2024 Vanuatu earthquake triggered rapid insurance payout for emergency response

Posted on by

Much of the world lacks earthquake insurance, but when a magnitude 7.3 earthquake struck Vanuatu, the Pacific island nation was protected.
 

By Ross S. Stein and Gabriel Lotto, Temblor, Inc., and Shinji Toda, Tohoku University
 

Citation: Stein, R.S., Lotto, G., and Toda, S., 2025, 2024 Vanuatu earthquake triggered rapid insurance payout for emergency response, Temblor, http://doi.org/10.32858/temblor.359
 

On November 1, 2024, the Pacific island nation of Vanuatu purchased streamlined, ‘parametric’ earthquake and tsunami insurance policies, modeled by Temblor for a regional insurer, the Pacific Catastrophe Risk Insurance Company (PCRIC). Six weeks later, on December 17, a magnitude 7.3 earthquake struck near the capital city of Port Vila.

The event took 14 lives and caused massive damage to the capital. About 2,500 people were displaced, while 20,000 were without water. The earthquake’s aftermath impacted some 6,000 workers from 200 businesses operating in Port Vila. The Vanuatu Recovery Operation Centre estimated that the recovery would cost $230 million USD. On December 28, an insurance check to support relief efforts was hand-delivered to Vanuatu’s prime minister by the CEO of PCRIC (Figure 1).
 

Figure 1. Less than ten days after the earthquake, Aholotu Palu, CEO of Pacific Catastrophe Risk Insurance Company (at left), handed the check to Hon. Charlot Salwai Tabimasmas, Prime Minister of Vanuatu (at right). Credit: PCRIC, used with permission
Figure 1. Less than ten days after the earthquake, Aholotu Palu, CEO of Pacific Catastrophe Risk Insurance Company (at left), handed the check to Hon. Charlot Salwai Tabimasmas, Prime Minister of Vanuatu (at right). Credit: PCRIC, used with permission

 

This approach — using parametric insurance — holds great promise to ensure speedy recovery. Such policies do not require claims adjustors after a calamity, so premiums are low and payout is quick. These factors make PCRIC’s parametric policies especially appealing for developing countries prone to natural hazards like quakes, tsunamis, tropical cyclones and floods — all of which are covered. Here, we explore details of these policies and discuss how Temblor’s models contributed. We also consider what happened, seismically speaking, during the December 2024 earthquake and what it might mean for the future.
 

What is parametric insurance?

Traditional (‘indemnity’) insurance typically covers damage to buildings or infrastructure. These policies have a high overhead because of the need to survey damage and negotiate payouts. This ‘claims’ process can be time-consuming, ranging from a few months to a few years, and is expensive.

On the other hand, ‘parametric’ insurance involves using a previously agreed-on threshold of some parameter — like shaking intensity, in the case of earthquakes — to define the payout. For governments of countries prone to hazards, these policies can be appealing for several reasons. They are cheaper than traditional insurance because of minimal overhead. Payout happens within weeks because there is no need to survey damage.

Further, in the event of an earthquake, tsunami, cyclone, flood or other major hazard, consequences are often much broader than the damage to individual buildings. For example, businesses and vital services are interrupted, and there may be a need for temporary housing, which are generally not covered by traditional insurance. So, for national governments dealing with emergency response, parametric insurance makes sense. This is especially true for remote locales or countries without an established insurance market and claims staff.

The disadvantage of parametric insurance is that the selected parameter may not be as well correlated with the consequences (say, the cost of repairs or the loss of life) as intended. So, parametric policy design is critical for a fair transaction.
 

Vanuatu’s hazards

The islands of the Pacific are enmeshed in a latticework of active faults, many of which are the products of the collision between the Pacific and Australian plates. The oceanic Pacific plate subducts, or is shoved, beneath the edge of the Australian plate at a rate of 100 millimeters per year (4 inches per year), one of the highest rates in the world (Figure 2). The high velocity of this 3,500-kilometer-long (2,200 mile) collision produces more than 15% of the world’s earthquakes. The larger and shallower of these Pacific quakes have launched frequent destructive tsunamis that can destroy vital harbors and low-lying coastal cities.

The islands themselves are a consequence of collision; volcanoes erupt at the edge of the Pacific plate, building islands upward. Yet, eruptions can have devastating impacts, including tsunamis. The 2022 Hunga Tonga-Hunga Ha’apai event (Edmonds, 2022) caused tsunamis that traveled around the globe, with massive waves inundating parts of Tonga. Such perils are the cost of living in these beautiful string-of-pearls tropical isles.
 

Figure 2. A century of magnitude 8 and larger shocks (upper panel) compared to a month of magnitude 4 and larger shocks in the Pacific islands (middle panel), and to two centuries of tsunamis (lower panel) reveal the power of forces unleashed by plate tectonics. Credit: Temblor, CC BY-NC-ND 4.0
Figure 2. A century of magnitude 8 and larger shocks (upper panel) compared to a month of magnitude 4 and larger shocks in the Pacific islands (middle panel), and to two centuries of tsunamis (lower panel) reveal the power of forces unleashed by plate tectonics. Credit: Temblor, CC BY-NC-ND 4.0

 

Foreseeing and reducing human impact

Temblor has constructed a worldwide earthquake simulation — what we call a stochastic event set — using seismic and geodetic observations. Our model uses these observations to simulate earthquakes ranging from magnitude 5.0 to 9.5 around the world for a time span of 50,000 years. Stochastic event sets are the backbone of insurance risk models.

Temblor’s stochastic event set includes the expected shaking footprint of each simulated quake. When we examine this simulation for Vanuatu, we see that in any given year, there is 1% chance that more than 7,500 people will experience very strong shaking — MMI 7 on the Modified Mercalli Intensity scale (equivalent to 20% of the force of gravity, which is strong enough to damage all but the best-constructed buildings).

When we look at the recent history of earthquakes and tsunamis in this region, we can assess the actual human impact. Temblor calculates that about 8,500 people in Vanuatu have experienced very strong shaking in the past century (Figure 3, left panel). So, the past century of observations is in good agreement with our simulations (7,500 people in the simulation vs. 8,500 estimated).

Interestingly, the shaking toll is not uniform across the Pacific islands. In our calculations, Tonga, to the east, has one-tenth the quake-impacted people, whereas the much more populated and vastly larger Papua New Guinea, to the west, has ten times the quake-impacted population than Vanuatu.

We also determined that about 3,000 people in Vanuatu live in the tsunami hazard zone, whose dwellings and businesses could be inundated by a 3 meter-high (10 feet) tsunami. As one can see (Figure 3, right panel), over the past two hundred years, about fifteen locations record a 3-meter or higher inundation, and so those 3,000 people are in harm’s way.
 

Figure 3. Earthquakes since 1905 that Temblor calculates to have caused very strong (MMI greater than or equal to 7, which is equivalent to 20% gravity) shaking in populated areas in Vanuatu (left panel). The smaller quakes tend to lie beneath population centers, whereas large quakes are typically farther away. Tsunami inundation has been widespread throughout the island chain, including at Port Vila (right panel). Credit: Temblor, CC BY-NC-ND 4.0
Figure 3. Earthquakes since 1905 that Temblor calculates to have caused very strong (MMI greater than or equal to 7, which is equivalent to 20% gravity) shaking in populated areas in Vanuatu (left panel). The smaller quakes tend to lie beneath population centers, whereas large quakes are typically farther away. Tsunami inundation has been widespread throughout the island chain, including at Port Vila (right panel). Credit: Temblor, CC BY-NC-ND 4.0

 

Creating parametric insurance policies

In June 2024, Temblor was contracted by the World Bank to build earthquake and tsunami models for the Pacific islands, which include Vanuatu. Our models were used by PCRIC, in collaboration with Willis Towers Watson (branded as WTW), as the foundation to price and structure insurance policies that could provide emergency response funds. The funds would be proportional to the population impacted by either hazard.

The shaking footprints of tens of thousands of simulated earthquakes enabled WTW to estimate the likelihood of populations experiencing various levels of shaking intensity over different time frames. This allowed PCRIC and WTW to price parametric earthquake policies for each country.

Our tsunami inundation model similarly helped PCRIC estimate the likelihood that NOAA’s Pacific Tsunami Warning Center would issue a Threat Message that a tsunami of more than 3 meters in height could reach any of the island nations. With that information, PCRIC priced the tsunami policies.

PCRIC then offered the parametric policy to the island nation governments, providing both coverage and guidance. Also included were tropical cyclone, modeled by Reask, and flood, modeled by JBA. The policies went into effect on November 1. The Global Shield Financing Facility, a fund hosted by the World Bank and financed by the Governments of Canada, Germany, Japan, Luxembourg, and the United Kingdom, subsidized the premiums the countries paid to PCRIC for their climate policies as a way to encourage their participation.
 

Vanuatu receives an earthquake insurance payout

The 17 December 2024 Port Vila shock struck on a vertical tear in the subduction zone, which is relatively uncommon (Figure 4). Fortunately, that means the quake launched only a modest tsunami, with a 0.6-meter peak-to-trough amplitude measured on a Port Vila harbor tide gage (Bradley and Hubbard, 2024), sparing the nearby islands from disastrous inundation. The quake nevertheless caused strong shaking, destroying the U.S. Embassy and many other buildings.
 

Figure 4. The Australian plate is subducting under Vanuatu, which results in frequent subduction earthquakes. The motion of the Australian plate is shown by the unfilled arrow. But offshore of Port Vila, there is a vertical tear in the plate, with the south side of the tear uplifted, and the north side subsided. The tear was unknown before the 2024 earthquake revealed its presence. The 2010 magnitude 7.3 shock was a more typical subduction event, whose motion and location are represented by the large grey arrow. Credit: Temblor, CC BY-NC-ND 4.0

Figure 4. The Australian plate is subducting under Vanuatu, which results in frequent subduction earthquakes. The motion of the Australian plate is shown by the unfilled arrow. But offshore of Port Vila, there is a vertical tear in the plate, with the south side of the tear uplifted, and the north side subsided. The tear was unknown before the 2024 earthquake revealed its presence. The 2010 magnitude 7.3 shock was a more typical subduction event, whose motion and location are represented by the large grey arrow. Credit: Temblor, CC BY-NC-ND 4.0

 

Vanuatu’s parametric insurance payment was based on whether a shaking intensity of MMI greater than or equal to 7 occurred within population clusters of 100 people. If any population cluster within a coverage zone — one of 66 administrative areas in the case of Vanuatu — experienced a shaking intensity greater than or equal to MMI 7, then the entire population in that coverage zone counts toward the total affected population. The total affected population defines the payout amount.

More seismic instruments would ensure a fairer policy

To determine whether any population cluster experienced the requisite shaking intensity, the policies use U.S. Geological Survey ShakeMaps, a data product produced by the USGS after major earthquakes. In this context, a ShakeMap functions as an authoritative independent source for who experienced shaking, and just how strong that shaking was.

ShakeMaps usually report shaking intensity based on a combination of data from strong ground motion instruments and felt reports provided by those who experienced shaking. The absence of strong ground motion instruments in the islands means that in this part of the world, ShakeMaps are based on models of typical earthquake behavior, modified by any felt reports submitted to the USGS by the local populace.

This is true for Vanuatu, where there are no strong ground motion instruments within 250 kilometers of the December quake’s epicenter (Figure 5). This means that the USGS ShakeMap for the Vanuatu event is more model than observation. If, for example, an earthquake shook much harder than usual for its magnitude and depth, this would likely be missed.

The installation of seismometers in the capital cities of these island nations would ensure better delivery of insurance, and better understanding of earthquake hazards. In other words, the introduction of insurance means there is both a scientific and financial reason for a denser seismic network.
 

Figure 5. The strength of shaking in the 2024 earthquake is not well known at Port Vila because there are no strong motion instruments within 250 kilometers. The USGS relied on several Did You Feel It? felt reports to augment the data on which its model is built, but these occupant reports are difficult to calibrate. Credit: Temblor, CC BY-NC-ND 4.0
Figure 5. The strength of shaking in the 2024 earthquake is not well known at Port Vila because there are no strong motion instruments within 250 kilometers. The USGS relied on several Did You Feel It? felt reports to augment the data on which its model is built, but these occupant reports are difficult to calibrate. Credit: Temblor, CC BY-NC-ND 4.0

 

So, how did our simulated 50,000 year-long record fare against such an unusual quake? After the earthquake, we searched the event set for similar shocks, and found a near-identical earthquake — one with the same magnitude, depth and location (Figure 6, left panel). When one compares the previously-simulated shaking in that event with the USGS model for the actual event (Figure 6, right panel), they are very similar, with the shaking slightly higher (about 0.5 MMI units) in the USGS model. That similarity suggests that the policies were priced fairly.
 

Figure 6. Temblor calculated the shaking caused by 50,000 years of simulated earthquakes (called a ‘stochastic event set’), with the goal to capture all possible quakes at their likely frequencies of occurrence. This calculation was needed to price Vanuatu’s earthquake insurance policy. After the 17 December 2024 earthquake, we sifted through our event set and found an event of the same magnitude, depth and location as the actual event (left panel). The modeled shaking from our event set quake had similar shaking to the USGS model for the actual event (right panel). Credit: Temblor, CC BY-NC-ND 4.0
Figure 6. Temblor calculated the shaking caused by 50,000 years of simulated earthquakes (called a ‘stochastic event set’), with the goal to capture all possible quakes at their likely frequencies of occurrence. This calculation was needed to price Vanuatu’s earthquake insurance policy. After the 17 December 2024 earthquake, we sifted through our event set and found an event of the same magnitude, depth and location as the actual event (left panel). The modeled shaking from our event set quake had similar shaking to the USGS model for the actual event (right panel). Credit: Temblor, CC BY-NC-ND 4.0

 

A previous quake, and future considerations

In 2010, another magnitude 7.3 earthquake struck just 18 kilometers (10 miles) to the north of the 2024 shock. This was a subduction event (Figure 4). Retrospectively, we calculate that the 2010 event brought the tear fault, which hosted the 2024 earthquake, closer to failure (Figure 7). Earthquake interaction in these calculations is dependent on the proximity of the two events, but also on their magnitudes, geometry, and sense of slip, which, fortunately, are monitored and reported by the USGS.

Why the 14-year wait between quakes? We don’t know. In general, stress increases greater than 0.1 bar turn on seismicity (red patches in Figure 7), and stress decreases by the same amount (blue patches in Figure 7) turn it off. The 2010 earthquake increased stress on the 2024 fault by ten times.
 

Figure 7. In retrospect, a 2010 magnitude 7.3 shock (‘source’) appears to have brought the tear fault on which the 2024 earthquake ruptured (‘receiver’) significantly closer to failure at and above the future 2024 epicenter. Credit: Temblor, CC BY-NC-ND 4.0
Figure 7. In retrospect, a 2010 magnitude 7.3 shock (‘source’) appears to have brought the tear fault on which the 2024 earthquake ruptured (‘receiver’) significantly closer to failure at and above the future 2024 epicenter. Credit: Temblor, CC BY-NC-ND 4.0

 

Looking to the future, the 2024 shock modified the conditions for failure in its surroundings, just as the 2010 shock did. The beachballs (technically, ‘focal mechanisms’ or ‘moment tensors’) reveal the faults on which past earthquakes have struck, and so we calculate the stress imparted to these faults by the 2024 magnitude 7.3 shock in Figure 8.

One sees a stress drop in the core of the 2024 rupture (where the beachballs turn blue), and a surrounding stress increase (where the beachballs turn red), particularly to the north and south along the subduction zone. Thus, we calculate that large subduction earthquakes are now promoted in the stress trigger zones. But since the 2010 shock already ruptured the subduction zone, our best guess is that another event to the south of the tear fault, is now more likely than one to the north.
 

Figure 8. We calculate that large earthquakes are now promoted in the stress trigger zones, and inhibited in the blue stress shadows. These effects will diminish with time over the next several decades, just as the rate of aftershocks does. This forecast strategy is explained more fully in Toda and Stein (2022). Credit: Temblor, CC BY-NC-ND 4.0
Figure 8. We calculate that large earthquakes are now promoted in the stress trigger zones, and inhibited in the blue stress shadows. These effects will diminish with time over the next several decades, just as the rate of aftershocks does. This forecast strategy is explained more fully in Toda and Stein (2022). Credit: Temblor, CC BY-NC-ND 4.0

 

Earthquake ‘doublets’ or pairs such as the 2010 and 2024 event are, rather surprisingly, common in the Vanuatu subduction zone. Cleveland et al. (2014) write, “The repeated occurrence of large earthquake doublets along the northern Vanuatu subduction zone is remarkable…” These authors attribute that behavior to the presence of small, irregularly spaced asperities (bumps, tears, and ridges) along the plate interface. Whether their explanation is right or wrong, the occurrence of doublets and progressive earthquake sequences along the Vanuatu subduction zone suggests that the 2024 event could be followed by another large shock during the next months to decades.
 

Solutions to island vulnerabilities

Pacific island nations — remote and with populations generally residing on low-lying swaths of land — will always be vulnerable to earthquake shaking and tsunamis. The best protection against these threats are stronger buildings, restricting buildings from tsunami inundation zones, and strengthening earthquake and tsunami monitoring and warning systems.

But parametric insurance offers a means to respond to the immediate aftermath of disasters to restore services and provide temporary food and housing. Although modest, the insurance payment for Vanuatu was a signal success. A similar program on the other side of the globe, the Caribbean Catastrophe Risk Insurance Facility, has been in place for more than a decade. These programs could be replicated in other seismically vulnerable regions worldwide.
 

Science editor: Dr. Alka Tripathy-Lang, Ph.D.
Reviewers: Nicolas Pondard, Ph.D. and Simon Young, Ph.D.

References

Bradley, K., Hubbard, J., 2024. Deadly Mw7.3 earthquake shakes Vanuatu. Earthquake Insights, https://doi.org/10.62481/76f004d0

Cleveland, K. M., C. J. Ammon, and T. Lay (2014), Large earthquake processes in the northern Vanuatu subduction zone, J. Geophys. Res. Solid Earth, 119, 8866–8883, doi:10.1002/2014JB011289.

Edmonds, M., 2022, Hunga-Tonga-Hunga-Ha’apai in the south Pacific erupts violently, Temblor, http://doi.org/10.32858/temblor.231

Salgado-Gálvez, M.A. Ordaz, M., Krishna Singh, S., Pérez-Campos, X., Huerta, B., Bazzurro, P., Fagà, E., 2023, Developing new parametric insurance models for Caribbean and Central American countries, Temblor, http://doi.org/10.32858/temblor.303

Toda, S., Stein, R.S. Central shutdown and surrounding activation of aftershocks from megathrust earthquake stress transfer. Nat. Geosci. 15, 494–500 (2022). https://doi.org/10.1038/s41561-022-00954-x

USGS (2010), Event Page for 25 December 2010 M 7.3 Vanuatu earthquake: https://earthquake.usgs.gov/earthquakes/eventpage/usp000hrw0/executive

USGS (2024), Event Page for 17 December 2024 M 7.3 Vanuatu earthquake: https://earthquake.usgs.gov/earthquakes/eventpage/us7000nzf3/executive

 

Copyright

Text © 2025 Temblor. CC BY-NC-ND 4.0

We publish our work — articles and maps made by Temblor — under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license.

For more information, please see our Republishing Guidelines or reach out to news@temblor.net with any questions.

Latest posts by Temblor (see all)