The topography of the 410 and 660 km seismic upper mantle discontinuities beneath the Korean Peninsula and southwestern Japan was determined using teleseismic receiver functions. The P receiver functions were migrated from delayed times to corresponding piercing (conversion) points of P-to-S converted phases, using one-dimensional (1-D) and three- dimensional (3-D) models. The receiver functions were then stacked using common conversion point (CCP) techniques to enhance signal-to-noise ratios to thereby reduce uncertainty (noise). The 410 and 660 km discontinuities were clearly imaged as positively valued amplitude peaks of the CCP stacked receiver functions in the study area. The topographic variations were roughly consistent with the low temperature of the subducting Pacific Plate. However, the complex structure of the subducting Pacific Plate produced distinct changes in the upper mantle discontinuities, which cannot be explained by the temperature variations alone. The depression of the 410 km discontinuity, observed in a wide region extending from the Korean Peninsula to Kyushu Island, may be related to trench rollback history. Furthermore, the topography of the 660 km discontinuity varies significantly with latitude. At latitudes higher than 38° N, its depth remains unchanged despite the presence of the stagnant slab, while significant depression has been observed at latitudes below 36° N. This may have been caused by differences in the angles of subduction of the Japan slab and the Izu-Bonin slab. However, the heterogeneity of the water content of the slabs may also have contributed to this topographical difference.
The topography of the lithosphere-asthenosphere boundary (LAB) beneath the Korean Peninsula was determined using teleseismic S-receiver functions. The receiver functions were migrated from the delayed time to the corresponding conversion points of S-to-p phases using the IASP91 model. Signals representing the low velocity layers beneath the Korean Peninsula were imaged by applying the CCP stacking method. The LAB beneath the Korean Peninsula was imaged at a depth range of 60 - 100 km. This shallow LAB, considering the Precambrian bedrock, implied the lithosphere thinning accompanied by the coeval subduction of the paleo-Pacific plate in the Jurassic. The distinct difference in the LAB structure across the tectonic boundary between the Gyeonggi Massif (GM) and the Okcheon Fold Belt (OFB) was recognized. The LAB at the southern part was shallower than that at the northern part. Furthermore, an additional negative discontinuity was observed below the LAB at the southern part. The structure resembles that produced by the continental collision as reported beneath Tibet. However, the continental collision had finished in the Early Triassic in the Korean Peninsula before the lithosphere thinning, which would have destroyed the lower lithosphere in the Jurassic. Therefore, the current LAB structure could not be a remnant of the continental collision. The subduction of the oceanic ridge between the Izanagi Plate and the Pacific Plate at ~60 Ma might have induced lithosphere delamination in the southern part. However, the corresponding evidence of magmatism, supporting the additional delamination in the southern part, is not reported. Unfortunately, none of existing tectonic models of the Korean Peninsula properly explain the LAB structure beneath the Korean Peninsula. Nevertheless, the presented LAB structure beneath the Korean Peninsula could contribute to establishing and understanding the tectonic history of the Korean Peninsula.