Kazuya Iida, Hikaru Iwamori, Yuji Orihashi, Taeho Park, Yong Joo Jwa, Sung Tack Kwon, Tohru Danhara, Hideki Iwano
Island Arc 24(2) 205-220 2015年6月1日 査読有り筆頭著者
The spatiotemporal distribution of Cretaceous-Paleogene granitic rocks in southwestern Japan is investigated to understand the origin of the granitic batholith belt and to reconstruct the tectonic setting of emplacement. New U-Pb zircon ages for 92 samples collected from a region measuring 50km (E-W) by 200km (N-S) reveals a stepwise northward younging of granitic rocks aged between 95 and 30Ma with an age-data gap between 60 and 48Ma. Based on the spatiotemporal distribution of granite ages, we examine two plausible models to explain the pattern of magmatic activity: (i) subduction of a segmented spreading ridge and subsequent slab melting (ridge-subduction model), and (ii) subduction with a temporally variable subduction angle and corresponding spatial distribution of normal arc magmatism (subduction angle model). We optimize the model parameters to fit the observed magmatism in time and space, and compare the best-fit models. As to ridge subduction model, the best-fit solution indicates that the spreading ridge started to subduct at approximately 100Ma, and involved a 45-km-wide section of the ridge segment, a subduction obliquity of 30°, and a slow migration velocity (∼1.6cm/y) of the ridge. These values are within the ranges of velocities observed for present-day ridge subduction at the Chile trench. On the other hand, the best-fit solution of subduction angle model indicates that the subduction angle decreases stepwise from 37° at 95Ma, 32° at 87Ma, 22° at 72Ma, to 20° at 65Ma, shifting magmatic region towards the continental side. These results and comparison, together with constraints on the geometry of the tectonic setting provided by previous studies, suggest that the ridge subduction model better explains the limited duration of magmatism, although both models broadly fit the data and cannot be ruled out.