Not every submarine earthquake generates a tsunami — and some generate catastrophic ones from thousands of kilometres away. The difference lies in fault geometry, rupture depth, and vertical seafloor displacement. Learn the physics from fault slip to shoreline run-up.
🌊 OPEN LIVE EARTHQUAKE MAPThe link between earthquakes and tsunamis seems intuitive — the ocean shakes, a wave forms. But the actual mechanism is far more selective. Every day, dozens of submarine earthquakes occur worldwide, yet tsunamis are rare. The generation of a destructive tsunami requires a specific combination of fault type, rupture geometry, seafloor displacement, and ocean depth. Understanding this selectivity is not academic: it is the basis of every tsunami warning system on Earth.
A submarine earthquake triggers a tsunami only when it fulfils three simultaneous conditions. First, the fault must have a significant vertical (dip-slip) component — thrust and normal faults displace the seafloor up or down, directly pushing a water column. Strike-slip faults (horizontal motion) rarely generate large tsunamis. Second, the event must be shallow — focal depths shallower than 70 km produce maximum seafloor deformation; deep events dissipate energy through the mantle and generate little surface displacement. Third, the rupture area must be large enough — the displaced water column must have horizontal extent comparable to ocean depth to avoid immediate radiation cancellation. A M8.0+ rupture may displace tens of thousands of square kilometres of seafloor simultaneously.
| PARAMETER | TSUNAMIGENIC THRESHOLD | WHY IT MATTERS |
|---|---|---|
| Magnitude | ≥ M7.5 | Minimum rupture area for significant water column displacement |
| Focal depth | < 70 km | Deeper events lose displacement energy before reaching seafloor |
| Fault mechanism | Dip-slip (thrust/normal) | Vertical component required to physically displace water column |
| Ocean depth over rupture | Any — deep preferred | Deep water allows long-wavelength wave to form intact |
| Rupture velocity | Slow ("tsunami earthquakes") | Anomalously slow ruptures generate larger waves per magnitude unit |
| Co-seismic landslides | Present in confined bays | Can amplify local wave height dramatically beyond seismic prediction |
A class of events called "tsunami earthquakes" poses particular danger: they generate tsunamis far larger than their seismic magnitude would predict. These events rupture slowly (sub-shear velocity) along shallow, sediment-rich portions of the subduction zone. The 1992 Nicaragua earthquake (M7.6) generated a 10 m tsunami that killed 170 people — a run-up disproportionate to its felt intensity onshore. Because tsunami earthquakes cause weak shaking, coastal populations often do not perceive the danger before the wave arrives. This is why NOAA's warning criterion uses ocean bottom pressure sensors (DART buoys), not just seismic magnitude alone.
The M9.0 Tōhoku earthquake ruptured ~500 km of the Japan Trench over approximately 3 minutes. Vertical seafloor displacement reached up to 7 m over an area of ~50,000 km². The resulting wave crossed the Pacific in under 24 hours, reaching California (~2 m) and Chile (~1.5 m). Run-up heights in northern Honshu exceeded 40 m. The event exceeded the design basis of the Fukushima Daiichi seawall by over 7 m — a consequence of underestimating the maximum credible rupture extent of the Japan Trench.
Key lesson: probabilistic seismic hazard models for tsunamigenic zones must account for the full statistical distribution of rupture scenarios, including rare, magnitude-saturating events that historical catalogs may not contain.
The global tsunami warning system combines two data streams. Seismic networks detect large earthquakes within minutes and compute magnitude, depth, and mechanism — enough for an initial advisory. But seismic data alone cannot confirm whether a tsunami has formed. That confirmation comes from DART buoys (Deep-ocean Assessment and Reporting of Tsunamis) — bottom pressure sensors moored in deep water that detect the sub-centimetre pressure change of a passing tsunami wave. DART data allows centres like the Pacific Tsunami Warning Center (PTWC) to confirm, upgrade, or cancel warnings within 20–30 minutes of generation.
For earthquakes occurring directly offshore a populated coast — Cascadia, Chile, Japan — warning time may be as short as 5–15 minutes. No warning system can overcome this latency for the nearest coastlines. This is why coastal communities in subduction zones practise vertical evacuation and have established the rule: strong shaking at the coast is itself the warning. Do not wait for an official alert. Move immediately to high ground.
Far-field tsunamis (crossing ocean basins) allow hours of warning time, making DART-based systems highly effective at protecting distant coasts — as demonstrated by the Pacific-wide response to Tōhoku in 2011.
1. Moment magnitude (Mw). Mw ≥ 7.5 triggers automatic evaluation; Mw ≥ 7.9 triggers automatic watch for surrounding ocean basins.
2. Focal depth. Events shallower than 35 km receive maximum concern; 35–70 km receive elevated concern. Events deeper than 100 km are deprioritised.
3. Focal mechanism (beach ball). Rapid centroid moment tensor solutions classify the fault as thrust, normal, or strike-slip within 10–15 minutes of origin time.
4. DART buoy confirmation. Nearest DART stations are polled for anomalous bottom pressure changes consistent with a propagating wave. Positive detection triggers a warning upgrade to "Warning" level (highest tier).
5. Historical analogue matching. Databases of past tsunamigenic events in the same tectonic segment inform probabilistic run-up estimates for threatened coastlines, driving evacuation zone activation.
Roughly 80% of all tsunamis originate in the Pacific — a consequence of the Ring of Fire's arc of active subduction zones. The highest-risk segments are the Cascadia Subduction Zone (last megathrust rupture: 1700 CE, estimated M9.0), the Chile Trench (M9.5 in 1960, the largest recorded earthquake), the Japan Trench (M9.0 in 2011), and the Hikurangi margin off New Zealand. The Indian Ocean, historically considered lower-risk, was catastrophically re-classified after 2004. Mediterranean tsunamis, while less frequent, are well-documented — the Hellenic Arc offshore Greece and the Calabrian Arc off southern Italy are the primary sources, capable of threatening coastlines within 5–20 minutes of rupture.
| ZONE | LAST MAJOR EVENT | EST. MAX Mw | NEAR-FIELD WARNING TIME |
|---|---|---|---|
| Cascadia (NW USA / Canada) | 1700 CE (M~9.0) | 9.2 | 15–30 min |
| Chile Trench | 1960 (M9.5) | 9.5+ | 10–25 min |
| Japan Trench | 2011 (M9.0) | 9.1 | 5–30 min |
| Sumatra–Andaman | 2004 (M9.1) | 9.3 | 15–30 min |
| Hellenic Arc (Greece) | 1956 (M7.7) | 8.5 | 5–20 min |
| Alaska–Aleutians | 1964 (M9.2) | 9.3 | 20–45 min |
While thrust faults are the dominant source, strike-slip earthquakes can and do cause tsunamis through secondary mechanisms. The 2018 Palu, Sulawesi event (M7.5, predominantly strike-slip) triggered submarine landslides that generated run-up heights exceeding 8 m locally — killing over 2,000 people in minutes. In confined geometries (bays, fjords, submarine canyons), even limited horizontal seafloor displacement can mobilise enough sediment to produce locally extreme waves. Warning algorithms increasingly account for co-seismic landslide potential in these environments.