December 26, 2004

2004 Indian Ocean earthquake and tsunami

Sumatra–Andaman earthquake generates the Indian Ocean tsunami

On Simeulue, stories handed down from a 1907 tsunami taught residents to run uphill after strong shaking. Seven people died there, among a population of nearly 80,000.

00:58:53 UTC, 26 December 2004

A great megathrust earthquake began off northern Sumatra at 00:58:53 UTC on 26 December 2004 — 07:58:53 in western Indonesia — at a depth of about 30 kilometres. USGS gives the official magnitude as Mw 9.1, while ultra-long-period analyses have produced estimates around Mw 9.3.

The rupture came from thrust faulting where the India Plate descends beneath the Burma microplate along the Sunda subduction zone. Vertical displacement extended northward for roughly 1,200 kilometres or more, over a duration the reviewed USGS event page puts at more than 500 seconds.

Twenty minutes to shore

The displaced seafloor generated waves travelling both toward nearby coasts and across the Indian Ocean. Parts of northern Sumatra were struck within about 20 minutes, and no official Indian Ocean warning system existed to alert the threatened coasts.

Thailand, Sri Lanka, India, the Maldives and East Africa were hit over the following hours. Maximum documented runup reached 51 metres in Aceh Province; more than 5,000 kilometres away in Somalia, surveyed runup approached 10 metres.

227,899

NOAA's database lists 227,899 people killed or missing and presumed dead across 15 countries, 167,540 of them in Indonesia. About 1.7 million people were displaced, and material losses were estimated at the time at approximately US$10 billion.

A tsunami is a sequence of waves, and later ones can be larger than the first. On Simeulue, oral accounts of a 1907 tsunami had taught residents to seek high ground after strong shaking; seven people died there among a population of nearly 80,000.

Sources

Researched 23 Aug 2026 9 sources 2 audit passes

How this was checked

Researched from the web into a fact sheet, rewritten from that sheet, then audited against it by a different model. How the pipeline works →

What the sources leave uncertain

USGS’s current preferred magnitude is Mw 9.1. Analyses of ultra-long-period normal modes have produced estimates around Mw 9.3, so the listed 9.1–9.3 range reflects different methods rather than a poorly known date.
Rupture-duration estimates depend on what part of the rupture or moment release is measured. USGS’s reviewed event page gives more than 500 seconds, another USGS model gives about eight minutes for the rupture front, and recent scholarly reviews round the total duration to approximately ten minutes.
USGS displays community reports reaching Modified Mercalli intensity IX, while its instrumental ShakeMap estimates intensity VIII. Mercalli intensity describes earthquake shaking, not tsunami runup.
Fatality records combine confirmed deaths with missing people presumed dead. NOAA also notes that earthquake and tsunami deaths cannot be separated precisely because the first waves arrived so quickly.