Erosion-Induced Backstepping and Reactivation of the Chelungpu Thrust: Implications for patterns of modern strain release in west-central Taiwan.

Restoration of fault-related folds and the geomorphology of the Taiwan thrust belt suggests the Chelungpu thrust was reactivated in late Quaternary time, perhaps as recently as ca. 145 ka. Evidence for reactivation includes: the morphology of young drainage channel networks uplifted between the Chelungpu and Shuangtung thrusts; folding of late Quaternary river terraces, geodetic data and erosion of a large fault-related fold in the hangingwall of the thrust. Previous work argued that reactivation of the Chelungpu thrust is caused by indentation of the thrust belt against the Peikang basement high, which is located in the foreland of the orogen southwest of the Pakuashan anticline. We argue, however, that reactivation of the thrust is due to erosional removal of material in its hangingwall where regional topography is 500-2000 m lower in elevation than further north and south. High sediment yields of suspended load in rivers incised in the fold belt hindward of the Chelungpu thrust support the idea of more rapid removal of material from the central highlands to the foreland in this region. To provide an initial test of this hypothesis, we undertook simple calculations of the orogen's response to erosion using critical taper theory based on the previous work of Dahlen. We calculated the change in width of the orogen by increasing the rate of erosion by 1.0 mm/yr to 6.5 mm/yr from an ambient rate of 5.5 mm/yr and used a 7 km thickness for the foreland basin and a shortening rate of 70 mm/yr. Our results suggest the orogen would step hindward ~14 km with an increase in erosion of 1 mm/yr. The distance between the Pakuashan and Tatuashan anticlines that mark the leading edge of the belt and the surface trace of the reactivated Chelungpu thrust is about 14 to 25 km. Differences between our results and the measured distance between the Chelungpu thrust and the frontal anticlines may be due to an underestimate of the erosion rate, or by our simple model of wedge geometry for this part of the Taiwanese mountain belt.