Restoration by a Simple Back-Stripping Method

As mentioned above, the basin-wide sedimentation process in late Quaternary was governed by sea level fluctuations along with changes of drainage system, so did the accumulation of Sungshan and Jingmei Formations in the Wuku Profile since local influences other than tectonics are not observed in sediment records. When sea level reached similar elevation to that of Taipei Basin plain (resembling the present configuration) since the recent major marine incursion in the basin at 10-9 ka (Teng et al., 2000b), sediments piled up across the entire profile, including the footwall, hanging wall and extensional fault block areas. Given the high rate of erosion and hence rapid sediment accumulation of the Taiwan Island in the present day (Dadson et al., 2003), the room of sedimentation created by rising eustacy is considered to be filled up contemporaneously, therefore producing flat topography similar to the modern one.

Under this assumption, the topography at beginning of the deposition for the C1 unit of the Sungshan Formation would be approximately flat. Because the C1 unit of SCF-1 lies in the footwall area, hence it would not involve significant vertical motion of the faulting. Referring to the growth normal faulting scheme, the bases of C1 units in the hanging wall side would be expected to be consequently downward displaced by dip-slips of the Shanchiao Fault which provided additional sedimentation rooms resulting in lateral thickening of the units. The elevation differences between the base of C1 unit at SCF-2 and SCF-1 (7.8 m), and SCF-2 and WK-1 (11.3 m) thus are interpreted to represent cumulative vertical displacements on the branch and main faults, respectively, from 8.4 ka till present (Fig. 3-6a). Furthermore, we can observe that the height of sea level e (around -16 m) at the onset of the C1 (~ 8.4 ka) is close to the depth of the base of the C1 unit (-14.5 m) in the footwall area (SCF-1).

Back to 8.4 ka, because that the top of the basin deposits (C2 unit) in the footwall appears to be near the sea level (level e), therefore the height of the top of the estuarine deposits of the C2 unit across the profile would be considered to be very much flat. Under this assumption, the fault zone stratigraphic configuration by then can be approximated by removing the C1 unit above and re-leveling the C2 unit tops in WK-1 (hanging-wall region) and SCF-2 (extension fault block) to the one at SCF-1 (footwall region) as restoration by ‘back-stripping’ (Fig. 3-6b). We calculate the remaining vertical differences of the depths of C2 unit between three holes as mentioned above. We then obtain 3.3 m and 7.4 m, which hence denote to be cumulative vertical offsets on the branch and main faults respectively during the 9 ka to 8.4 ka period (Fig. 3-6b). These relative large amounts of vertical offset at a short time span of 600 years appear to be consistent with two possible earthquake events

49

occurred during this period inferred by previous study (Huang et al., 2007). We will discuss it in more details in the later section.

The above approach became much difficult to apply to deeper sediments of the Sungshan Formation because the 10 ka datum is obscure at SCF-2 and WK-1, and synchroneity of sedimentation along the Wuku profile was absent during 23 ka to 10 ka. However, the Jingmei Formation alluvial fan deposits of 25 – 23 ka resulted in a relatively flat topography at the end of the river-capture event. Provided the comparable thicknesses of the Jingmei conglomerates (C4 unit) at SCF-2 and WK-1, the ancient topography of the Wuku profile is considered to be roughly even at around 23 ka, therefore permitting the ‘back-stripping’ of C3 unit and evaluation of fault vertical displacements during the deposition of the C3 unit. Under this assumption, we then obtain that the Shanchiao branch fault slipped 13.6 m vertically and the main fault 32.2 m during ~23 ka to 9 ka according to constraints from top horizons of the C3 unit and the Jingmei Formation (Fig. 3-6c).

Details of sediment accumulation and fault throws between ~23 ka to 9 ka cannot be accurately resolved, while the stacking pattern of the deposits does show westward onlapping of layers which is closely tied to tectonic modification on local geomorphology and base level changes controlled by eustacy. Regional sedimentation rate (modulated by tectonic movements) was zero at the LGM as the sea level (level a) was low at about -140 m compared to the present day level. Since LGM the sedimentation rate is expected to be rapidly increasing in pace with the rising sea level, and again dropped to trivial numbers when eustacy stabilized to elevations similar to that of today at ~6 ka (Fig. 3-5).

In summary, the total vertical offset across the Shanchiao fault zones since 23 ka can be yielded by summing the above results during different periods, as shown in Fig.

3-6. We thus obtain a total vertical offset of 75.6 m, with 24.7 m and 50.9 m for the branch fault and the main fault, respectively, since the LGM of 23 ka. It can translate to an average vertical fault slip rate of 3.3 mm/yr across the Shanchiao fault zone in Wuku profile, with 1.1 mm/yr for the branch fault and 2.2 mm/yr for the main fault, since 23 ka.

50

Fig. 3-6. Back-stripping method and reconstruction of cumulative deformation on the Shanchiao Fault since the LGM of about 23 ka. Three periods of 8.4 ka – present (a), 9 – 8.4 ka (b), and 23 - 9 ka (c), which correspond to depositional ages of Units C1, C2, and C3, respectively, are adopted for explaining the evolution of the growth faulting of the Shanchiao Fault. Numbers in black color: thickness of stratigraphic units. Numbers in red color: vertical differences of thickness for each stratigfaphic unit between drilling holes. Levels a to e are paleo-sea levels trough time, which are denoted in Fig. 3-5.

51