- Email: [email protected]
Effect of Tsunamis generated in the Manila Trench on the Gulf of Thailand
Journal of Asian Earth Sciences 36 (2009) 56–66
Contents lists available at ScienceDirect
Journal of Asian Earth Sciences journal homepage: www.elsevier.com/locate/jaes
Effect of Tsunamis generated in the Manila Trench on the Gulf of Thailand Anat Ruangrassamee a,*, Nopporn Saelem b a b
Department of Civil Engineering, Chulalongkorn University, Phayathai Road, Patumwan, Bangkok 10330, Thailand Technip Engineering (Thailand) Ltd., 20F Sun Tower Building A, 123 Vibhavadee-Rangsit Road, Jatujak, Bangkok 10900, Thailand
a r t i c l e
i n f o
Article history: Received 16 July 2008 Received in revised form 21 December 2008 Accepted 23 December 2008
Keywords: Tsunami Tsunami modeling South China Sea
a b s t r a c t Tsunamis generated in the Manila Trench can be a threat to Thailand. Besides runup of tsunamis along the eastern coast, infrastructures in the Gulf of Thailand, for example, gas pipelines and platforms can be affected by tsunamis. In this study, the simulation of tsunamis in the Gulf of Thailand is conducted. Six cases of fault ruptures in the Manila trench are considered for earthquakes with magnitudes of 8.0, 8.5, and 9.0. The linear shallow water wave theory in spherical coordinate system is used for tsunami simulation in the large area covering Southeast Asia while the nonlinear shallow water wave theory in Cartesian coordinate system is used for tsunami simulation in the Gulf of Thailand. It is found that tsunamis reach the southern part of Thailand in 13 h after an earthquake and reach Bangkok in 19 h. The tsunami amplitude is largest in the direction towards the Philippines and Vietnam. The southern part of China is also severely affected. The Gulf of Thailand is affected by the diffraction of tsunamis around the southern part of Vietnam and Cambodia. The tsunami amplitude at the southernmost coastline is about 0.65 m for the Mw 9.0 earthquake. The current velocity in the Gulf of Thailand due to the Mw 9.0 earthquake is generally less than 0.2 m/s. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction The 26 December 2004 Indian Ocean tsunami was generated by the Mw 9.0 + earthquake off the shore of northwestern Sumatra. The first wave of the tsunami struck the western coast of Thailand in the Andaman Sea about 2 h after the earthquake. The tsunami caused human loss and devastating damage to civil engineering structures along the west coast of southern Thailand (TCLEE, 2005; Lukkunaprasit and Ruangrassamee, 2008). The event has drawn alerts in the engineering community to re-evaluate the tsunami hazard in Thailand. Studies on seismic activities in South East Asia have shown that the Philippines are seismically active with subduction earthquakes along the Manila trench (Zhu et al., 2000; Michel et al., 2000; Kreemer et al., 2000; Bautista et al., 2001; Torregosa et al., 2001). Tsunamis generated in the Manila trench can be a threat to Thailand. Besides runup of tsunamis along the eastern coast, infrastructures in the Gulf of Thailand, for example, gas pipelines and platforms can be affected by tsunamis. Tsunami modeling has been developed as a tool to capture generation, propagation, and inundation of tsunamis (Shuto et al., 1986; Imamura, 1992). In this study, the simulation of tsunamis generated in the Manila trench is conducted to investigate the arrival time, tsunami amplitude, and current velocity with an
* Corresponding author. Tel.: +66 2 218 6571; fax: +66 2 251 7304. E-mail addresses: [email protected] (A. Ruangrassamee), [email protected] (N. Saelem). 1367-9120/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jseaes.2008.12.004
emphasis on the Gulf of Thailand. The findings give an insight in the effects on Thailand and neighboring areas. 2. Scenario earthquakes for numerical modeling Bautista et al. (2001) studied the geometry of subducting slabs in the Philippines. The focal mechanism data of earthquakes in the subduction zone were analyzed. A number of historical major events are associated with the Manila trench in the western part of the Philippines. Hypocenters are mainly located at the depth less than 100 km along the subducting slab of the Eurasian plate beneath the Manila trench. Based on collected information and analysis, they proposed a new model of the subducting slab in the Manila trench. In this study, events of earthquakes generated around the Manila trench and the Philippines are collected from Advanced National Seismic System (ANSS) from 1963 to 2006 (ANSS, 2006). Fig. 1 shows the seismicity map of collected events. In selecting magnitudes of scenario earthquakes for further tsunami modeling, it is important to estimate return periods of the earthquakes. Based on the Gutenberg–Richter recurrence law, the annual rate of exceedance and magnitude relation is analyzed as presented in Fig. 2 (Gutenberg and Richter, 1954). The return period which is related to the annual rate of exceedance is summarized in Table 1. Since a return period of about 500 years is typically considered in seismic design of structures, magnitudes of 8.0, 8.5, and 9.0 are used in this study (AASHTO, 2004).
57
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Annual rate of exceedance of Mw
100.00 10.00 1.00 0.10 0.01 5.0 5.2 5.4 5.6 5.8 6.0 6.2 6.4 6.6 6.8 7.0 7.2 7.4 Mw Fig. 2. Relation between the annual rate of exceedance and magnitude.
Table 1 Return period for each magnitude. Magnitude (Mw)
Return Period (Years)
7.0 7.5 8.0 8.5 9.0
6 19 63 205 667
Table 2 Predicted dimension and displacement of faults. Magnitude Mw
Length (km)
Width (km)
Dislocation (m)
8.0 8.5 9.0
162 305 575
71 102 145
2.2 4.5 9.5
Fig. 1. Seismicity in the Philippines from the ANSS database.
Fig. 3. Location of faults for six cases in analysis.
58
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Fig. 4. Cross section of the sea bottom deformation due to different magnitudes of earthquakes.
3. Fault parameters The uplift of sea bottoms caused by earthquakes generates disturbance of water surface which leads to tsunami propagation. Fault parameters are mainly related to earthquake magnitudes (Wells and Coppersmith, 1994; Papazachos et al., 2004). Papazachos et al. (2004) proposed empirical equations developed from global earthquakes classified into strike-slip faults, dip-slip continental faults, and dip-slip faults in subduction regions. The equations for dip-slip faults in subduction regions are expressed as:
log L ¼ 0:55Mw 2:19 log W ¼ 0:31Mw 0:63 log A ¼ 0:86Mw 2:82
ð1Þ
log D ¼ 0:64M w 2:78 where L is a fault length (km), W is a fault width (km), A is a fault area (km2), D is a dislocation (cm), and Mw is a moment magnitude. The dimension and dislocation of faults for the earthquake used in the study are predicted and summarized in Table 2. Six cases of fault ruptures are considered for earthquakes with magnitudes of 8.0, 8.5, and 9.0. The fault planes for all six cases are located on the subducting slab in the Manila trench as shown in Fig. 3. Two locations of fault ruptures caused by Mw 8.5 earth-
quakes and three locations of fault ruptures caused by Mw 8.0 earthquakes are considered. A slip angle of 90° which causes the worst scenario for tsunamis is used. In the study by Bautista et al. (2001), hypocenters are located mostly in the upper 50 km. Hence, the focal depth of 25 km is assumed. And the dip angle is 30° at the depth. Then, vertical sea bottom deformation is determined using the solution by Mansinha and Smylie (1971). Fig. 4 shows the cross section of the sea bottom deformation for three magnitudes of earthquakes. The Mw 9.0 earthquake generates an uplift of about 4.5 m while the Mw 8.5 earthquake causes an uplift of about 2 m. 4. Analytical models Numerical simulation is done using the TUNAMI model which was developed using the shallow water wave theory (Shuto et al., 1986; Imamura, 1992). Fig. 5 shows the area covered in the analysis. The Gulf of Thailand is about 2000 km from the Manila trench. The curvature of the Earth needs to be taken into account. Hence, the linear shallow water wave theory in spherical coordinate system is used for tsunami simulation in the Southeast Asia region (Region 1). Sea bottom friction is not considered in the linear model. The ETOPO2 bathymetric data is used for the Southeast Asia
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Fig. 5. Regions for analysis.
Elevation (m) 2000 1000 0 -1000 -2000 -3000 -4000 -5000 -6000 100.00
104.00
108.00
112.00
Longitude (degree E) Fig. 6. Cross section of Region 1.
116.00
120.00
59
60
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66 Table 3 Locations of observation. No.
1 2 3 4 5 6 7 8 9
Coordinate
Water depth (m)
Latitude (° N)
Longitude (° E)
13.3333 13.0000 11.7667 11.6667 9.6333 9.2000 7.2000 8.0000 11.6000
100.8333 100.2333 99.9667 100.9667 101.0667 100.1000 100.8000 102.4000 102.5333
14 19 28 46 66 21 22 75 13
Table 4 Fault parameters for the 2004 Indian Ocean tsunami.
40
Elevation (m)
20 0 -20
Parameters
Southern segment
Northern segment
Length (km) Width (km) Strike (°) Dip (°) Slip (°) Depth (km) Dislocation (m) Location
500 150 329 15 90 10 11 94.8° E 2.5° N
400 150 358 15 90 10 11 92.0° E 6.5° N
-40 -60 -80 99.20
100.00
100.80
101.60
102.40
103.20
104.00
104.80
105.60
Longitude (degree E)
Fig. 7. Cross section of Region 2.
low water wave theory in Cartesian coordinate system is used for tsunami simulation in the Gulf of Thailand (Region 2). The grid size in Region 2 is 15 s. The bathymetry in Region 2 is digitized from the nautical chart by Hydrographic Department, Royal Thai Navy. Nested grids are used to couple the analysis of two regions (IOC, 1997). Perfect reflection at coastlines is assumed in the analysis. The time steps of 5 and 1 s are used in Region 1 and 2, respectively to satisfy the stability condition. Fig. 8 illustrates nine locations of observation. The locations close to coastlines have a water depth of about 20 m. The locations of observation and corresponding water depths are summarized in Table 3. 5. Validation of tsunami simulation
Fig. 8. Locations of observation points.
region with a grid size of 2 min (Smith and Sandwell, 1997). The cross section of the South China Sea and the Gulf of Thailand is shown in Figs. 6 and 7. The sea depth is about 4 km at the Manila trench while that in the Gulf of Thailand is about 70 m. When tsunami propagates into the Gulf of Thailand where water is relatively shallow, nonlinearity needs to be considered. The nonlinear shal-
The numerical simulation with the numerical scheme, the grid size, and the time step mentioned in the previous section is validated with observation data in the 2004 Indian Ocean tsunami. After the 26 December 2004 tsunami, researchers conducted surveys in affected areas and recorded the maximum water level at various locations in the southern part of Thailand. The data is used to validate the numerical simulation. The fault parameters proposed by Koshimura and Takashima (2005) are used in this study as summarized in Table 4. The fault parameters were obtained by analysis of satellite altimetry data. Two types of analyses are conducted. In Analysis 1, two regions of computations with grid sizes of 2 min and 15 s are used in Region 1 and 2, respectively. Tsunami amplitudes along coastlines are computed and compared with observation data. The purpose of this analysis is to validate the model used in the simulation of tsunamis in the South China Sea and the Gulf of Thailand. In Analysis 2, four regions of computations with grid sizes of 2 min, 15, 5, and 1.67 s are used in Regions 1, 2, 3, and 4, respectively. This analysis is used to observe the effect of grid sizes on simulation results. Inundation is considered in Region 4. Fig. 9 shows Region 1 and 2 in Analysis 1. Fig. 10 shows Regions 1, 2, 3, and 4 in Analysis 2 with focal areas around Khao Lak and Phuket. In both analyses, the tide
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
61
at the time of tsunami occurrence is included. The results will show tsunami amplitudes relative to the mean sea level. Fig. 11 shows the comparison of observation data and analytical results from Analysis 1. It is seen that the results obtained from Region 2 is close to the observed data. The simulation can capture high tsunami amplitudes in Khao Lak (at about 8.7° N). Fig. 12 shows the results from Analysis 2 compared with observation data. The tsunami amplitudes in Region 4 are slightly higher then those in Regions 2 and 3 which have larger grid sizes. Overall, the simulation results agree well with observation data in all regions. Based on the validation of simulation, the numerical scheme and parameters in Analysis 1 are used for the analysis of tsunamis in the South China Sea and the Gulf of Thailand. 6. Analytical results A series of analyses is conducted for all six cases to obtain the arrival time, tsunami amplitude, and current velocity. 6.1. Arrival time Fig. 13 shows time histories of tsunamis at Points 1, 6, and 7 for Case 2. It is seen that the tsunami reaches the southernmost coast of Thailand in about 13 h and takes about 6 h more to reach Bangkok. Since the reverse slip in the Manila trench causes an uplift of the sea bottom in the west and a depression in the east, the leading elevation wave arrives first. Table 5 summarizes arrival time for all six cases. The Mw 9.0 earthquake causes tsunamis to reach Thailand slightly faster than other cases. However, the difference is insignificant. Comparing between Cases 2 and 3; or among Cases 4, 5, and 6, it is obvious that fault planes located southernmost cause tsunamis to arrive earlier. Fig. 14 shows the contour map of tsunami arrival time for the Mw 9.0 earthquake. Since the Gulf Fig. 9. Regions of computation in Analysis 1.
Fig. 10. Regions of computation in Analysis 2.
62
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Fig. 11. Comparison of tsunami amplitudes in Analysis 1 with observation data. (See above-mentioned references for further information).
Fig. 12. Comparison of tsunami amplitudes in Analysis 2 with observation data. (See above-mentioned references for further information).
63
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Water Level (m)
Water Level (m)
0.1
0.1
0.05
0.05
0
0
-0.05
-0.05
-0.1 10
12
14
16
18
-0.1 10
20
12
14
Time (hrs)
16
18
20
Time (hrs)
(a) Point 1
(b) Point 6
Water Level (m) 0.1
0.05
0
-0.05
-0.1 10
12
14
16
18
20
Time (hrs)
(c) Point 7 Fig. 13. Time histories of tsunamis at Points 1, 6, and 7 (Case 2).
Table 5 Arrival time for six cases (h). Mw
Case
Point 1
Point 3
Point 6
Point 7
9.0 8.5
1 2 3 4 5 6
19.2 19.5 19.3 19.8 19.8 19.8
16.6 16.8 16.7 17.2 16.9 16.8
14.8 15.0 14.8 15.3 15.0 14.9
13.5 13.8 13.6 14.0 13.8 13.8
8.0
of Thailand has shallow depths, the propagation speed of the tsunami in the gulf is low. 6.2. Tsunami amplitude The tsunami amplitude is defined as the water level at the time of tsunami occurrence measured relative to the reference water level which is the mean sea level in this study. The effect of earthquake magnitudes on tsunami amplitudes is investigated at various observation points. The tsunami amplitude at Point 7 is presented in Table 6 as an example. It is obvious that the earthquake magnitude contributes significantly to the tsunami amplitude. Note that tsunami amplitudes presented in Table 6 is obtained at a sea depth of 22 m. Fig. 15 shows the shoaling effect as tsunamis approach the coastline near Point 7. The tsunami amplitude increases by 50% from the amplitude at the sea depth of 22 m. The tsunami amplitude at the coastline becomes about 0.65 m for the Mw 9.0 earthquake. Note that the amplitude can be underestimated because the result is obtained from a large grid size. For hazard and risk assessment in coastal zones, inundation
Fig. 14. Tsunami arrival time (h) for the Mw 9 earthquake.
analysis with smaller grid sizes is required. Fig. 16 shows the maximum tsunami amplitudes in the South China Sea for Case 3. The tsunami amplitude is largest in the direction towards the Philippines and Vietnam. The continental shelf on the southeastern part of China causes the amplification of tsunamis in the area. The Gulf
64
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Table 6 Tsunami amplitude at Point 7 for different cases. Mw
Cases
Tsunami amplitude (m)
9.0 8.5
1 2 3 4 5 6
0.42 0.07 0.10 0.02 0.02 0.03
8.0
Coastline
Point 7
Tsunami Amplitude (m)
0.7 case 1 case 3 case 6
0.6 0.5 0.4
location for ships. There are a number of off-shore platforms and gas pipelines in the Gulf of Thailand. In this study, current velocity is computed at the sea depth greater than 15 m with an assumption that the distribution of velocity is uniform throughout the depth. Time histories of current velocity at Point 7 are shown in Fig. 18 for Cases 1, 3, and 6. The positive value is presented because current velocity is computed from vector summation of discharges in two perpendicular directions. As the earthquake magnitude increases from 8.0 to 8.5, the current velocity increases by about 3 and when the earthquake magnitude increases from 8.5 to 9.0, the current velocity increases by about 2. The earthquake magnitude significantly affects current velocity. The distribution of current velocity in the Gulf of Thailand is presented in Fig. 19 for the Mw 9.0 earthquake. The maximum current velocity is 0.27 m/s at a sea depth of 16 m in the southernmost province. In the middle of the gulf, the velocity is about 0.1 m/s. The Gulf of Thailand is slightly affected by currents from tsunamis generated in the Manila trench.
0.3
7. Conclusions
0.2 0.1 0 100.6
100.7
100.7 100.8 Longitude (degree E)
100.8
Fig. 15. Amplification of tsunamis at the coastline near Point 7.
of Thailand is affected by the diffraction of tsunamis around the southern part of Vietnam and Cambodia. The distribution of maximum tsunami amplitudes in the Gulf of Thailand is illustrated in Fig. 17 for earthquake magnitudes of 8.5 and 9.0. It is seen that as the earthquake magnitude increases from 8.5 to 9.0, the tsunami amplitude in the Gulf of Thailand increases by about 3. 6.3. Current velocity Current velocity is an important parameter for evaluating the effect of tsunamis on off-shore structures and determining safe
The effect of tsunamis generated in the Manila trench is investigated in this study. The propagation of tsunamis is analyzed for earthquakes with magnitudes of 8.0, 8.5, and 9.0. The arrival time, tsunami amplitude, and current velocity are computed with an emphasis on the Gulf of Thailand. From the study, the following conclusions can be deduced: (1) Tsunamis reach the southernmost coast of Thailand in about 13 h and take about 6 h more to reach Bangkok. Since the reverse slip in the Manila trench causes an uplift of the sea bottom in the west and a depression in the east, the leading elevation wave arrives first. The effect of earthquake magnitudes on the arrival time is insignificant. (2) The tsunami amplitude is largest in the direction towards the Philippines and Vietnam. The continental shelf on the southeastern part of China causes the amplification of tsunamis in the area. The Gulf of Thailand is affected by the diffraction of tsunamis around the southern part of Vietnam
Fig. 16. Distribution of tsunami amplitudes in the South China Sea for Case 3.
65
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
0.3
0.3
0.25
0.25
0.2
0.2
Current (m/s)
Current (m/s)
Fig. 17. Distribution of maximum tsunami amplitudes.
0.15 0.1 0.05
0.1 0.05 0
0 -0.05 10
0.15
12
14
16
18
-0.05 10
20
12
14
16
Time (hrs)
Time(hrs)
(a) Case 1
(b) Case 3
18
20
0.3
Current (m/s)
0.25 0.2 0.15 0.1 0.05 0 -0.05 10
12
14
16
18
20
Time (hrs)
(c) Case 6 Fig. 18. Time histories of current velocity at Point 7.
and Cambodia. As the earthquake magnitude increases from 8.5 to 9.0, the tsunami amplitude in the Gulf of Thailand increases by about 3. The tsunami amplitude at the southernmost coastline is about 0.65 m for the Mw 9.0 earthquake.
(3) The current velocity in the Gulf of Thailand due to the Mw 9.0 earthquake is generally less than 0.2 m/s. As the earthquake magnitude increases from 8.0 to 8.5, the current velocity increases by about 3. And when the earthquake
66
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Fig. 19. Current velocity for the Mw 9 earthquake.
magnitude increases from 8.5 to 9.0, the current velocity increases by about 2. The earthquake magnitude significantly affects current velocity. References Advanced National Seismic System (ANSS), accessed in 2006. http:// www.ncedc.org/anss/. American Association of State Highway and Transportation Officials (AASHTO), 2004. AASHTO LRFD Bridge Design Specifications, third ed. USA. Bautista, B.C., Baulista, M.L.P., Oike, K., Wu, F.T., Punongbayan, R.S., 2001. A new insight on the geometry of subducting slabs in Northern Luzon, Philippines. Tectonophysics, 279–310. Gutenberg, B., Richter, C.F., 1954. Seismicity of the Earth and Associated Phenomena. Princeton Univ. Press. Princeton, NJ, USA. Imamura, F., 1992. Review of tsunami simulation with a finite difference method. In: Yeah, H., Lin, P., Synolakis, C. (Eds.), Long-Wave Runup Models. World Scientific, pp. 25–42. Intergovernmental Oceanographic Commission (IOC), 1997. IUGG/IOC TIME Project – Numerical Method of Tsunami Simulation with the Leap-Frog Scheme. UNESCO. Koshimura, S., Takashima, M., 2005. Remote Sensing, GIS, and Modeling Technologies Enhance the Synergic Capability to Comprehend the Impact of Great Tsunami Disaster. Third International Workshop on Remote Sensing for Post Disaster Response, Chiba University, Japan. Kreemer, C., Holt, W.E., Goes, S., Govers, R., 2000. Active deformation in Eastern Indonesia and the Philippines from GPS and seismicity data. Journal of Geophysical Research 105 (B1), 663–680. Lukkunaprasit, P., Ruangrassamee, A., 2008. Building damage in Thailand in the 2004 Indian Ocean tsunami and clues for tsunami-resistant design. The IES Journal Part A: Civil & Structural Engineering 1 (1).
Mansinha, L., Smylie, D., 1971. The displacement fields of inclined faults. Bulletin of the Seismological Society of America 61 (5), 1433–1440. Michel, G.W., Becket, M., Angermann, D., Reigber, C., Reinhart, E., 2000. Crustal motion in E-and SE-Asia from GPS measurements. Earth Planets Space 52, 713– 720. Ministry of Natural Resources and Environment, (2005). Report on Evaluation of Damage due to the Tsunami and Recommendations on Prepardness (in Thai). Papazachos, B.C., Scordilis, E.M., Panagiotopoulos, D.G., Papazachos, C.B., Karakaisis, G.F., 2004. Global relations between seismic fault parameters and moment magnitude of earthquakes. Bulletin of the Geological Society of Greece 36, 1482–1489. Poovarodom, N., Limsamphancharoen, H., 2005. Post-Tsunami Survey along the Northern Coastlines of the Affected Areas in Thailand. In: Proceedings of the Tenth National Convention on Civil Engineering, Engineering Institute of Thailand, Thailand. Shuto, N., Suzuki, T., Hasegawa, K., Inagaki, K., 1986. A study of numerical technique on tsunami propagation and run-up. Science of Tsunami Hazards 4 (2), 111– 124. Smith, W.H.F., Sandwell, D.T., 1997. Global sea floor topography from satellite altimetry and ship depth soundings. Science 277, 1956–1962. Technical Council on Lifeline Earthquake Engineering (TCLEE), 2005. Sumatra– Andaman Islands Earthquake and Tsunami of 26 December, 2004 – Lifeline Performance, ASCE-TLCEE Monograph No. 29, American Society of Civil Engineers. Torregosa, R.F., Sugto, M., Nojima, N., 2001. Strong motion simulation for the Philippines based on seismic hazard assessment. Journal of Natural Disaster Science 23 (1), 35–51. Wells, D.L., Coppersmith, K.J., 1994. New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement. Bulletin of the Seismological Society of America 84 (4), 974–1002. Zhu, W., Wang, X., Cheng, Z., Xiong, Y., Zhang, Q., Ye, S., Ma, Z., Chen, J., Hsu, H., Wei, Z., Lai, X., Liu, J., Jin, B., Ren, J., Wang, J., 2000. Crustal motion of Chinese mainland monitored by GPS. Science in China 43, 394–400.
Contents lists available at ScienceDirect
Journal of Asian Earth Sciences journal homepage: www.elsevier.com/locate/jaes
Effect of Tsunamis generated in the Manila Trench on the Gulf of Thailand Anat Ruangrassamee a,*, Nopporn Saelem b a b
Department of Civil Engineering, Chulalongkorn University, Phayathai Road, Patumwan, Bangkok 10330, Thailand Technip Engineering (Thailand) Ltd., 20F Sun Tower Building A, 123 Vibhavadee-Rangsit Road, Jatujak, Bangkok 10900, Thailand
a r t i c l e
i n f o
Article history: Received 16 July 2008 Received in revised form 21 December 2008 Accepted 23 December 2008
Keywords: Tsunami Tsunami modeling South China Sea
a b s t r a c t Tsunamis generated in the Manila Trench can be a threat to Thailand. Besides runup of tsunamis along the eastern coast, infrastructures in the Gulf of Thailand, for example, gas pipelines and platforms can be affected by tsunamis. In this study, the simulation of tsunamis in the Gulf of Thailand is conducted. Six cases of fault ruptures in the Manila trench are considered for earthquakes with magnitudes of 8.0, 8.5, and 9.0. The linear shallow water wave theory in spherical coordinate system is used for tsunami simulation in the large area covering Southeast Asia while the nonlinear shallow water wave theory in Cartesian coordinate system is used for tsunami simulation in the Gulf of Thailand. It is found that tsunamis reach the southern part of Thailand in 13 h after an earthquake and reach Bangkok in 19 h. The tsunami amplitude is largest in the direction towards the Philippines and Vietnam. The southern part of China is also severely affected. The Gulf of Thailand is affected by the diffraction of tsunamis around the southern part of Vietnam and Cambodia. The tsunami amplitude at the southernmost coastline is about 0.65 m for the Mw 9.0 earthquake. The current velocity in the Gulf of Thailand due to the Mw 9.0 earthquake is generally less than 0.2 m/s. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction The 26 December 2004 Indian Ocean tsunami was generated by the Mw 9.0 + earthquake off the shore of northwestern Sumatra. The first wave of the tsunami struck the western coast of Thailand in the Andaman Sea about 2 h after the earthquake. The tsunami caused human loss and devastating damage to civil engineering structures along the west coast of southern Thailand (TCLEE, 2005; Lukkunaprasit and Ruangrassamee, 2008). The event has drawn alerts in the engineering community to re-evaluate the tsunami hazard in Thailand. Studies on seismic activities in South East Asia have shown that the Philippines are seismically active with subduction earthquakes along the Manila trench (Zhu et al., 2000; Michel et al., 2000; Kreemer et al., 2000; Bautista et al., 2001; Torregosa et al., 2001). Tsunamis generated in the Manila trench can be a threat to Thailand. Besides runup of tsunamis along the eastern coast, infrastructures in the Gulf of Thailand, for example, gas pipelines and platforms can be affected by tsunamis. Tsunami modeling has been developed as a tool to capture generation, propagation, and inundation of tsunamis (Shuto et al., 1986; Imamura, 1992). In this study, the simulation of tsunamis generated in the Manila trench is conducted to investigate the arrival time, tsunami amplitude, and current velocity with an
* Corresponding author. Tel.: +66 2 218 6571; fax: +66 2 251 7304. E-mail addresses: [email protected] (A. Ruangrassamee), [email protected] (N. Saelem). 1367-9120/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jseaes.2008.12.004
emphasis on the Gulf of Thailand. The findings give an insight in the effects on Thailand and neighboring areas. 2. Scenario earthquakes for numerical modeling Bautista et al. (2001) studied the geometry of subducting slabs in the Philippines. The focal mechanism data of earthquakes in the subduction zone were analyzed. A number of historical major events are associated with the Manila trench in the western part of the Philippines. Hypocenters are mainly located at the depth less than 100 km along the subducting slab of the Eurasian plate beneath the Manila trench. Based on collected information and analysis, they proposed a new model of the subducting slab in the Manila trench. In this study, events of earthquakes generated around the Manila trench and the Philippines are collected from Advanced National Seismic System (ANSS) from 1963 to 2006 (ANSS, 2006). Fig. 1 shows the seismicity map of collected events. In selecting magnitudes of scenario earthquakes for further tsunami modeling, it is important to estimate return periods of the earthquakes. Based on the Gutenberg–Richter recurrence law, the annual rate of exceedance and magnitude relation is analyzed as presented in Fig. 2 (Gutenberg and Richter, 1954). The return period which is related to the annual rate of exceedance is summarized in Table 1. Since a return period of about 500 years is typically considered in seismic design of structures, magnitudes of 8.0, 8.5, and 9.0 are used in this study (AASHTO, 2004).
57
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Annual rate of exceedance of Mw
100.00 10.00 1.00 0.10 0.01 5.0 5.2 5.4 5.6 5.8 6.0 6.2 6.4 6.6 6.8 7.0 7.2 7.4 Mw Fig. 2. Relation between the annual rate of exceedance and magnitude.
Table 1 Return period for each magnitude. Magnitude (Mw)
Return Period (Years)
7.0 7.5 8.0 8.5 9.0
6 19 63 205 667
Table 2 Predicted dimension and displacement of faults. Magnitude Mw
Length (km)
Width (km)
Dislocation (m)
8.0 8.5 9.0
162 305 575
71 102 145
2.2 4.5 9.5
Fig. 1. Seismicity in the Philippines from the ANSS database.
Fig. 3. Location of faults for six cases in analysis.
58
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Fig. 4. Cross section of the sea bottom deformation due to different magnitudes of earthquakes.
3. Fault parameters The uplift of sea bottoms caused by earthquakes generates disturbance of water surface which leads to tsunami propagation. Fault parameters are mainly related to earthquake magnitudes (Wells and Coppersmith, 1994; Papazachos et al., 2004). Papazachos et al. (2004) proposed empirical equations developed from global earthquakes classified into strike-slip faults, dip-slip continental faults, and dip-slip faults in subduction regions. The equations for dip-slip faults in subduction regions are expressed as:
log L ¼ 0:55Mw 2:19 log W ¼ 0:31Mw 0:63 log A ¼ 0:86Mw 2:82
ð1Þ
log D ¼ 0:64M w 2:78 where L is a fault length (km), W is a fault width (km), A is a fault area (km2), D is a dislocation (cm), and Mw is a moment magnitude. The dimension and dislocation of faults for the earthquake used in the study are predicted and summarized in Table 2. Six cases of fault ruptures are considered for earthquakes with magnitudes of 8.0, 8.5, and 9.0. The fault planes for all six cases are located on the subducting slab in the Manila trench as shown in Fig. 3. Two locations of fault ruptures caused by Mw 8.5 earth-
quakes and three locations of fault ruptures caused by Mw 8.0 earthquakes are considered. A slip angle of 90° which causes the worst scenario for tsunamis is used. In the study by Bautista et al. (2001), hypocenters are located mostly in the upper 50 km. Hence, the focal depth of 25 km is assumed. And the dip angle is 30° at the depth. Then, vertical sea bottom deformation is determined using the solution by Mansinha and Smylie (1971). Fig. 4 shows the cross section of the sea bottom deformation for three magnitudes of earthquakes. The Mw 9.0 earthquake generates an uplift of about 4.5 m while the Mw 8.5 earthquake causes an uplift of about 2 m. 4. Analytical models Numerical simulation is done using the TUNAMI model which was developed using the shallow water wave theory (Shuto et al., 1986; Imamura, 1992). Fig. 5 shows the area covered in the analysis. The Gulf of Thailand is about 2000 km from the Manila trench. The curvature of the Earth needs to be taken into account. Hence, the linear shallow water wave theory in spherical coordinate system is used for tsunami simulation in the Southeast Asia region (Region 1). Sea bottom friction is not considered in the linear model. The ETOPO2 bathymetric data is used for the Southeast Asia
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Fig. 5. Regions for analysis.
Elevation (m) 2000 1000 0 -1000 -2000 -3000 -4000 -5000 -6000 100.00
104.00
108.00
112.00
Longitude (degree E) Fig. 6. Cross section of Region 1.
116.00
120.00
59
60
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66 Table 3 Locations of observation. No.
1 2 3 4 5 6 7 8 9
Coordinate
Water depth (m)
Latitude (° N)
Longitude (° E)
13.3333 13.0000 11.7667 11.6667 9.6333 9.2000 7.2000 8.0000 11.6000
100.8333 100.2333 99.9667 100.9667 101.0667 100.1000 100.8000 102.4000 102.5333
14 19 28 46 66 21 22 75 13
Table 4 Fault parameters for the 2004 Indian Ocean tsunami.
40
Elevation (m)
20 0 -20
Parameters
Southern segment
Northern segment
Length (km) Width (km) Strike (°) Dip (°) Slip (°) Depth (km) Dislocation (m) Location
500 150 329 15 90 10 11 94.8° E 2.5° N
400 150 358 15 90 10 11 92.0° E 6.5° N
-40 -60 -80 99.20
100.00
100.80
101.60
102.40
103.20
104.00
104.80
105.60
Longitude (degree E)
Fig. 7. Cross section of Region 2.
low water wave theory in Cartesian coordinate system is used for tsunami simulation in the Gulf of Thailand (Region 2). The grid size in Region 2 is 15 s. The bathymetry in Region 2 is digitized from the nautical chart by Hydrographic Department, Royal Thai Navy. Nested grids are used to couple the analysis of two regions (IOC, 1997). Perfect reflection at coastlines is assumed in the analysis. The time steps of 5 and 1 s are used in Region 1 and 2, respectively to satisfy the stability condition. Fig. 8 illustrates nine locations of observation. The locations close to coastlines have a water depth of about 20 m. The locations of observation and corresponding water depths are summarized in Table 3. 5. Validation of tsunami simulation
Fig. 8. Locations of observation points.
region with a grid size of 2 min (Smith and Sandwell, 1997). The cross section of the South China Sea and the Gulf of Thailand is shown in Figs. 6 and 7. The sea depth is about 4 km at the Manila trench while that in the Gulf of Thailand is about 70 m. When tsunami propagates into the Gulf of Thailand where water is relatively shallow, nonlinearity needs to be considered. The nonlinear shal-
The numerical simulation with the numerical scheme, the grid size, and the time step mentioned in the previous section is validated with observation data in the 2004 Indian Ocean tsunami. After the 26 December 2004 tsunami, researchers conducted surveys in affected areas and recorded the maximum water level at various locations in the southern part of Thailand. The data is used to validate the numerical simulation. The fault parameters proposed by Koshimura and Takashima (2005) are used in this study as summarized in Table 4. The fault parameters were obtained by analysis of satellite altimetry data. Two types of analyses are conducted. In Analysis 1, two regions of computations with grid sizes of 2 min and 15 s are used in Region 1 and 2, respectively. Tsunami amplitudes along coastlines are computed and compared with observation data. The purpose of this analysis is to validate the model used in the simulation of tsunamis in the South China Sea and the Gulf of Thailand. In Analysis 2, four regions of computations with grid sizes of 2 min, 15, 5, and 1.67 s are used in Regions 1, 2, 3, and 4, respectively. This analysis is used to observe the effect of grid sizes on simulation results. Inundation is considered in Region 4. Fig. 9 shows Region 1 and 2 in Analysis 1. Fig. 10 shows Regions 1, 2, 3, and 4 in Analysis 2 with focal areas around Khao Lak and Phuket. In both analyses, the tide
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
61
at the time of tsunami occurrence is included. The results will show tsunami amplitudes relative to the mean sea level. Fig. 11 shows the comparison of observation data and analytical results from Analysis 1. It is seen that the results obtained from Region 2 is close to the observed data. The simulation can capture high tsunami amplitudes in Khao Lak (at about 8.7° N). Fig. 12 shows the results from Analysis 2 compared with observation data. The tsunami amplitudes in Region 4 are slightly higher then those in Regions 2 and 3 which have larger grid sizes. Overall, the simulation results agree well with observation data in all regions. Based on the validation of simulation, the numerical scheme and parameters in Analysis 1 are used for the analysis of tsunamis in the South China Sea and the Gulf of Thailand. 6. Analytical results A series of analyses is conducted for all six cases to obtain the arrival time, tsunami amplitude, and current velocity. 6.1. Arrival time Fig. 13 shows time histories of tsunamis at Points 1, 6, and 7 for Case 2. It is seen that the tsunami reaches the southernmost coast of Thailand in about 13 h and takes about 6 h more to reach Bangkok. Since the reverse slip in the Manila trench causes an uplift of the sea bottom in the west and a depression in the east, the leading elevation wave arrives first. Table 5 summarizes arrival time for all six cases. The Mw 9.0 earthquake causes tsunamis to reach Thailand slightly faster than other cases. However, the difference is insignificant. Comparing between Cases 2 and 3; or among Cases 4, 5, and 6, it is obvious that fault planes located southernmost cause tsunamis to arrive earlier. Fig. 14 shows the contour map of tsunami arrival time for the Mw 9.0 earthquake. Since the Gulf Fig. 9. Regions of computation in Analysis 1.
Fig. 10. Regions of computation in Analysis 2.
62
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Fig. 11. Comparison of tsunami amplitudes in Analysis 1 with observation data. (See above-mentioned references for further information).
Fig. 12. Comparison of tsunami amplitudes in Analysis 2 with observation data. (See above-mentioned references for further information).
63
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Water Level (m)
Water Level (m)
0.1
0.1
0.05
0.05
0
0
-0.05
-0.05
-0.1 10
12
14
16
18
-0.1 10
20
12
14
Time (hrs)
16
18
20
Time (hrs)
(a) Point 1
(b) Point 6
Water Level (m) 0.1
0.05
0
-0.05
-0.1 10
12
14
16
18
20
Time (hrs)
(c) Point 7 Fig. 13. Time histories of tsunamis at Points 1, 6, and 7 (Case 2).
Table 5 Arrival time for six cases (h). Mw
Case
Point 1
Point 3
Point 6
Point 7
9.0 8.5
1 2 3 4 5 6
19.2 19.5 19.3 19.8 19.8 19.8
16.6 16.8 16.7 17.2 16.9 16.8
14.8 15.0 14.8 15.3 15.0 14.9
13.5 13.8 13.6 14.0 13.8 13.8
8.0
of Thailand has shallow depths, the propagation speed of the tsunami in the gulf is low. 6.2. Tsunami amplitude The tsunami amplitude is defined as the water level at the time of tsunami occurrence measured relative to the reference water level which is the mean sea level in this study. The effect of earthquake magnitudes on tsunami amplitudes is investigated at various observation points. The tsunami amplitude at Point 7 is presented in Table 6 as an example. It is obvious that the earthquake magnitude contributes significantly to the tsunami amplitude. Note that tsunami amplitudes presented in Table 6 is obtained at a sea depth of 22 m. Fig. 15 shows the shoaling effect as tsunamis approach the coastline near Point 7. The tsunami amplitude increases by 50% from the amplitude at the sea depth of 22 m. The tsunami amplitude at the coastline becomes about 0.65 m for the Mw 9.0 earthquake. Note that the amplitude can be underestimated because the result is obtained from a large grid size. For hazard and risk assessment in coastal zones, inundation
Fig. 14. Tsunami arrival time (h) for the Mw 9 earthquake.
analysis with smaller grid sizes is required. Fig. 16 shows the maximum tsunami amplitudes in the South China Sea for Case 3. The tsunami amplitude is largest in the direction towards the Philippines and Vietnam. The continental shelf on the southeastern part of China causes the amplification of tsunamis in the area. The Gulf
64
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Table 6 Tsunami amplitude at Point 7 for different cases. Mw
Cases
Tsunami amplitude (m)
9.0 8.5
1 2 3 4 5 6
0.42 0.07 0.10 0.02 0.02 0.03
8.0
Coastline
Point 7
Tsunami Amplitude (m)
0.7 case 1 case 3 case 6
0.6 0.5 0.4
location for ships. There are a number of off-shore platforms and gas pipelines in the Gulf of Thailand. In this study, current velocity is computed at the sea depth greater than 15 m with an assumption that the distribution of velocity is uniform throughout the depth. Time histories of current velocity at Point 7 are shown in Fig. 18 for Cases 1, 3, and 6. The positive value is presented because current velocity is computed from vector summation of discharges in two perpendicular directions. As the earthquake magnitude increases from 8.0 to 8.5, the current velocity increases by about 3 and when the earthquake magnitude increases from 8.5 to 9.0, the current velocity increases by about 2. The earthquake magnitude significantly affects current velocity. The distribution of current velocity in the Gulf of Thailand is presented in Fig. 19 for the Mw 9.0 earthquake. The maximum current velocity is 0.27 m/s at a sea depth of 16 m in the southernmost province. In the middle of the gulf, the velocity is about 0.1 m/s. The Gulf of Thailand is slightly affected by currents from tsunamis generated in the Manila trench.
0.3
7. Conclusions
0.2 0.1 0 100.6
100.7
100.7 100.8 Longitude (degree E)
100.8
Fig. 15. Amplification of tsunamis at the coastline near Point 7.
of Thailand is affected by the diffraction of tsunamis around the southern part of Vietnam and Cambodia. The distribution of maximum tsunami amplitudes in the Gulf of Thailand is illustrated in Fig. 17 for earthquake magnitudes of 8.5 and 9.0. It is seen that as the earthquake magnitude increases from 8.5 to 9.0, the tsunami amplitude in the Gulf of Thailand increases by about 3. 6.3. Current velocity Current velocity is an important parameter for evaluating the effect of tsunamis on off-shore structures and determining safe
The effect of tsunamis generated in the Manila trench is investigated in this study. The propagation of tsunamis is analyzed for earthquakes with magnitudes of 8.0, 8.5, and 9.0. The arrival time, tsunami amplitude, and current velocity are computed with an emphasis on the Gulf of Thailand. From the study, the following conclusions can be deduced: (1) Tsunamis reach the southernmost coast of Thailand in about 13 h and take about 6 h more to reach Bangkok. Since the reverse slip in the Manila trench causes an uplift of the sea bottom in the west and a depression in the east, the leading elevation wave arrives first. The effect of earthquake magnitudes on the arrival time is insignificant. (2) The tsunami amplitude is largest in the direction towards the Philippines and Vietnam. The continental shelf on the southeastern part of China causes the amplification of tsunamis in the area. The Gulf of Thailand is affected by the diffraction of tsunamis around the southern part of Vietnam
Fig. 16. Distribution of tsunami amplitudes in the South China Sea for Case 3.
65
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
0.3
0.3
0.25
0.25
0.2
0.2
Current (m/s)
Current (m/s)
Fig. 17. Distribution of maximum tsunami amplitudes.
0.15 0.1 0.05
0.1 0.05 0
0 -0.05 10
0.15
12
14
16
18
-0.05 10
20
12
14
16
Time (hrs)
Time(hrs)
(a) Case 1
(b) Case 3
18
20
0.3
Current (m/s)
0.25 0.2 0.15 0.1 0.05 0 -0.05 10
12
14
16
18
20
Time (hrs)
(c) Case 6 Fig. 18. Time histories of current velocity at Point 7.
and Cambodia. As the earthquake magnitude increases from 8.5 to 9.0, the tsunami amplitude in the Gulf of Thailand increases by about 3. The tsunami amplitude at the southernmost coastline is about 0.65 m for the Mw 9.0 earthquake.
(3) The current velocity in the Gulf of Thailand due to the Mw 9.0 earthquake is generally less than 0.2 m/s. As the earthquake magnitude increases from 8.0 to 8.5, the current velocity increases by about 3. And when the earthquake
66
A. Ruangrassamee, N. Saelem / Journal of Asian Earth Sciences 36 (2009) 56–66
Fig. 19. Current velocity for the Mw 9 earthquake.
magnitude increases from 8.5 to 9.0, the current velocity increases by about 2. The earthquake magnitude significantly affects current velocity. References Advanced National Seismic System (ANSS), accessed in 2006. http:// www.ncedc.org/anss/. American Association of State Highway and Transportation Officials (AASHTO), 2004. AASHTO LRFD Bridge Design Specifications, third ed. USA. Bautista, B.C., Baulista, M.L.P., Oike, K., Wu, F.T., Punongbayan, R.S., 2001. A new insight on the geometry of subducting slabs in Northern Luzon, Philippines. Tectonophysics, 279–310. Gutenberg, B., Richter, C.F., 1954. Seismicity of the Earth and Associated Phenomena. Princeton Univ. Press. Princeton, NJ, USA. Imamura, F., 1992. Review of tsunami simulation with a finite difference method. In: Yeah, H., Lin, P., Synolakis, C. (Eds.), Long-Wave Runup Models. World Scientific, pp. 25–42. Intergovernmental Oceanographic Commission (IOC), 1997. IUGG/IOC TIME Project – Numerical Method of Tsunami Simulation with the Leap-Frog Scheme. UNESCO. Koshimura, S., Takashima, M., 2005. Remote Sensing, GIS, and Modeling Technologies Enhance the Synergic Capability to Comprehend the Impact of Great Tsunami Disaster. Third International Workshop on Remote Sensing for Post Disaster Response, Chiba University, Japan. Kreemer, C., Holt, W.E., Goes, S., Govers, R., 2000. Active deformation in Eastern Indonesia and the Philippines from GPS and seismicity data. Journal of Geophysical Research 105 (B1), 663–680. Lukkunaprasit, P., Ruangrassamee, A., 2008. Building damage in Thailand in the 2004 Indian Ocean tsunami and clues for tsunami-resistant design. The IES Journal Part A: Civil & Structural Engineering 1 (1).
Mansinha, L., Smylie, D., 1971. The displacement fields of inclined faults. Bulletin of the Seismological Society of America 61 (5), 1433–1440. Michel, G.W., Becket, M., Angermann, D., Reigber, C., Reinhart, E., 2000. Crustal motion in E-and SE-Asia from GPS measurements. Earth Planets Space 52, 713– 720. Ministry of Natural Resources and Environment, (2005). Report on Evaluation of Damage due to the Tsunami and Recommendations on Prepardness (in Thai). Papazachos, B.C., Scordilis, E.M., Panagiotopoulos, D.G., Papazachos, C.B., Karakaisis, G.F., 2004. Global relations between seismic fault parameters and moment magnitude of earthquakes. Bulletin of the Geological Society of Greece 36, 1482–1489. Poovarodom, N., Limsamphancharoen, H., 2005. Post-Tsunami Survey along the Northern Coastlines of the Affected Areas in Thailand. In: Proceedings of the Tenth National Convention on Civil Engineering, Engineering Institute of Thailand, Thailand. Shuto, N., Suzuki, T., Hasegawa, K., Inagaki, K., 1986. A study of numerical technique on tsunami propagation and run-up. Science of Tsunami Hazards 4 (2), 111– 124. Smith, W.H.F., Sandwell, D.T., 1997. Global sea floor topography from satellite altimetry and ship depth soundings. Science 277, 1956–1962. Technical Council on Lifeline Earthquake Engineering (TCLEE), 2005. Sumatra– Andaman Islands Earthquake and Tsunami of 26 December, 2004 – Lifeline Performance, ASCE-TLCEE Monograph No. 29, American Society of Civil Engineers. Torregosa, R.F., Sugto, M., Nojima, N., 2001. Strong motion simulation for the Philippines based on seismic hazard assessment. Journal of Natural Disaster Science 23 (1), 35–51. Wells, D.L., Coppersmith, K.J., 1994. New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement. Bulletin of the Seismological Society of America 84 (4), 974–1002. Zhu, W., Wang, X., Cheng, Z., Xiong, Y., Zhang, Q., Ye, S., Ma, Z., Chen, J., Hsu, H., Wei, Z., Lai, X., Liu, J., Jin, B., Ren, J., Wang, J., 2000. Crustal motion of Chinese mainland monitored by GPS. Science in China 43, 394–400.