In conversation with Assistant Professor Judith Hubbard, Principal Investigator at the Earth Observatory of Singapore
1. Why it is important to understand tsunami generation? Recent earthquakes in New Zealand triggered tsunami warnings along coastal communities, yet no tsunamis resulted. Why was this so?
These recent events in New Zealand highlight the fact that tsunami warning systems have to do two things: first, they have to identify which earthquakes can produce tsunamis, and second, figure out which earthquakes won’t. This is difficult to do because the process of tsunami generation is complex and depends on a lot of factors.
In general, when people talk about an earthquake, they are referring to the shaking that we feel as a result of a fault slipping. Most faults are in the region between two tectonic plates, and they periodically slip as the plates move. When a fault slips, it releases a lot of energy in the form of ground shaking, which travels very far from the source. This shaking is felt by people and detected by seismic equipment. For tsunami warning systems, this sets off the first alarm bells.
The tsunami is a secondary effect. A tsunami is produced when slip on a submarine fault changes the shape of the seafloor. When the fault slips, the rocks on either side move. If the rocks move sideways, there will usually not be a lot of change to the seafloor. However, in other cases, rocks move vertically; when this happens, there is a chance of a tsunami. In the case of two of the recent New Zealand earthquakes, one side of the fault went up by several metres compared to the other side, so these earthquakes passed that first test.
When the seafloor changes shape, the ocean will be affected. A vertical change in the seafloor will create a bulge in the ocean. This bulge won’t last long - gravity will work to return the ocean to flat. To do this, the bulge will start to collapse towards the sides, resulting in waves traveling out quickly from the source across the ocean. In deep water, these waves can travel very fast, reaching speeds of 800 kilometres per hour. When the waves reach shallower water closer to shore, they slow down.
This is why the first call to action is to move quickly to higher ground, up hillsides, or to take shelter in an appropriate building that is built tall enough and strong enough to withstand the wave. The warning time will depend on how far you are from the source – you may have minutes to hours to evacuate.
2. Was there anything different about the recent earthquakes in New Zealand?
On March 5th 2021, three earthquakes occurred along the fault zone that extends north of New Zealand, with magnitudes-7.3, 7.4, and 8.1. Two of these earthquakes are the type that we would expect to produce a tsunami: they involved vertical movement of the rocks in the Earth’s crust. They also occurred in the shallower part of the crust, which will have more effect on the seafloor. However, in these events, it turned out that the warning was larger than the actual tsunami.
In reviewing the data, we have to look at each event to understand why this happened. The largest event has the biggest potential risk since it involved the most slip. However, in this event, the seismic waves that were emitted from the earthquake show that even though the earthquake started shallow in the Earth’s crust, the slip extended primarily downwards, going deeper into the Earth. Because of this, the deformation of the seafloor occurred over a broader zone, lessening the overall effect.
The magnitude-7.4 earthquake was different: the slip moved from deep to shallow. However, this event started much deeper, so the slip patch was pretty deep anyway. This, combined with its smaller magnitude, blunted its effect.
The magnitude-7.3 earthquake was actually much closer to New Zealand and therefore more of a risk to that region, but it was not the right type of earthquake – its slip was generally horizontal and so would not have caused vertical movement of the seafloor. Still, we have seen other earthquakes like this one trigger strong local tsunamis, due to secondary effects like land sliding. It is always a good idea to move away from the coast if you feel the ground shaking from an earthquake.
3. What are the additional factors that help you determine the generation of a tsunami?
A lot of factors go into predicting what a tsunami will look like because it depends on how deep the slip is, how narrow or wide the slip area is, how deep the water is where the earthquake starts, how the wave travels through different ocean basin shapes, and how far away the event occurs.
Some of these are things we can find out before an earthquake happens, in order to be prepared. We can map ocean basins in detail and figure out where faults are in advance. Other pieces of information, like the location and style of the earthquake, cannot be known until the earthquake happens. And some information cannot be assessed precisely at all, like the amount of fault slip at the tips of subduction zones, or whether a given earthquake will cause submarine landslides, which can produce their own tsunamis.
We can improve our ability to forecast tsunamis by getting more accurate and faster data about these different components. By improving our mapping of ocean basins, finding areas that have produced submarine landslides in the past, and installing enhanced instrumentation to rapidly detect earthquakes and slip patterns, we can be more prepared for the next events.
4. Why do we need detailed ocean floor maps to better understand tsunami generation?
When a wave travels across ridges, valleys, or mountains on the ocean floor, it changes the shape of the tsunami and its speed. To date, we do not have complete detailed ocean floor maps. We have coarse resolution maps in most places, with strips of detailed maps in places where ships tend to travel – along shipping routes, or in harbors. In the case of tsunamis, the most important information is along the coasts, since the shape of the wave that hits the coast depends very much on that last bit of seafloor. Mapping the coastal regions is hard work because the water is shallow, so the scanners we use to map the seafloor can only see a small piece of the seafloor at a time. This is therefore a very time-consuming task and can only be done with agreements from the relevant country.
In the end, though, we must be careful about relying too much on tsunami models that only take into account fault slip. While tsunamis that affect very large coastal areas are typically caused directly by earthquakes, local damaging tsunamis can occur by unexpected mechanisms. For example, the highest tsunami wave that we have seen in recorded history happened when an earthquake of magnitude-7.7 in Alaska triggered a rockfall that landed in a lake. This created a huge sloshing effect, and the wave travelled almost a kilometre up the hill on the other side. This was larger than by a factor of 100 compared to other tsunamis and had nothing to do with the slip on the fault.
5. How will your research at EOS help us better understand tsunami generation in Southeast Asia and keep communities safe?
In my research group at EOS, we work on understanding how different fault systems can contribute to tsunami generation, and how unmapped fault systems may pose invisible hazards. Our goal is to be able to understand how different factors affect tsunami generation in order to be able to identify the key factors that are the most useful or important.
This sort of work is important because it can be done before a tsunami occurs. Often after a tsunami hits, there is a lot of discussion about whether or not warnings were used effectively. To help with this, in some regions, tsunami warning systems have been set up to try to give effective warning time. These are sets of buoys that can detect a tsunami out in the oceans. But in Southeast Asia, these systems will not help people who live very close to the faults that pose a hazard. If the buoys are out in the Pacific or the Indian Ocean, they may not feel the tsunami until after it hits people on land.
The systems will also not work for local tsunamis that have high waves but only over a small area. Developing better models of tsunami generation and propagation can help us figure where these events may occur and what they might look like, to support preparedness and education in those regions.
For example, there was an earthquake in 2018 in Palu in Indonesia. This event had horizontal slip and would not typically generate a tsunami. However, just a few minutes after the ground shaking, a tsunami came rushing in. We saw news of this event from people filming on their mobile phones and broadcasting it live. This tsunami was probably caused by submarine landslides. Because the tsunami was mostly in an enclosed bay and the waves hit so quickly, a tsunami warning system would not have been effective. Instead, residents should be educated to evacuate immediately following an earthquake without waiting for directions.
Southeast Asia is ringed by large fault systems and contains many other faults in its interior, both on land and under the water. We are still working on mapping these faults and understanding the hazards that they pose. This work will help us understand how different regions in Southeast Asia should prepare for natural hazards, including earthquakes, volcanoes, landslides, and tsunamis.
For the people of Southeast Asia who are exposed to all these different hazards, it is important that we learn about how they relate to one other, and their frequency. While we cannot predict earthquakes, we can often provide warnings of incoming tsunamis, or be aware that the shaking from an earthquake is a warning that we should listen to. We aim to provide useful information for people in affected areas so they can take cover, move to higher ground, and protect lives, communities, making Southeast Asia a safer place to live.
(Thumbnail source: TopPhotoImages/Canva.com)