Case Study · Tsunami · 1960
May 23, 1960. Fifteen hours after the largest earthquake ever recorded in Chile, a tsunami arrived at Hilo Bay, Hawaii. The Pacific Tsunami Warning System had worked. Sirens had sounded 6 hours earlier. Yet 61 people died — because they watched the first small waves and concluded it was safe to return. Eight waves arrived in sequence over the next several hours. The third wave was the most destructive. It arrived after many people had come back to the waterfront.
Hilo Bay, Hawaii · May 23, 1960
On May 22, 1960, the largest earthquake ever instrumentally recorded struck the coast of Chile — magnitude 9.5, rupturing a fault zone approximately 1,100 kilometers long along the South American coast. The earthquake itself killed more than 2,000 people in Chile and left 2 million homeless. But the tsunami it generated was Pacific-wide. Traveling at speeds exceeding 400 miles per hour, the wave train moved north and west across the open ocean. The Pacific Tsunami Warning System, established in 1948 after the deadly 1946 Aleutian tsunami, detected the event and issued warnings. Sirens sounded in Hilo, Hawaii, approximately 6 hours before the first wave was expected to arrive.
The tsunami reached Hilo Bay on the big island of Hawaii approximately 15 hours after the earthquake — on the morning of May 23, 1960. The Hawaii Emergency Management Agency documents the arrival: "Fifteen hours later, eight waves began arriving in Hawaiian waters between twelve and twenty minutes apart, and crashed in the town of Hilo, Waiakea, and Shimachi downtown area." The first waves were smaller. The USGS/NOAA account of the Hilo event captures what happened at 12:12 AM in a famous photograph: a policeman and curious citizens stand watching the water arrive at the Wailoa River. The caption reads: "The tsunami had been predicted to hit at 12 midnight; according to the clock on the wall of the building, this photo was taken at 12:12 AM. The largest and most damaging third wave had not yet arrived." The wave that would kill most of the night's 61 victims had not yet come. And yet people were standing by the water, watching — either because they had not evacuated, or because they had evacuated, seen the first waves, and returned.
May 23, 1960
Date
8 waves
In Sequence
61 deaths
Hawaii
35 feet
Max Wave Height
6 hrs early
Warning Issued
The Moodys analysis of the event documents the reasons for the deaths: "The majority of these casualties occurred because people did not evacuate, either due to misunderstanding or not taking the warnings seriously. Many remained in the Waiakea peninsula area, which was perceived to be safe due to the minimal damage experienced there during the event triggered by the 1946 Aleutian Islands earthquake. Others initially evacuated to higher ground but returned before the event had finished." The history of the previous 1946 tsunami created false confidence in multiple ways: those who remembered 1946 knew where the 1946 waves had and hadn't reached, and used that geography as a false guide. And those who knew the 1960 warning suggested the waves would be smaller than 1946 (which they were in most of Hawaii) concluded that the risk was lower than it turned out to be in Hilo Bay specifically — where bay geometry created a "cumulative bounce" that made the waves more destructive than elsewhere in the islands. The warning system worked. The waves came in sequence, with the first being smaller. And the people who came back to see what was happening when the first waves were smaller than expected died in the third.
The Science
Think of a large earthquake rupturing a fault zone 1,100 kilometers long as a stone dropped into a lake — not a single point, but a long line. The resulting wave train is not a single wave but a series of waves with different periods, heights, and arrival patterns depending on the exact shape of the sea floor displacement, the distance traveled, and the shape of the receiving bay. The SMS Tsunami Warning analysis of the 1960 event documents this directly: "The first wave is not necessarily the largest and each wave may crest higher at different locations. Each tsunami may also have its maximum crest at a different location." For the Hilo 1960 tsunami, eight waves arrived between 12 and 20 minutes apart. The first wave was not the largest. The third wave was. A person who saw the first wave arrive and recede and concluded the event was finished was killed by the third wave — 24–40 minutes after the first one. This is the defining physics lesson of every tsunami case study: a tsunami is not a single wave. It is a train of waves, and the event is not over until official all-clear is given.
The Britannica account of the 1960 Chilean earthquake notes that "Hilo's position in the bay caused a cumulative bounce of tsunami waves far more destructive to Hilo than to other more exposed areas of Hawaii." Elsewhere in Hawaii, wave heights were in the 3–17 foot range — significant but not catastrophic. In Hilo Bay, they reached 35 feet. The bay geometry — a curved coastline at the end of a bay that faces directly into the direction of wave arrival from the Pacific — focuses wave energy. As waves compress into a narrowing bay, their height increases. The shape of the seafloor further amplifies the effect. Hilo has been hit by damaging tsunamis in 1946, 1960, and several other events precisely because its geography functions as a natural wave amplifier. The warning system could not inform individual households about this bay-specific amplification in 1960, which is one reason the death toll was concentrated there despite the city-wide warning.
The Moodys analysis of the 1960 event specifically addresses the tendency to underestimate far-field (tele-tsunami) risk: "A series of waves is a common feature of far field tsunamis, with the first wave typically not being the largest... It can be easy to forget the risks at potential far field sites." A far-field tsunami is one generated by a distant earthquake — not the local quake you felt, but one that struck thousands of miles away. For communities on the US Pacific Coast, Hawaii, and coastal US territories, the primary tsunami threat is from subduction zone earthquakes along the Alaska-Aleutian chain, the Cascadia Subduction Zone, South America, and the Japan Trench — all far-field sources. The Pacific Tsunami Warning System provides 2–15+ hours of warning for far-field tsunamis. But those hours of warning are only valuable if people leave and stay gone until the all-clear.
Timeline
01
May 22, 1960, 3:11 PM local (Chile): M9.5 earthquake strikes near Valdivia, Chile — the largest ever recorded. 1,100 km of fault ruptures. 2,000+ dead in Chile; 2 million homeless. Tsunami generated immediately, travels outward in all directions across the Pacific. Pacific Tsunami Warning System detects the earthquake and activates. Warnings transmitted across the Pacific. Sirens sound in Hilo, Hawaii, approximately 6 hours before expected first wave arrival.
02
Evening of May 22: Many Hilo residents either don't evacuate or evacuate and return. Reasons: misunderstanding the siren (some thought it was a preliminary signal requiring another signal to evacuate); confidence based on 1946 experience (the previous tsunami had missed the Waiakea peninsula); assumption the 1960 event would be smaller and less dangerous than 1946. By midnight: people in the Waiakea peninsula area, the Shimachi area, and along the waterfront — either having never left or having returned after the initial smaller waves.
03
May 23, midnight: First waves begin arriving at Hilo Bay. 12:12 AM: A famous photograph captures a policeman and citizens watching the water arrive from the waterfront. Clock visible: 12:12 AM. Caption: "The largest and most damaging third wave had not yet arrived." Waves arrive 12–20 minutes apart. Third wave: most destructive; 35 feet at Hilo. 61 dead; 282 injured. 229 dwellings destroyed; 308 businesses destroyed. $75 million in damage. The third wave finds people who thought the tsunami was over.
04
Japan, 22 hours after earthquake: 138 killed, 1,600+ homes destroyed. Philippines: 32 dead or missing. West coast of US: $3.7M damage, no deaths. Total global death toll: ~2,223 (including Chile). The event prompted massive post-mortem review of why the warning failed to save more lives in Hilo — and produced the USGS "Surviving a Tsunami" guide with explicit lessons: "Expect Many Waves," "Head for High Ground and Stay There." Hilo redesigned its waterfront as parkland after 1960.
Human Decisions
Why people didn't stay gone
Hilo had been hit by a Pacific-wide tsunami in 1946, which killed 159 people in the town. Survivors of 1946 knew where that tsunami had and hadn't reached. The Waiakea peninsula had minimal damage in 1946 — so many residents there felt safe staying. The 1960 warning also suggested the waves might be smaller than 1946 (which they were in most of Hawaii). People used their prior experience — the map of the 1946 disaster — as their safety guide for 1960. That map was wrong. The 1960 waves were not 1946 waves. The same place that had been safe before was not safe this time. Prior experience with a disaster provides calibration; it does not provide a guarantee.
The USGS/NOAA account documents a critical communication failure: "It appears many of those killed or injured thought the siren was a preliminary warning that would be followed by a second evacuation signal, instead of understanding that it was a signal to evacuate immediately without further notice." The siren system was correctly operated. The message was ambiguous in the minds of recipients. People waited for a second signal to tell them to actually leave — and no second signal came, because the first one meant "leave now." This communication failure has directly shaped how emergency alert systems are designed since 1960: clarity about what each alert means, not just that an alert is sounding.
The curiosity that kills
The famous 12:12 AM photograph captures the moment precisely: people are watching the water arrive. The tsunami had been predicted at midnight; the clock shows it's now 12 minutes past. The first waves have arrived and are visible. The people in the photograph are watching from what they believe is a safe distance. They have seen the water come in and they are watching. They are treating the tsunami as a spectacle to observe rather than a danger to flee. And the third wave — larger than the first, larger than the second — had not yet arrived. The curiosity to watch a disaster rather than flee it is a documented, consistent, fatal human response to tsunami warnings when the first waves don't appear immediately life-threatening.
No official all-clear was issued after the first waves in Hilo. Yet many people acted as if one had been. They made an individual judgment — the waves seemed small, the situation seemed under control — and acted on it. The tsunami did not match the catastrophic image they had of what a tsunami should look like. The mismatch between expectation and first observation caused them to revise their risk assessment downward and return. The USGS lesson that emerged directly from 1960: don't leave the evacuation zone until you receive an official all-clear from emergency authorities. Not until the waves seem small. Not until nothing has happened for 20 minutes. Until officials say it's over.
The cascade lesson
The 1960 Hilo tsunami is the defining case study for the "return before the all-clear" failure mode — the pattern that reappeared in Crescent City four years later, that the USGS enshrined in its tsunami survival guide, and that continues to kill people in every major tsunami event where the first wave is not the largest. The tsunami warning system gave Hilo six hours. The waves arrived in eight sequences over multiple hours. Every person who died in Hilo had time to be safe. The failure was not the warning. It was the human response to a smaller-than-expected first wave: the judgment that the danger had passed, the curiosity about what was happening, the pull of belongings and buildings and familiar places. The USGS lesson from the survivors' accounts is as simple as it is hard to follow in practice: when you get a tsunami warning, go to high ground. Stay there. Don't come back until an official all-clear is issued.
What You Can Do Now
Every tsunami survival lesson from the 1960 Hilo disaster comes down to one principle: get to high ground when the warning sounds, and don't come back until officials say it's safe. These five actions make that principle actionable.
The 1960 Hilo deaths include people who interpreted the siren as a preliminary signal requiring a second alert before they should actually leave. A tsunami warning siren is a "leave now" signal. In the US, NOAA's Pacific and Alaska Tsunami Warning Centers issue tsunami watches, advisories, and warnings — a Warning means leave coastal and low-lying areas immediately. Don't wait for the waves to arrive to decide how serious it is. By the time you can see the wave, you are usually out of time.
Tsunami warning response guideThis is the lesson extracted directly from the Hilo 1960 survivors' accounts. The third wave killed people who had gone back to watch after the first smaller waves arrived. The all-clear in a far-field tsunami event may come 6–12+ hours after the first wave, because wave trains can continue arriving for many hours. The "nothing has happened for 20 minutes" judgment you make yourself is not the same as an official all-clear. Go to high ground, have supplies to wait, and wait for the official signal.
Tsunami evacuation guideThe people who survived in Hilo in 1960 had gone to high ground before the waves arrived and stayed there. Knowing in advance where the designated high-ground evacuation areas are in your community — and having a route to reach them — means you can act in the first minutes of a warning without searching for information. NOAA and state emergency management agencies publish tsunami evacuation zone maps for all coastal US communities. Check yours and know your route.
Tsunami zone identification guideThe residents of Waiakea peninsula believed they were safe because the 1946 tsunami had missed them. The 1960 tsunami didn't. Tsunamis from different source locations, different earthquake magnitudes, and different approaching angles can produce drastically different wave heights in the same bay. The 1960 tsunami was smaller than 1946 everywhere in Hawaii except Hilo, where bay geometry amplified it unexpectedly. Your personal experience of a previous tsunami tells you what that specific tsunami did. It doesn't tell you what the next one will do.
Tsunami risk and wave behavior guideThe 1960 Hilo survivors who stayed safe on high ground were there for hours. An evacuation that requires you to return to your house to gather supplies before leaving — or that leaves you without water, medications, or communications tools while waiting for an all-clear — creates pressure to go back too soon. A prepared go-bag with water, snacks, medications, phone charger, and a battery-powered radio means you can evacuate immediately and wait safely without creating a reason to return before the all-clear.
Tsunami evacuation go-bag guideTsunami case study series
The 2004 Indian Ocean Tsunami covers the world without a warning system. Crescent City 1964 covers the fourth wave that arrived after people came back. Tōhoku 2011 covers the seawalls that didn't hold. Palu 2018 covers the tsunami no one expected from a strike-slip fault.
Full tsunami case study seriesSources