PART TWO OF TWO PARTS

The giant April 1, 1946, tsunami that devastated Hilo and hit other islands spawned a sophisticated Hawaii emergency alert system that is continuing to improve the collection and transmission of information in a potential disaster.

The 1946 waves, generated by a magnitude-7.4 quake in the Aleutian Islands, caught Hawaii residents off guard because there was no extensive warning system in place.

Today, isle residents would have several hours' warning ahead of a tsunami from Alaska or Chile, the source of deadly waves in 1960.

But a locally generated quake could cause waves that strike within minutes.

That continues to be a challenge for emergency officials. The effort is three-pronged: getting accurate information fast, disseminating it fast and educating the public not to ignore it.

On the first front, the Pacific Tsunami Warning Center at Ewa Beach is working on a Hawaii Integrated Seismic Network with real-time stations reporting information to calculate the location, depth and size of a locally generated earthquake.

Geophysicist Gerard Fryer said the center's goal is to have “;12 super-duper broadband stations”; as high quality as instruments in the advanced national seismic system.

When the earth shook off Waianae on Aug. 13, for instance, data from the station at Kekaha, Kauai, was critical in determining the magnitude was small, about 3, Fryer said.

But the development of these stations has been slowed, Fryer said, “;not just by a shortage of funds, but by delays in getting site leases approved and in getting approval for communications expenses.”; The National Weather Service has assigned a staff member to help facilitate lease agreements, he said.

The completed network will include broadband seismometers, strong-motion sensors and short-period seismometers run collaboratively by the National Oceanic and Atmospheric Administration, U.S. Geological Survey, University of Hawaii and state Civil Defense, Fryer said.

“;It will mean the ability to assess the tsunami hazard from an earthquake anywhere in the Hawaiian Islands within two minutes,”; he said.

“;We managed three minutes for the Kiholo Bay earthquake (in October 2006), but speed was possible only because the earthquake was in the best possible place relative to our instruments,”; he said. “;With more instruments we'll do at least that well for any earthquake.”;

To get tsunami information out fast, Civil Defense officials are looking at methods of “;mass notification”;—possibly via cell phones or computers, said Alvin Kaku, director, city Department of Emergency Management.

“;If you're not watching or next to a TV, most people are carrying cell phones,”; he said. “;Redundancy is important.”;

There was no way to notify people that no tsunami was generated after the widely felt 6.7-magnitude Kiholo Bay temblor, said Jim Weyman, director of the Central Pacific Hurricane Center and meteorologist in charge of the Honolulu Forecast Office. So it was decided to use a code for a civil emergency message, requiring a voice-over, for earthquakes between 5.0 and 6.7, he said.

But glitches developed March 9 in the minutes after a quake off the Big Island.

A message was issued by the forecast office to inform people no tsunami was generated.

Oceanic Cable put out a trailer to broadcasters saying people should turn to Channels 48 or 6 for additional information, but there was no voice message on Channel 48.

The lapse was puzzling because a flash-flood warning was issued that night, and Channel 48 played the voice message three times.

Officials of all agencies involved got together with some broadcasters and identified a coding problem.

Weyman said there are 36 numerical codes for different weather events that the forecast office sends to broadcasters when it issues a watch or warning. Radio and TV stations have black boxes to receive and decode the signals for messages, which they carry voluntarily.

Ray Lovell, state Civil Defense spokesman, said the civil emergency message was successful except with the cable channels. He said the code had too much information, and it clogged Oceanic's system.

“;The problem was unearthed because we went through this exercise,”; Fryer said. “;The only way we would have found out is by testing it. There are so many pieces to the system.”;

Tsunami exercises generally are done the first working day of October and April, and the system is tested monthly to identify and fix any glitches.

He said the recent earthquake off Laupahoehoe also was valuable because it showed “;we have to make little changes in our processing here (at the warning center).”;

The earthquake was 4.4 in magnitude, and the center overestimated it at 5.2.

“;The reason is there is a very strong depth dependence,”; he said, explaining the earthquake was deep down in solid rock. “;We just had the interface between solid rock and weaker rock. We didn't have that correct for the location.”;

Some new instruments developed by Dan Walker, tsunami adviser to the Oahu Department of Emergency Management, will be installed soon in Hawaii Volcanoes National Park to send a signal to the warning center when the instruments get flooded from tsunami run-up.

Four instruments that operate with a satellite will be installed in the park's campsites, Walker said, with more planned later for broader coverage. “;It will help in determining if a wave is propagating towards Hilo,”; he said.

Walker developed another device he is putting at four heavily used shoreline campsites that will immediately sound a siren when they are flooded.

“;If they hear a siren, they'd better wake up quick and head to higher ground,”; he said.