Inversion of strong motion seismograms for the source process of the Naganoken-Seibu earthquake of 1984

Abstract

The source process of the Noganoken-Seibu earthquake of 1984 ( M = 6.8) has been investigated in detail by using strong motion seismograms recorded at stations within 100 km of the epicenter. A new inversion method was applied to analyze the strong motion seismograms. We parameterized the problem by using a model in which the fault was composed of a certain number of subfaults which started rupturing with arbitrary magnitude at arbitrary time, and we have successfully obtained the dislocation distribution and the characteristics of rupture propagation on the fault. The rupture starts at the shallowest point on the central part of the fault and then propagates over the whole source region within about 3 s. The total seismic moment estimated from the final solutions is 3 × 10 25 dyne · cm. The dislocation is inhomogeneously distributed on the fault plane and reaches a maximum of about 1.5–2 m. The largest dislocation occurs in the near surface region. On the deeper part of the fault, the dislocation seems to be small. There were few aftershocks at the center and on the shallow part of the fault; most of the large aftershocks ( M > 4.0) were located on the deeper part of the fault. These results suggest that the aftershock activity was highest on the unruptured regions of the fault where stresses were concentrated. From the distribution of dislocations on the fault, the size of the source area is estimated to be approximately 12 km (length) × 6 km (depth). The main shock ground motions, having periods longer than l s were estimated at sites where no recordings exist by using the final source model. At the summit of Mt. Ontake, the maximum displacement and velocity are 15 and 12 cm/s respectively. The maximum short-period acceleration at the period of 0.1 s is about 300 Gal at the summit of Mt. Ontake.