The lighthouse at Indira Point is twenty-three metres tall, stands on the southernmost tip of Great Nicobar, and after the earthquake of 26 December 2004 it was leaning by about a millimetre. That is worth pausing on. The ground under it had dropped by something close to three metres. The tower tilted by a thousandth of a metre, because the drop was not a hinge — it was the whole block going down, spread across fifteen to twenty-five kilometres of island, so the angle changed by almost nothing even though the height changed by a great deal.
That is the first thing to understand about these islands: they move as blocks, not as shapes. And the second is that they do not all move the same way.
One rupture, two directions
The Sumatra–Andaman earthquake ruptured about 1,600 kilometres of the Sunda megathrust, from northern Sumatra to Preparis, and the Andaman segment was roughly two-thirds of that length. Where a megathrust slips, the overriding plate is usually pushed up on one side of a hinge and dropped on the other, and the hinge — the pivot line — ran through these islands.
North of it, the Andamans rose. At the northern tip of North Andaman the uplift reached about three metres; at Diglipur and at Ariel Jetty it was around one metre. South of it, the Nicobars sank: about 1.2 metres along Car Nicobar, and close to three metres at the southern end of Great Nicobar.
Two consequences follow directly. The first is that "sea level" stopped being a single number for the territory: on one island the sea is now a metre further down the beach than it was, and on another it is a metre further up. The second is that the change was not only vertical — the rupture also shifted the islands horizontally, and tilted blocks that had been level.
The terraces that say this has happened before
On the western margin of Interview Island, in the middle Andamans, there are at least two older raised terraces above the one the 2004 earthquake produced. They are the remains of earlier coseismic uplifts — the same mechanism, earlier instalments. Similar paired uplift-and-subsidence patterns are recorded for the 1881 Car Nicobar earthquake, which also raised the west coast and dropped the east.
This is the part that changes how the islands should be read. A raised coral terrace is a staircase of earthquakes. Each step is a century or a few centuries of strain released in under a minute, and the flight of steps is the record. An island here is not a shape that happens to sit at a height. It is a level that keeps being reset, and the record of the resets is written in the rock as a series of ledges that no longer have sea in them.
The arc does more than move vertically. After the earthquake a mud volcano in Middle Andaman became active, throwing out grey mud and colourless gas and building a roughly circular mound about seventy metres across; it had erupted before, in 1983, 1996 and 2003. Deep cracks opened in the ground nearby, wide enough to split tree trunks and ruin stretches of road. What a subduction zone looks like when it adjusts is a set of unrelated-looking events: one end up, one end down, mud in the middle, pavement broken.
What a metre does to a coast
In the north, a metre of uplift is not a disaster. It gives you land: exposed reef flat, new beach, dry ground where there was shallow water. It also kills the reef that was lifted, because coral needs to be within a narrow band of the tide — too high and it dries, too deep and it starves for light. A reef lifted clear of the tidal range is a dead reef with a very good view.
In the south, a metre of subsidence is worse. On Car Nicobar the drop put rice paddies and village ground below the reach of the tide, and salt water in the soil does not leave on the next tide; it stays until it is flushed, which on an island with a two-metre water table means it stays a while. Jetties were built to a level that no longer exists. Roads that were a metre above the highest tide are now washed at springs.
None of this required a change in sea level. It required one minute.
The map is a claim with a date on it
Charts and topographical sheets for the archipelago were drawn, like all maps, to a datum — a reference level agreed at the time of survey. After 2004 those datums were wrong, in places by a metre or more, and the error was not uniform across the territory, so it could not be corrected with one offset. The islands had to be re-surveyed, which is expensive, slow, and much less exciting than a satellite image.
This is the ordinary condition of a tectonic coast, and it is worth stating plainly because the climate conversation has trained everyone to think of coastal change as a slow slope measured in millimetres a year. On a subduction arc, the same coast can move a metre or three in a minute, in either direction, and then sit still for two centuries. Anyone planning a jetty here is planning against both curves at once: the slow one and the step.
Why the step is the harder one
Gradual change can be adapted to incrementally — you raise a road, you move a village inland, you rebuild the embankment a little higher each generation. A step cannot. It arrives before the plan, it moves different places in opposite directions, and it invalidates the survey that the plan was drawn on.
The islands will keep doing this. The arc is a subduction zone converging at about five centimetres a year; the slip of 2004 released something like four centuries of that motion in a single rupture. Four hundred years of accumulated push, spent in under a minute, and the count starts again from wherever the ground now is.
So the useful way to hold these islands in mind is not as dots on a chart but as a set of levels under negotiation. The negotiation is slow, then instant, then slow again — and the people who live there have to keep rebuilding to a datum that the ground itself keeps revising.
Sources and checks
- U.S. Geological Survey Briggs et al., Uplift and subsidence associated with the great Aceh–Andaman earthquake of 2004, JGR Solid Earth, 2006
- Indian Institute of Technology Kanpur Rai et al., Landscape changes in the Andaman and Nicobar Islands after the December 2004 earthquake and tsunami
- Geological Survey of India Seismotectonics of the Andaman–Nicobar arc and the Andaman Sea spreading centre
- Annual Review of Earth and Planetary Sciences Satake and Atwater, Long-term perspectives on giant earthquakes and tsunamis at subduction zones, 2007
- Journal of Asian Earth Sciences Curray, Tectonics and history of the Andaman Sea region, 2005
- Survey of India / National Hydrographic Office Topographical sheets and nautical charts for the Andaman and Nicobar Islands, pre- and post-2004 editions
Last checked 2 November 2026 Yellowsea Huang Corrections