of heading southwest, the tsunami had been directed due west, toward Sri Lanka, and due east, toward Thailand.

Sixty-five minutes after the earthquake, the warning center revised its magnitude to 8.5. Even then the tsunami might be small. The center warned of a possible tsunami in its second bulletin, received by both Thailand and Indonesia. But it did not know the true size of the earthquake, magnitude 9.0, or the dreadful fact of the tsunami for another hour. By then 150,000 were dead.

If the Sumatran earthquake had instead happened in the Pacific, the initial misinterpretations would have been corrected and warnings given well before the tsunami reached a distant shore. For shorelines close to the source, however, the warning would have come too late. That is an important message for Hawaii.

Our tsunami hazard is not just from distant earthquakes: The tsunami from the 1975 Kalapana earthquake killed two and caused $4 million in damage. A tsunami from an earthquake in South Kona (a similar geological setting to Kalapana) would be vastly more damaging. It would reach the Kona Coast in five minutes and flood the southern coasts of Maui and Oahu in only 30 minutes. While the earthquake would trigger warnings for Hawaii and Maui counties, five minutes of delay inherent in the system means that for much of the Kona coast, the only useful warning would be the shaking of the ground itself. A statewide warning would have to be delayed until the tsunami is confirmed, but with existing instruments -- seismometers, tide gauges and flooding sensors along the Big Island coast -- that delay could be eight or nine minutes.

Enter real-time seismology. Techniques exist to exploit the data from modern seismometers to work out earthquake characteristics while rupture is actually occurring. With real-time seismology, delays could be eliminated; tsunami warnings could be issued within tens of seconds rather than minutes. For Kona, ground shaking would be followed a few seconds later by the sirens -- a powerful motivation to evacuate. The rest of the state would gain back those precious lost minutes. Such rapid warning, however, requires new seismometers, at least one on every island.

In the wake of the Indian Ocean tsunami, a dramatic enhancement of the Pacific-wide tsunami warning system is planned. We must make sure that our legitimate claim to improved local warning is not ignored.

---