Ground-motion prediction equations (GMPEs) for the horizontal and vertical spectral accelerations (SAs) from the offshore area off the Sagami Bay is presented in this article. To compare the ground motions on the seafloor with those at onshore sites, an onshore GMPE was derived for the onshore stations adjacent to the offshore sites and including those in the Kanto basin sites. The offshore dataset includes 738 three-component records, and the onshore dataset includes 3775 records; both datasets are derived from the same set of 233 earthquakes selected for this study. The local site conditions of the offshore sites are investigated by implementing the mean horizontal-to-vertical response spectral ratios because soil condition data are unavailable. Individual site correction terms are used for the offshore sites due to differences among the site conditions, whereas site-class terms are used for the onshore sites based on the travel-time-averaged shear-wave velocity to 30 m (VS30). A comparison between the offshore and onshore models shows that the horizontal SAs for the whole periods and vertical SAs for moderate and long periods of the offshore motions are considerably larger than those of the onshore motions due to the deep deposition layer comprising soft sediment blanketing the seafloor; the vertical SAs of the offshore motions are slightly smaller than those of the onshore motions at short spectral periods. The inconsistency of the vertical motions at short periods may be due to the combined effect of the deposition layer and water layer. However, the vertical site amplification for short periods is independent of the water depth, and the reason of high-frequency deficiency in vertical components need further investigation.

No abstract

Damping modification factors (DMFs) are used to scale the 5%-damped response spectra for the seismic design of structures with different damping ratios. However, there is no DMF model for offshore engineering response spectra at present. Using 892 offshore and 4033 onshore ground-motion records from the Kyoshin net, the difference between offshore and onshore DMFs are revealed by DMF ratios and t tests and whether site conditions need to be incorporated in offshore DMF models are determined. The between-event residual distribution shows that DMFs depend on the focal depth. The results show that the DMFs from the offshore and onshore ground motions differ significantly for most spectral periods, and the impact of the site condition on offshore DMF is very slight. The offshore DMF models for the horizontal and vertical acceleration spectra from the Japan Sagami Bay region are presented from spectra of 34 spectral periods and 10 damping ratios, in which only spectral periods and damping ratios are considered as variables. The uncertainty in the DMF model mainly depends on the path and site effects, and the standard deviations of DMFs for the vertical component are slightly less than those for the horizontal component. All standard deviations are modeled by a set of quadratic functions related to damping ratios. A comparison between the offshore DMF model in this study and those derived from onshore ground motions indicates that the differences are significant. The spectra from the Hu et al. (2020) study for offshore ground motions adjusted by the DMF models are used as our example applications.

A prediction model for offshore vertical-to-horizontal (V/H) spectral ratios of peak ground acceleration (PGA) and 5%-damped elastic response spectra for periods ranging from 0.01 to 10 s for an offshore area off Sagami Bay was developed. To compare differences between offshore and onshore ground motions in the V/H spectral ratios, an onshore V/H prediction model was derived for onshore stations adjacent to the studied offshore sites. The offshore dataset includes 738 three-component records, and the onshore dataset includes 2219 records; both datasets are derived from the same set of 233 earthquakes selected for this study. The moment magnitude, hypocentral distance, focal depth, tectonic source type, and individual site correction term are used as independent variables in the V/H models. A comparison drawn between the offshore and onshore models shows that the V/H spectral ratios of offshore ground motions are obviously smaller than those of the onshore motions over short periods (< 1.0 s) but are comparable for periods of longer than 1.0 s. Water layer reduces vertical ground motions, especially over short periods, and as the periods increase, the effect of water layer decreases. The effect of the moment magnitude on offshore V/H ratios is stronger than that of onshore ratios, especially for long periods; the effect of hypocentral distance and focal depth is considerable for long periods for offshore V/H ratios, and the V/H ratios between offshore sites are very different due to complex local site geologies beneath the offshore stations.

No abstract