In particular, the midfrequency active sonar that has attracted all the opposition is the same device, operating at the same power, that has been used for a generation.

Conrad Erkelens, a senior scientist with Kaya Associates Inc., which is consulting on the EIS, said Monday that nobody noticed any harm to marine mammals all those years, The Maui News reported. It was only when President George W. Bush ordered the Navy to conduct comprehensive environmental reviews of training ranges around the world did anyone become concerned about the whales.

The Hawaii Range Complex is the furthest along in the EIS review.

The strongest evidence that the Navy kills whales came in 2004, when RIMPAC ships were using sonar several dozen miles off the coast of Kauai. Shortly thereafter about 150 melon-headed whales congregated in Hanalei Bay.

Local boaters hung strings in the water, and used canoes and kayaks to herd the whales back to sea. One immature whale was found dead. That is the basis of the claim that the Navy strands and kills whales in Hawaii.

Professor Joe Mobley of the University of Hawaii and other researchers found the dead whale was undernourished and did not conclude that it was killed by the Navy.

A National Oceanic and Atmospheric Administration report completed in April 2006 on the stranding said the sonar exercises are "a plausible, if not likely, contributing factor to the animals entering and remaining in the bay." But it also did not find that the sonar was the cause of the calf's death.

At the time of the fuss, July 2004, the Navy and local researchers were unaware that at the same time, several hundred melon-headed whales had congregated in a bay on Rota, 3,000 miles from Kauai.

The news, with video, arrived later, courtesy of a tour boat captain, who reported there were no Navy or large commercial vessels around.

Erkelens and Frank Stone, another researcher working on the EIS, agree that the congregations of the whales is important, but they say nobody has any real idea what was going on. In any case, the cause hardly could have been sonar.

The Navy used to use passive listening devices of very low frequency to seek out submarines. Like the famous rumblings of African elephants – too low for humans to hear – that apparently carry messages many miles across the veldt, the low-frequency passive sonar had an immense range.

In fact, a station in California was able to cover the entire Pacific Ocean.

That was in the days of Soviet nuclear submarines, which were so noisy that not only could passive listening devices hear them, each sub could be identified by its sound signature.

Since then, China, North Korea and Iran have accumulated fleets of more than 300 diesel-electric submarines. Without the noisy plumbing of nuclear reactors, these subs are said to be much harder to detect. (China has some nuclear subs, which the Navy says are quieter than the Soviet subs.)

During the last few decades, the range at which passive sonar could hope to detect a sub has shrunk, while the range at which the subs' weapons can reach has expanded. Thus, the desire of the Navy to use active "pinging" sonar.

The preferred range is 1 kilohertz to 10 kHz, and the power can be as high as 240 decibels. This has led to claims that the Navy sonar is radiating at "millions of times" the power known to injure marine mammals.

This, say Erkelens and Stone, is not correct. However, they also say there is no simple conversion factor that allows a sonar ping underwater to be compared with a familiar loud noise in the air (a jet airliner is a favorite comparison).

Sound behaves differently in the water. In general, and very roughly, the scientists say, sound rates in decibels underwater compares with sound rated 60 dB less in air. (The decibel scale is logarithmic; so a difference of 10 between any two numbers indicates the same interval. In other words, a rise from 50 dB to 60 dB is not an increase of about 20 percent but of 100 percent. A rise from 150 dB to 160 dB is also a rise of 100 percent.)

Since the sound radiates away from the sonar on a ship's hull in a rapidly expanding hemisphere, its power attenuates rapidly.

How rapidly depends on the temperature, saltiness of the water, even the height of the waves.

Roughly, a source rated 235 dB at the hull will attenuate by 60 dB within a kilometer (half a mile). A 235 dB sonar ping is fairly close to the click of a sperm whale, the loudest natural sound source in the ocean at about 230 dB.

A twin-engine Zodiac might hit 160-190 dB, a humpback whale's song 173-195 dB.

But it also makes a difference what frequency the sound has and how long the sound lasts. The blast of a jet airliner is continuous. A sonar ping lasts about a second, repeated after 30 seconds.

Ships don't like to ping, because it gives their position away to a lurking submarine, but Erkelens says a good sonar operator can often identify the presence of a sub with a single ping, then establish its distance and bearing with one more.

A good deal is known about how loud sounds affect human hearing, but not much research has been done on marine mammals.

In a tank in California, bottlenose dolphins and beluga whales have been exposed to sounds as loud as 195 dB (at one micropascal energy), Erkelens says.

"At that level, they experienced a temporary threshold shift," which is what we experience at a loud rock concert. For a time, the range we can hear shifts downward, but it returns.

A permanent hearing shift can occur at higher exposures, but no one knows how high for whales.

For humans, permanent damage sets in around 215 dB.

There is also the question of whether Navy ships far at sea will adhere to whatever standstill measures an EIS imposes. (Proposed mitigation measures, already in use, include shutting down sonar when a marine mammal is spotted within 2,000 yards of a ship or halving the power of the ping.)

Capt. Aaron Cudnohufsky, commander of the Pacific Missile Range Facility and the Hawaii Range Complex, says he is "very confident. We have a very disciplined military."

He says there is an extensive debriefing after each exercise, and every use of a weapon, whether practice or tactical, is inventoried.

The days when Navy pilots (Cudnohufsky is an aviator) "flat hatted" are over, he says. Today, violations of regulations lead to courts-martial, and for aviators, a number have lost their wings.