Tsunamis and Volcanoes
Everybody, it seems, knows that active volcanoes make poor neighbors. In Hawaii, they disgorge hot lava that starts forest fires and buries homes and highways. Here in Alaska, Mt. St. Augustine, which is on Augustine Island in Cook Inlet, spews clouds of ash into international air routes. The great explosive eruption of Mt. St. Helens in Washington State and the accompanying landslide devastated thousands of acres within minutes.
What happens when such a landslide hits a sizable body of water? The answer can be a tsunami.
Popularly but erroneously called a tidal wave, a tsunami is actually a displacement wave. When a mass of land shifts under (or into) the sea because of submarine landslides, earthquake-caused slumps or uplifts of the sea floor, or volcanic activity, a mass of water must move. Anyone with a bathtub and a lively child is probably familiar with this general principle. Typically, Mt. St. Augustine's eruptions involve block and ash flows and debris avalanches rather than lava flows, so this volcano could cause a tsunami.
A little over a hundred years ago, apparently it did. On the morning of October 6, 1883, the daily log of the Alaska Commercial Company at English Bay, on the shore of the Kenai Peninsula straight across Cook Inlet from Augustine Island, noted "Eruption of the active volcano" with a "rain of finely Powdered Brimstone Ashes" and four tidal waves, "the sea rising 20 feet above the usual Level." The lobe of the debris avalanche of 1883 has been found; it extends at least three kilometers offshore of the north coast of Augustine Island.
An enormous avalanche like the one of 1883 comes not just from eruptive products but from edifice collapse; as at Mt. St. Helens, a portion of the volcano itself hurtled down. The possibility of a similar collapse during the 1986 eruptions of Mt. St. Augustine was a real concern, especially since the shores of Cook Inlet are now far more populated.
Zygmunt Kowalik of the University of Alaska's Institute of Marine Science, T.S. Murty of the Institute of Ocean Sciences in Sidney, British Columbia, and I set out to produce a computer model of the 1883 event. With the historical record and the geologic evidence to go on, we would have a "postdiction" that could be verified. That would give a check on our methods and information to let us know if we could then make a reasonable prediction of things like tsunami size and travel times for different-size avalanches entering the sea on different sides of the island.
A full report on our work was published in the June 12, 1987 issue of Science, the magazine of the American Association for the Advancement of Science, but in essence it turned out that our computer model matched the real event pretty well - well enough so that the calculations were useful for emergency planning during last year's eruptions. The model indicated, incidentally, that should an event identical to the one of 1883 occur, Anchorage would have some four hours' time to prepare for a very small wave that would cause no problems at low tide. Homer, on the other hand, would have about an hour to get ready for a tsunami near the size of the one reaching English Bay. That could inundate a lot of land if it came at a time of high tide.
