The apparent incompatibility of the classical "bar door" opening model of the Japan Sea based on paleomagnetic studies and the pull-apart geometry based on the observation of shear zones in and around Japan is symptomatic of a lack of understanding of the tectonic history of the Japan Sea. After a critical review of paleomagnetic data and a discussion on the possible error bars we present information concerning the internal strain of the Japan arc during the opening. We show that it is possible to integrate both sets of data in a single model, provided that some of the paleomagnetic rotations are due to distributed deformation of SW and NE Japan and not to a rigid body rotation of 600 km long blocks. Rotations occur at all scales and the 50 ø of clockwise rotation of southwest Japan is the sum of 30 ø of rotation of southwest Japan as a whole and of 20 ø due to internal deformation. The amount of paleomagnetic rotation is then compatible with the kinematics based on the internal structure of the Japan Sea and the geometry of major dextral shear zones. The rigid body rotation of southwest Japan is accommodated by extension and left-lateral motion along the Median Tectonic Line which is a second order fault between the two major dextral shear zones that bound the Japan Sea to the east and to the west. The problem of the apparent instantaneous rotation of southwest Japan as opposed to the progressive opening of the Japan Sea probably results from the way paleomagnetic data and radiometric dates are averaged.
PROBLEM SETTINGThe past ten years have seen the development of a debate on the opening mechanism of the Japan Sea ( Figure 1). An important breakthrough was made by Otofuji et al.short period, less than 2 m.y. long, about 15 Ma. The model is that of a double rotation about two nearby poles and a double fan-shaped opening of the Japan Sea. The rigid behavior of the two continental blocks, each of them being more than 500 km long, implies that oceanic crust formation in the Japan Sea was contemporaneous with the rotation. A rapidly opening model was then postulated in concert with rapid rates of motion of the rotating blocks.Based on different data, Lallemand and Jolivet [ 1985] postulated that major dextral strike slip faults guided the opening of the Japan Sea in a pull-apart manner, in opposition with the fan-shaped kinematics of Otofuji et al.Later Jolivet et al. [1989, 1991] and Jolivet and Tamaki [1992] introduced block rotations in the pull-apart model, describing the whole region as a wide dextral shear zone (Figure 2). 356 PALEOMAGNETIC ROTATIONS AND THE JAPAN SEA OPENING JOLIYET ET AL. 357 JOLIVET ET AL. 359