Résumé
This paper presents the results of a detailed survey combining Seabeam mapping, gravity and geomagnetic measurements as well as single-channel seismic reflection observations in the Japan Trench and the juncture with the Kuril Trench during the French-Japanese Kaiko project (northern sector of the Leg 3) on the R/V "Jean Charcot". The main data acquired during the cruise, such as the Seabeam maps, magnetic anomalies pattern, and preliminary interpretations are discussed. These new data cover an area of 18,000 km 2 and provide for the first time a detailed three-dimensional image of the Japan Trench. Combined with the previous results, the data indicate new structural interpretations. A comparative study of Seabeam morphology, single-channel and reprocessed multichannel records lead to the conclusion that along the northern Japan Trench there is little evidence of accretion but, instead, a tectonic erosion of the overriding plate. The tectonic pattern on the oceanic side of the trench is controlled by the creation of new normal faults parallel to the Japan Trench axis, which is a direct consequence of the downward flexure of the Pacific plate. In addition to these new faults, ancient normal faults trending parallel to the N65 ° oceanic magnetic anomalies and oblique to the Japan trench axis are reactivated, so that two directions of normal faulting are observed seaward of the Japan Trench. Only one direction of faulting is observed seaward of the Kuril Trench because of the parallelism between the trench axis and the magnetic anomalies. The convergent front of the Kuril Trench is off~t left-laterally by 20 km relative to those of the Japan Trench. This transform fault and the lower slope of the southernmost Kuril Trench are represented by very steep scarps more than 2 km high. Slightly south of the juncture, the Erimo Seamount riding on the Pacific plate, is now entering the subduction zone. It has been preceded by at least another seamount as revealed by magnetic anomalies across the landward slope of the trench. Deeper future studies will be necessary to discriminate between the two following hypothesis about the origin of the curvature between both trenches: Is it due to the collision of an already subducted chain of seamounts? or does it correspond to one of the failure lines of the America/Eurasia plate boundary?