Abstract
Events detected by the KIER microearthquake network operated in the Southern Part of Korea for 265 days in 1982~1984 were reviewed, and some of them were identified to be a dynamite explosion from several construction sites. The purpose of the present work is to determine the crustal structure of the Southern Korea using the time-destance data obtained from such explosion seismic records. The time·distance data can be well explained by a crustal model composed of four horizontal layers of which thickness, p and s-wave velocity ($V_p$ and $V_s$) are characterized as follows. 1st layer (surface) ; 0~2km, $V_p=5.5km/sec$, $V_s=3.3km/sec$ 2nd layer (upper crust) ; 2~15km, $V_p=6.0km/sec$, $V_s=3.5km/sec$ 3rd layer (lower crust) ; 15~29km, $V_p=6.6km/sec$, $V_s=3.7km/sec$ 4th layer (upper mantle) ; 29km~ , $V_p=7.7km/sec$, $V_s=4.3km/sec$ The relatively shallow crust·mantle boundary and low $P_n$ velocity compared with the mean values for stable intraplate region are noteworthy. Supposedely, it is responsible for the high heat flow in the South-eastern Korea or an anomalous subterranean mantle. The mean $V_p/V_s$ ratio calculated from the relation between p-wave arrival and s-p arrival times appears to be 1.735 which is nearly equivalent to the elastic medium of ${\lambda}={\mu}$. However, the ratio tends to be slightly larger with the depth. The ratio is rather high compared with that of the adjacent Japanese Island, and the fact suggests that the underlying crust and upper mantle in this region are more ductile and hence the earthquake occurrences are apt to be interrupted. As an alternative curstal model, a seismic velocity structure in which velocities are successively increased with the depth is also proposed by the inversion of the time·distance data. With the velocity profile, it is possible to calculate a travel time table which is appropriate to determine the earthquake parameters for the local events.