The Landslide Blog is written by Dave Petley, who is widely recognized as a world leader in the study and management of landslides.
On 9 December 2019, a small phreatic eruption occurred on Whakaari/White Island in Aotearoa/New Zealand at a time when there were 47 people, most of whom were tourists, on the site. Tragically, 22 people were killed and many others were seriously injured. The tragedy has led to considerable focus on risk-to-life in Aotearoa/New Zealand with respect to tourist activities – this is a country with abundant natural hazards. It is generally agreed that the risk-to-life at Whakaari/White Island was unacceptable, and a number of prosecutions followed.
However, if it is agreed that the risk-to-life was unacceptable, then a similar question must be asked for other tourist sites in New Zealand. There is a really interesting new paper (Davies and Dykstra 2025) in the International Journal of Disaster Risk Reduction that starts to ask some very interesting questions around this topic, with a particular focus on the Piopiotahi/Milford Sound area. The analysis is fascinating and unsettling.
Piopiotahi/Milford Sound is a major tourist attraction in southern New Zealand. Consisting of a spectacular fjord landscape, it is located in an area with high seismic hazard. There is a distinct possibility that a major earthquake would trigger collapses of the fjord walls, which in turn would have the potential to generate highly destructive local tsunami. Clearly, this would be an exceptionally dangerous situation for anyone located in the area.
Davies and Dykstra (2025) note that, in 2019, Piopiotahi/Milford Sound was visited “…by road by about 870,000 people, plus about 60,000 by air and about 300,000 (passengers and crew) in cruise ships. While most visitors spend only part of a day at Milford, some stay overnight, and about 200 people work there and remain overnight; cruise ships are present for about 9 h on average. Most of the day visitors take a small-boat cruise on the fiord, and those who do not almost always remain within about 10 m of sea level.”
The paper seeks to calculate the individual risk-to-life (i.e. how likely is it that you, as an individual, would be killed if you were to visit) and the societal risk-to-life (i.e. what is the aggregate risk for all visitors over time) for both Whakaari/White Island and Piopiotahi/Milford Sound. This is, in itself, very challenging, involving a cascade of models (the likelihood of an earthquake, the resultant likelihood of landslide, the characteristics of the tsunami, etc.). The authors have developed a rigorous methodology to allow these analyses.
For Whakaari/White Island, the individual risk-to-life was marginally acceptable given that visiting was a voluntary activity (assuming that the visitors were made aware that the site had a significant level of risk – many of the survivors contend that this level of information was not provided). On the other hand, the societal risk-to-life was most definitely not acceptable, and the individual level of risk carried by the guides (who visited the site on multiple occasions) was also unacceptable. Tourist visits to Whakaari/White Island ae no longer permitted, and rightly so.
So what about Piopiotahi/Milford Sound? The individual risk-to-life is most definitely in the acceptable range, and is considerably lower than at Whakaari/White Island. However, the number of visitors is vastly greater. Davies and Dykstra (2025) have calculated that an “average” event at Piopiotahi/Milford Sound would see 1,006 people exposed to the hazard, resulting in 750 fatalities. Thus, such an event would be a true disaster. The resultant societal risk is considerably higher than for Piopiotahi/Milford Sound (0.75 fatalities per year) than it is for Whakaari/White Island (0.3 fatalities per year).
Davies and Dykstra (2025) examine whether the risk could be managed by, for example, forecasting of events. However, earthquakes cannot be predicted, and the proximity of the fault to the site means that alert systems based on the detection of seismic waves would provide little warning. The conclusion is that the only feasible way to manage the risk-to-life is to reduce vulnerability – i.e. to reduce visitor numbers.
Thus, Aotearoa/New Zealand is left with a substantial dilemma. Davies and Dykstra (2025) report that visitors to Piopiotahi/Milford Sound spend about NZ$200 million per year. Thus, the financial impact of cutting visitor numbers is potentially severe, and in simple economic terms cannot be justified. However, should an event occur that led to 750 fatalities, there is little doubt that there would be a view in the aftermath that the risk was unacceptable.
The decision as to whether this level of risk should be tolerated is one for society, not for scientists. However, taking a decision to change the status quo would be highly contested, and it is extremely difficult to explain concepts of societal risk to a wide audience.
This situation is not unique to Aotearoa/New Zealand of course, but most other areas with high levels of societal risk associated with tourism have not had a Whakaari/White Island type event to bring the issue into focus.
Reference
Davies, T. and Dykstra, J. 2025. Societal risk-to-life from natural hazards: Assessments, acceptability and actions at Whakaari/White Island and Piopiotahi/Milford Sound, Aotearoa/New Zealand. International Journal of Disaster Risk Reduction, 119, 105325. https://doi.org/10.1016/j.ijdrr.2025.105325
