- Frontier letter
- Open Access
Dynamic faulting on a conjugate fault system detected by near-fault tilt measurements
© Fukuyama; licensee Springer. 2015
- Received: 31 October 2014
- Accepted: 19 February 2015
- Published: 10 March 2015
There have been reports of conjugate faults that have ruptured during earthquakes. However, it is still unclear whether or not these conjugate faults ruptured coseismically during earthquakes. In this paper, we investigated near-fault ground tilt motions observed at the IWTH25 station during the 2008 Iwate-Miyagi Nairiku earthquake (M w 6.9). Since near-fault tilt motion is very sensitive to the fault geometry on which the slip occurs during an earthquake, these data make it possible to distinguish between the main fault rupture and a rupture on the conjugate fault. We examined several fault models that have already been proposed and confirmed that only the models with a conjugated fault could explain the tilt data observed at IWTH25. The results support the existence of simultaneous conjugate faulting during the main rupture. This will contribute to the understanding of earthquake rupture dynamics because the conjugate rupture releases the same shear strain as that released on the main fault, and thus it has been considered quite difficult for both ruptures to accelerate simultaneously.
- Conjugate rupture
- 2008 Iwate-Miyagi Nairiku earthquake
- Near-fault tilt
1987 East Chiba earthquake
Fukuyama (1991) investigated the rupture process of the 1987 earthquake that occurred off the east coast of Chiba (M 6.7) based on the relocated aftershock distribution. The rupture was initiated at the deep southern corner where a conjugate fault could be recognized from the relocated aftershock distribution. In this model, the rupture propagated unilaterally along the main fault as well as on the conjugate fault.
1997 Kagoshima earthquake
Horikawa (2001) analyzed the May 1997 Kagoshima earthquake (M 6.0) by inverting strong ground motions based on the fault model constructed by the aftershock distribution (Miyamachi et al. 1999). He concluded that the rupture was initiated close to the junction of two fault segments and propagated bilaterally.
2000 western Tottori earthquake
Fukuyama et al. (2003) investigated a detailed aftershock distribution of the 2000 western Tottori earthquake (M w 6.6). They found that there was aftershock activity along the conjugate fault trace within 1 h after the main shock at the northern end of the coseismic main rupture region close to the main shock hypocenter. This suggests that a possible conjugate rupture occurred simultaneously with the main rupture as expected by the fault model of Iwata and Sekiguchi (2002).
2007 Niigata Chuetsu-Oki earthquake
Yukutake et al. (2008) relocated the aftershocks of the 2007 Chuetsu-Oki earthquake (M w 6.6) and found that there was a conjugate fault close to the hypocenter region and that the hypocenter was located close to the junction. Aochi and Kato (2010) investigated the possibility of a spontaneous rupture along the conjugated fault using the boundary integral equation method.
Category 2 1987
Superstition hills earthquake
The Superstition Hills earthquake (M 6.8) occurred 12 h after the Elmore Ranch fault earthquake (M 6.2) (Hudnut et al. 1989; Magistrale et al. 1989; Wald et al. 1990). The Superstition Hills fault is a conjugate of the Elmore Ranch. It should be noted that the Elmore Ranch earthquake was initiated northeast of the fault and the rupture propagated toward the junction. The Superstition Hills earthquake was initiated at the junction and propagated toward the southeast.
2000 Wharton basin earthquake
The June 2000 Wharton Basin earthquake (M w 7.8) occurred in the region of diffuse deformation between the Indian and Australian plates, and it is considered as an intraplate earthquake. Robinson et al. (2001) reported, from the analysis of teleseismic waveforms, and relocated aftershock distribution that the Wharton Basin earthquake consisted of two subevents, which were conjugate but they ruptured sequentially. The second rupture initiated around the hypocenter of the first rupture.
2012 Sumatra earthquake
The April 2012 Sumatra earthquake (M w 8.6) occurred at the northwestern end of the Wharton Basin. This is an intraplate earthquake similar to the 2000 Wharton earthquake. According to the back projection of teleseismic data from European and Japanese seismic arrays, it consisted of four subevents that were conjugate and were cascaded together; the rupture propagated sequentially (Meng et al. 2012).
Category 2 earthquakes (1987 Superstition Hill, 2000 Wharton Basin, and 2012 Sumatra) ruptured sequentially along the conjugate fault system, while in the category 1 earthquakes (1987 East Chiba, 2000 Western Tottori, and 2007 Niigata Chuetsu-oki), there is the possibility that the rupture propagated simultaneously along both conjugate faults. The aim of this paper is to investigate the category 1 earthquakes in detail.
Based on theoretical modeling of dynamic fault ruptures, several investigations were made on fault steps (e.g., Harris and Day 1993; Kase and Kuge 1998) and fault branches (e.g., Poliakov et al. 2002; Kame et al. 2003). The above observations are consistent with the theoretical investigation of Kase and Kuge (1998) where they found that the conjugate fault could be initiated where the main rupture terminated. Xu et al. (2015) systematically investigated the physical mechanism of this conjugated faulting using two dimensional (2-D) dynamic rupture simulations. Actually, slip on the two conjugate faults releases the same strain so that it becomes difficult for the rupture to propagate simultaneously on both faults when the main rupture is propagating. The only exception could be at the initiation of the rupture or after the termination of the previous rupture.
It should, however, be noted that all of the above results are based on the relocated aftershock distribution as well as the fault slip distribution estimated by waveform inversions. As can be seen in the fault model database (Mai 2004), waveform inversion results are not unique and may include a certain amount of uncertainty. Thus, there remains the possibility that conjugate faulting did not occur simultaneously during the main rupture. Therefore, we need a solid observation showing that conjugate faulting has occurred during the main rupture.
The Iwate-Miyagi Nairiku earthquake (M w 6.9, MJMA 7.2) occurred on 13 June 2008 at 23:43 (UT). The hypocenter location was estimated to be at a latitude of 39.0269° N and a longitude of 140.8779° E; its depth was 6.5 km (Sekine et al. 2008). This was one of the largest intraplate earthquakes with reverse faulting that had occurred at a shallow depth inland in northeast Japan. According to the F-net moment tensor estimated by regional broadband seismograms, the estimated fault strike, dip, and rake angles were N209° E, 51°, and 104°, respectively, and the seismic moment was estimated to be 2.72 × 1019 Nm (NIED 2008a). Based on the aftershock distribution estimated by the Hi-net microseismic network, the fault length and width were roughly 30 and 10 km, respectively (NIED 2008b).
Before the earthquake, no active fault traces had been recognized in the source region. However, immediately after the earthquake, several fault surface breaks were observed (e.g., Ishiyama et al. 2008; Maruyama et al. 2009), suggesting a west-northwest dipping reverse fault with north-northeast strike.
In this paper, in order to investigate the conjugate faulting during the main rupture, we analyzed the near-fault accelerograms to estimate the coseismic tilt motion in addition to the tiltmeter collocated as well as Global Positioning System (GPS) data. Through this analysis, we constrained the fault slip model of the 2008 Iwate-Miyagi Nairiku earthquake and clarified whether a conjugated fault slip occurred during the main rupture based on the near-fault tilt data.
Tilt measured by accelerograms
We first computed ground displacements on the surface as well as at the bottom of the borehole of the IWTH25 station by numerically integrating the accelerograms two times in the time domain. In this integration procedure, we only removed the initial offset of acceleration in the original records and we kept the other offsets in the following processes. It should be noted that we did not apply any baseline corrections to the accelerograms except for the initial offset for the following reason.
It is well known that strong motion accelerograms sometimes include baseline shifts caused by local tilts. When an accelerogram tilts, its axes deviate from the gravity axis, which causes additional acceleration. Several techniques for baseline correction have been proposed (e.g., Iwan et al. 1985; Boore 2001; Graizer 2006; Kinoshita 2008). However, as pointed out by Boore et al. (2002), translation motion and rotation motion observed by pendulum-type seismometers cannot intrinsically be separated. Thus, it is theoretically impossible to extract the rotation component from an accelerogram. Therefore, in the present analysis, we did not correct the baselines to obtain the displacements. We judged that a baseline shift occurred (i.e., rotational motion occurred) only when the displacements behaved as t2-type displacements.
Observed and predicted displacements and tilts at IWTH25 and ICNS
Disp. East [m]
Disp. North [m]
Disp. up [m]
Tilt E-down [×10 −3 radian]
Tilt N-down [×10 −3 radian]
> ± 0.1
Ohta et al. (2008)7
Takada et al. (2009) 8
Suzuki et al. (2010) 9
Suzuki + conj. fault10
Regarding the NS component, the surface displacement started to behave as t2-type after 5 s (Figure 4) so that in this time window the NS displacement at the surface might include surface tilt around the sensor. We could see similar features in the UD and EW components. Thus, we could not estimate the NS tilt motion correctly and we were unable to determine whether it was positive or negative.
The error in the tilt estimates depends on how precisely the displacements were estimated at both locations, and a typical uncertainty in displacement is on the order of 3 cm. Therefore, the uncertainty in tilt motion reached 1 × 10−4 radians in this case.
Static tilt measured by other instruments
Some other observations were made of the static tilt change at and around the IWTH25 station. Static tilt change represents the tilt change measured by the difference between the tilt before and after the earthquake, thus it might include post-seismic deformation in addition to the coseismic one. A pendulum-type tiltmeter was installed at the bottom of the borehole in the same capsule as the borehole accelerometer. The static tilt change was observed to be 0.23 × 10−3 radians down to the east and 0.14 × 10−3 radians down to the north (Figure 3a, Table 1). This is more or less consistent with the DIF tilt observed by the accelerometers. Although the magnitude of the EW tilt change measured by the tiltmeter was smaller, the tilt direction was consistent with those of the accelerograms taking into account the uncertainty in the NS component.
From the accelerogram on the surface, we could estimate the surface tilt by measuring the static offset appearing in the surface accelerogram, which was 2.2 × 10−3 radians down to the east (Figure 3b, Table 1). We call this tilt SUF tilt to avoid potential confusion with DIF tilt. The SUF tilt change was larger than the DIF tilt change, but the tilt direction was again consistent with the DIF tilt. In the borehole accelerogram, the static offsets were not consistent among the three components, suggesting that these offsets were not caused only by tilt motion. Accordingly, we did not use the offset of the borehole accelerogram for the tilt estimation.
In addition, we could roughly estimate the tilt change using the GPS displacement at the station named ICNS (39.01206° N, 140.86314° E, 468.5 m, Figure 2b) with respect to the displacement at IWTH25. The coseismic displacements at ICNS were estimated to be 0.439 m to the east, 0.339 m to the north, and 1.557 m upward (Ohta et al. 2008, Table 1). Since the ICNS station is located about 320 m N6.5° W from IWTH25, and it is 83 m higher than that of IWTH25 (Figure 2b), a tilt motion of 3 × 10−3 radians down to the east can be estimated (Table 1). This is again consistent with other observations.
Interpretation of tilt data
The tilt estimated above should be quite useful for interpreting the dynamic faulting of this earthquake. Several kinematic fault models have been proposed for this earthquake. Ohta et al. (2008) constructed a fault model consisting of two planar faults by inverting GPS data around the source region. Takada et al. (2009) estimated a fault model to fit the Interferometric Synthetic Aperture Radar (InSAR) data. Takada’s model consists of five planar reverse faults, one of which has a conjugate dip. Abe et al. (2013) updated Takada’s model by assuming two nonplanar fault segments. Asano and Iwata (2008), Suzuki et al. (2010), and Cultrera et al. (2013) estimated the kinematic fault motion using strong motion seismograms. Hikima et al. (2008) used strong motion and geodetic data. Yokota et al. (2009) used 1-Hz GPS data. Asano and Iwata (2008), Hikima et al. (2008), Yokota et al. (2009), Suzuki et al. (2010), and Cultrera et al. (2013) assumed one planar fault divided by approximately 2 × 2 km2 subfaults, and the strike- and dip-slip developments were estimated at each subfault.
Here, we compute the ground deformation at the IWTH25 station using the models by Ohta et al. (2008), Takada et al. (2009), and Suzuki et al. (2010). Ground displacements and ground tilts were computed using the crustal deformation code from Okada (1992). The results are summarized in Table 1.
As shown in Ohta et al. (2008), the model by Ohta et al. could not fit the GPS data at ICNS which is close to the IWTH25 station. Thus, apparently, this model could not explain the deformation at IWTH25 as shown in Table 1. This is because they used only two planar rectangular faults, on each of which uniform slip was assumed. Therefore, the fault model estimated using GPS data was not suitable for the present purpose.
An interesting consequence of this result is that the slip on the conjugate fault occurred simultaneously during the dynamic rupture of this earthquake on the main fault because the tilt change at IWTH25 occurred coseismically during the earthquake rupture as indicated above. If both the west-dipping main fault and the east-dipping conjugate fault slipped simultaneously during the earthquake, the initiation mechanism for the conjugate fault rupture might be of interest. It should be pointed out that the conjugate fault and the branched fault had opposite slip directions at the junction. In the future, it would be interesting to investigate the rupture dynamics for the simultaneous rupture of conjugate faults in the future.
At the IWTH25 station, near-field tilt motion was observed by two sets of accelerometers vertically collocated on the surface and at the bottom of a 260-m-deep borehole during the 2008 Iwate-Miyagi Nairiku earthquake. The amount of tilt was estimated to be 0.8 × 10−3 radians down to the east. To explain this tilt motion based on pre-existing fault models, we concluded that a slip along the conjugate fault coseismically occurred during the earthquake. This represents an important direct observation of a conjugated fault rupture in the field.
S. Aoi provided with the detailed site condition data at IWTH25. H. Hirose provided the tilt data at IWTH25, which was originally stored in the Hi-net data center at the National Research Institute for Disaster Prevention (NIED). W. Suzuki provided his slip model. Accelerograms at IWTH25 were provided by the KiK-net data center at NIED. Discussions with Y. Takada, S. Hok, S. Aoi, W. Suzuki, and N. Pulido were quite valuable. Comments by two anonymous reviewers were also very helpful. I appreciate A. Nishizawa for her editorial work on this manuscript. This work was supported by a NIED research project titled “Development for Crustal Activity Monitoring and Forecasting.”
- Abe T, Furuya M, Takada Y (2013) Nonplanar fault source modeling of the 2008 Mw6.9 Iwate-Miyagi inland earthquake in northeast Japan. Bull Seismol Soc Am 103(1):507–518, doi:10.1785/0120120133View ArticleGoogle Scholar
- Aochi H, Kato A (2010) Dynamic rupture of crosscutting faults: a possible rupture process for the 2007 Mw 6.6 Niigata-ken Chuetsu-oki earthquake. J Geophys Res 115, B05310, doi:10.1029/2009JB006556Google Scholar
- Asano K, Iwata T (2008) Kinematic source rupture process of the 2008 Iwate-Miyagi Nairiku earthquake, a Mw6.9 thrust earthquake in northeast Japan, using strong motion data. Eos Trans AGU Fall Meet Suppl Abstract 89(53):S23B–S1890BGoogle Scholar
- Boore DM (2001) Effect of baseline corrections on displacements and response spectra for several recordings of the 1999 Chi-Chi, Taiwan, earthquake. Bull Seismol Soc Am 91:1199–1211View ArticleGoogle Scholar
- Boore DM, Stephens CD, Joyner WB (2002) Comments on baseline correction of digital strong-motion data: examples from the 1999 Hector Mine, California, earthquake. Bull Seismol Soc Am 92:1543–1560View ArticleGoogle Scholar
- Cultrera G, Ameri G, Saraò A, Cillera A, Emolo A (2013) Ground-motion simulations within ShakeMap methodology: application to the 2008 Iwate-Miyagi Nairiku (Japan) and 1980 Irpinia (Italy) earthquakes. Geophys J Int 193:220–237, doi:10.1093/gji/ggs074View ArticleGoogle Scholar
- Fukuyama E (1991) Inversion for the rupture details of the 1987 east Chiba earthquake, Japan, using a fault model based on the distribution of relocated aftershocks. J Geophys Res 96(B5):8205–8217View ArticleGoogle Scholar
- Fukuyama E, Ellsworth WL, Waldhauser F, Kubo A (2003) Detailed fault structure of the 2000 western Tottori, Japan, earthquake sequence. Bull Seismol Soc Am 93(4):1468–1478View ArticleGoogle Scholar
- Graizer VM (2006) Tilts in strong motion. Bull Seismol Soc Am 96:2090–2102View ArticleGoogle Scholar
- Harris RA, Day SM (1993) Dynamics of fault interaction: parallel strike-slip faults. J Geophys Res 98(B3):4461–4472View ArticleGoogle Scholar
- Hikima K, Miyazaki S, Koketsu K (2008) Rupture process of the 2008 Iwate-Miyagi Nairiku earthquake (Mj7.2), Japan, inferred from strong motion and geodetic data. Eos Trans AGU Fall Meet Suppl Abstract 89(53):S51D–S1789DGoogle Scholar
- Horikawa H (2001) Earthquake doublet in Kagoshima, Japan: rupture of asperities in as stress shadow. Bull Seismol Soc Am 91(1):112–127View ArticleGoogle Scholar
- Hudnut KW, Seeber L, Pacheco J (1989) Cross-fault triggering in the November 1987 superstition hills earthquake sequence, southern California. Geophys Res Lett 16(2):199–202View ArticleGoogle Scholar
- Ishiyama T, Imaizumi T, Koshiya S, Sugito N, Tsutsumi H, Hirouchi D, Marushima N (2008) Surface earthquake fault scarp appearing at Hanokidachio, Ichinoseki-city, during the 2008 Iwate-Miyagi Nairiku earthquake. Active Fault Res 29:iv, in JapaneseGoogle Scholar
- Iwan WD, Moser MA, Peng CY (1985) Some observations on strong-motion earthquake measurement using a digital accelerograph. Bull Seismol Soc Am 75:1225–1246Google Scholar
- Iwata T, Sekiguchi H (2002) Source process and near-source ground motion during the 2000Tottori-ken Seibu earthquake. In: Project Report of the MEXT Grant-in-aid for Scientific Research (B) No. 11694134 “Assessment of Seismic Local-Site Effect at Plural Test Sites”. pp 231–241Google Scholar
- Kame N, Rice JR, Dmowska R (2003) Effects of prestress state and rupture velocity on dynamic fault branching. J Geophys Res 108(B5):2265, doi:10.1029/2002JB002189View ArticleGoogle Scholar
- Kase Y, Kuge K (1998) Numerical simulation of spontaneous rupture processes on two non-coplanar faults: the effect of geometry on fault interaction. Geophys J Int 135:911–922View ArticleGoogle Scholar
- Kinoshita S (2008) Tilt measurement using broadband velocity seismograms. Bull Seismol Soc Am 98:1887–1897View ArticleGoogle Scholar
- Magistrale H, Jones L, Kanamori H (1989) The superstition hills, California earthquake of 24 November 1987. Bull Seismol Soc Am 79(2):239–251Google Scholar
- Mai PM (2004) SRCMOD: a database of finite-source rupture models. http://equake-rc.info/SRCMOD/ Accessed 14 Oct 2014
- Maruyama T, Toda S, Yoshimi M, Omata M (2009) Application of airborne LiDAR to mapping surface ruptures and active faults: an example from the 2008 Iwate-Miyagi Nairiku, Japan, earthquake (Mw6.9). Active Fault Res 30:1–12 (in Japanese with English abstract)Google Scholar
- Meng L, Ampuero JP, Stock J, Duputel Z, Luo Y, Tsai VC (2012) Earthquake in a Maze: Compressional rupture branching during the 2012 Mw 8.6 Sumatra earthquake. Sci 337:724–726View ArticleGoogle Scholar
- Miyamachi H, Iwakiri K, Yakiwara H, Goto K, Kakuta T (1999) Fine structure of aftershock distribution of the 1997 northwestern Kagoshima earthquake with a three-dimensional velocity model. Earth Planets Space 51:233–246View ArticleGoogle Scholar
- NIED (2008a) http://www.fnet.bosai.go.jp/event/tdmt.php?ID=20080613234200&LANG=en. Accessed on 7 Jan 2015
- NIED (2008b) http://www.hinet.bosai.go.jp/topics/iwate-miyagi080614/?LANG=en. Accessed on 7 Jan 2015 (in Japanese)
- Ohta Y, Ohzono M, Miura S, Iinuma T, Tachibana K, Takatsuka K, Miyao K, Sato T, Umino N (2008) Coseismic fault model of the 2008 Iwate-Miyagi Nairiku earthquake deduced by a dense GPS network. Earth Planets Space 60:1197–1201View ArticleGoogle Scholar
- Okada Y (1992) Internal deformation due to shear and tensile faults in a half-space. Bull Seismol Soc Am 82(2):1018–1040Google Scholar
- Poliakov AN, Dmowska R, Rice JR (2002) Dynamic shear rupture interactions with fault bends and off-axis secondary faulting. J Geophys Res 107(B11):2295, doi:10.1029/2001JB000572View ArticleGoogle Scholar
- Robinson DP, Henry C, Das S, Woodhouse JH (2001) Simultaneous rupture along two conjugate planes of the Wharton Basin earthquake. Sci 292:1145–1148View ArticleGoogle Scholar
- Scholz CH (2002) The mechanics of earthquake faulting. Cambridge University Press, Cambridge, 496 pp. ISBN 0521655404View ArticleGoogle Scholar
- Sekine S, Shiomi K, Takeda T, Asano Y, Obara K, (2008) Complex seismic activity within the source region of 2008 Iwate-Miyagi Nairiku earthquake on the basis of double-difference analysis. Paper presented at The 7th general assembly of Asian Seismological Commission and the 2008 fall meeting of Seismological Society of Japan, Tsukuba International Conference Hall, Tsukuba, 24-27 November 2008, X1–017 (in Japanese)Google Scholar
- Suzuki W, Aoi S, Sekiguchi H (2010) Rupture process of the 2008 Iwate-Miyagi Nairiku, Japan, earthquake derived from near-source strong-motion records. Bull Seismol Soc Am 100(1):256–266, doi:10.1785/0120090043View ArticleGoogle Scholar
- Takada Y, Kobayashi T, Furuya M, Murakami M (2009) Coseismic displacement due to the 2008 Iwate-Miyagi Nairiku earthquake detected by ALOS/PALSAR: preliminary results. Earth Planets Space 61:e9–e12View ArticleGoogle Scholar
- Wald DJ, Helmberger DV, Hartzell SH (1990) Rupture process of the 1987 superstition hills earthquake from the inversion of strong motion data. Bull Seismol Soc Am 80(5):1079–1098Google Scholar
- Xu S, Fukuyama E, Ben-Zion Y, Ampuero JP (2015) Dynamic rupture activation of backthrust fault branch. Tectonophysics 644–645:161–183, doi:10.1016/j.tecto.2015.01.011View ArticleGoogle Scholar
- Yokota Y, Koketsu K, Hikima K, Miyazaki S (2009) Ability of 1-Hz GPS data to infer the source process of a medium-sized earthquake: the case of the 2008 Iwate-Miyagi Nairiku, Japan, earthquake. Geophys Res Lett 36:L12301, doi:10.1029/2009GL037799View ArticleGoogle Scholar
- Yukutake Y, Takeda T, Obara K (2008) Well-resolved hypocenter distribution using the double-difference relocation method in the region of the 2007 Chuetsu-oki earthquake. Earth Planets Space 60:1105–1109View ArticleGoogle Scholar
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.