M5.6, M5.4 earthquakes strike south of Hualien, Taiwan
A possible reinvigoration of the swarm? Maybe not.
若要閱讀這篇文章的中文版(由 Google 翻譯),請點擊此處。
Earthquake Insights is an ad-free newsletter written by two earthquake scientists. Our posts are written for a general audience, with some advanced science thrown in! To get these posts delivered by email, become a free subscriber. If you would like to support our work here, please also consider a paid subscription.
A pair of earthquakes shook Taiwan in the evening hours of May 6, 2024. The events, located just south of Hualien, are a reminder of the M7.4 earthquake that caused deaths and damage in the same region on April 3rd. Since then, aftershocks have caused further shaking - most notably, a swarm of seismicity around the southern half of the rupture that started on April 22nd and lasted for about a day. The two earthquakes that shook the country in the evening of May 6 - first a M5.6, and then seven minutes later a M5.4 - occurred in the same region as the swarm. (We note that Taiwan’s Central Weather Administration reports these as two magnitude 5.9 events. The disparity is because of the type of magnitude scale used - we used the moment magnitude Mw; they use the local Richter magnitude, ML.)
The intense swarm behavior stopped on April 23rd; since then, the swarm area has produced more typical “aftershock” behavior - smaller events, becoming less frequent with time. Swarms are usually associated with some kind of progressive movement - for instance, the migration of pressurized fluids along fractures - so when the behavior transitioned from repeated M5-6 earthquakes to smaller, sporadic events, it seemed like those fluids (if they existed) might have stopped migrating.
These two new earthquakes occurred in the same region as the swarm. Does that mean that we expect a renewal of swarm-like behavior?
We cannot predict earthquakes - no one can. However, we can explore the data and make some educated guesses.
The plots below show the earthquake sequence in a variety of ways. At top is a map view, with earthquakes colored by depth, from 0 (red) to 30 (blue) kilometers. Those same earthquakes are projected to the right, onto a timeline going from April 1 to May 10. (Of course, we don’t know what earthquakes will occur in the future - that part of the timeline is blank for that reason.) Below the map, we have included three cross-sections (A-A’, B-B’, C-C’), which show depth profiles of the earthquake sequence from north (A-A’) to south (C-C’).
So, what do we see in these plots? Let’s look at the timeline first. It shows several things.
(1) First, the mainshock rupture initiated in the south and propagated toward the north. Aftershocks occurred across a long section of fault - presumably revealing the extent of the rupture - mostly at depths of 10-30 km. The complex pattern of the aftershocks seems to indicate that multiple faults were affected, either by direct rupture or by stress changes surrounding the rupture. There were more aftershocks to the north than to the south, and the events to the north were a bit shallower. The rate of seismicity decreased, until…
(2) The swarm started on April 22. The swarm was intense, and also significantly shallower (0-10 km) than the bulk of the other seismicity. It was also confined to the southern part of the rupture. Although the maximum magnitude was well below that of the mainshock, four buildings in the area partially collapsed. Fortunately, the swarm died down about a day later, after causing quite a stir in Taiwan.
(3) On April 26, two more large-ish aftershocks occurred. These were to the north, and pretty deep (25-30 km), and were not related to the swarm.
(4) On May 6, the two newest earthquakes occurred (M5.6, M5.4). These events were to the south, around the swarm area. However, they were not shallow, occurring instead at 15-20 km depth.
You can track this sequence of events by looking at the map and cross-sections. You might also notice some other interesting patterns - for instance, aside from the swarm, the seismicity seems to generally illuminate an east-dipping fault plane - perhaps the Longitudinal Valley Fault, which we highlighted in our post immediately after the M7.4 earthquake. However, the seismicity to the north (profile A-A’) is more complicated, and the central profile also suggests a second west-dipping fault.
So, do these new earthquakes indicate that the swarm might be starting up again? We think probably not. First, there hasn’t been any swarm behavior in the six hours since the two events. Second, although in map view the earthquakes overlap with the swarm, in cross-section it is clear that they did not occur in the same place. The swarm was concentrated at shallow depths; the recent earthquakes are well below. While fluids can migrate vertically, they generally move upwards rather than down, so it is unlikely that fluids associated with the swarm would have moved down to the location of these new events.
To be sure we could see what was going on, we made a few more plots. Here is the same set of plots, but with earthquakes colored by time rather than depth. Is it interesting that the aftershocks on April 26 occurred deep below the northwestern edge of the rupture - not a match for an east-dipping fault.
And here is seismicity only from April 22 onward, capturing the swarm and the two more recent earthquakes. The timeline shows that the swarm propagated both northward and southward from its starting point.
Zooming in on the swarm (April 22 to 25) shows that it grew laterally in pulses, before decaying away. It will be interesting to see whether anyone will choose to study the dynamics of this swarm in detail.
Should people in the area be concerned about these recent earthquakes? It is surely true that the faults to the south were stressed by the mainshock, and to a lesser extent by the following aftershock activity. In some cases, earthquakes can trigger other large events on nearby faults, and that remains true here. This is not common, but possible. Even if no large (M7+) event is triggered, regular aftershocks can also cause damage (as was seen in the swarm), especially to buildings that were weakened by the first earthquake. Aftershocks can be expected to continue for months to years, becoming less frequent with time.
However, it does not appear that the physical mechanisms we called on for the swarm - progressive movement of materials - are responsible for these two most recent earthquakes, which look more like delayed aftershocks. Of course, only time will tell how this interesting and complex earthquake sequence evolves.
References:
Hubbard, J. and Bradley, K., 2024. Deadly M7.4 earthquake strikes Taiwan. Earthquake Insights, https://doi.org/10.62481/c4a3297f
Hubbard, J. and Bradley, K., 2024. Seismic swarm near southern tip of April 3 M7.4 in Taiwan. Earthquake Insights, https://earthquakeinsights.substack.com/p/seismic-swarm-near-southern-tip-of