What caused the M=6.4 Taiwan earthquake?

By David Jacobson, Temblor

Check your hazard rank

Updated: 1 p.m. Thursday Feb. 8, 2018

hualien-earthquake-building-damage
At least seven buildings are confirmed to have collapsed in Tueday’s M=6.4 earthquake on Taiwan’s eastern coast near the city of Hualien. This photo shows a building tilting precariously in Hualien. (Photo from: Central News Agency, via Associated Press)

 

In Tuesday’s M=6.4 earthquake on Taiwan’s eastern coast, at least six people died and numerous buildings collapsed in the city of Hualien. Now that it has been a little over 48 hours since the quake, we know more about it, and where damage is concentrated, allowing us to develop hypotheses about what may have happened.

temblor-meilun-fault
This Temblor map shows the location of the two recent large earthquakes off the coast of Taiwan. Also labeled are the major faults in the region. What should be pointed out is that both the Longitudinal Valley Fault and Meilun Fault are predominantly compressional faults, with some left-lateral strike-slip motion.

 

An astonishing foreshock sequence

In a typical aftershock sequence, the rate of earthquakes steadily declines. However, in the figure below, a time series shows that approximately 12 hours prior to the M=6.4 on Tuesday, the rate of earthquakes began increasing. While the exact reason for this increase is unclear, it should be noted that there is no increase in GPS velocities associated with the building in quake rates.

This figure shows an earthquake time series beginning about a day before Sunday’s M=6.1 earthquake. Unlike a typical aftershock decay, the one following the M=6.1 shows an unusual quake buildup beginning about 12 hours prior to Tuesday’s M=6.4. It should be pointed out that there is no increase in GPS velocities associated with this increase in rate of earthquakes.
This figure shows an earthquake time series beginning about a day before Sunday’s M=6.1 earthquake. Unlike a typical aftershock decay, the one following the M=6.1 shows an unusual quake buildup beginning about 12 hours prior to Tuesday’s M=6.4. It should be pointed out that there is no increase in GPS velocities associated with this increase in rate of earthquakes. (Figure from: Taiwan Earthquake Center)

 

An additional factor regarding the aftershock sequences is the distribution of earthquakes. The figure below highlights the location of aftershocks following both Saturday’s M=6.1 and Tuesday’s M=6.4. While the aftershocks after the M=6.1 are more of a smear, there is a clear directionality to the southwest in the quakes following the M=6.4. This leads us to believe they occurred on different faults, which lends itself to hypotheses over which fault may have been the cause of Tuesday’s deadly quake.

The figure above show the aftershock sequences following both M=6+ earthquakes off the eastern coast of Taiwan in the last 4 days. What is important to note is that while the aftershock sequence following the M=6.1 on Feb. 4 is more of a smear, with most earthquakes occurring at a depth of 10-15 km, the aftershocks from Tuesday’s M=6.4 are shallower and have a directionality to the southwest.
The figure above show the aftershock sequences following both M=6+ earthquakes off the eastern coast of Taiwan in the last 4 days. What is important to note is that while the aftershock sequence following the M=6.1 on Feb. 4 is more of a smear, with most earthquakes occurring at a depth of 10-15 km, the aftershocks from Tuesday’s M=6.4 are shallower and have a directionality to the southwest. (Figure from: Taiwan Earthquake Center)

 

Crime scene

Based on the location of Tuesday’s M=6.4 earthquake there are numerous candidate faults. On Tuesday we pointed out that the recent quakes struck at the intersection of the Ryukyu Trench and the northern extension of the Longitudinal Valley Fault, a 150 km-long fault that snakes its way across eastern Taiwan. Because this is a junction of two major systems, there are likely many unmapped faults, many of which could be capable of generating a M=6.4 earthquake like the one on Tuesday. Additionally, the ‘yin-yang’ of seismology means that there are two (mutually perpendicular) fault planes possible for every earthquake focal mechanism, making it still harder to read the tea leaves. In Tuesday’s event, the two possible solutions are left-lateral strike-slip and right –lateral strike-slip. So, here we focus on a major fault that many Taiwanese scientists believe could be responsible, weighing the arguments for and against its role in the quake.

meilun-fault-map
This figure from Shyu, J.B.H., et al. shows an interpretation of the major faults around Hualien, with Tuesday’s M=6.4 earthquake shown. This figure shows the Meilun Fault trending to the east offshore of Hualien. Such an orientation appears to suggest that a fault other than the Meilun Fault was responsible for the quake on Tuesday. (Figure from: Shyu, J.B.H., et al., Seismotectonic characteristics of the northernmost Longitudinal Valley, eastern Taiwan: Structural development of a vanishing suture, Tectonophysics (2016), http://dx.doi.org/10.1016/j.tecto.2015.12.026

 

Could the Meilun Fault be responsible?

One of the major active structures in Eastern Taiwan is the Meilun Fault, which runs right underneath the city of Hualien. Some scientists speculate that the Meilun Fault is just part of the major Longitudinal Valley structural system while others believe it is an independent feature. While it has been quiet during the past two decades, it ruptured in a M=7.3 earthquake in 1951 (Chen et al., 2008), and is capable of generating still-larger quakes because of the length of the Longitudinal Valley Fault.

In Tuesday’s earthquake, one of the many pictures which caught our attention was one of a damaged road (see below). In this picture there is evidence of left-lateral motion (whichever side of the fault you are on, the other side has moved to the left). Even though the Meilun Fault is often mapped as a reverse (compressional) fault, it does have a significant left-lateral strike-slip component. Such strike-slip motion matches the dominant movement shown in the focal mechanism produced by both the USGS and GFZ Potsdam for Tuesday’s M=6.4 event. Additionally, in the 1951 M=7.3 earthquake, the co-seismic vertical offset was ~1.2 m, while the left-lateral offset was ~2 m. Therefore, oblique ruptures, when there is both strike-slip and dip-slip motion, are possible along the Meilun Fault.

buckled-roads-strike-slip
This photo of buckled roads appears to show left-lateral strike-slip offset. Such offset is partially consistent with the movement seen on the Meilun Fault, which runs right underneath Hualien.

 

Further supporting the idea that the Meilun Fault was responsible for Tuesday’s earthquake is the distribution of collapsed buildings. In the figure below, five of the collapsed buildings (red dots) are shown in relation to the fault (red dashed line). All of these buildings are within a few hundred meters of the fault, where peak ground accelerations reached nearly 1.0 g, which the USGS considers extreme shaking, with the potential to cause very heavy damage.

collapsed-buildings-taiwan-earthquake
This figure shows the location of collapsed buildings following Tuesday’s M=6.4 earthquake in Hualien. The dashed red line in this figure represents the Meilun Fault. The graph on the right shows spectral acceleration. What should be pointed out from that graph is that there is a large pulse at 2.5 seconds, which could have been enough to topple buildings.

 

Arguments against the Meilun Fault

While all of these points appear to support the idea that the Meilun Fault was responsible for Tuesday’s earthquake, there are counterarguments. First, as was pointed out, the Meilun Fault has a large compressional component, which was not evident in the focal mechanisms produced. Additionally, in both the Temblor and J. Bruce H. Shyu maps shown above, the Meilun Fault is not in the vicinity. In fact, in the Temblor map, Tuesday’s quake is on the wrong side of the fault, assuming the fault dips to the east. And, in the J. Bruce H. Shyu, the fault is shown to trend to the east. Therefore, not all facts point in the same direction.

Furthermore, while five of the collapsed buildings are in very close proximity to the fault, their demise could be explained by two other factors, resonance and liquefaction. In the figure showing collapsed buildings, there is graph on the right-hand side that shows a pulse of energy at the 2.5 second mark that reached nearly 1.0 g. Such ground shaking could have toppled the buildings. Additionally, all the buildings are close to bodies of water, where there is an elevated liquefaction potential, according to the Central Geological Survey.

rescue-taiwan-earthquake
This photo shows rescue workers as they search for survivors following Tuesday’s M=6.4 earthquake in Hualien. (Photo by Kyodo News/Sipa USA)

 

This earthquake sequence should not be considered surprising

What all of these facts show is that there remains uncertainty over what fault caused Tuesday’s M=6.4 earthquake. While ground investigation could lead to greater knowledge, we also may never know, as the epicenter was location offshore. Nonetheless, this earthquake should not be considered surprising as the entire regions is extremely susceptible to large magnitude earthquakes. Therefore, the immediate concern is recovering those trapped, and understanding how to better prepare the region.

taiwan-earthquake-damage
Rescue services search for survivors in a damaged building in Hualien, following Tuesday’s M=6.4 earthquake. So far, there are only 6 confirmed casualties. (Photo from: EPA-EFE/RITCHIE B. TONGO)

 

References
Central Weather Bureau
USGS
EMSC
New York Times
CNN
Kate Huihsuan Chen, Shinji Toda, and Ruey-Juin Rau, A leaping, triggered sequence along a segmented fault: The 1951 ML 7.3 Hualien-Taitung earthquake sequence in eastern Taiwan, JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 113, B02304, doi:10.1029/2007JB005048, 2008
Shyu, J.B.H., et al., Seismotectonic characteristics of the northernmost Longitudinal Valley, eastern Taiwan: Structural development of a vanishing suture, Tectonophysics (2016), http://dx.doi.org/10.1016/j.tecto.2015.12.026
Taiwan Earthquake Research Center Report – Link

  • Afroz Ahmad Shah

    “This photo of buckled roads appears to show left-lateral strike-slip offset. Such offset is partially consistent with the movement seen on the Meilun Fault, which runs right underneath Hualien.”
    Thank you for the information.
    Do you have any information about the strike of the fault that shows offset of the road. If it goes parallel to LVF or MF that may help in finding of the source.
    Left-lateral with reverse seems more likely.

  • Ross Stein

    In the Taiwan Earthquake Center report linked at the end of the article, you will see a map of left-lateral, right lateral, and compressional surface faulting on page 12, followed by about 25 photos of these offsets on succeeding pages. These do appear to correspond to the mapped Meilun Fault.

  • Bob Avakian

    The displaced road seems to be the up-slope lanes of a multilevel system and on the side of a hill as per the retaining wall on the right side of the picture. In addition the right side of the break seems to be wider than the left side. Could this just be the result of a slight bit of subsidence in the upper lanes?