Pingtung Earthquakes and Potential Tsunami Hazards for Taiwan

Size: px

Start display at page:

Download "Pingtung Earthquakes and Potential Tsunami Hazards for Taiwan"

估念糜
5 years ago
Views:

1 Pingtung Earthquakes of Taiwan and Geodynamics David A Yuen Boris Kaus Yingchun Liu Yaolin Shi Erik Sevre

2 Outline Pingtung Earthquake Event, Dec. 26, 2006 The Depths of Epicentres and seismic constraints Geodynamical Modelling

3 The Focal Mechanisms of Two Earthquakes 12:26:00 PM FIRF Lat Long Magnitude Depth (Km) Strike Dip Slip Harvard 21.83N E USGS N E BATS N E :34PM Havard USGS 22.03N E BATS 21.95N E

5 Which epicentral depth is reasonable? Why?? Here we take one way to check it: The co-seismic GPS data The co-seismic displacement field on the surface The simulation of earthquake Displacement

6 Seismic Profiles across Taiwan and the Western Philippine Sea The preferred Moho depth is 30-32km. Vs Vp (Kwang-Hee Kim, et al, 2004) (Kirk McIntosh, 2005)

7 The Model of Crust 4 int: no_model_lines; # no depth[km] vp[km/s] vs[km/s] rho[kg/m^3] eta[pa*s] relax E E E E E Depth 10 Km 32 Km Half space Maxwell model Elastic Dislocation theory 120 Km Rongjian Wang,, Potsdam,2004

8 The displacement field of Uz Maximum Displacement Uz in vertical direction from the focal mechanism of three official websites: USGS : Uzmax = 1.12m Uzmin= m HCMT: Uzmax = 0.616m Uzmin= m BATS : Uzmax = m Uzmin= m (Wen-Tzong Liang, 2007)

9 Cumulative Co-seismic Deformation Uzmax = 0.043m

10 Broken Cable CS Lee, SK Hsu, HW Chen, F Wu, 2007

11 Conclusion(1) The first earthquake is normal fault. From the report of website, only half a meter wave height was heading for Philippines. The earthquakes could have induced the landslide. Above all, the focal mechanism of BATS is more reasonable. The epicenter could be 40km deep.

12 Geophysical Background (1) Chang, et al, 2001

13 The Geodynamic model Why this earthquake occurred in the lower Moho, or upper mantle? Not the upper hard crust?

14 Seismic Profile Geophysical Background (2) Hengchun Peninsula Upper Crust Lower Crust Ocean Crust Upper Mantle Based on C. P. CHANG et al., 2005

15 The Numerical Model

16 Numerical Method - SloMo 2D/3D Thermo-mechanical Lagrangian FEM. Rheology: Maxwell viscoelastic with Mohr Coulomb non-associated plasticity. Bulk deformation: compressible elastic/incompressible viscous Velocity-Pressure formulation (Q 1 P 0+ elements). Implicit plasticity. Markers to track material properties. Remeshing for large deformations. True free surface. Buiter et al. (2006)

17 Model setup Interest: Stress-field caused by mantle flow and mantle/lithosphere interaction (compressional/extensional) Effect of rheological layering? Slab break-off, weak lower crust? Therefore: Models ran for very short timescale only (to ensure isostatic re-equilibrium). Maxwell timescale artificially reduced (<30 years). Limitation: Viscosity constant in each of the layers (Tdependent underway, not finished yet).

18 Horizontal Normal Stress Field Compression Horizontal compression direction = thrust EQ Vertical compression direction = normal EQ Eurasia Continent Plate Upper continental crust Moho Mantle Extension Lower continental crust Oceanic crust Mantle lithosphere Serpentinite (weak) Philippine Sea Plate > Extensional stresses below the Moho due to bending of subducting (oceanic) slab.

19 Continuous lower crust

20 With weak crustal wedge

21 Weak crustal wedge & overriding PSP (8cm/yr)

22 With overriding PSP (8 cm/yr)

23 0 cm year 8 cm year

24 With overriding PSP & weak ML

25 Slab Breakoff 100 km)

26 Vertical slab

27 Vertical weak slab

28 Strong lower crust

29 Strong LC & crustal wedge

30 Strong LC & crustal wedge & overriding PSP

31 Strong LC, overriding PSP (8cm/yr)

32 Strong LC, overriding PSP (8cm/yr), weak slab

33 Strong LC & slab breakoff

34 Strong LC, weak slab (Crème Brulee)

35 Strong LC, weak vertical slab

36 Historical Earthquakes Distribution

37 Geological Tectonics (1)

38 Manila Trench Luzon Trough Geological Tectonic (2)

39 Conclusion 1 In this case, the lower mantle layer must be decoupled from the upper crust by a very soft layer (lower crust). The upper crust is in compressional environment. The upper mantle lithosphere is extensional because of bending due to subducted slab, and because decoupled with upper crust compression

40 Conclusion 2 Normal faulting EQs below Moho below southern Taiwan can be explained by slab subduction of the oceanic slab attached to the China plate. A weak lower crust & sufficiently strong mantle lithosphere are required

41 To be done Include thermal effects (modifies the result but is unlikely to change the main conclusions).

43 David Spring Yaolin Erik Boris

14 17.,., Fig. 1 Bougeur gravity map of research region 2 Fig. 2 Magnetic map of research region

14 17.,., Fig. 1 Bougeur gravity map of research region 2 Fig. 2 Magnetic map of research region 1 7 1 Vol. 17 No. 1 2002 3 (13 23) PROGRESS IN GEOPHYSICS March 2002 (,,100101) [ ],, 39.,,. [ ] ; ; [] P738 [] A [ ]100422903 (2002) 0120013211 0,,.,,, NE [1 ],.., ( NE, EW,NE ),.,.,,. 1 20 20 km ( 10