Soon after going to work on the bright morning of April 1, 1946, the seismologist in charge of the earthquake detection instruments at the University of California in Berkeley started to develop the photographic records of the preceding 24 hours of observation. Without unusual interest he watched the wavy traces appearing on the wet paper. There was evidence of a strong earthquake some 2,200 miles distant. Earthquakes, however, were a frequent occurrence, and this must have been in a deserted place—the morning radio news had mentioned no death nor destruction. Probably it was under the sea.
Seismologists in other places were developing their records, and noting the early morning shock. As is the way with these scientists, an exchange of readings soon began. Since it appeared this was a very strong shock, interest grew and wide use of wire and radio was made. It became apparent that a new rupture had occurred under 2,000 fathoms in the North Pacific nearly a hundred miles southeast of Unimak Pass, in the Aleutians.
On the cliffs of Unimak Island, on land 32 feet above the water, stood the six-year-old Scotch Cap Lighthouse, so-called because of the striking profile of the headland. The flashing navigational light, in a tower atop the concrete bunkhouse, stood 60 feet high, watching over the turbulent currents and tide rips of the pass joining the Pacific and Bering Sea. Suddenly at 0238 the waters rose in the dark night and dashed against the cliff. The man on watch must have heard the angry rush of water—four others must have been asleep. In a few moments the structure and its associated building were engulfed, then the waters fell back, carrying much of what was there, including a radio tower swept from its base 100 feet above the water. What was left was a shambles of broken concrete and twisted steel. All five men were lost in those few moments, but their fate was not known until daylight next morning, when Coast Guard planes, investigating the loss of radio contact, flew over the strewn wreckage.
By noon the newsrooms of the big city papers on the coast were filled with teletype messages telling of trouble at Scotch Cap, and of breaking waves on the shores of the Aleutian Islands and the Alaska Peninsula. Radio news reporters told breathless and highly colored tales of great waves racing across the North Pacific. Fliers of the Air Force and Navy skippers were sent on scouting missions to observe these menacing waves. They never found anything but evidences of surf and flooding in low places along the coast.
At Honolulu, where morning comes two hours later than on the coast and the serenity of the never-changing weather is counted a local blessing, early risers were starting their day’s routine. The officer in charge of the Coast and Geodetic Survey Magnetic Observatory at Ewa, near Pearl Harbor, preparing for breakfast at his beach home, looked in amazement at the sea. It seemed to have withdrawn! The whole bay was bare, with flopping fish everywhere. He stifled an impulse to run out to see what this was all about, when he heard a distant rumbling and then saw a line of combers out beyond the bay. This rapidly approached, first filling the bay, then sending a seething mass of water over the land. The house shook with the impact. Furniture floated away. The rooms were filled waist deep. Nearby houses were washed away. After a breathless struggle to hold things together till the water receded, he rested wearily, trying to recall what he had read of tsunamis—seismic sea- waves, so-called by the Japanese to distinguish them from other types of seawaves, commonly called tidal waves.
Just before 0700 the water receded from the bay area at Hilo, on the big southern island of Hawaii, 2,375 miles from the submarine earthquake that had disturbed seismographs the world around. Children rushed out, fascinated by the stranded marine life on the flats. Preoccupied, they failed to see the rapidly approaching crest of the tsunami. There was sudden turmoil. The entire waterfront collapsed before millions of tons of water. There was a recession, then two still larger waves, then others in a long diminishing series about fifteen minutes apart. Sad survivors of Hilo found the cost of the morning’s events to have been 173 lives and 25 million dollars—the most disastrous wave in Hawaiian history!
At 0136 next morning the tide gage at Valparaiso, Chile, 8,066 miles and 18 hours distant from the now notorious Aleutian earthquake, showed a rise of five feet, followed by a slowly diminishing series of waves that lasted many hours. Tide gages up and down the west coast of the Americas and in the islands of the western Pacific and Japan, all in their own times, reacted to the giant sloshing of the water in the Pacific basin. This sloshing—the tsunami—is perhaps one of the most fearsome effects of nature.
It is relatively infrequent, though by no means rare, and little is known about it outside scientific circles. About one small wave can be found each year on the tide gage records of Honolulu. Damage has occurred at rare intervals—sometimes from waves traced to earthquakes under South American and Japanese coastal waters. It has been estimated that one damaging tsunami occurs somewhere on earth each year. Nicholas Heck, in his catalog of historic tsunamis through 1946, lists 270 which are sufficiently documented to be considered proven—most of them in the past 200 years. The earliest wave mentioned is listed without details at Potidaea, in 479 BC. Three later inundations, all in Greece, preceded the Christian era. The three disasters causing the greatest death tolls occurred in Japan in 1293, when 30,000 died; in Italy in 1783, with an equal toll; and again in Japan, on June 15, 1$96, when 27,000 were caught on wave-swept shores. Some cities, including Concepcion, Chile, have been destroyed time and again. The U. S. Navy Gunboat Wateree was lost in 1868 by being swept a quarter mile inshore by a wave estimated at seventy feet near Arica, on the west coast of South America.
Obviously much loss of life and property damage could be avoided if the people at danger points had timely warnings. Some weeks after the 1946 events, the Commerce Department’s Coast and Geodetic Survey was asked about this by a member of Congress. “How did it happen,” he asked, “that the disaster at Hilo happened without warning from you, since it began with an earthquake in Alaska, and the wave took 4£ hours to reach Hilo, and meanwhile you had the seismograph reading to give warning?”
Of course this was not so. Most earthquake records at that time were made on photographic paper and developed once a day. Obviously eight o’clock in the morning was too late. This explained Hilo, but it was certainly time for the Survey to develop and install instrumental warning devices at its seismograph stations, arrange a communications network connecting earthquake and tide stations with a surveillance center at Honolulu, and thenceforth to provide the desired warnings.
Organization of such a warning service required the solution of several problems. Relatively few submarine quakes produce tsunamis. Shepard, Macdonald, and Cox, in analyzing the 1946 wave, noted that it originated on the north slope of the Aleutian Trench and called attention to the fact that most tsunami-producing quakes are those in the deep ocean trenches. This is a natural conclusion since these trenches are active zones of vertical crustal readjustment. Since, however, many seaquakes cause no tsunamis, the surveillance system has to call on tide and wave reporting stations to observe actual tsunamis before definite alerts are possible. The arrival times at Hawaii from a known earthquake epicenter require estimation based on travel-time charts, compiled by the Survey on theoretical grounds for the determination of wave speeds over any particular course.
Accuracy of time predictions is limited by our incomplete knowledge of sea depths, which control wave velocities. In the case of the 1946 wave arrival at Valparaiso, the wave beat the prediction by 30 minutes, “finding a better path,” as the oceanographers put it, “than we could!”
Earthquakes result from impulsive splitting or shearing of rock strata in the earth’s crust, the break occurring when the rocks are stressed beyond their strength by crustal deformation. Vibratory waves of various types radiate from such a shock, passing by several paths through the solid earth and producing the seismograph records read by seismologists to locate and analyze the disturbance. These waves traverse the earth in a few minutes. When a vertical motion of the strata comprising the sea bottom is involved, the whole ocean is displaced vertically, imparting tremendous potential energy to the water. Subsequent levelling of the water surface generates spreading waves on the water surface which have great kinetic energy. Since they are caused along fracture lines and not at points, they are most effective in directions normal to the fault line. These spreading waves are strange indeed.
Their physical characteristics are surprising. The initial amplitude, or height, is determined by the amount of sea bottom displacement—never more than a few feet. This amplitude does not change much as long as the wave crosses deep ocean waters. It is probable that the 1946 tsunami crossed the North Pacific with heights not much more than two feet! If the height is slight however —not so the velocity! The speed of a long gravity wave in deep water, regardless of its height, varies with the square root of an amount which is the product of the depth and the gravitational constant g. It may reach more than 500 knots in some Pacific areas. The average velocity of the Hilo wave during its 4½ hour trip was 420 knots. We have one more attribute of the wave to consider—its period. Since a considerable time is required for the settling down of hundreds of square miles of elevated ocean, the waves are of long period, generally between twelve and twenty minutes.
These seemingly unrelated and disproportionate characteristics result in some interesting effects. A wave travelling at 500 knots with a period of fifteen minutes will be nearly 150 statute miles from crest to crest. Such a wave two feet high is not an impressive sight. It is actually an imperceptible swelling of the sea surface no more visible to a ship or plane overhead than the tide itself. This unobtrusiveness prevents detection until it reaches a coast. Once there it is seldom unobtrusive!
Why, then, does the power of the wave unleash itself in such terrifying manner upon reaching land? We have seen that the wave velocity depends upon the depth. As it reaches shoal water and the bottom slope leading up to a coastline, the velocity is more and more reduced. But the energy has to go somewhere. Friction dissipates some of it, but most of it goes to build greater and greater wave height. Just as a lazy ground swell shortens and steepens near a beach, and becomes a foaming breaker before its destruction, so does this long, low, and fast-moving swell slow down, grow higher, and end as a destructive super-breaker. Heights of sixty and seventy feet have often been reported. Walls of water strike waterfronts, great bores race up estuaries, and swirling floods cover low lands. That is what killed the watch at Scotch Cap, set the Wateree up on the shore, and made a shambles of the Hilo waterfront.
Like numerous other types of disturbances, the tsunami is propagated as a wave, which dies out gradually as the energy is slowly lost in internal friction and energy transfer to the earth and atmosphere. While the first crest is undoubtedly the highest at the source, there is a tendency for later waves, perhaps the third or fourth, to be the greatest at some distant point. The energy loss is so slow that two or three days are sometimes required for the rise and fall of the water surface to come to an end. In 1946, the period appeared to be twelve to fifteen minutes at Hawaii and throughout the North Pacific, but periods of twenty minutes at South American points were the rule—indicating a tendency for the waves to lengthen with distance. In many places the wave forms become complicated. Irregularities in height and period occur, particularly in places where local seiche activity is stimulated.
Tsunamis have sometimes arrived in the form of troughs or recessions before any rise of the water occurs. The 1946 wave in Hawaii actually began with a small rise which escaped wide notice, so that the following deep recession seemed like a beginning—a bad tiling for unwary spectators who may be lured into danger. Initial recessions might be explained as resulting from sinking, instead of rising, of the ocean floor at the source, and it could even be surmised that both upheaval and subsidence of the bottom might occur on opposite sides of a fracture line. The first arrivals might then have opposite phases on opposite sides of the line, although this has not actually been observed.
Wave reflections from continental shores have been noted. At Hanasaki, Japan, there were waves in 1946 which must have been reflected from the North American coast. Late arrivals at Honolulu were apparently reflected from the submerged continental slopes off the coast of Asia. The results of the reflections, of seiche interferences, and of the effects of beach slope, angle of approach, and the state of the tide and sea, so confuse the effect that, with present knowledge, the warning system can be used only to predict arrival times—not wave heights nor effects.
Notwithstanding its limitations, the warning system has given a good account of itself. Numerous submarine quakes have triggered alarms and started periods of feverish activity in the quarters of the network seismologists. Earthquake reports have gone ahead of other traffic to Honolulu. Plottings once made, the Honolulu workers have queried tide observers at strategic points, then studied travel-time charts. In rare cases, when necessary, warnings have been issued to civil and military authorities to clear beaches and make preparations.
A predicted wave costing $800,000 swept the shores of the islands in 1952 after an earthquake near Kamchatka, without loss of life. The dramatic quakes of March 9-15, 1957, in the Aleutian Trench near Adak, proved again the usefulness of the system. The initial shock caused a moderate tsunami which caused damage of three million dollars, claiming no lives but arousing widespread alarm. The recurring shocks in turn renewed the apprehension, which was quickly stilled by repeated assurances through the system that no additional waves existed.
The tired workers at Honolulu were finally relieved by the cessation of the heavy shocks and a welcome message:
CINCPACFLT EXTENDS A WELL DONE ON THE ALERT AND EFFICIENT MANNER IN WHICH THE SEISMIC WAVE WARNING WERE EXECUTED ON 9 MARCH X THE TIMELY WARNING SAVED MANY LIVES AND ASSISTED MATERIALLY IN MINIMIZING PROPERTY DAMAGE