The Lituya Bay tsunami of July 9, 1958, produced the tallest wave run-up ever recorded in human history: 524 meters (1,720 feet) of water surging up a mountainside in a narrow Alaskan fjord. That height is taller than the Empire State Building. The event was not caused by an open-ocean earthquake like most tsunamis people picture, but by an enormous rockslide slamming into the water at the head of the bay. The mechanics of how that happened, what the wave actually looked like, and what it means for hazard science today are all more interesting and more unsettling than the headline number alone suggests.
What Triggered the Wave
Lituya Bay is a T-shaped fjord on the coast of southeast Alaska, roughly 11 kilometers long and just over 3 kilometers wide at its broadest point. The Fairweather Fault, one of the most active tectonic features in North America, runs directly through the head of the bay. On July 9, 1958 (local time), a magnitude 7.8 earthquake struck along this fault. The shaking itself did not create the tsunami. Instead, the earthquake displaced the ground enough to destabilize a massive slab of rock on the steep eastern wall of Gilbert Inlet, one of the two arms at the head of the bay. That rock mass broke free and dropped into deep water below.
1Bulletin of the Seismological Society of America. The Alaska earthquake of July 10, 1958: Giant wave in Lituya BayThe rockslide itself was staggering. The original geological survey estimated the sliding mass at roughly 30.6 million cubic meters, tumbling from elevations between about 230 and 915 meters above sea level. Later researchers working with physical models and numerical simulations revised the effective bulk volume upward to around 51 million cubic meters once they accounted for the air voids within the broken rock, giving the mass a bulk density of about 1,620 kilograms per cubic meter. Picture a block of shattered mountain more than 130 meters thick at its thickest point, dropping from nearly a kilometer high into a narrow inlet of cold, deep water.
2Natural Hazards and Earth System Sciences. The 1958 Lituya Bay tsunami – pre-event bathymetry reconstruction and 3D numerical modelling utilising the computational fluid dynamics software Flow-3DHow a Rockslide Makes a Tsunami
Most people think of tsunamis as products of underwater earthquakes, and that is true for the large open-ocean waves that strike coastlines thousands of kilometers away. The Lituya Bay event was fundamentally different. When the rockslide hit the water at high velocity, it acted like a colossal piston being slammed into a bathtub. The impact created a massive air cavity behind the plunging rock, and the water displaced outward formed the initial wave. Then the air cavity collapsed, and the chaotic mixing of rock, air, and water amplified the energy being transferred into the wave.
3Natural Hazards and Earth System Sciences. The 1958 Lituya Bay tsunami – pre-event bathymetry reconstruction and 3D numerical modelling utilising the computational fluid dynamics software Flow-3DThis two-step process, impact followed by cavity collapse, is what makes landslide-generated impulse waves so violent compared to earthquake-generated tsunamis. An earthquake-generated tsunami in the open ocean might have a wave height of only a meter or two while carrying enormous energy across a vast wavelength. A landslide-generated impulse wave concentrates that energy into a much shorter wavelength and a much taller crest, especially when it forms inside an enclosed body of water like a fjord. The confining walls of the inlet acted like a funnel, directing the energy rather than letting it dissipate across an open ocean.
Researchers have also found that the viscosity of the sliding mass changes as it enters the water. The rock mass becomes saturated, which thickens the flow and reduces how quickly material enters the water. That rate of entry, the landslide flux, is one of the key factors determining how tall the resulting wave gets.
4PubMed Central. Landslides and tsunamis predicted by incompressible smoothed particle hydrodynamics (SPH) with application to the 1958 Lituya Bay event and idealized experimentThe 524-Meter Run-Up
The most frequently cited number about Lituya Bay, 524 meters, is a run-up height, not a wave height. That distinction matters. Run-up is the vertical elevation the water reaches on a slope, measured from sea level. The actual wave in the open water of the bay was far shorter than 524 meters. What happened is that the initial surge, propelled by the rockslide directly behind it, raced across Gilbert Inlet and struck the spur ridge on the opposite side. The wave ran up the steep, forested slope on that ridge, stripping every tree and scouring the earth down to bare bedrock up to an elevation of 524 meters. A sharp trim line, the boundary between destroyed forest and surviving forest, marks that height to this day.
5Geophysical Research Letters. Hybrid modeling of the mega‐tsunami runup in Lituya Bay after half a centuryNumerical simulations have reconstructed the timing in startling detail. The wave reached its peak amplitude of about 272 meters just eight seconds after the rockslide hit the water. Within 22 seconds it struck the base of the spur ridge across the inlet. The maximum run-up of roughly 524 meters was reached at about 39 seconds after the slide began. In under a minute, a wave had climbed more than half a kilometer up a mountainside.
6Natural Hazards and Earth System Sciences. The Lituya Bay landslide-generated mega-tsunami – numerical simulation and sensitivity analysisWhat Happened as the Wave Traveled Down the Bay
After the initial run-up at Gilbert Inlet, the wave did not simply vanish. It propagated southward down the length of Lituya Bay, losing height as it spread but still carrying devastating energy. Near Mudslide Creek on the south shoreline, the wave caused run-up close to 180 meters about 70 seconds after the slide. A reflected wave bounced off the southern shore and actually produced a second, taller crest heading back across the bay, reaching about 32 meters when it struck the east coast near Cenotaph Island roughly two minutes after the initial slide.
7Natural Hazards and Earth System Sciences. The Lituya Bay landslide-generated mega-tsunami – numerical simulation and sensitivity analysisAt Cenotaph Island, near the middle of the bay, the wave split into two branches flowing through the channels on either side of the island. Waves exceeding 25 meters hit the northern shoreline and pushed more than a kilometer inland near Fish Lake. Along the southern shore, waves of 40 to 50 meters struck the coast and penetrated roughly a kilometer inland on the flat terrain. By the time the wave reached La Chaussee Spit, the narrow gravel bar at the bay’s mouth, it was still more than 15 meters high. It washed over the spit and out into the open Pacific about five minutes and fifty seconds after the rockslide.
8Natural Hazards and Earth System Sciences. The Lituya Bay landslide-generated mega-tsunami – numerical simulation and sensitivity analysisThe rapid attenuation from 272 meters near the source to around 20 meters at the bay’s midpoint to 15 meters at the mouth illustrates something important about landslide-generated waves. They are terrifyingly powerful in confined spaces but lose amplitude quickly with distance, especially when the geometry of the basin allows the energy to spread laterally and dissipate through turbulence and bottom friction. If this same slide had occurred on an open coastline, the resulting wave would have been dramatically smaller.
Survivors on the Water
Three fishing boats were anchored in Lituya Bay that evening. Howard Ulrich and his seven-year-old son were aboard the Edrie near the bay’s entrance. They saw the wave coming, rode over it as it lifted them above the tree line, and survived. Bill and Vivian Swanson on the Badger were carried by the wave over the spit at the bay’s mouth and deposited in the ocean; their boat sank, but they were rescued. Orville and Mickey Wagner on the Sunmore were not as fortunate. Their boat was caught by the wave and neither they nor the vessel were ever found.
The Swansons’ account, in particular, has become central to the scientific record. Bill Swanson reported watching the wave strip the forest from the slopes as it swept down the bay, describing a wall of water that seemed to stretch from shore to shore. Eyewitness descriptions like these helped geologists calibrate their initial estimates of wave height before modern modeling tools existed.
Was It One Landslide or Two?
The standard account of the 1958 event describes a single massive rockslide as the sole wave source. But some researchers have argued that one rockslide alone does not fully explain the size of the wave beyond Gilbert Inlet. A closer look at eyewitness accounts and the trim lines along the bay revealed that a rockslide-only tsunami seemed “somewhat lacking in size” once it left the immediate inlet. Between 1926 and 1959, roughly 300 million cubic meters of sediment had accumulated in the deepest parts of Lituya Bay. The suggestion is that the earthquake and the initial rockslide may have also mobilized a large volume of this underwater sediment, essentially producing a second, submarine landslide that contributed additional wave energy as the tsunami moved down the bay.
9Journal of Earthquake and Tsunami. THE 1958 LITUYA BAY LANDSLIDE AND TSUNAMI — A TSUNAMI BALL APPROACHThis “two-source” hypothesis remains a subject of active research rather than settled consensus. Most numerical models can reproduce the 524-meter run-up in Gilbert Inlet using the subaerial rockslide alone, but some of the wave heights farther down the bay are harder to match without invoking additional energy input. The debate matters beyond academic curiosity: if submarine sediment mobilization can amplify a landslide tsunami significantly, then hazard assessments for fjords need to account for the condition of the seafloor, not just the stability of the visible slopes above the waterline.
How Scientists Reconstructed the Event
No instruments were recording inside Lituya Bay in 1958. Everything known about the wave comes from post-event geological surveys and, in the decades since, from increasingly sophisticated physical and numerical models. In the early 2000s, researchers built a scaled physical model of Gilbert Inlet and dropped granular slides into it at speeds matching the estimated rockslide velocity. Those experiments confirmed the two-step mechanism of impact and cavity collapse and reproduced the extreme run-up on the opposite slope.
10Natural Hazards and Earth System Sciences. The 1958 Lituya Bay tsunami – pre-event bathymetry reconstruction and 3D numerical modelling utilising the computational fluid dynamics software Flow-3DMore recent work has used computational fluid dynamics software to simulate the entire event in three dimensions, from the moment the rock detaches to the wave’s exit through the bay mouth. One research group reconstructed the pre-event bathymetry of the bay floor, since the tsunami itself rearranged enormous volumes of sediment, and used that as the starting condition for their simulation. Their results closely matched the observed trim lines and run-up heights around the bay, lending confidence to the physical understanding of the event.
Separately, a two-layer numerical model successfully reproduced the wave for a full 250 seconds of real-time travel, capturing multiple run-ups on both shorelines and six distinct wave crests and troughs in the channel. That kind of long-duration simulation had not been achieved before, and it confirmed that the reflected waves bouncing between the bay walls were an important part of the overall hazard, not just the initial surge.
11Advances in Water Resources. A two-layer model for landslide generated impulse wave: Simulation of the 1958 Lituya bay landslide impact wave from generation to long–duration transportLituya Bay’s History of Giant Waves
The 1958 event was not the first mega-tsunami in Lituya Bay. Geological evidence shows that similar events occurred in roughly 1853 and 1936, both caused by landslides triggered by earthquakes on the Fairweather Fault. The 1936 wave produced run-up estimated at around 150 meters. The bay’s geometry, a narrow fjord with steep, unstable walls and a major fault running through its head, makes it essentially a natural tsunami machine. Every large earthquake on the Fairweather Fault risks destabilizing the slopes above Gilbert Inlet or Crillon Inlet and launching another wave.
Local Tlingit oral histories describe the bay as a dangerous place, and early European explorers noted the barren shorelines where waves had stripped vegetation in the recent past. The trim lines from these older events are still visible alongside the 1958 line, creating a layered record of destruction on the hillsides that geologists have used to reconstruct the bay’s hazard history.
Why Fjords Are Particularly Dangerous
Lituya Bay’s combination of traits, steep unstable slopes, deep water at the base, narrow confining walls, and tectonic activity, is not unique. Similar geometry exists in fjords throughout Alaska, Norway, Greenland, Chile, and New Zealand. What makes fjords especially hazardous for landslide-generated waves is the combination of confined water and steep-sided entry points. A rockslide entering a wide, shallow bay would lose most of its energy to lateral spreading. The same slide entering a deep, narrow fjord transfers far more energy into the wave because the water has nowhere to go except along the channel.
This risk is increasingly relevant because of two converging trends. First, glacial retreat is exposing rock walls that were previously buttressed by ice, leaving them more prone to collapse. Second, cruise ship tourism in fjord regions has grown substantially. A 2025 study documented a 481-meter-high landslide tsunami in an Alaska fjord that is frequented by cruise ships, describing it as a “near-miss event” and warning that climate change is making similar events more likely.
12PubMed. A 481-meter-high landslide-tsunami in a cruise ship-frequented Alaska fjordThe fact that a modern event came close to matching the scale of the 1958 Lituya Bay wave, in a location visited by large passenger ships, has sharpened the urgency around monitoring and early warning systems for fjord environments.
Toward Early Warning in Fjords
Traditional tsunami warning systems are designed for earthquake-generated tsunamis in the open ocean, where seismic detection gives minutes to hours of lead time before waves reach distant coastlines. Landslide-generated tsunamis in fjords offer almost no warning by that standard. At Lituya Bay, the wave reached its maximum run-up 39 seconds after the slide began. A community or vessel inside the bay would have had virtually no time to react, and no existing warning system would have helped.
Researchers have been working on approaches better suited to these confined environments. One promising direction uses seismic data to identify landslides in near real time by analyzing whether the source mechanism looks like an earthquake or a gravitational collapse. A coupled simulation study of the 2017 Nuugaatsiaq landslide tsunami in Greenland, which killed four people in a small coastal settlement, demonstrated that seismic monitoring could detect the landslide and estimate tsunami propagation quickly enough to provide some warning to communities farther down the fjord. The researchers noted that turning this concept into an operational alert system will require testing in real environments, but the fundamental physics appears workable.
13Natural Hazards and Earth System Sciences. Coupled simulation of landslide, tsunami, and ground displacement for the 2017 Nuugaatsiaq event in GreenlandFor locations very close to a slide, like the fishing boats in Lituya Bay in 1958, no alert system can overcome the physics of a wave arriving in seconds. Warning systems for fjords are more about protecting settlements and vessels farther down the channel, where even a few minutes of advance notice could save lives. Given that a single event in 1958 killed two people in an essentially uninhabited bay, the stakes in a populated or tourist-heavy fjord are obvious.
The Landscape Left Behind
Lituya Bay today still bears visible scars from 1958. The trim line on the spur ridge opposite Gilbert Inlet, where the forest was shaved to bare rock up to 524 meters, remains the event’s most dramatic physical marker. Below the trim line, younger-growth forest has filled in over the past six decades, creating a stark visual contrast with the older-growth trees above. The shoreline topography was rearranged: sediment was stripped in some places and deposited in others, the bay floor was reshaped, and the delta at the head of Gilbert Inlet was partially rebuilt by glacial sediment in the years that followed.
The bay remains largely wild and uninhabited. It is technically accessible by boat, though the entrance through La Chaussee Spit is notoriously hazardous due to tidal currents and breaking waves over the shallow bar. A small number of kayakers and fishing vessels enter the bay each summer. There is no permanent settlement, no road access, and no infrastructure. In a sense, the bay’s remoteness is what kept the 1958 event from being a major humanitarian disaster rather than a geological curiosity. Had the same wave occurred in a populated Norwegian fjord or a cruise-ship anchorage, the outcome would have been catastrophic on a completely different scale.