Tsunamis and atolls: why warnings must be locally meaningful

You live on a coral atoll, knowing well most of its square centimetres. It takes seven minutes to wander across your island's broadest point at a leisurely pace, which includes the apex of 1.2 metres above sea level. You can stretch out the high-tide circumnavigation to ninety minutes if you really try.

Your mobile phone pings with a warning that a trans-oceanic tsunami is projected to hit in 50 minutes at high tide with a local wave height of up to 4.5 metres. What do you do?

This situation is not hypothetical. On 26 December 2004, Maldives was inundated by the Indian Ocean tsunami about three hours after the wave-generating earthquake. Over several minutes, the ocean rose by up to 3-4 metres, sweeping over all the islands. Over 80 people died, some being dragged into the sea and never recovered, and buildings were destroyed.

After submarine earthquakes on 29 September 2009, inhabitants of Tokelau in the Pacific had enough time to take precautions. The tsunami was not that severe for Tokelau, so reaching the limited high ground helped them to survive the waves.

Both 11 March 2011 and 30 July 2025 witnessed trans-Pacific tsunamis. On each day, few major impacts were reported from the numerous atolls dotted across thousands of kilometres. It remains unclear how bad it could have been.

The usual tsunami advice to move inland and to move to high ground might not be possible for many atolls. Generic, international, trans-oceanic warnings need to become locally meaningful. What advice should be given for surviving tsunamis which strike an atoll?

How do tsunamis affect atolls?

For most atolls, the seabed on the side of the ocean (rather than on the side of the internal lagoon) drops steeply, often vertically, offshore. Rather than tsunami waves slowing down and piling up as they approach the shoreline - which increases the breaking height - single waves can strike an atoll's reef wall straight on, washing across the reef platform and across each island.

Given the low elevation of atoll islands, the small tidal range - of 0.6 metres or less in the Central and South Pacific, and around one meter in the central Indian Ocean - contributes to people's survival. It still makes a difference at which point in the tidal cycle the tsunami strikes, with King Tides being the most perilous time.

When tsunami waves strike, their energy might dissipate, perhaps leaving the lagoon and islands on the other side relatively sheltered. As such, a tsunami's full force is perhaps not felt across an entire atoll. Yet a non-breaking tsunami of under one metre can flood islands, bring down structures, and knock people off their feet. The 2004 tsunami devastated the Maldives, and people there had limited preparedness or readiness.

Tsunami actions on atolls

Climbing a tree gambles that you can reach higher than the tsunami - and that the tree stays up. Same with racing to the top of a building. On 11 March 2011 in (non-atoll) Japan, people perished on roofs because the tsunami flooded higher.

Ambling over to the side of the island - or even across the entire atoll - opposite the expected tsunami arrival direction might help. Just hope that the tsunami is not powerful enough to rip across! Even if it does, the opposite side will still typically be safer. On some atolls, the small amount of elevation available could suffice to avoid the tsunami's worst, so gathering there is a roll-the-dice option.

If a central lagoon is large enough, then being on boats there might help to ride out the tsunami. Alternatively, take boats beyond the reef, in any direction, as far to the deep sea as possible. Tsunamis sport little destructive power in the open ocean, so the steep drop-off from most atolls could keep people safe within sight of land.

Immediate practical problems relate to having enough serviceable boats for the open ocean for everyone - with adequate supplies for a few hours, which might include fuel if wind or paddling are not the travel modes. These logistics would never be easy, including if many boats were out fishing when the tsunami warning arrives.

Moreover, the open ocean is not always safe. On 30 July 2025, the Central Pacific featured a Category 3 tropical cycloneand a tropical storm at the same time as the tsunami.

Finally, it is not straightforward to know when it would be safe to return to shore after the main tsunami has passed. Near-shore currents and waves can be turbulent for a day or two afterwards, while aftershocks can produce further tsunamis.Even the original tsunami typically consists of multiple successive waves, not one single wave.

The most effective actions might change with each tsunami, depending on the direction from which the waves arrive, wave sizes and strengths, timings within the daily-to-decadal tidal cycles, and any coast and land changes such as harbours and coral reef or mangrove health. Plans would need to be updated and tested regularly. Then, a bespoke tsunami warning would be needed for almost every atoll - and ideally every single island of an atoll.

Start preparing now

Atolls around the world have been inhabited for millennia. Despite the dearth of narratives about large tsunamis in these places - perhaps those most affected were wiped out or perhaps people just moved on or rebuilt afterwards - evidence of and models for major waves exist. Also, parallels emerge with storm surges and waves from tropical cyclones, plus potentially some King Tides.

Local anecdotes relate how many islanders know about tsunamis and other waves, even if contemporary science has not fully recognised or acknowledged this wisdom. For example, in January 1899, witness reports described large waves washing across Pacific islands in the absence of nearby storms. They may have been tsunami-related from a distant earthquake, volcanic eruption, meteorite strike, or landslide.

As another example, field observations from Yemga or Egom, a small atoll just southwest of Muyuw (Woodlark Island), Papua New Guinea, suggest that people have built their houses in places which are highly protected from waves - including tsunamis - while ensuring that mangroves in other parts of the island absorb wave energy. For Muyuw, people know that post-tsunami, they should head for the sago swamp to collect stranded fish and smoke them before they rot. A tsunami, if not too big, brings them food - and they know exactly how to use the opportunity.

For a tsunami (or other wave) warning system for atolls to be viable in the long term, readiness needs to start now, so that people know what to do and they are willing to do it - whether they have twelve minutes or twelve hours and whether they receive a warning or observe signs in the ocean. Especially since so much about warnings is premised on working equipment, such as charged phones with coverage or a radio with batteries or reliable electricity.

The challenge remains in knowing what do to - or, at minimum, offering a set of options. Even if the next large tsunami does not strike an atoll for several generations, people will at least be confident in how to act. And it could hit today.

Ilan Kelman is Professor of Disasters and Health at University College London, England and a Professor II at the UiT The Arctic University of Norway. His overall research interest is linking disasters and health, integrating climate change into both. He can be reached on his website, Instagram/Threads/X: @ILANKELMAN, Bluesky: ilankelman.bsky.social.

Frederick H. Damon is Emeritus Professor of Anthropology at the University of Virginia. He has conducted research on and written about the Kula Ring in Papua New Guinea since 1973 and now has extensive ethnographic and literature experience in the PRC. He writings deal with exchange theory, ritual and environmental structures across the Indo-Pacific and calendrical/astronomical matters.

Dirk H. R. Spennemann (Charles Sturt University, Albury, Australia) researches and teaches Cultural Heritage Management and Environmental History with an emphasis on the Indo-Pacific Region. His research foci are heritage theory, especially heritage futures and emerging forms of heritage(s), as well as the physical manifestation of human responses to the Indo-Pacific environment during the nineteenth and early twentieth centuries. He can be reached at [email protected]

Julius Riese is an independent researcher based in Berlin, Germany. In his transdisciplinary work, he attempts to bridge diverse temporal and spatial scales at the interface of human cultural evolution and the natural environment. He can be reached at [email protected]