Framework for Selection of Representative Earthquake Scenarios for Regional Risk Simulation leveraging National Seismic Hazard Model
This paper presents a framework that leverages National Seismic Hazard Models to select representative earthquake scenarios for regional risk simulation, demonstrated through the identification of two specific scenarios for Wellington, New Zealand, and validated against hazard distributions and criteria of consistency, plausibility, and transparency.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Earthquakes are a force of nature that cannot be stopped, but their impact on human life can be managed. To do this effectively, city planners, engineers, and emergency managers need to understand what a future earthquake might actually look like in their specific neighborhood. For decades, scientists have used two main ways to study this risk. One approach, called probabilistic analysis, calculates the statistical likelihood of shaking at any given spot over a long period, accounting for every possible earthquake that could happen. It produces complex charts showing how often a certain level of shaking might occur. The other approach, known as deterministic scenario planning, imagines a single, specific earthquake event. It asks a straightforward question: "If this specific fault breaks right now, what happens?" This method is powerful for communication because it tells a clear story about damage, recovery time, and displaced populations, helping communities visualize the consequences of a disaster rather than just seeing abstract numbers. However, choosing which single earthquake to imagine has often been a matter of guesswork, relying on past history or a fear of the worst possible case, without a clear guide on whether that imagined event truly represents the real risks facing a region.
A new study led by researchers from University College London, the University of Auckland, and Earth Sciences New Zealand offers a structured way to solve this problem. The team developed a framework that uses the most advanced national seismic hazard models available to select the most representative earthquake scenarios for a region. They tested this method in Wellington, New Zealand, a city known for its high seismic activity and complex geology. Instead of picking a scenario based on a gut feeling or a single historical event, the researchers used data from the country's national model to identify which types of earthquakes contribute most to the shaking risk at a specific location. They then selected two specific rupture events to serve as their representative scenarios: a magnitude 7.3 earthquake on the local Wellington-Hutt Valley fault and a magnitude 8.1 earthquake on the distant Hikurangi subduction zone. By grounding these choices in rigorous data, the researchers ensured that the stories they tell about future disasters are consistent with the actual physics of the region.
The process began by looking at the "ingredients" of risk for Wellington. The national model considers dozens of fault lines and different types of tectonic activity, including shallow crustal faults and deep subduction zones where one tectonic plate slides under another. The researchers needed to find out which of these sources were most likely to cause significant shaking in Wellington, specifically for the level of shaking used in building design codes. They analyzed the data to find the combination of earthquake size and distance that contributed most to the risk. They found that while many different ruptures could happen, a few specific combinations of magnitude and distance were the primary drivers of the hazard. For the local fault, a magnitude 7.3 event was the most significant contributor. For the deeper subduction zone, a magnitude 8.1 event stood out. The team then looked at the specific rupture instances in the national database that matched these sizes and distances. When multiple ruptures fit the description, they chose the one that occurred most frequently in the model's data, ensuring the selected scenario was not just possible, but statistically probable.
Once the two scenarios were chosen, the researchers mapped out what the ground shaking would look like across the greater Wellington region. They used computer simulations to generate maps showing the intensity of shaking for both events. The results revealed a distinct difference between the two types of earthquakes. The magnitude 7.3 earthquake on the local fault produced highly variable shaking. The intensity was extremely high right next to the fault line and dropped off sharply as you moved away. This creates a situation where some neighborhoods could face catastrophic shaking while others just a few miles away experience much less. In contrast, the magnitude 8.1 earthquake from the Hikurangi subduction zone, which is much deeper underground, produced a more uniform layer of shaking across the entire region. While the peak intensity was lower than the worst-case local fault scenario, the shaking was felt more evenly over a wider area.
To ensure these scenarios were truly useful, the researchers compared the shaking predicted by their chosen scenarios against the statistical estimates from the national hazard model. They checked if the shaking from their specific "story" matched the expected shaking for a design-level earthquake. They found that the magnitude 7.3 local fault scenario produced shaking levels that aligned very closely with the median expectations from the national model. However, the magnitude 8.1 subduction scenario produced shaking that was generally lower than the national model's median estimate, though it still captured the extreme upper limits of what might happen. This comparison is crucial because it validates that the chosen scenarios are not just random guesses but are firmly rooted in the scientific understanding of the region's hazard. It also highlights that different scenarios tell different parts of the risk story: the local fault scenario is critical for understanding extreme, localized damage, while the subduction scenario is essential for understanding widespread, regional disruption.
The value of this work lies in its ability to provide a clear, defensible foundation for risk reduction. By using a transparent method to select scenarios, the researchers have created a tool that city planners and engineers can trust. These scenarios can now be used to test how buildings, bridges, and water systems would perform under stress, to plan for landslides and liquefaction, and to communicate risks to the public in a way that is both realistic and actionable. The study confirms that while the selection of an earthquake scenario involves some human judgment, that judgment can be guided by robust data to ensure the resulting picture of the future is both plausible and consistent with the best available science. This approach allows communities to move beyond vague fears of "the big one" and instead prepare for specific, representative events that reflect the true nature of the seismic threats they face.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.