Cyclic upper crustal collision and relaxation at the junction of collision to subduction in northern Hualien, Taiwan

Continuous GPS (CGPS), borehole strainmeter, and seismic data from northern Hualien in Taiwan reveal that the convergence between the Eurasian plate and the Philippine Sea plate produces abundant background seismicity and cyclically leading to more seismicity containing larger events with opposite direction of the plate convergence. We discover unusually large horizontal displacement variations of annual main east-westward cycle with peak-to-peak more than 20 mm from 15 CGPS time series only in northern Hualien. Most stations move eastward during May to October, and move westward during November to April every year. Several studies have shown that CGPS vertical component is typically influenced by seasonal changes of environment factors, and CGPS horizontal components are one-third to one-tenth those of the vertical displacement. We test the perception with newly acquired geodetic and seismological data and found it inadequate for the case of surface displacements and borehole strain in northern Hualien. The evidence for contribution of environment factors in the ground displacements is very weak in this region. We thus calculated the average velocity of northwestern and reversal southeastern motion; the observed displacements are nearly parallel to the direction of convergence of the Eurasian plate and the Philippine Sea plate. The CGPS areal strain also implies high correlation between occurrences of local earthquakes Mw > 5.0 and the strain transition of compression to extension, also the borehole strain is larger than surface areal strain. All observations clearly demonstrate a cycle of the converging and relaxation behaviors of the Eurasian plate and the Philippine Sea plate within the junction of collision to subduction in northern Hualien. Moreover, the decoupling seems to exist in the crust because lack of normal faulting along the plate boundary, thus the cyclic motion might be limited in the upper crust. As an alternative explanation, the cyclic movement linking to upper crustal deformation is proposed for the junction area of subduction and collision, instead of the common perception for influence of environmental factors.