(p4)

1. Stress change before the 1976 M=7.8 Tangshan earthquake (China)

Levelling results during 1971 and 1973 show significant height changes in the Tangshan region. The vertical deformation several years before the 1976 Tangshan earthquake could be caused mainly by the movement of the active faults in the region. Inversion analysis of the levelling data indicates that the observed vertical deformation can be well interpreted by the movement of the Jiyunhe fault which truncates the Tangshan earthquake to the southwest. Further stress analysis reveals that shear stress concentration occurred around the Jiyunhe fault at least 3~4 years before the 1976 Tangshan earthquake.

 Implications: (1). Aseismic fault movement occurred before the 1976 Tangshan earthquake; (2). The activity of the faults was mainly aseismic and could not be detected by present seismometers; (3). The active fault segments were found beneath the seismogenic layer.

 Back to upper level
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

P5

2. Stress change before the 1995 M=7.2 Kobe earthquake (Japan)

Significant horizontal deformation were detected by GPS survey before the 1995 Kobe earthquake in a larger area surrounding the Kobe region. The heterogeneous pattern of the observed horizontal deformation could not be interpreted by a uniform strain model. An inversion analysis of the observed horizontal deformation indicates that an active blind-fault might be present in the region. The blind-fault (model) can explain the observed horizontal deformation reasonably well, and its activity could lead to significant stress localization in the area close to the Kobe earthquake fault. The activity of the blind-fault could be related to the reactivation of the pre-existing rupture planes of historical earthquakes and the aftershocks following the 1946 M=8.2 Nankaido earthquake, or alternatively, the blind-fault could be a portion of the upper face of the subducted plate-slab in the region.

   Back to upper level
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 P6

3. Stress change before the 1996 M=7.0 Lijiang earthquake (China)

Inversion analysis of observed GPS baseline and height changes between 1985 and 1992 reveals two active fault segments at the Red River fault zone, Yunnan Province of Southwest China. The modelled active fault segments lead to a clear interpretation of the observed GPS baseline and height changes, and the geometric characteristics of the fault segments are in good agreement with those from geological investigations. The significant activity of the fault segments during this period might be transient, and may reflect the macroscopic, ductile deformation in the mid- to lower crust, which could be responsible for the stress localization in this region. The 3 February 1996 M_L = 7.0 Lijiang earthquake and its aftershocks occurred in a north-south band about 45-km long and 16-km wide, truncated by the Jianchuan-Lijiang fault to the south. The aseismic slip of the Red River and Jianchuan-Lijiang faults during 1988-1991 resulted in significant Coulomb stress concentration (at a depth of 10km) around the 1996 earthquake fault (up to 6.0 bar) and southwestern portion of the Jianchuan-Lijiang fault segment (up to 9.5 bar), and a rise in the Coulomb failure stress by 2.6 bar at the hypocenter of the 1996 Lijiang earthquake. These results demonstrate that the 1996 Lijiang earthquake was possibly triggered or induced by aseismic movement of the Jianchuan-Lijiang fault.

Back to upper level
 
 

.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

The Red Rive fault zone and the 1996 Lijiang earthquake

Map showing the geological faults (dashed curves), epicenters of the Lijiang main shock (star, M_L =7.0) and aftershocks (dots for M_L ³ 4.0; the plus sign for the M_L = 6.0 aftershock), the rupture segment (thin solid-line marked by EF), active segments (bold lines for lower margins) of the Red River (RR) and Jianchuan-Lijiang (JL) faults estimated from the joint inversion of GPS and levelling data. The lower-left inset is a sketch map of the Chinese mainland which shows the location (star) of the 1996 Lijiang earthquake.

See (Dx2ya.gif)

Back to text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

(p13)

Distribution of GPS baselines and levelling stations. (also shown are the Red River (RR) and Jianchuan-Lijiang (JL) faults estimated from the joint inversion of GPS and levelling data (bold lines). The stars denote the epicenters of the 1996 earthquakes (M_L³ 6.0). Also shown are the levelling points (dots), GPS stations (circles), and baselines (solid lines).

See dx21b.gif

 Back to text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Observed (left) and predicted (right) height changes at the Red River fault zone (1985 - 1992). The contour interval is 15 mm. Also shown are the epicenters (stars) of the 1996 earthquakes (M_L ³ 6.0). The squares (left) denote levelling points, and the straight lines (right) represent the two active segments (marked by RR and JL, respectively) estimated from the joint inversion.

See  hsp2ar.gif

Back to text
 
 
 
 
 
 
 
 
 
 
 
 

Back to text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Locations of the active segments at depth for (a) the Red River fault and (b) the Jianchuan-Lijiang fault. Also shown in (b) are hypocenters (circles) of the M_L³ 4.0 aftershocks in the 1996 Lijiang earthquake sequence which occurred within 10 km of the JL fault trace. The star denotes the hypocenter of the M_L = 6.0 aftershock which occurred on the JL fault. The standard error in hypocentral locations is estimated to be less than ± 5km.

See dxdep1.gif

Back to text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Coulomb stress (CCFF) generated by aseismic movement of the Red River (RR) and Jianchuan-Lijiang (JL) faults during 1988-1991 (at a layer of 10km deep). The 1996 earthquake fault (straight line marked by EF) is represented as a north-south trending dislocation (normal dip-slip) extending from the surface to a depth of 10 km. Also shown are the epicenters of the main shock (star) and aftershocks (triangles, M_L ³ 4.0), and active fault segments estimated from the joint inversion (straight lines marked by RR and JL, respectively). The plus sign denotes the forecasted epicenter in 1993 (Zhao et al., 1993). The contour interval is 1.5 bar.

See dxcl10r.gif

Back to text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Annual horizontal deformation around Japanese islands

Some remarks:

The horizontal deformation along the Japanese islands has been well determined by GPS. The data set can be directly used to evaluate the plate subduction models which were commonly used for the Japan subduction zone. In the past, some kinematic models for simulating plate subduction along the Japanese islands were mainly constrained by vertical deformation observed along the Japan subduction zone. Recent studies show that neither could these models interpret the observed horizontal deformation, nor predict correctly the stress change assiciated with the plate subduction process. Future kinematic or mechanic models for simulating plate subduction along the Japanese islands should be able to account for the pattern of the horizontal deformation observed by GPS.
 
 

View horizontal deformation. Please select an area:

1. Whole Japan

2. Kyushu region

3. Chugoku and Kinki region

4. Tokai and Kanto region

5. Northern Honshu (Tohoku)

6. Hokkaido

 Back to topor Go to stress map
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Annual horizontal deformation -whole Japan(t1)

Observational data were from the Japanese permanent GPS network (www.gsi-mc.go.jp). Note that the station Tsusima was used as a fixed point.

See (jpg97b.gif)
 
 

 Back to upper level
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Annual horizontal deformation - Kyushu region (t2)

jzh4.gif
 
 

Note: The short solid-lines denote the displacements computed with a plate subduction model. See the Section of Stress Change for more details.

 Back to upper level
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Annual horizontal deformation - Chugoku and Kinki region (t3)

(nank7.gif)
 
 

Note: The short dashed-lines denote the displacements computed with a plate subduction model. See the Section of Stress Change for more details.

 Back to upper level
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Annual horizontal deformation - Tokai and Kanto region(t4)

See (tk7a.gif)
 
 

Note: The short solid-lines denote the displacements computed with a plate subduction model. See the Section of Stress Change for more details.

 Back to upper level
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Annual horizontal deformation - northern Honshu (Tohoku)(t5)

See (hok12.gif)

 Note: The short dotted-lines denote the displacements computed with a plate subduction model. See the Section of Stress Change for more details.

 Back to upper level
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

  Annual horizontal deformation - Hokkaido(t6)

See  (bht5.gif)

 Note: The short dotted-lines denote the displacements computed with a plate subduction model. See the Section of Stress Change for more details.

 Back to upper level
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Japan Stress Map Version 1.0

(Annual stress change around Japanese islands)

By S. Zhao, 1998.

Some remarks:

Efforts have been paid to establish a 3D-plate subduction model to explain the horizontal deformation observed along the Japanese islands. The model was constructed based on the geometry of the subducted slabs and the rates of the plate motion in the region. The parameters of the model were estimated by inversion of the observed horizontal deformation. The model has provided a reasonable fit (within ± 5mm on an average) to the horizontal deformationcomponents observed at 504 GPS stations and is in good agreement with the principle of the present plate tectonic theory, which justify the model as well as the resultant stress estimates. However, the geometry of the subducted slabs at depth, which were inferred mainly from the depth distribution of earthquakes, could be subject to some large uncertainties, hence the stress changes evaluated by the model as well as the stress maps displayed below may only be taken as semi-quantitative results.

In general, the stress map may be used as a preliminary guide to the assessment of the long-term seismic hazards along the Japan subduction zone. In addition, the stress map would be further improved and updated with input of more GPS data as well as refined geometry of the subducted slabs at depth.

Note that the activity of inland-faults was not considered in the present study, and hence one should be cautious when trying to use the results to explain the local stress changes or seismicity.

Acknowledgements: Part of the research work was done in Kyoto University during the author held a visiting fellowship from the Japan Society for the Promotion of Science (JSPS). The author would like to thank Prof. Shuzo Takemoto and other colleagues in the Department of Geophysics, Kyoto University for their kind assistance.

References

1. Whole Japan

2. Kyushu region

3. Chugoku and Kinki region

4. Tokai and Kanto region

5. Northern Honshu (Tohoku)

6. Hokkaido

 Back to top
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Back to top

Coming soon.

 Sorry, this subject is under construction.

 Back to top
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Recent publications

Zhao S. et al., Deformation and stress localization in southwest Japan, Earth Planet. Sci. Lett., 2003, in press.

Zhao S. & R. D. Mueller, Three-dimensional finite element modelling of tectonic stress field in continental Australia, Joint Publication of Geological Societies of America and Australia, 2003, in press.

Zhao S. et al., 3D finite element modelling of the Northwest Australian stress field, Proceedings of West Australian Basins Symposium (WABS) III, October 20-23, Perth, 2002. 

Zhao S. & R. D. Mueller, Effect of crustal heterogeneities on deformation and stress change associated with faulting, Proceedings of the International Workshop on Physics of Active Fault, E. Fukuyama and R. Ikeda (Eds.), 26-27 Feb., 2002, Tsukuba, Japan, pp. 333-340, 2002.

Zhao S. & Takemoto S., Deformation and stress change associated with plate interaction at subduction zones, Geophys. J. Int, 142, 300-318, 2000.

Zhao S. & S. Takemoto, Aseismic fault movement before the 1976 Tangshan earthquake detected by levelling survey: Implication for preseismic stress localization? Geophys. J. Int., 136, 68-82, 1999.

Zhao S., D. Chao, X. Lai, J. Xu, and G. Seeber, The 1996 M_L=7.0 Lijiang earthquake, Yunnan, China: An anticipated event, J. Geodynamics, 27, 529-546, 1999.

Zhao S. & S. Takemoto, Aseismic fault movement before the 1995 Kobe earthquake detected by GPS survey:
Implication for preseismic stress localization? Geophys. J. Int., 135, 595-606, 1998.

Zhao S. & S. Takemoto, Fault geometry and stress release pattern of the 1995 Hyogo-ken Nanbu earthquake,
Japan estimated by inversion of GPS data, J. Geod. Soc. Japan, 43, 33-43, 1997.

Zhao S., Joint inversion of observed gravity and GPS baseline changes for the detection of the active fault segment at the Red River fault zone. Geophys. J. Int., 122, 70-88, 1995.
 
 

Back to top