Case Study · Tsunami · 1964
March 27, 1964. The Good Friday Earthquake struck Alaska. Four hours later, tsunami waves reached Crescent City, California. After the second wave, the ocean went quiet for 40 minutes. Survivors went back to check on their businesses. The harbor then drained completely — people could see the seafloor in the moonlight. The fourth wave was the largest. It flooded 29 city blocks. Eleven people died. They were the people who had come back.
Crescent City, California · March 28, 1964
On March 27, 1964 — Good Friday — the largest earthquake ever recorded in North America struck Prince William Sound, Alaska at 5:36 PM. Magnitude 9.2. The earthquake itself was devastating: it destroyed much of downtown Anchorage, triggered massive local tsunamis in Alaska fjords that killed 82 people, and leveled coastal towns including Valdez and Seward. But it also generated an open-ocean tsunami that traveled south along the Pacific Coast at the speed of a jetliner. British Columbia, Washington, Oregon, and California all sustained damage. The hardest-hit community outside Alaska was Crescent City, California — a small town of 3,000 people that juts into the Pacific Ocean near the Oregon border.
The Redwood Coast Tsunami Work Group's detailed analysis of the Crescent City event documents the wave sequence in precise terms. The first surge arrived. Then the second. Then, approximately 40 minutes after the second surge, the ocean went quiet. The account is specific: "This is the point when some survivors reported returning to the coastal area. They assumed the tsunami was over, in part based on what they had experienced in the 1960 tsunami." The people who went back were applying the pattern of the 1960 event — which had also struck Crescent City — to the 1964 event. In 1960, the wave sequence had been different. The quiet after the 1964 second surge seemed, to those who returned, like the confirmation that it was over. It was not. The RCTWG analysis continues: "At a little before 1 AM, a third surge begins and the water quickly rises, surpassing the first surge and knocking the tide gauge housing over. After this point, the tsunami behavior is uncertain. Most survivors described four major surges and the fourth was the largest. After the third surge, the harbor drained completely and people could see the seafloor in the moonlight."
The fourth surge. The largest. The NOAA account of the Alaska tsunami's California effects documents the outcome: "Waves more than 14 feet above tidal height flooded 29 blocks of Crescent City, California, and killed 11." The RCTWG provides the specific amplitude: the fourth surge was "just high enough to top the sea barrier on the west side of Crescent Harbor, an amplitude of 15.7 feet." With the still-high tide boosting the peak water elevation: nearly 22 feet. Eleven Californians died. The Earth Magazine analysis is specific about the character of their deaths: the four waves arriving up to 7 meters in height destroyed the main business district and ignited a Texaco oil tank farm, which burned for three days. The entire 11 deaths were among those who had come back — either those who had never left, or those who returned during the 40-minute quiet. The people on high ground survived.
Mar 27–28, 1964
Date
4 surges
Wave Sequence
40 mins
False Quiet After Wave 2
22 feet
Wave 4 Peak Height
11 deaths
All After Return
The Science
Think of a tsunami wave as oscillating — it doesn't just push water toward shore, it also pulls water away. When a tsunami is about to send its largest surge toward the coast, it often first pulls water dramatically away from shore — a phenomenon called the drawback or withdrawal. This can expose the seafloor for hundreds of meters beyond the normal low-tide line. Historically, people who observe the drawback and have never been warned about it see an unprecedented opportunity: the seafloor is exposed, boats are grounded on mud, fish are visible. Some go toward the exposed seafloor out of curiosity. The drawback in Crescent City that night — the harbor draining to reveal the seafloor in the moonlight after the third surge — was the visible signal that the largest wave was drawing its breath for the largest push. The people who were in or near the harbor to witness the draining were killed when the fourth surge arrived.
Far-field tsunamis (generated by distant earthquakes) produce wave trains where the relationship between wave order and wave height is not predictable. A 40-minute gap between surges is not unusual; the period between individual waves in a tsunami train can range from 10 minutes to 90 minutes or more. The gap doesn't mean the tsunami is over. It means the next wave is somewhere between 10 and 90 minutes away and its height cannot be predicted from the heights of the preceding waves. The 1964 Crescent City event — like the 1960 Hilo event — documents the specific danger of interpreting a quiet period as an all-clear. The people who returned to Crescent City during the 40-minute gap were using experiential reasoning (nothing is happening, it must be safe) to make a life-or-death decision that required scientific knowledge they didn't have.
The NPR account of Crescent City's tsunami history notes that local emergency officials describe "something out there off our coast that magnifies the tsunamis as they come into Crescent City." This is not unusual for small coastal communities: underwater topography, bay geometry, and the angle of wave arrival can create focusing effects that make some locations more susceptible than others to specific tsunami sources. Crescent City has been struck by multiple tsunamis from both Alaska and the 2011 Tōhoku earthquake — each time receiving attention from the same geographic focusing that amplifies waves approaching from the north Pacific. The 2011 Tōhoku tsunami also caused damage in Crescent City, killing one person — the only California fatality from that event. The same geography that made Crescent City the most hard-hit California location in 1964 makes it the most vulnerable California location to future northern Pacific tsunamis.
Timeline
01
March 27, 1964, 5:36 PM: M9.2 earthquake strikes Prince William Sound, Alaska — the largest US earthquake ever recorded. Destroys downtown Anchorage. Alaska continental shelf rises 30+ feet. Local tsunamis kill 82 in Alaska fjords. Valdez, Seward, Kodiak devastated. Open-ocean tsunami generated, travels south at jetliner speed. Pacific Coast states issued tsunami warnings. Crescent City, California — a town that juts into the Pacific near Oregon border — directly in path.
02
Late night, March 27: First surge arrives at Crescent City — water rises, then recedes. Second surge. Then approximately 40 minutes elapses with no wave action. "This is the point when some survivors reported returning to the coastal area. They assumed the tsunami was over, in part based on what they had experienced in the 1960 tsunami." People walk back to check on boats and businesses. Some had never left. The town is partially flooded but not catastrophically damaged yet.
03
Before 1 AM: Third surge begins, surpasses first surge, knocks over tide gauge housing. After third surge: harbor drains completely. The seafloor is visible in the moonlight. "People could see the seafloor in the moonlight." To those in the harbor: unprecedented. To those with tsunami knowledge: the fourth wave is drawing its breath. People who have come back are now in the harbor, in the business district, at what they think is the receded sea — standing on or near the ocean floor.
04
~1 AM: The fourth surge arrives — the largest. 15.7 feet amplitude, topping the sea barrier on the west side of Crescent Harbor. With the tide still relatively high: nearly 22 feet total water elevation. Waves 14+ feet above tidal height flood 29 city blocks. Main business district destroyed. Texaco oil tank farm ignited — burns for three days. 11 people killed in California. 29 blocks flooded. $25–30M damage. The fourth wave finds the people who came back.
Human Decisions
The pattern matching that killed
The RCTWG analysis explicitly documents that the people who returned during the 40-minute quiet did so "in part based on what they had experienced in the 1960 tsunami." The 1960 Chilean tsunami had also struck Crescent City — and its wave sequence had been different. Those who had experienced 1960 had a mental model of what a tsunami in Crescent City looks like. They used that model to interpret the 40-minute quiet as the end of the event. The 1964 Alaskan tsunami did not match the 1964 Chilean tsunami's pattern. Every tsunami has a different wave sequence depending on the source earthquake's location, magnitude, and mechanism. Prior experience with one tsunami is not a predictor of the wave sequence of the next.
The RCTWG analysis notes that the fourth surge's destructive height was in part because "the tide was still relatively high and boosted the peak water elevation to nearly 22 feet." This is a critical detail for coastal preparedness: a tsunami that arrives at high tide is more destructive than the same tsunami arriving at low tide. The effective height of a tsunami wave at a specific location is the sum of the tsunami wave amplitude plus the tidal height at arrival. The 1964 event at Crescent City arrived during a high tide period, which pushed the fourth surge over the sea barrier that might have contained it at a lower tide. Tsunami preparedness must account for the compound risk of a major tsunami coinciding with high tide.
The warning that existed — and its limits
The NPR 2005 account of Crescent City's tsunami history notes directly: "Back then, the only real warnings were the tsunami alerts on radio and TV." The town had no sirens in 1964. The Pacific Tsunami Warning Center had issued alerts and warnings, which were broadcast over radio. But no audible outdoor warning was immediately available to people who might be outside, asleep, or not monitoring radio at 1 AM. In the 41 years between 1964 and 2005, Crescent City installed a tsunami siren system — tested monthly at 10 AM on the first Tuesday of every month, as the NPR account documents. The 1964 deaths contributed directly to the installation of that warning system.
The NPR 2006 account of a much smaller Hokkaido earthquake tsunami documents Crescent City's vulnerability persisting: the city was "the only California community to suffer major damage" from that event. The same geographic factors — the bay shape and underwater topography that scientists describe as "something out there off our coast that magnifies the tsunamis" — make Crescent City the most tsunami-vulnerable California community. The 2011 Tōhoku tsunami caused one death and significant harbor damage in Crescent City — 47 years after 1964. The Redwood Coast Tsunami Work Group exists specifically to address this vulnerability. It is one of the most active community tsunami preparedness programs in the US because the community knows its specific exposure.
The cascade lesson
The Crescent City 1964 tsunami adds a specific lesson to the 1960 Hilo lesson: not only is the first wave often not the largest, but wave sequences can include long quiet intervals that appear to be the end of the event. The 40-minute silence between surges 2 and 3 in Crescent City was enough to convince people who had survived both earlier waves that it was over. They were wrong. And the visible physical signal that should have warned them — the harbor completely draining, the seafloor exposed in the moonlight — was not recognized as the precursor to the fourth wave because they didn't know what the drawback meant. The USGS lesson from the 1964 Crescent City disaster: expect many waves; the event is not over until an official all-clear is issued; if you see the ocean drain dramatically, get to high ground immediately. The ocean draining is not the end. It is the next wave drawing its breath.
What You Can Do Now
The 1964 Crescent City deaths add a specific lesson beyond 1960 Hilo: the ocean can go quiet for 40 minutes between waves, and that quiet is not an all-clear. If you see the ocean drain dramatically, it is the most urgent warning to run.
The drawback or withdrawal — when the ocean recedes dramatically before a tsunami wave arrives — is one of the most dangerous phenomena in a tsunami event because it looks unusual and draws curiosity. The people who saw the Crescent City harbor drain completely had never seen anything like it. The correct response, now well-documented: if the ocean recedes dramatically and unusually, don't go to look at it. Run inland to high ground as fast as possible. This is a natural warning sign that requires no official alert — it is the next wave drawing water toward itself.
Natural tsunami warning signs guideThe 40-minute gap between the second and third surges at Crescent City fooled people who had experienced the 1960 event into thinking the 1964 event was over. A 40-minute gap is not unusual in a tsunami wave train. Between-wave periods can range from 10 to 90+ minutes. Silence does not mean it's safe to return. Only an official all-clear from emergency management authorities means it's safe to return. The 1964 lesson and the 1960 lesson are the same: stay on high ground until officials say it's over.
Tsunami wave sequence guideFor local tsunamis (generated by a nearby earthquake), official warnings may not arrive before the waves. The three natural warning signs that require immediate high-ground evacuation without waiting: (1) strong or prolonged ground shaking near the coast, (2) an unusual and dramatic withdrawal of the ocean, and (3) a loud roaring sound from the ocean similar to a freight train. Any one of these is sufficient reason to run to high ground immediately. The earthquake in Alaska that caused the Crescent City waves was felt as far south as the US Pacific Northwest — that felt shaking is the natural first warning for teleseismic tsunamis close enough to generate local impact.
Natural tsunami warning guideThe 1964 Crescent City fourth wave reached 22 feet total because the tide was still relatively high when it arrived, boosting the tsunami amplitude over the sea barrier. Tsunami preparedness awareness should include recognition that a tsunami's effective inundation height at any specific location is the sum of the wave amplitude plus the tidal height at arrival. A tsunami that arrives at high tide will flood further inland and to greater heights than the same tsunami arriving at low tide. This doesn't change the evacuation response — evacuate whenever a warning is issued — but it informs why the same warning can produce different outcomes at different tidal phases.
Tsunami and tidal interaction guideCrescent City established the Redwood Coast Tsunami Work Group specifically because of its documented, persistent vulnerability. The RCTWG produces evacuation maps, community drills, and detailed analysis of Crescent City's specific tsunami exposure. Most US coastal communities with documented tsunami risk have equivalent programs — often through their county emergency management office or through NOAA's TsunamiReady program. Knowing your community's specific vulnerability, evacuation routes, and designated high-ground shelters is the preparedness investment that makes the difference when a tsunami warning arrives at 2 AM.
Community tsunami preparedness guideTsunami case study series
The 2004 Indian Ocean Tsunami covers the world without a warning system. Hilo 1960 covers people watching the first waves and coming back. Tōhoku 2011 covers the seawalls that didn't hold. Palu 2018 covers the tsunami from a fault type that wasn't supposed to generate one.
Full tsunami case study seriesSources