Back to Taiwan Main PagePresent Seismicity
| Figure 17: 3-D plot of seismicity below Taiwan. After Academia Sinica (1997). |
Two striking features show from seismic maps of Taiwan. First, the deep seismicity outlines the subducting slabs to a depth of 300 km (figure 17). Second is the seismic gap that exists under the Central Ranges in the central part of the island (Lin, 1998; Lin, 2000) (figure 18). Lin (1998) conjectured that this gap must either signify no strain in this area at all (i.e. it is a rigid block) or it is an area that is experiencing aseismic deformation. He also noted that this zone has sub-vertical boundaries (see figure 18; C-C').
Crustal Thickness and Structure
| Figure 18: Epicenters with magnitudes greater than 3.0 and three hypocenter projections recorded by the Central Weather Bureau Seismic Network (CWBSN) in the Taiwan area from 1992-1996. Dotted lines in the A-A' and B-B' cross-sections mark the Benioff zone in the northeast and the south of Taiwan, respectively. Dashed lines in section C-C' delineate the boundary between seismic and aseismic zones in central Taiwan. The sub-vertical lines in this profile emphasizes the orientation of the boundaries of the aseismic gap in the middle. After Lin (2000). |
Receiver function analysis of Taiwan is somewhat suspect, though it generally agrees with tomographic studies of the island (Tomfohrde and Nowack, 2000). Tomfohrde and Nowack's (2000) study only utilized teleseismic events from two directions with a 90 degree separation, and neither their study or Ma and Song's (1997) study accounted for the possible effects of a dipping layer (the subducting slab) or anisotropy (which is documented by (Rau et al., 2000)). Values for Moho depth are between 20-32 km for the Coastal Plains, between 25 and 35 km for the Western Foothills and 34-43 km for the Coastal Range (Ma and Song, 1997; Rau and Wu, 1995). Ma and Song (1997) used a two-layer model with a crustal P wave velocity of 6.35 km/s and a Moho velocity of 8.52 km/s, which also roughly agrees with tomographic studies (Rau and Wu, 1995). Tomfohrde and Nowack note that Moho depths from these studies largely in contradict critical wedge theories that would only place the Moho under the Central Range at 20-25 km.
Tomographic images of the Taiwan lithosphere also show quite a different picture from the critical wedge model. Multiple topographic sections are shown in figure 19. Some simple observations from these sections are a low velocity zone under the Ilan Plane in northern Taiwan which is interpreted as due to opening of the Okinawa Trough (Rau and Wu, 1995) (profile not shown). A low velocity upper crust thickens and then thins from north to south (profile not shown). This low velocity upper crust also thickens from west to east (figure 19; C-C', D-D'). However, a high velocity zone exists under the Central Ranges (figure 19; A-A') that is not associated with volcanic arc rocks, which are clearly visible in the profiles as a distinct body in the eastern part of the high-resolution sections. This is clearly in disagreement with the classic model, which would have the thickest pile of sediments and upper crustal rocks under the Central Ranges. Rau and Wu interpret this as an up-arching of basement and lower crustal materials.
| Figure 19: Tomographic cross sections of Taiwan. Profiles A-A' and B-B' are of higher resolution but narrower scope. See text for discussion. This figure compiled from the following sources: Academia Sinica (2001); Lin (1998); Rau and Wu (1995). |
| || It is interesting to note that the principal stress field under the Eastern Central Ranges is vertical, as shown by focal mechanisms for this area (Lin, 1998) (figure 20). This vertical stress field is not just a surficial effect of uplift, but extends down to the brittle-ductile transition (Crespi et al., 1996). |
Figure 20 (left): Focal mechanisms located in the Eastern Central Ranges. After Lin (1998).
| || Anisotropy is shown to exist in the lithosphere under Taiwan (Rau et al., 2000) (figure 21). It spans the depths of 25-230 km and its orientation closely follows the trend of structure at the surface (figure 22). Though the depth of the anisotropy is less than clear, one can deduce, simply from the trend of the anisotropy, that the lithosphere is reacting to tectonic stresses coherently with the upper crust. |
Figure 21 (left): Shear wave splitting parameters for both S and ScS phases plotted as vectors at station sites. Bars represent the fast direction. Delay time is proportional to bar length, as shown in the scale. Scale at the bottom is for elevation. After Rau et al. (2000).
Offshore Southern Taiwan Seismicity and Bathymetry
| Figure 22: (a) Location map with markers A, B, C, and D for orientation. (b) Splitting parameters plotted as a function of position along strike of the island. Open and dark circles represent S and ScS measurements, respectively. Error bars are also given. After Rau et al. (2000). || |
| || As noted by Chang et al. (2000), since present deformation is highest off the coast of Taiwan to the south, morphology and seismicity in this area should be considered along with its relation to onshore structures. The accretionary prism at the Manila trench widens westward and involves the Hengchun Ridge and Kaoping Slope (Kao et al., 2000) (HR and KC in figure 4). Since the Hengchun Ridge appears to be a southern continuation of the Central Ranges, it follows that the Central Ranges and every thing west thereof could be interpreted as the sole accretionary wedge of the subducting Eurasian plate. |
Another interesting is between the Hengchun Ridge and the Luzon arc. Topographic features here include the Southern Longitudinal Trough, the Huatang Ridge, and the Taitung Trough (SLT, HR, and TT, respectively in figure 4). The Huatang Ridge is underlain by an east-verging thrust that connects the Coastal Range to the north with the Hengchun ridge to the south (Kao et al., 2000) (figure 4). This is interpreted as the active collapsing zone of the forearc basin as depicted in figure 23. It also happens to be the zone of highest seismicity in this part of the orogen (Tang and Chemenda, 2000) (figure 24).
Figure 23 (left): Schematic diagram showing the current tectonic configuration in the Taiwan-northern Luzon arc region. Note the interpreted position of the closure of the forearc basin coincides with the position of the Huatang ridge in figure 4. After Kao et al. (2000).
| Figure 24: Seismicity across southern and offshore Taiwan. (a) Seismicity map shows earthquakes of magnitude greater than 3.0 between 0 and 30 km depth located by the TTSN and CWBSN for the period of 1974-1997. Cross sections are shown in (b) and (c). After Tang and Chemenda (2000). || |
An interpreted seismic section from east to west across this part of the orogen further substantiates this equivalence-of-structures interpretation. A comparison of the seismic section to the structural cross section of Lee et al. (1997) shows a striking similarity of structures (figure 25).
| Figure 25: Interpreted seismic profile across offshore southern Taiwan compared to the interpreted structural section across central Taiwan. Seismic section after Tand and Chemenda (2000) and structural section after Lee et al. (1997). WF = Western Foothills; HR = Hsuehshan Range; CR = Central Range. Note the similarity of structures even though scale, age, and location are very different. |
| || The most striking thing about the Bouguer gravity anomaly map for Taiwan (figure 26) is that the gravity anomaly does not in anyway follow the topography. This indicates that Taiwan is not compensated by normal isostasy. Figure 27 further illustrates that an isostatically compensated model using Taiwan's topography in no way mimics Taiwan's observed gravity anomaly. The modeled density structure that does fit the gravity actually matches the tomographic results above very well (figure 28). Note that the Moho depths agree with the seismically derived Moho depths of Rau and Wu (1995), and that the upwarping of the Moho under the Eastern Central Ranges and Longitudinal Valley roughly matches with the high velocity structure under the Central Ranges. |
Figure 26: Bouguer gravity contours and topography of Taiwan. Gravity map after Yen et al. (1998) and topography after Academia Sinica (2001). Warm colors on the topographic map indicate high elevations. Note that Bouguer gravity lows do not match with topographic highs in Taiwan. Location of modeled section in figure 28 is shown labeled C-C'.
| Figure 27: The gravity effect of an isostatically compensated crust according to the Airy model. Calculated anomaly does not match the observed anomaly. After Yen et al. (1998). || |
| || Figure 28: Two dimensional gravity model for the profile C-C' in figure 26. CP = Coastal Plain; WF = Western Foothills; HR = Hsuehshan Range; BR = Backbone Range (Central Range); LV = Longitudinal Valley; CR = Coastal Range. See text for interpretation. After Yen et al. (1998). |
Taiwan Thermal Structure
| Figure 29: Gravity anomaly produced from the critical wedge model of (Suppe, 1981). After Yen et al. (1998). |
| || Surficial heat flow of the Taiwan island is quite high and roughly mimics topography (Lin, 2000) (figure 30). Lin (2000) uses the coincidence of a thermal high and low seismicity under the Central Ranges to build a thermal model of orogenesis in Taiwan. The model that fits the observed heat flow and seismicity involves a late stage of rapid exhumation (Lin, 1998; Lin, 2000) (figure 31). With this model of uplift, temperatures of 250-450 degrees C and 300-400 MPa is demonstrated under the Central Ranges, and seismicity is limited by the 350 degree isotherm. |
Figure 30 (left): Surface heat flow map of Taiwan. Circles are the location of measurements and values marked at the contours represent the heat flow in mW/m2 units. After Lin (2000).
| Figure 31: (a) Comparison of the calculated and observed heat flows across Taiwan. (b) The thermal structure. White lines are at intervals of 100 degrees C; the dotted line along the 350 degree isotherm delineates the brittle-ductile transition boundary. (c) Seismicity as in figure 18, for comparison. After Lin (2000). || |
A possible model for this rapid, late-stage uplift under the Central Ranges is the uncoupling of subducted continental crust from the down-going oceanic slab (figure 9). The near-vertical boundaries on the seismic gap seem to call for a fault-bounded model for uplift rather than a wedge-dynamic type uplift (Lin, 1998). Additionally, thermal models involving a critical wedge also do not show the sharp peak in heat flow under the Central Ranges (figure 32).
| Figure 32: Thermal model for the critical wedge model. Cross section extends across the entire island. Note that the isotherms' intersection with the surface does not show a sharp peak under the Central Ranges-it smoothly increases all the way to the eastern side of the island. After Barr and Dahlen (1989). |
The late exhumation by crustal decoupling should be easy to test. Late stage uplift of the Central Range as compared to surrounding ranges must be demonstrated. Stress fields should also differ sharply between the Central Ranges and surrounding areas because of the decoupling of this range along near-vertical faults.
Fission-track and K-Ar dating show uplift in the Central Ranges over geologic time at 7.5-15 mm/yr. This is a much faster rate than for the Coastal Ranges for the same period (Lin, 2000). Recent uplift as measured with leveling data shows uplift at a rate of up to 36-42 mm/yr (Lin, 1998). It is possible that this great increase is showing an acceleration in uplift over the last million years or so. It is also interesting to note the correlation of this area of rapid uplift with focal mechanisms which show a vertical principal stress axis in the Central Ranges down to the brittle-ductile zone (Crespi et al., 1996; Lin, 1998). The rapid uplift could be causing this extensional regime.
Orogenic Systems Home | U of A Geosciences | Megan Anderson's Page
Last edited 5/3/2001 by Megan Anderson email@example.com