Skip to main content
U.S. flag

An official website of the United States government

Why do some earthquakes trigger large tsunamis, and others don’t? Learn how earthquakes produce tsunamis, how scientists predict tsunami size and arrival times, and the differences in two recent small tsunamis.

Map of oceans and continents with 3-D effect, with earthquake epicenter and cities and countries labeled.
Map shows location of the September 7, 2017 Mexico earthquake epicenter (star). km, kilometers; mi, miles, with respect to North and South America.(Credit: Basemap is the intellectual property of Esri and is used herein under license. Copyright © 2017 Esri and its licensors. All rights reserved.)

One of the largest earthquakes ever recorded along Mexico’s south coast struck just before midnight, local time, on September 7, 2017. Centered about 100 kilometers (60 miles) off the coast of the state of Chiapas, the magnitude 8.2 quake killed nearly 100 people and caused widespread damage (see “M 8.2 - 101km SSW of Tres Picos, Mexico”). The temblor also triggered a tsunami that reached the Pacific shores of Mexico and Guatemala and was detected as far away as Hawaiʻi and New Zealand.

Map shows land and ocean with 3-D features of relief and depth and labeled locations such as cities and earthquake epicenter.
Map of the south coast of Mexico and part of Guatemala. Star marks the epicenter of the magnitude 8.2 earthquake that struck at 11:49p.m.(Credit: Basemap is the intellectual property of Esri and is used herein under license. Copyright © 2017 Esri and its licensors. All rights reserved.)

Even though the tsunami crossed the Pacific Ocean, it “was smaller than we would expect from the magnitude of the earthquake,” said USGS geophysicist Eric Geist. Tide gauges showed wave heights of 1.2 meters (4 feet) at Salina Cruz, and 1.7 meters (5.5 feet) at Puerto Chiapas, with most of the damage confined to a small boat harbor there.

Nearly five months later, something similar happened in Alaska. Just after midnight on January 23, 2018, a magnitude 7.9 earthquake occurred beneath the Pacific Ocean about 280 kilometers (175 miles) southeast of Kodiak Island (see “Magnitude 7.9 Earthquake Gulf of Alaska”). The National Oceanic and Atmospheric Administration (NOAA) issued a tsunami warning for Alaska and Canada, and a tsunami watch for Washington, Oregon, and California. Sirens wailed in Alaskan coastal communities, cell phones sounded emergency alerts, and numerous residents fled to higher ground. A few hours later, coastal tide gauges and deep-ocean tsunami detectors had recorded waves less than a foot high and the alerts were canceled. Neither the shaking nor the tsunami caused serious damage, and what might have been a disaster had become a drill.

Both tsunamis could have been much larger. Why weren’t they?

How earthquakes produce tsunamis

“To generate a significant tsunami,” said Geist, “an earthquake needs to meet three criteria: First, the part of the fault that breaks needs to be at least partly underneath the ocean; typically, the deeper the water, the bigger the tsunami waves.” Second, the earthquake must permanently move large areas of the seafloor up and down—the bigger the movement and area, the bigger the tsunami. Finally, longer and wider fault breaks create larger magnitude earthquakes, which trigger more destructive tsunamis.

“Most tsunamis are generated at subduction zones, where one tectonic plate dives beneath another,” said Geist. “And the most devastating ones, like the 2004 Indian Ocean and 2011 Japan events, typically result from large-magnitude earthquakes on the interplate thrust”—the fault that separates the two plates. (View a diagram of a subduction zone.)

Recent large tsunamis

Line drawing of cross-section of oceanic crust showing the location and kind of earthquake that may or may not cause a tsunami.
Generalized diagram of a subduction zone, showing the approximate position of tsunami-generating earthquakes on the interplate thrust fault.(Credit: Eric Geist, USGS.)

It’s easy to imagine a large earthquake on an interplate thrust fault, which might be hundreds of miles long and wide.

That’s what happened in 2004 off the coast of Sumatra, Indonesia. A magnitude 9.1 earthquake broke an area of the interplate thrust fault that was 1,300 km (800 miles) long by 150 km (93 miles) wide, triggering a tsunami that killed about 250,000 people around the Indian Ocean (see “M 9.1 - off the west coast of northern Sumatra”). The devastating 2011 Japanese tsunami also started with a magnitude 9.1 earthquake on the thrust fault between a subducting and an overriding plate (see “M 9.1 - near the east coast of Honshu, Japan”). Both earthquakes pushed large areas of the seafloor higher and lower, the way opening a horizontal fist simultaneously raises your fingertips and lowers your knuckles. These motions moved very large quantities of seawater, creating a series of catastrophic tsunami waves.

Line drawing of cross-section of oceanic crust showing the location and kind of earthquake that may or may not cause a tsunami.
The 2009 Samoa earthquake and tsunami resulted primarily from slip on a normal fault in the “outer rise,” a bulge in the seafloor caused by bending of the oceanic plate.(Credit: Eric Geist, USGS.)

The story was slightly different for a tsunami that hammered Samoa, American Samoa, and Tonga in 2009 (see “Preliminary Analysis of the September 29, 2009 Samoa Tsunami, Southwest Pacific Ocean”). It was generated by a magnitude 8.1 earthquake seaward of the trench—the place on the seafloor where the two plates meet (see diagram). Because the quake was close to the surface, there was a large movement of the seabed and the water above it.

This earthquake was on a normal fault, where rocks on one side of the fracture drop down. (Watch a normal fault animation.) It triggered a pair of large aftershocks on the interplate thrust, which are thought to have contributed to the tsunami ( According to Geist, however, “most of the tsunami energy came from the outer-rise normal fault earthquake.” The Samoa earthquake was one of a few normal-fault earthquakes that have caused deadly tsunamis (see “Samoa Disaster Highlights Danger of Tsunamis Generated from Outer-Rise Earthquakes”).

Let’s look at the factors that made the recent earthquakes in Mexico and Alaska different from these earlier events.

Illustration of ocean and land south of central America with outlines of tectonic plates, symbols show locations of earthquakes.
Map showing the Cocos Plate in relation to nearby tectonic plates. Yellow star locates the 2017 Chiapas earthquake. Red triangles are volcanoes; black dots and circles are epicenters of previous earthquakes.Modified from “This Dynamic Planet” (USGS Geologic Investigations Map I-2800) which contains a detailed key.

Chiapas—Deep earthquake, shallow water

The Chiapas earthquake occurred in the Middle America subduction zone, where the Cocos oceanic plate dives beneath Central America.

Line drawing of cross-section of oceanic crust showing the location and kind of earthquake that may or may not cause a tsunami.
The 2017 Chiapas earthquake and tsunami occurred on a normal fault inside the Cocos oceanic plate.(Credit: Eric Geist, USGS.)

Like the 2009 Samoa temblor, the Chiapas quake occurred not on the interplate thrust but on a normal fault in the oceanic plate. It was slightly larger than the Samoa earthquake; in fact, at magnitude 8.2, the Chiapas event was one of the largest normal fault earthquakes ever recorded. But it was landward of the trench and much deeper than the Samoa earthquake. It produced less movement of the seafloor and a much smaller tsunami. In both Chiapas and Samoa, the normal faults most likely developed from tension (pulling apart) stresses caused by gravity bending and dragging the oceanic plate into the Earth.

In contrast to the Samoa event, the Chiapas earthquake broke beneath a region of shallow water of the continental shelf. Tsunami waves grow higher and steeper as they move from deep water into shallower water. The deeper the water where they start, the more their heights increase as the waves approach the shore. That’s another reason why the Samoa earthquake, centered under much deeper water, produced a larger tsunami than the Chiapas quake.

Kodiak—Horizontal fault motion

Illustration of the ocean south of Alaska with symbols to show locations of earthquakes and volcanoes.
Boundary between the Pacific and North American tectonic plates. Yellow star locates the 2018 Kodiak earthquake. Red triangles are volcanoes; black dots and circles are epicenters of previous earthquakes.Modified from “This Dynamic Planet” (USGS Geologic Investigations Map I-2800), which contains a detailed key.

The Kodiak earthquake occurred in deep water seaward of the Aleutian subduction zone (see map and diagram, below), but it was on a third type of fault: a strike-slip fault. The movement on such faults is mainly horizontal and does not produce much vertical movement of the seabed. The Great San Francisco Earthquake of 1906, for example, occurred on the strike-slip San Andreas fault. Although it was a very large earthquake, it produced a tsunami that measured just 10 centimeters (6 inches) at a San Francisco Bay tide gauge station (see “Analysis of the tsunami generated by the Mw 7.8 1906 San Francisco earthquake,” by Geist and Zoback, Geology (1999) vol. 27, no. 1, p. 15-18).

Immediately after the January 2018 earthquake off Kodiak, it was not clear what kind of fault had slipped. The earthquake’s high magnitude (initially estimated at 8.0 and later revised to 7.9) and its location near coastal communities prompted NOAA’s National Tsunami Warning Center to issue alerts for Alaska, Canada, and the U.S. west coast. As scientists analyzed the seismograph data, they quickly realized that the quake had occurred on a strike-slip fault—their first hint that the tsunami danger might not be as serious as feared. A few hours later, when measurements from coastal tide gauges and deep-ocean tsunami detectors had reported tsunami wave heights of less than 1 foot, the alerts were cancelled.

Line drawing of cross-section of oceanic crust showing the location and kind of earthquake that may or may not cause a tsunami.
The 2018 Kodiak earthquake and tsunami occurred on a strike-slip fault in the Pacific oceanic plate. T is side of fault moving toward viewer; A, away from viewer.(Credit: Eric Geist, USGS.)

Details about the earthquakes off Alaska and Mexico—including fault motions, depth of the earthquakes beneath the seafloor, depth of the water above the earthquakes—made their tsunamis much smaller than one would expect from their magnitudes. That was a lucky thing for communities near and far from the epicenters.

Predicting tsunami size and arrival times

Scientists use the close relationship between earthquakes and tsunamis to create computer models simulating the destructive waves. Geist used seafloor maps plus detailed USGS information about the Chiapas earthquake (see “Preliminary Finite Fault Results for the Sep 08, 2017 Mw 8.2 15.0286,-93.8070 Earthquake (Version 1)”) to create tsunami simulations: “Preliminary simulation of the tsunami from the September 8, 2017 M=8.1 intermediate-depth earthquake offshore of Chiapas, Mexico.”

The animations show that a tsunami wave reflected off the edge of the continental shelf and returned to shore. Reflected and “coastal-trapped” waves continued striking the coast many hours after the first wave (see “Life of a Tsunami).”

Cartoon shows a transparent ocean surface and the ocean floor as well as land features, and what happens to a tsunami wave.
The largest wave (center) came directly from the seafloor movement caused by the earthquake. Left of that is a smaller wave that bounced off the shore, and then off the edge of the continental shelf, and back toward the coast. Screenshot from a tsunami simulation. Wave heights enlarged for better visibility.(Credit: Eric Geist, USGS.)

These simulations show how the earthquake’s characteristics (the amount and direction of movement on the fault, for example) affected the resulting tsunami and its impact on the nearby coast. Similar computer models can predict how fast tsunami waves will cross the ocean and reach distant shores. Immediately after a large undersea earthquake, NOAA scientists use the magnitude and epicenter to simulate the likely tsunami, then fine-tune the model with actual wave measurements from deep-ocean tsunami detectors. The size and arrival time of distant tsunamis can now be projected with life-saving accuracy; NOAA’s National Weather Service uses those predictions for tsunami alerts. Unfortunately, local tsunamis, triggered by a nearby earthquake, often arrive too fast for calculations and alarms. The strong shaking from an earthquake itself is the best warning.

“If you feel strong shaking at the coast,” said Geist, “head for higher ground.”

USGS scientists study many aspects of tsunamis—from field studies of recent and ancient tsunamis to computer modeling and forecasting—all of which can help us better prepare for these rare but destructive events.

Explore the links below to learn more:

EERI Preliminary Notes on Tsunami Damage and Response: Tsunami Generated by MW8.1 Chiapas, Mexico, Earthquake on September 7, 2017

January 23, 2018 M7.9 Gulf of Alaska Earthquake and Tsunami

International Team Studies Tsunami Deposits in Japan to Improve Understanding and Mitigation of Tsunami Hazards

SAFRR—Science Application for Risk Reduction: Tsunamis

Some Coastal Communities May Not Have Time for Tsunami Evacuation

Tsunami and Earthquake Research at the USGS

Probabilistic Forecasting of Earthquakes, Tsunamis, and Earthquake Effects in the Coastal Zone

Get Our News

These items are in the RSS feed format (Really Simple Syndication) based on categories such as topics, locations, and more. You can install and RSS reader browser extension, software, or use a third-party service to receive immediate news updates depending on the feed that you have added. If you click the feed links below, they may look strange because they are simply XML code. An RSS reader can easily read this code and push out a notification to you when something new is posted to our site.