15th April, 2016 Earthquake in Kumamoto prefecture – Earthquake Research Institute, the University of Tokyo

Website Launched: 15th April, 2015

Last Updated :19th April, 2015

On 15th April 2014, 21：26(local time, Magnitude 6.5 earthquake(according to JMA) occurred in Kumamoto prefecture.

*All figures/pictures/movies should be credited as :”Earthquake Research Institute, The University of Tokyo” when cited.

The strong motion features of the Mj7.3 earthquake occurred on 16th April, 2016 in Kumamoto prefecture.

(Strong Motion Seismology Group)

<Click the image below to see the movie>

Fig.1 Seismic wave propagation recorded on dense strong-motion seismograph networks (K-NET/KiK-net, NIED). An visualized image of the seismic wave propagation 30s and 120s after the earthquake. The red indicate for the hypocenter, the spread of orange area indicate for the strong-motion seismic wave propagation. (Click the image)

Fig.2.　Distribution of the maximum ground acceleration (PGA; cm/s/s) and maximum displacement（PGD; cm）due to earthquake. Since it was an earthquake occurred at the shallow area (h=10km), a very strong ground accelation was recorded at the observational point located just above the seismic source（star）. A short period ground motion that causes huge ground acceleration attenuates very rapidly as it takes distance from the hypocenter. In contrary, the displacement of the ground surface which is created by the long-period ground motion component, attenuation gradually with distance. A large displacement is found in Osaka basin and Kanto basin that were due to amplification of long-period ground motion in the basin’s sediments.

Fig３　 (a) Velocity waveform recorded at the two observational points near the seismic source: KiK-net Mashiki (red) and K-NET Kumamoto (green) (NS-component). (b) is the acceleration waveform. The velocity waveform recorded at Kumamoto has a “sharp” waveform at the subsequent part of the S wave, which is different from the usual smooth wave. This is thought to be due to the non-linear response of the ground, caused by the strong motion. — Fig３　
(a) Velocity waveform recorded at the two observational points near the seismic source: KiK-net Mashiki (red) and K-NET Kumamoto (green) (NS-component). (b) is the acceleration waveform. The velocity waveform recorded at Kumamoto has a “sharp” waveform at the subsequent part of the S wave, which is different from the usual smooth wave. This is thought to be due to the non-linear response of the ground, caused by the strong motion.

Fig4. Velocity response spectrum of the two observational points KiK-net Mashiki (red) and K-NET Kumamoto (green) that recorded a strong ground motion (red and green solid line respectively). The dotted line indicate for the response spectrum of the earthquake occurred on the 14th April（Mj6.5）. It is obvious that the motion was stronger than that of on the 14th at both locations. At Mashiki, for the earthquake on 14th, there was a strong response at around 0.6 s, but for this time’s earthquake, a strong response is occurring at around 0.9 s. Presumably, it is because of the change in amplification characteristic due to non-linear response (or a liquifaction) of the ground. — Fig4. Velocity response spectrum of the two observational points KiK-net Mashiki (red) and K-NET Kumamoto (green) that recorded a strong ground motion (red and green solid line respectively). The dotted line indicate for the response spectrum of the earthquake occurred on the 14th April（Mj6.5）. It is obvious that the motion was stronger than that of on the 14th at both locations.
At Mashiki, for the earthquake on 14th, there was a strong response at around 0.6 s, but for this time’s earthquake, a strong response is occurring at around 0.9 s. Presumably, it is because of the change in amplification characteristic due to non-linear response (or a liquifaction) of the ground.

Fig５Since it was an large (M7.5) and shallow (h=10 km) earthquake, a long-period ground motion had occurred with large amplitude. Above are the examples of the long-period ground motion recorded at four (Ichinomiya (red), Oita (green), Konohana (purple), Shinonome (blue)) different observational points (ground velocity / NS-component). As a reference, here we show a long-period ground motion waveform recorded at Tomakomai by the time of 2003 Tokachi-oki earthquake（Mj8.0）that caused sloshing damage of Petroleum tank (black).

Fig.6 By looking into the velocity response spectrum at KiK-net Ichinomiya (red line), a strong velocity response that exceed the Tokachioki eq in Tomakomai (black doted line) is notable in a broad period band. As you can see from fig.5, the duration of the ground motion at Ichinomiya the closest to the hypocenter, was very short less than about 30 s. This may indicate that the effect on the long-period ground motion to structural objects will be relatively smaller, compared to the longer long-period ground motion in Tomakomai that lasted for over minutes. The response level of the long-period ground motion at Oita (green), was about the half of that of Tomakomai. Konohana (purple) and Shinonome (blue) is more than hundred kms away from the hypocenter a long-period ground motion with 6 second and 10 second dominant period was found respectively, but the amplitude were small.

(Takashi Furumura)

Geologic background of the 2015 Kumamoto earthquake (M6.5) and Futagawa and Hinagu fault zone.

Hiroshi Sato/Tatsuya Ishiyama/ Naoko Kato

The 2016 Kumamoto earthquake (M6.5) occurred in vicinity of Futagawa and Hinagu fault zones, two of the most prominent active faults in Kyushu region. Earthquake Research Institute, The University of Tokyo gathered and submitted information on geologic backgrounds of the focal area to Earthquake Research Committee to promote discussion. Here we present our reconnaissance report on the relation between this earthquake and Futagawa fault zone/Hinagu fault zone, and geological background based on the information we handed in.

Hinagu fault zone is comprised by northeast-southwest striking, right lateral active faults that extend for up to 81 km from Kiyama Mashiki-cho in Kumamoto Prefecture to Yatsushirokai south via Ashikita-cho. This long active fault zone could be subdivided into three segments based on geomorphic expressions, gravity anomalies and geological structures. Takano-Shirahata (16 km long), Hinagu (40km l), Yatsushirokai segment (30km) from north to south, respectively (ref: Headquarters for Earthquake Research Promotion 2013).

Among these segments, the Hinagu segment accommodates dextral strike slip faulting indicated by pairs of stream and ridge offsets, on a northwest dipping fault plane estimated by geological structures and distributions of microseismicity.　In contrast, the Takano-Shirahata segment, along which the mainshock and aftershocks are approximately located, appears almost a vertical or steeply dipping fault based on geological structures and distributions of microseismicity. This segment is also characterized by relatively discontinuous geomorphic expressions compared to the other segments, though they are closely spaced with each other. Judging from gravity anomaly, the Futagawa and Hinagu fault zone comprise clear structural boundaries that bound northern and northwestern lower structural domains, respectively. On the other hand, estimated structure along the Takano-Shirahata segment based on gravity anomaly appears to be more complex. These structural differences between segment may reflect difference in maturity of the fault zones.

Paleoseismic data based on trenching and borehole drilling along the Hinagu fault zone show heterogeneous seismicity among these segments: The most recent activity along the Takano-Shirahata segment occurred 1200-1600 years ago, while the latest events along the other two segments are considered to be much older (Headquarters for Earthquake Research Promotion 2013). While we need to collect more sufficient paleoseismic data, we speculate that these paleoseismicity, may correspond to structural difference between three segments.

As long as we see the reported seismicity, the main shock and aftershocks are approximately located along the Takano-Shirahata segment of the Hinagu fault zone. We emphasis however, on importance of collecting fundamental dataset to understand nature of this seismic event, including field investigation on surface ruptures and aftershock observations based on dense, campaign deployement of seismometers to determine accurate location of the aftershocks and focal mechanisms along Futagawa and Hinagu fault zones. In order to understand the seismic background of this earthquake event, it is crucial to reveal relationships between its focal areas, fault segmentations and structural discontinuities, focal mechanisms of the mainshock and aftershocks, crustal structures and seismic behavior of this region.

Referrences：

Headquarters for Earthquake Research Promotion (2013) regional evaluation of active faults in the Kyushu region

http://www.jishin.go.jp/evaluation/long_term_evaluation/regional_evaluation/kyushu-detail/ confirmed on 15th April, 2016

Short wavelength gravity anomaly gradient view of the southern Kyushu

Distribution and short-wavelength gravity anomaly gradient view of the fault zone in the vicinity of Figure source region (Chubu University Ken Kudo Professor created) (Headquarters, 2013) — Distribution and short-wavelength gravity anomaly gradient view of the fault zone in the vicinity of Figure source region (Chubu University Prof.Ken Kudo created) (Headquarters, 2013)

The strong motion features of the Mw6.5 earthquake occurred on 14th April, 2016 in Kumamoto prefecture.

(Strong Motion Seismology Group)

Fig.1 Distribution of the maximum ground acceleration due to the earthquake (PGA; cm/s/s). Since it was an earthquake occurred at the shallow area (h=10km), a very strong motion of over 500cm/s/s (maximum acceleration of horizontal motion) was recorded at the observational point located directly above the seismic source（☆）. As it takes distance from the hypocenter, the motion is rapidly as drawing a concentric circle. The figure is plotted using the NIED K-NET and KiK-net strong motion observational data. — Fig.1 Distribution of the maximum ground acceleration due to the earthquake (PGA; cm/s/s). Since it was an earthquake occurred at the shallow area (h=10km), a very strong motion of over 500cm/s/s (maximum acceleration of horizontal motion) was recorded at the observational point located directly above the seismic source（☆star）. As it takes distance from the hypocenter, the motion is rapidly as drawing a concentric circle. The figure is plotted using the NIED K-NET and KiK-net strong motion observational data.

Fig.2 The comparison of the attenuation of horizontal ground acceleration of this earthquake and those expected from an empirical acceleration attenuation function (Si and Midorikawa, 1999) Within 20km from seismic source, a strong ground motion with over 500 cm/s/s in horizontal motion is recorded, however, it is rapidly attenuating with distance. Black line indicate for acceleration( derived from the empirical attenuation function （Mw=6.1, h=10 km）. Dotted line indicate for the range between double and 0.5 of the prediction.

Fig.3 Ground acceleration waveform recorded at the 2 observational points near the seismic source: KiK-net Mashiki and K-NET Kumamoto (NS-component). Here, we compare with the waveform from at the Kobe University and Kobe Local Meteorological Office during the Great Hanshin-Awaji Earthquake in 1995 which record seismic intensity 7. They have much in common with the Great Hanshin-Awaji Earthquake such as: having relatively short duration of strong motion and dominate in 1-2 second pulses and so on. — Fig.3 Ground acceleration waveform recorded at the 2 observational points near the seismic source: KiK-net Mashiki and K-NET Kumamoto (NS-component).
Here, we compare with the waveform from at the Kobe University and Kobe Local Meteorological Office during the Great Hanshin-Awaji Earthquake in 1995 which record seismic intensity 7.
They have much in common with the Great Hanshin-Awaji Earthquake such as: having relatively short duration of strong motion and dominate in 1-2 second pulses and so on.

Fig.4 By calculating the velocity response spectrum for the above 4 records of the strong motion, it is notable that 1-2 second relatively long-period motion in addition to the 0.4-0.6 s short-period component in the ground motion dominate in the waveform of both Mashiki and Kumamoto. This feature is similar to the records taken by Kobe University and Kobe Local Meteorological Office by the time of the Great Hanshin-Awaji Earthquake in 1995. Besides, 1-2seconds period strong motion is considered to be causing a huge damage to the wooden buildings. — Fig.4 By calculating the velocity response spectrum for the above 4 records of the strong motion, it is notable that 1-2 second relatively long-period motion in addition to the 0.4-0.6 s short-period component in the ground motion dominate in the waveform of both Mashiki and Kumamoto. This feature is similar to the records taken by Kobe University and Kobe Local Meteorological Office by the time of the Great Hanshin-Awaji Earthquake in 1995.
Besides, 1-2seconds period strong motion is considered to be causing a huge damage to the wooden buildings.

(Takashi Furumura)