Ground-motion simulations within ShakeMap methodology: application to the 2008 Iwate-Miyagi Nairiku (Japan) and 1980 Irpinia (Italy) earthquakes

ShakeMap package uses empirical ground motion prediction equations (GMPEs) to estimate the ground motion where recorded data are not available. The GMPEs, however, account only for average characteristics of source and wave propagation processes and the ground motion estimate can fail in the near-source area when few stations are available.In this study, we investigate the performance of ShakeMap in the near-fault area when source effects are included at different levels of complexity. We focus on the 2008, Mw 7.0, Iwate-Miyagi Nairiku (Japan) earthquake because of the large amount of recording stations which contribute to the definition of a reference shakemap. After shutting off some stations from the original data set, we evaluate the resulting shakemaps bias as if the earthquake was recorded at a smaller number of receivers. We then compute the shakemaps replacing the missing records with synthetic seismograms from a hybrid deterministic-stochastic method for extended fault. We suppose an increasing knowledge of seismic source approximation and of the slip history on the fault, obtained both from the expeditious inversion of teleseismic data and, afterwards, from strong-motion data inversion. In particular, a non-linear kinematic inversion technique allowed us to retrieve a complete kinematic description of the source process on the fault plane.Our results reveal that the integration of real data with synthetics is quite efficient, providing reliable shaking maps mainly when near source recordings are scarce. However, the accuracy of the fault plane position plays a major role in increasing the effectiveness of the results.We then apply the methodology to a poorly instrumented earthquake of similar magnitude, the 1980, Ms 6.9, Irpinia (Southern Italy) earthquake. When the peak motions inferred from synthetic seismograms are included in the database, the fit with respect to the observed Mercalli–Cancani–Sieberg intensities improves.